CN118114855B

CN118114855B - A method and system for intelligently controlling community digital information

Info

Publication number: CN118114855B
Application number: CN202311715741.3A
Authority: CN
Inventors: 毛春淼; 侯义古
Original assignee: Community Rubik's Cube Hunan Digital Technology Co ltd
Current assignee: Shaanxi Longrui Weiyuan Technology Co ltd
Priority date: 2023-12-14
Filing date: 2023-12-14
Publication date: 2025-06-17
Anticipated expiration: 2043-12-14
Also published as: CN118114855A

Abstract

The present invention relates to the field of data processing technology, and in particular to a method and system for intelligent control of community digital information. The method comprises the following steps: obtaining historical community equipment data; optimizing the data interface design of the equipment based on the historical community equipment data to generate an equipment identification data interface; using the equipment identification data interface to optimize the data reception of the community equipment for data collection tasks to obtain effective community equipment data; extracting unidentified personnel behavior data from the effective community equipment data to generate unidentified personnel behavior data; performing unidentified personnel behavior anomaly analysis on the unidentified personnel behavior data to generate unidentified personnel behavior anomaly analysis data; transmitting the unidentified personnel behavior anomaly analysis data to the automated early warning engine to perform automated early warning control tasks. The present invention realizes more intelligent control of community digital information.

Description

Intelligent control method and system for community digital information

Technical Field

The invention relates to the technical field of data processing, in particular to an intelligent control method and system for community digital information.

Background

With the rapid development of the information technology of the present society and the complexity of community management, the popularization of digital technology and communities face huge and diverse information flows, and the aspects of residents, equipment, safety, environment and the like are involved, so that the effective management of the information is important to the improvement of the operation efficiency of the communities and the life quality of the residents. The community operation condition is known more comprehensively through intelligent control of community digital information, reasonable distribution of resources and timely early warning of risks are achieved, accordingly community management is scientific and efficient, participation and satisfaction of community residents are promoted, intelligent information control is applied, residents can acquire community services and participate in community activities more conveniently, sharing and interaction of information are achieved, recognition and satisfaction of the residents to communities are improved, and a tighter community group is formed. And through real-time monitoring and analysis of safety information inside and outside the community, potential safety risks can be responded quickly, emergency response capability of the community is improved, and accordingly life and property safety of residents is guaranteed. However, the traditional intelligent control method for the digital information of the community does not monitor whether the community equipment has abnormal operation conditions in time, so that the community equipment cannot collect the digital information of the community comprehensively, and moreover, unidentified external personnel need to judge whether abnormal behaviors exist manually, so that manpower and time are consumed, and early warning feedback of the community is not timely enough.

Disclosure of Invention

Based on the above, the invention provides a method and a system for intelligently controlling digital information of a community, so as to solve at least one of the technical problems.

In order to achieve the purpose, the intelligent control method for the community digital information comprises the following steps:

step S1, acquiring historical community equipment data, designing a data interface of equipment based on the historical community equipment data, and generating an equipment data interface;

Step S2, designing the equipment identity of the equipment data interface according to the historical community equipment data to generate equipment identity data;

step 3, when the community equipment executes a data acquisition task, acquiring equipment signals of the community equipment to generate equipment signals, dividing the community equipment with abnormal signals according to the equipment signals to respectively obtain effective community equipment and ineffective community equipment;

Step S4, performing data receiving of a data acquisition task on the optimized community equipment by utilizing an equipment identification data interface to obtain effective community equipment data, performing real-time acquisition on monitoring data of the community equipment on the effective community equipment data to generate multi-mode monitoring data, performing unidentified face image data extraction on the multi-mode monitoring data to generate unidentified face image data, performing unidentified personnel behavior data extraction on the multi-mode monitoring data according to the unidentified face image data to generate unidentified personnel behavior data;

Step 5, acquiring a historical personnel behavior anomaly training sample, establishing a relationship model for unidentified personnel behavior anomaly analysis by utilizing a convolutional neural network algorithm and the historical personnel behavior anomaly training sample to generate a behavior anomaly analysis model, transmitting unidentified personnel behavior data to the behavior anomaly analysis model for unidentified personnel behavior anomaly analysis, and generating unidentified personnel behavior anomaly analysis data;

and step S6, an automatic early warning engine is built based on the unidentified personnel behavior abnormality analysis data, and the unidentified personnel behavior abnormality analysis data is transmitted to the automatic early warning engine to execute an automatic early warning control task.

According to the method and the device, the historical running condition of the community equipment and the historical data collected by the community equipment are comprehensively known by acquiring the historical community equipment data. The device data interface design is carried out based on the historical community device data, the device data interface is generated, the standardization and unified management of the device data are facilitated, the data collected by different brands of devices can be managed in a unified mode, the expandability and the interoperability of the system are improved, a foundation is provided for real-time community data collection, and meanwhile an accurate reference basis is provided for device management and maintenance. And the device identity identification design of the device data interface is carried out according to the historical community device data, so that the device identity identification data with the unique identification is generated, and the identification degree and the management accuracy of the system to community devices are improved. The identity of the equipment data interface is optimized based on the equipment identity data, and the equipment identity data interface is generated, so that the readability and the relevance of the equipment data are further enhanced, the data interface request of community equipment can be effectively identified and responded, acquired data are acquired from the community equipment, the intelligent degree and the operation efficiency of the whole system are improved, the data interaction capability is enhanced, and the community equipment with multiple brands is integrated into the system in a unified manner for processing. The method has the advantages that the community equipment is subjected to data acquisition tasks, timely acquisition of equipment signals is achieved, detailed equipment signal data are further generated, invalid equipment with abnormal signals in the community equipment can be effectively divided through analysis of the signals, meanwhile, normal effective community equipment is obtained, the invalid community equipment is fed back to the terminal to execute repairing operation tasks, automatic detection and repairing of the abnormal equipment are achieved, stability and reliability of the community equipment are improved, real-time monitoring and management of the state of the community equipment are achieved, influence of potential faults on stability of the whole system is effectively reduced, and reliability and maintenance efficiency of the community digital information operation system are improved. The method comprises the steps of utilizing a device identification data interface to perform data receiving of a data acquisition task on optimized community equipment, successfully generating accurate and effective community equipment data, enabling multi-mode data of communities to be acquired in real time, further performing real-time acquisition of multi-mode monitoring data, including monitoring frequency, audio and other data sources of the community, further realizing extraction of unidentified face image data by processing the multi-mode monitoring data, further providing the capability of identifying unidentified personnel, providing a basis for subsequent unidentified personnel behavior anomaly analysis, and further mining unidentified personnel behavior information in the multi-mode data, and realizing real-time monitoring of community equipment and intelligent security protection through the deep data processing process, the high-level functions in the aspects of abnormal behavior detection and the like provide sufficient data support, and the intelligent level and the safety of the community digital information control system are further improved. By acquiring historical personnel behavior anomaly training samples, an analysis basis is laid for unidentified personnel behavior anomalies, a high-efficiency relation model is established by utilizing the training samples through a convolutional neural network algorithm, a behavior anomaly analysis model is formed, the unidentified personnel anomaly can be identified more accurately by establishing the model, the community activity deep analysis is realized, unidentified personnel behavior anomaly analysis data are included in the model, behavior anomaly data are included in the model, behavior normal data are included, the system is more comprehensive when identifying anomalies, the model is trained through the training samples in training, and the characteristics of different behaviors can be learned by the model, so that the accuracy and sensitivity of the anomaly behavior are improved, and the unidentified personnel anomaly can be identified more accurately. An automated early warning engine is successfully established based on the unidentified personnel behavioral anomaly analysis data. By transmitting detailed unidentified personnel behavior anomaly analysis data to an early warning engine, real-time monitoring and identification of the abnormal behavior are realized, the engine is combined with a behavior anomaly analysis model, can rapidly respond to potential risks and execute an automatic early warning control task, and because the engine is established on the basis of a deep learning model, the engine can accurately identify the abnormal behavior from a large amount of data, improves the automatic early warning capability of the potential risks, through real-time transmission analysis data, the engine can timely feed back abnormal conditions, and then can take rapid and effective measures, including sending out alarms, linking other safety devices and the like, so that the overall safety of communities is enhanced, real-time monitoring and early warning of behavior abnormality of unidentified personnel are realized, powerful safety guarantee is provided for community management, and the community management can more intelligently cope with potential safety risks. Therefore, the intelligent control method for the community digital information can timely monitor whether the community equipment has abnormal operation conditions, timely remind management personnel of repairing the community equipment through the terminal, comprehensively collect the community digital information, intelligently judge whether unidentified external personnel have abnormal behaviors, reduce manpower to judge the behaviors of the external personnel, and automatically perform early warning feedback of the community according to the judged abnormal behaviors.

The present disclosure provides a community digital information intelligent control system, configured to execute the community digital information intelligent control method described above, where the community digital information intelligent control system includes:

the device data interface design module is used for acquiring historical community device data, carrying out data interface design of the device based on the historical community device data and generating a device data interface;

The device data interface optimization module is used for carrying out device identity design of a device data interface according to historical community device data to generate device identity data;

The community equipment anomaly analysis module is used for acquiring equipment signals of the community equipment to generate equipment signals when the community equipment performs a data acquisition task, dividing the community equipment with abnormal signals according to the equipment signals to respectively obtain effective community equipment and ineffective community equipment, and feeding the ineffective community equipment back to the terminal to perform an ineffective community equipment repairing control task;

The system comprises an unidentified personnel behavior data acquisition module, a multi-mode monitoring data acquisition module, an unidentified face image data acquisition module, a unidentified personnel behavior data generation module and a multi-mode monitoring module, wherein the unidentified personnel behavior data acquisition module is used for receiving data of a data acquisition task of optimized community equipment by utilizing an equipment identification data interface so as to obtain effective community equipment data;

The system comprises a non-identified personnel behavior data analysis module, a relation model building and behavior abnormality analysis module, a behavior abnormality analysis module and a behavior analysis module, wherein the non-identified personnel behavior data analysis module is used for acquiring a historical personnel behavior abnormality training sample;

and the automatic early warning control module is used for establishing an automatic early warning engine based on the unidentified personnel behavior abnormality analysis data, and transmitting the unidentified personnel behavior abnormality analysis data to the automatic early warning engine to execute an automatic early warning control task.

The method has the advantages that the historical community equipment data are obtained and format analysis of the equipment data is carried out, so that the system can better understand and process the structures of different equipment data, the equipment data interface design is carried out based on the equipment format data and the data transmission protocol, and standardized and normalized transmission of the equipment data is realized. The system has the advantages that the capability of fully utilizing historical data is improved, different equipment types and data formats can be more intelligently adapted through format analysis and interface design, so that maintainability and integrity of the system are improved, the system is beneficial to more efficiently collecting, monitoring and analyzing data of community equipment of different brands, and a solid foundation is provided for intelligent control of community digital information. The equipment identity design of the equipment data interface is carried out through the equipment data source information, equipment identity data with source identification is generated, different equipment is identified and identified more accurately, the accuracy and the integrity of the equipment identity information are enhanced, and the data interface identity is optimized, so that the system is more flexibly adapted to the equipment identity information at the data interface level, and the identification and response capability of different equipment is improved. Through deep knowledge of the data source of the equipment and establishment of accurate identity, the data interface is more intelligent, so that the identification and adaptability of the equipment are improved, and the processing capacity of diversified equipment data is enhanced. The device signal acquisition is carried out, the device state information can be timely acquired, the Fourier transformation technology is utilized to carry out spectrogram conversion on the device signal and the abnormality detection is carried out by utilizing the device signal spectrogram abnormality detection algorithm, so that the system is helped to more comprehensively understand the frequency spectrum characteristics of the device signal, the abnormal condition of the device signal can be distinguished in real time, the community device is divided into effective or invalid community devices, the automatic classification of the device state is realized, the invalid community device is fed back to the terminal to execute the invalid community device repairing operation task, the action can be timely taken to process the invalid device, and the overall reliability of the system is improved. By means of real-time monitoring and automatic classification of equipment signals, interference of invalid equipment to normal operation of the system is reduced, efficiency and response speed of community equipment management are improved, the system is facilitated to maintain community equipment better, and smooth operation of the digital information control system is ensured. And the device identification data interface is utilized to perform data receiving of a data acquisition task on the effective community device, so that a foundation is provided for subsequent relevance analysis. according to the optimized community equipment, the community equipment relevance analysis is carried out, the multi-mode community relevance data matrix is established, the relation among different equipment can be known in depth, the integration and relevance of the multi-mode data of different community equipment are realized, the face image data extraction of the monitoring data is carried out on the multi-mode community relevance data matrix, the monitoring face image data is generated, extracting unidentified face image data from the monitoring face image data according to the acquired owner identification face image data, so as to analyze which external people are unidentified people not in the community, analyze the behavior of unidentified people, provide further guarantee for the management of the community, and obtain the information of the unidentified people through comprehensive data acquisition, The association and analysis realize intelligent monitoring and management of community equipment and personnel behaviors, are beneficial to improving community safety and management efficiency, and provide more comprehensive and intelligent support for intelligent control of community digital information. The method comprises the steps of establishing a mapping relation of behavior anomaly analysis of unidentified personnel by using a convolutional neural network algorithm, generating an initial behavior anomaly analysis model, and performing model training by using a historical personnel behavior anomaly training sample, so that the model can learn and understand normal modes of behaviors of different personnel, a more accurate and highly-adaptive behavior anomaly analysis model is generated, behavior data of unidentified personnel are transmitted to the behavior anomaly analysis model to perform behavior anomaly analysis of unidentified personnel, behavior anomaly analysis data of unidentified personnel are generated, and abnormal behaviors of unidentified personnel can be identified and analyzed in real time, so that subsequent automatic early warning control is performed according to the analysis data. By establishing the deep learning model, the behaviors of unidentified personnel are intelligently analyzed and monitored in real time, so that the accuracy and instantaneity of the system on abnormal behaviors in communities are improved, and the sensing and processing capacity on abnormal conditions are enhanced. This provides a powerful support for the security management of communities. An automatic early warning engine is established based on the unidentified personnel behavior anomaly analysis data, early warning operation can be automatically executed, and the automatic user early warning engine discriminates whether to trigger the early warning operation or not according to the unidentified personnel behavior anomaly analysis data. When the data received by the engine corresponds to abnormal behavior, the equipment closest to the monitored abnormal behavior is subjected to early warning, and the positioning capability of the early warning is improved. On the other hand, when the behavior of the unidentified personnel corresponding to the behavior normal data is normal, the unidentified personnel behavior data corresponding to the behavior normal data is transmitted to the terminal to perform personnel abnormal behavior feedback, so that real-time monitoring of unidentified personnel is eliminated, and an administrator and monitoring equipment can monitor other unidentified personnel. Through automatic early warning engine, realized real-time supervision and automatic response to unusual action, improved automation early warning ability and response speed, through feeding back the normal data of action to the terminal, realized not having the timely confirmation of unusual action to personnel, reduced the false alarm rate for community digital information intelligent control system is more intelligent, accurate and reliable.

Drawings

FIG. 1 is a schematic flow chart of steps of an intelligent control method for community digital information;

FIG. 2 is a flowchart illustrating the detailed implementation of step S3 in FIG. 1;

FIG. 3 is a flowchart illustrating the detailed implementation of step S4 in FIG. 1;

FIG. 4 is a flowchart illustrating the detailed implementation of step S6 in FIG. 1;

The achievement of the objects, functional features and advantages of the present invention will be further described with reference to the accompanying drawings, in conjunction with the embodiments.

Detailed Description

The following is a clear and complete description of the technical method of the present patent in conjunction with the accompanying drawings, and it is evident that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present invention without making any inventive effort, are intended to fall within the scope of the present invention.

Furthermore, the drawings are merely schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and thus a repetitive description thereof will be omitted. Some of the block diagrams shown in the figures are functional entities and do not necessarily correspond to physically or logically separate entities. The functional entities may be implemented in software or in one or more hardware modules or integrated circuits or in different networks and/or processor methods and/or microcontroller methods.

It will be understood that, although the terms "first," "second," etc. may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another element. For example, a first element could be termed a second element, and, similarly, a second element could be termed a first element, without departing from the scope of example embodiments. The term "and/or" as used herein includes any and all combinations of one or more of the associated listed items.

In order to achieve the above objective, referring to fig. 1 to 4, the present invention provides an intelligent control method for digital information of a community, comprising the following steps:

In the embodiment of the present invention, as described with reference to fig. 1, the method for intelligently controlling the digital information of the community according to the present invention includes the following steps:

In the embodiment of the invention, historical community equipment data is acquired from a community database, the historical community equipment data comprises data related to community equipment such as historical acquisition data of community equipment, and monitoring cameras of a certain community are taken as an example to collect past historical monitoring data comprising information such as videos, images, equipment running states and the like. The data interface design of the equipment is carried out to ensure that the data of the historical community equipment can be efficiently utilized by the system, a flexible and extensible RESTful API design is selected to meet the access requirements of different types of equipment, a data exchange format, a protocol, an access endpoint and the like are defined based on the data of the historical community equipment, and a unified data interface is established for the community equipment. This includes the receipt of camera video streams, the acquisition of sensor data, and the interaction of other device information, generating a device data interface.

In the embodiment of the invention, the historical community equipment data is analyzed, key identity information of equipment such as equipment model, manufacturer, unique identification code and the like is identified, and taking a monitoring camera as an example, the information such as the position, model, installation time and the like of the camera can be focused, and the information forms the basis of equipment identity identification. A device identity data structure containing these key identity information is designed to include a JSON object containing device model fields, manufacturer fields, unique identification code fields, etc., so that the device identity information can be clearly represented and transferred. In order to optimize the identity of the data interface, the equipment identity data is embedded into the equipment data interface to form the equipment identity data interface, the RESTful API specification is adopted, and the equipment identity information is transmitted as part of the HTTP header or the request body, so that the data of the specific equipment can be tracked and managed more easily, and the maintainability of the data is improved.

In the embodiment of the invention, when community equipment executes a data acquisition task, equipment signal acquisition is implemented for each equipment, and taking early warning equipment as an example, data receiving and signal transmission and the like can be acquired, or data flow signals and the like are acquired in a monitoring camera, and the acquired data form equipment signals. And dividing the signal abnormality of the community equipment according to the equipment signal. By setting a suitable threshold and historical signal conditions, the signals can be converted into spectrograms by utilizing a Fourier transform technology so as to be easier to perform signal point analysis and calculation, signals which are contrary to normal operation can be detected, the devices are divided into effective community devices and ineffective community devices, and for example, when the received signals of the early warning device are obviously higher than those of the normal conditions, the ineffective devices can be determined. And (3) feeding back the information to the terminal by implementing a feedback mechanism for the invalid community equipment, wherein in a terminal system, a task queue can be established, the invalid community equipment identification is transmitted to a module for executing an invalid community equipment repairing operation task, and the module can comprise an automatic repairing program, notify related personnel or trigger maintenance personnel to repair. In order to improve accuracy and timeliness, equipment signals can be acquired in real time, and the equipment signals are checked regularly, so that the division and repair of abnormal equipment can be completed in the shortest time.

In the embodiment of the invention, the standardized protocol for communication with the community equipment is established through the equipment identification data interface, so that stable transmission of data is ensured, and the equipment identification data interface is used for connecting with the community equipment in a data receiving stage, so that the data acquisition task of optimizing the community equipment, such as accessing a monitoring camera, is realized, and video stream and other related data are acquired. In the process of carrying out monitoring data real-time acquisition on effective community equipment data, a real-time image processing technology such as an open source image processing library (OpenCV) is utilized to extract key information such as personnel positions, moving directions and the like so as to ensure timeliness of the data, and the step of generating multi-mode monitoring data comprises integrating different types of data such as video, temperature, humidity and the like to form a comprehensive monitoring data set. The method is helpful for comprehensively knowing the state of community equipment and providing more information for subsequent behavioral exception analysis. The extraction of unidentified face image data is performed on the multimodal monitoring data by image processing techniques, such as face recognition models. The method comprises the steps of face detection, feature extraction and the like to obtain unidentified face images in the monitoring data. Extraction of unidentified person behavior data is performed based on unidentified face image data, and detailed data about unidentified person behavior, such as movement trajectories, dwell times, and the like, is obtained by analyzing information of actions, positions, and the like of persons in the monitored data.

In the embodiment of the invention, in the aspect of acquiring historical abnormal personnel behavior training samples, monitoring data of past time in communities is collected, wherein the monitoring data comprise video clips of personnel behaviors, the video clips are labeled by an administrator to design labels of the abnormal behaviors, for example, the abnormal behaviors comprise entering forbidden areas, abnormal movement patterns or continuous stay of the personnel, and the samples form a training data set. The method comprises the steps of selecting a Convolutional Neural Network (CNN) as an algorithm of behavioral exception analysis, preprocessing a training sample, including video frame extraction, image enhancement and the like, so as to ensure the quality of input data, designing a structure of a behavioral exception analysis model based on the structure of the convolutional neural network, including a convolutional layer, a pooling layer and a full-connection layer, so as to capture spatial characteristics and time sequence information in images, training the behavioral exception analysis model by adopting an open source deep learning framework (such as TensorFlow or PyTorch), repeatedly training the behavioral exception analysis model through a training data set, and adjusting model parameters so as to capture normal and abnormal modes of personnel behaviors to the greatest extent, thereby obtaining the behavioral exception analysis model. The model can analyze the behaviors of unidentified personnel in real time through the monitoring video, detect whether the abnormal situation exists by setting a threshold value, and trigger an alarm or take corresponding control tasks when the abnormal probability output by the model exceeds the threshold value. The unidentified personnel behavior data is transmitted to the model for analysis, including pedestrian behaviors, vehicle movement tracks and the like captured by the monitoring cameras, and the model evaluates the data to identify possible abnormal behaviors such as running, abnormally long stay time and the like.

In the embodiment of the invention, an automatic early warning engine is built based on unidentified personnel behavior abnormality analysis data, a real-time flow processing framework such as APACHE FLINK or SPARK STREAMING is adopted for the built automatic early warning engine to ensure quick response to abnormal behaviors, an automatic early warning rule, an automatic early warning decision and the like are built based on the unidentified personnel behavior abnormality analysis data and used for analyzing and judging unidentified personnel behavior abnormality, the automatic early warning rule is built based on the unidentified personnel behavior abnormality analysis data, an alarm is triggered when conditions such as abnormal speed, abnormal stay time or abnormal aggregation are monitored, and the rules are integrated into an automatic rule. And if the automatic early warning rule is triggered to execute the early warning task of the automatic early warning decision, executing the safety feedback task of the automatic early warning decision for triggering the automatic early warning rule. The method comprises the steps of transmitting unidentified personnel behavior anomaly analysis data to an automatic early warning engine in real time, and realizing an automatic early warning decision based on an automatic early warning rule after the engine receives the data, wherein the automatic early warning decision comprises an early warning control task and a safety feedback task, the early warning control task comprises triggering of real-time warning notification or real-time monitoring enhancement of an anomaly area, and the safety feedback task comprises normal behavior of the monitored unidentified personnel so as to cancel monitoring. The engine can also generate detailed reports recording the time, place and related information of occurrence of the abnormal event for subsequent analysis and optimization of the system.

Preferably, step S1 comprises the steps of:

Step S11, acquiring historical community equipment data;

step S12, carrying out format analysis on the equipment data according to the historical community equipment data to generate equipment format data;

and step S13, designing a data interface of the equipment based on the equipment format data and the data transmission protocol, and generating an equipment data interface.

According to the method and the device, the historical running condition of the community equipment and the historical data collected by the community equipment are comprehensively known by acquiring the historical community equipment data. And carrying out format analysis on the device data according to the historical community device data, and further designing based on the device format data when designing a data interface by analyzing the data format of the data collected by the community device, so that community devices of different brands can be uniformly processed in the system. Through carrying out equipment data interface design based on equipment format data and data transmission protocol, standardization and standardization of equipment data are realized, so that the data can be better understood and processed in the system, high-efficiency transmission of the equipment data and good butt joint of different community equipment are ensured, and the stability and expandability of the whole system are improved.

In the embodiment of the invention, historical community equipment data of various community equipment are obtained from a community database, and monitoring equipment, early warning equipment and the like are taken as examples. The data may include video streams of the camera, signal transmission traffic of the video, pre-warning signal transmission traffic, etc. collected data, ensuring that there is enough historical information for subsequent analysis and interface design. For the early warning audio data of the early warning device, the vibration size, the sampling frequency and the like of the audio data can be analyzed, and the specific characteristics of each type of device data can be determined, so that standardized device format data, such as camera information expressed in a JSON format, can be obtained, besides the video format data of the camera, the camera information also comprises key information such as position, model, manufacturer and the like, and the device format data becomes the basis of subsequent interface design. The design of the data interface is performed by using the device format data, which selects a proper data transmission protocol, such as HTTP or MQTT, according to the universality of the data interface, and defines corresponding API endpoints, for example, for camera data, an API endpoint capable of acquiring real-time video streams is designed, so that the data of different types of devices in the community can be received and processed by the system in a unified mode, and the device data interface is generated.

Preferably, step S2 comprises the steps of:

s21, extracting equipment data source information from historical community equipment data to generate equipment data source information;

s22, designing equipment identity of an equipment data interface according to the equipment data source information to generate equipment identity data;

step S23, data interface identity optimization is carried out on the equipment data interface based on the equipment identity data, and the equipment identity data interface is generated.

According to the method, the device data source information is extracted from the historical community device data, which community device the device data is acquired from is comprehensively known, and a finer reference is provided for subsequent identity design. The equipment identity identification design of the equipment data interface is carried out through the equipment data source information, the equipment identity identification data with the source identification is generated, the equipment can be identified through associating the equipment data with the source information, the source of the data can be traced more accurately, and the tracing precision and traceability of the system to the data are improved. The data interface identification optimization is carried out on the equipment data interface based on the equipment identification data, the equipment identification information is more flexibly adapted to the data interface level through the data interface identification optimization, the identification and response capacities of different equipment are improved, and the data traceability of the real-time data acquired later is improved.

In the embodiment of the invention, the data source information of the equipment is extracted from the historical community equipment data, and the method comprises the step of determining the data source of each equipment, such as a camera, a sensor, an access control system and the like. In an embodiment, a device data source information table is obtained that contains device IDs and corresponding data source types for subsequent identification designs. The device data interface is designed with device identity according to the device data source information, and a device identity data structure containing the two information is created by taking the device ID and the data source type as examples, and the structure can be a unique identifier, so that the uniqueness of the device identity in the system is ensured. Optimizing the identity of the device data interface using the device identity data includes adding device identity information to the API request header or body, such as in the RESTful API design, including the device ID and data source type in each request to enable accurate identification and processing of the data for the particular device.

Preferably, step S3 comprises the steps of:

step S31, when community equipment executes a data acquisition task, equipment signal acquisition is carried out on the community equipment, and equipment signals are generated;

s32, performing spectrogram conversion of the equipment signal on the equipment signal by utilizing a Fourier transform technology to generate an equipment signal spectrogram;

S33, performing spectrogram anomaly detection calculation on the equipment signal spectrogram by using an equipment signal spectrogram anomaly detection algorithm to generate spectrogram anomaly detection data;

Step S34, community equipment is divided according to the spectrogram anomaly detection data, when the spectrogram anomaly detection data is smaller than a preset anomaly spectrogram detection threshold value, community equipment corresponding to the spectrogram anomaly detection data is marked as effective community equipment, and when the spectrogram anomaly detection data is not smaller than the preset anomaly spectrogram detection threshold value, community equipment corresponding to the spectrogram anomaly detection data is marked as ineffective community equipment;

and step S35, feeding back the invalid community equipment to the terminal to execute an invalid community equipment repairing operation task.

The method and the system acquire the device signals of the community devices, so that the operation signals of the devices can be acquired in real time, and a data base is provided for subsequent analysis of the community devices. The Fourier transform technology is utilized to perform the spectrogram conversion of the equipment signal, the distribution condition of the equipment signal on the frequency domain is reflected through the spectrogram, the deeper understanding of the signal characteristics and the clearer reflection of the abnormal condition of the signal are facilitated, and the analysis precision of the equipment signal is improved. The equipment signal spectrogram anomaly detection algorithm is utilized to carry out spectrogram anomaly detection calculation on the equipment signal spectrogram, and if the frequency of the spectrogram deviates greatly from the frequency of the spectrogram under the conventional condition, the result of spectrogram anomaly detection data output by the algorithm is higher, the algorithm realizes anomaly detection on the equipment signal, and the sensitivity to equipment state change is further improved. The community equipment is divided according to the spectrogram anomaly detection data, so that the anomaly equipment can be accurately identified, and the intelligent discrimination capability of the equipment running state is improved. The invalid community equipment is fed back to the terminal to execute the repairing and controlling task of the invalid community equipment, and the information of the invalid equipment is fed back to the terminal automatically, so that the maintenance of the equipment state is realized, the abnormal time of the manual monitoring equipment is reduced, the maintainability and the stability of the whole system are improved, and the health management and maintenance efficiency of the community equipment are improved. The method realizes comprehensive monitoring, anomaly detection and automatic restoration of the state of the community equipment, and improves the real-time performance, intelligence and automation level of the community digital information control system.

As an example of the present invention, referring to fig. 2, a detailed implementation step flow diagram of step S3 in fig. 1 is shown, where step S3 includes:

In the embodiment of the invention, when community equipment executes a data acquisition task, the data are acquired by using corresponding equipment such as a sensor, a monitoring camera and the like. Taking a monitoring camera as an example, we collect video transmission signals of the camera, which forms the basis of the device signals.

In the embodiment of the invention, the Fourier transform technology is applied to perform spectrogram conversion on the equipment signal, the equipment signal is converted from the time domain to the frequency domain, the spectrogram of the equipment signal is obtained, and for the video transmission signal spectrogram of the monitoring camera, the analysis of the frequency domain information of signal transmission, such as the peak value, the average value and the like of the transmission signal, is facilitated.

in the embodiment of the invention, the spectrogram is calculated by using the equipment signal spectrogram anomaly detection algorithm, wherein the equipment signal spectrogram anomaly detection algorithm is used for calculating the deviation condition of frequency data in the spectrogram and frequency data of the spectrogram under normal conditions, and considering the spectrogram frequency error caused by external signal interference, so as to judge whether the equipment signal spectrogram is abnormal or not, generate spectrogram anomaly detection data, and the conventional spectrogram anomaly detection algorithm can also be used for carrying out spectrogram anomaly detection calculation, but has lower accuracy of data accuracy.

in the embodiment of the invention, community equipment is divided based on spectrogram anomaly detection data. When the spectrogram anomaly detection data is smaller than a preset anomaly spectrogram detection threshold value, marking community equipment corresponding to the smaller spectrogram anomaly detection data as effective community equipment, namely that the anomaly condition of the community equipment is in an anomaly error range; when the spectrogram anomaly detection data is not smaller than a preset anomaly spectrogram detection threshold, marking the community equipment corresponding to the smaller spectrogram anomaly detection data as invalid community equipment, namely, the anomaly condition of the community equipment is out of the anomaly error range.

In the embodiment of the invention, the information marked as the invalid community equipment is fed back to the terminal, so that a task queue is set in the terminal, the invalid community equipment identification is transmitted to a module for executing the repairing operation task of the invalid community equipment, and the module comprises the steps of sending a notification to related personnel, triggering maintenance tasks or performing other targeted repairing operations to ensure the normal operation of the community equipment.

Preferably, the device signal spectrogram anomaly detection algorithm in step S33 is as follows:

Wherein D (f) is represented as spectrogram anomaly detection data, f is represented as a frequency parameter of the device signal spectrogram, σ is represented as a standard deviation of a conventional frequency parameter, μ is represented as a conventional frequency parameter, α is represented as a correction adjustment value of an external interference frequency parameter of the device signal spectrogram, γ is represented as a frequency attenuation rate of the device signal spectrogram, δ is represented as an offset of an attenuation frequency parameter of the device signal spectrogram, and η is represented as an attenuation frequency parameter of the device signal spectrogram.

The invention utilizes an abnormality detection algorithm of a device signal spectrogram, which comprises a frequency parameter f of the device signal spectrogram, a standard deviation sigma of a conventional frequency parameter, a conventional frequency parameter mu, a correction adjustment value alpha of an external interference frequency parameter of the device signal spectrogram, a frequency attenuation rate gamma of the device signal spectrogram, an offset delta of an attenuation frequency parameter of the device signal spectrogram, an attenuation frequency parameter eta of the device signal spectrogram and an interaction relation among functions to form a functional relation:

That is to say, The function relation considers the abnormal value of the signal spectrogram of the equipment by considering the frequency parameter of the signal spectrogram of the equipment and the difference between the signal spectrograms corresponding to the conventional equipment in history, so as to judge whether the equipment has abnormal conditions, for example, if the deviation between the transmission signal of the equipment and the signal of the conventional equipment in the same class is larger, the result of the abnormal detection data of the corresponding generated spectrogram is higher.

The expected distribution of the spectrogram under the normal condition is described, whether the frequency of the spectrogram is abnormal or not is further preliminarily judged, the external interference possibly existing in the spectrogram is corrected by introducing a cosine term, alpha is a parameter for adjusting the correction amplitude, the algorithm is helped to correct the external interference factors in the spectrogram, the robustness to the interference frequency is improved, and the algorithm is more robust under the condition of noise or interference.The frequency attenuation factors are described, wherein gamma, delta and eta are used for adjusting specific forms of attenuation, so that the attenuation characteristics of a spectrogram can be sensitively captured, and the frequency attenuation factors have certain adaptability to time variation of the spectrogram. The whole functional relation realizes the anomaly detection of the whole spectrogram by carrying out integral operation on various spectrogram characteristics in the whole frequency domain range, so that the comprehensive response of the algorithm to different frequency components is more comprehensive, and the analysis capability of the algorithm to the complex spectrogram is enhanced. The functional relation comprehensively considers a plurality of factors such as normal distribution of a spectrogram, external interference, frequency attenuation and the like in the aspect of anomaly detection, so that the abnormal situation can be detected more accurately in a complex environment, and a comprehensive and flexible solution is provided for anomaly monitoring of equipment signals.

Preferably, step S4 comprises the steps of:

s41, performing data receiving of a data acquisition task on effective community equipment by utilizing an equipment identification data interface so as to obtain effective community equipment data;

step S42, carrying out community equipment relevance analysis according to the optimized community equipment to generate community equipment relevance data;

S43, establishing a multi-mode community association matrix node according to community equipment association data, transmitting effective community equipment data to the multi-mode community association matrix node for data filling, and generating a multi-mode community association data matrix;

S44, monitoring data of community equipment is acquired in real time for the multi-mode community association data matrix, and multi-mode monitoring data are generated;

Step S45, extracting face image data of monitoring data of the multi-mode monitoring data to generate monitoring face image data;

step S46, acquiring owner identification face image data of a community database;

step S47, extracting unidentified face image data from the monitoring face image data according to the owner identification face image data to generate unidentified face image data;

And S48, extracting unidentified personnel behavior data from the multi-mode monitoring data according to the unidentified face image data, and generating unidentified personnel behavior data.

According to the invention, the data of the data acquisition task is received by the device identification data interface, so that the real-time information acquisition of the effective community device is realized, the latest data of the effective community device can be accurately and timely acquired, the data acquired by community devices of different brands are uniformly processed, and a reliable data base is provided for subsequent analysis. The community equipment relevance analysis is carried out according to the optimized community equipment, community equipment relevance data are generated, the relation among the equipment can be better understood through the relevance analysis, therefore, the integration of multi-mode data is optimized, the understanding and the utilization of the community equipment relevance are improved, abnormal monitoring behaviors of unidentified personnel are monitored in the follow-up steps, the equipment with the largest relevance can be found out by monitoring equipment for collecting the monitoring data, accurate early warning feedback is carried out, and management personnel can obtain early warning positions, related information and the like. The multi-mode community association matrix nodes are established according to the community equipment association data, and the effective community equipment data are transmitted to the multi-mode community association matrix nodes to be filled with the data, so that the integration and association of the multi-mode data are realized, the association between the community equipment and the corresponding data can be more comprehensively understood, and the management capability of the overall community condition is improved. The monitoring data of the community equipment is acquired in real time by the multi-mode community association data matrix, so that the monitoring data of the community equipment can be acquired more timely, and the real-time monitoring of the state of the community equipment is realized. The facial image data extraction of the monitoring data is carried out on the multi-mode monitoring data, and a basis is provided for subsequent facial image analysis. And acquiring owner identification face image data of the community database, and facilitating identification and association of the system in the multi-mode monitoring data. And extracting unidentified face image data from the monitoring face image data according to the owner identification face image data, and effectively identifying the face image of the non-owner, so that the monitoring precision of the community non-owner personnel is improved. The unidentified personnel behavior data is extracted from the multi-mode monitoring data according to unidentified face image data, so that behavior modes of unidentified personnel can be deeply known, and powerful support is provided for subsequent abnormal behavior analysis.

As an example of the present invention, referring to fig. 3, a detailed implementation step flow diagram of step S4 in fig. 1 is shown, where step S4 includes:

In the embodiment of the invention, the data acquisition is carried out on the effective community equipment through the equipment identification data interface. The method comprises the step of calling an API endpoint to receive relevant data collected by effective community equipment so as to obtain real-time effective community equipment data, such as monitoring video data and the like.

In the embodiment of the invention, the optimized community equipment is subjected to relevance analysis, and the data relevance among the equipment is analyzed, so that the judgment can be carried out through the positions among the equipment and the same type of the equipment, for example, the relevance of the monitoring equipment of the same type is larger, and the relevance of the monitoring equipment and the closer early warning equipment is larger, thereby being convenient for carrying out early warning operation through the early warning equipment with larger relevance when the monitoring equipment monitors abnormal personnel in the subsequent steps, further obtaining community equipment relevance data and being used for establishing a subsequent multi-mode relevance matrix.

In the embodiment of the invention, a multi-mode community association matrix node is established according to the community equipment association data, the effective community equipment data is transmitted to the matrix node for data filling, and a multi-mode community association data matrix is generated, wherein the matrix comprises the association among all the equipment, and a basis is provided for the follow-up monitoring data extraction.

in the embodiment of the invention, the real-time monitoring data acquisition is performed based on the multi-mode community association data matrix, and comprises the steps of acquiring real-time monitoring data, such as monitoring data of a monitoring camera, monitoring data of a door lock camera and the like, from each device to form multi-mode monitoring data.

In the embodiment of the invention, the image processing technology, such as a face detection algorithm, is used to extract the face image contained in the monitoring data from the multi-mode monitoring data, and the face image may include a face detection library (such as OpenCV, dlib, etc.) using a deep learning model or an open source. The extracted facial image is used for subsequent analysis to generate monitoring facial image data.

in the embodiment of the invention, the identification face image data of the owner is obtained from a community database, which can be the face image data of the owner which is input into the system in advance and is used for comparing with the face image in the monitoring data to identify the identity of the owner.

In the embodiment of the invention, the face image data is identified by the owner, the previously extracted monitoring face image data is compared and matched, so that unidentified face image data outside the owner is filtered, the unidentified face image data is realized by means of a face matching algorithm or a feature matching technology (such as FaceNet, arcFace and the like) of an open source resource library of python, and the generated unidentified face image data is used for extracting follow-up unidentified personnel behavior data.

In the embodiment of the invention, based on the extracted unidentified face image data, behavior analysis algorithms or models, such as convolutional neural networks, can be used for analyzing and extracting behaviors of unidentified personnel, including identifying information such as actions, postures, moving paths and the like of the personnel, so as to generate unidentified personnel behavior data.

Preferably, step S5 comprises the steps of:

s51, establishing a mapping relation of unidentified personnel behavior anomaly analysis by using a convolutional neural network algorithm, and generating an initial behavior anomaly analysis model;

Step S52, acquiring a historical personnel behavior abnormality training sample;

s53, performing model training on the initial behavior anomaly analysis model by using a historical personnel behavior anomaly training sample to generate a behavior anomaly analysis model;

And S54, transmitting the behavior data of the unidentified personnel to a behavior anomaly analysis model to perform behavior anomaly analysis of the unidentified personnel, and generating behavior anomaly analysis data of the unidentified personnel, wherein the behavior anomaly analysis data of the unidentified personnel comprise behavior anomaly data or behavior normal data.

According to the invention, the mapping relation of the behavior anomaly analysis of the unidentified personnel is established by utilizing the convolutional neural network algorithm, so that the initial behavior anomaly analysis model is generated, the complex mapping relation of the behavior of the unidentified personnel can be automatically learned and captured, and the deep understanding of the behavior of the unidentified personnel is facilitated. The historical personnel behavior anomaly training samples are obtained, and representative historical data are provided for the system, so that the accuracy of the model pair is improved during model training. And carrying out model training on the initial behavior abnormality analysis model by using the historical personnel behavior abnormality training sample to generate a behavior abnormality analysis model. Model parameters are continuously optimized through training, so that the model parameters are better adapted to the characteristics of historical data, and the generalization capability and accuracy of the model are improved. The behavior data of the unidentified personnel are transmitted to the behavior anomaly analysis model to conduct behavior anomaly analysis of the unidentified personnel, the behavior anomaly analysis data of the unidentified personnel are generated, the behavior of the unidentified personnel can be monitored in real time, and potential risks and the abnormal behaviors are identified through the anomaly analysis data output by the analysis model. Through combining deep learning modeling and historical data training, accurate analysis and anomaly detection of unidentified personnel behaviors are achieved, the perception capability of potential threats is improved, unidentified personnel behaviors in different scenes can be well adapted and understood, and intelligent and accurate monitoring and analysis support is provided for community safety.

In the embodiment of the invention, a Convolutional Neural Network (CNN) algorithm is used for establishing a mapping relation of unidentified personnel behavior anomaly analysis, an initial behavior anomaly analysis model is generated, and a neural network structure comprising a plurality of convolutional layers, pooling layers and full-connection layers can be designed for extracting features from behavior data of unidentified personnel and carrying out anomaly analysis. In the aspect of acquiring historical abnormal personnel behavior training samples, monitoring data of past time in communities is collected, wherein the monitoring data comprise video fragments of personnel behaviors, the video fragments are labeled by an administrator to design labels of the abnormal behaviors, such as abnormal behaviors and the like, including entering forbidden areas, abnormal movement patterns or continuous stay of the personnel, and the samples form a training data set. Model training is carried out on the initial behavior anomaly analysis model by using a historical personnel behavior anomaly training sample, and model parameters are adjusted by a back propagation algorithm and an optimizer, so that normal and abnormal behavior modes can be accurately identified, and a behavior anomaly analysis model is generated. And transmitting the behavior data of the unidentified person acquired in real time to a trained behavior anomaly analysis model for analysis, wherein the behavior anomaly analysis model can judge whether the input behavior data is normal or abnormal according to the learned mapping relation. The generated unidentified personnel behavior abnormality analysis data can comprise behavior abnormality data and identification of behavior normal data, and a basis is provided for subsequent automatic early warning.

Preferably, the executing the automatic early warning operation task includes executing the early warning operation task on the target early warning device or transmitting unidentified personnel behavior data corresponding to the normal behavior data to the terminal for personnel abnormal behavior feedback, and the step S6 includes the following steps:

Step S61, an automatic early warning engine is built based on unidentified personnel behavior abnormality analysis data;

Step S62, transmitting unidentified personnel behavior abnormality analysis data to an automatic early warning engine for automatic early warning control, executing step S63 when the automatic engine receives the behavior abnormality data corresponding to the unidentified personnel behavior abnormality analysis data, or transmitting unidentified personnel behavior data corresponding to the behavior abnormality analysis data to a terminal for personnel non-abnormal behavior feedback when the automatic engine receives the behavior abnormality data corresponding to the unidentified personnel behavior abnormality analysis data;

And step 63, marking the target early warning equipment for the optimized community equipment according to unidentified personnel data corresponding to the behavior anomaly data so as to obtain the target early warning equipment, and executing an early warning control task for the target early warning equipment.

The invention establishes an automatic early warning engine based on unidentified personnel behavior anomaly analysis data, and lays a foundation for subsequent automatic early warning operation by identifying unusual personnel behavior modes through deep learning and model establishment of the anomaly analysis data. And transmitting the unidentified personnel behavior abnormality analysis data to an automatic early warning engine for automatic early warning control, so that the automatic early warning engine can intelligently judge abnormal behaviors and take different treatment measures under different conditions. And marking the target early warning equipment of the optimized community equipment according to unidentified personnel data corresponding to the behavior abnormal data so as to obtain the target early warning equipment, executing an early warning control task on the target early warning equipment, rapidly positioning the related equipment according to the abnormal behavior data, and rapidly responding and processing abnormal conditions through the marking and the control task of the target early warning equipment. By establishing an automatic early warning engine, the abnormal behavior can be accurately judged, the corresponding control task can be automatically executed, the perception and response speed of the system to the potential risk are improved, and more comprehensive and intelligent guarantee is provided for community safety.

As an example of the present invention, referring to fig. 4, a detailed implementation step flow diagram of step S6 in fig. 1 is shown, where step S6 includes:

In the embodiment of the invention, an automatic early warning engine is built based on unidentified personnel behavior abnormality analysis data, a real-time flow processing framework such as APACHE FLINK or SPARK STREAMING is adopted for the built automatic early warning engine to ensure quick response to abnormal behaviors, an automatic early warning rule, an automatic early warning decision and the like are built based on the unidentified personnel behavior abnormality analysis data and used for analyzing and judging unidentified personnel behavior abnormality, the automatic early warning rule is built based on the unidentified personnel behavior abnormality analysis data, an alarm is triggered when conditions such as abnormal speed, abnormal stay time or abnormal aggregation are monitored, and the rules are integrated into an automatic rule. And if the automatic early warning rule is triggered to execute the early warning task of the automatic early warning decision, executing the safety feedback task of the automatic early warning decision for triggering the automatic early warning rule.

In the embodiment of the invention, unidentified personnel behavior anomaly analysis data are transmitted to an automatic early warning engine in real time, and an automatic early warning decision is realized based on an automatic early warning rule after the engine receives the data, wherein the automatic early warning decision comprises an early warning control task and a safety feedback task, the early warning control task comprises triggering of real-time warning notification or real-time monitoring enhancement of an anomaly area and the like, and the safety feedback task comprises that the monitored unidentified personnel is normal in behavior so as to cancel monitoring.

In the embodiment of the invention, when the behavior anomaly data corresponding to the unidentified personnel behavior anomaly analysis data is received by the automation engine, the automatic user early warning engine executes an early warning control task to locate the target early warning equipment associated with the corresponding monitoring equipment according to the monitored data, and the target early warning equipment executes the corresponding early warning control task, such as triggering an alarm, adjusting monitoring parameters, feeding back the position information of the early warning equipment by a related manager and the like, so as to cope with the detected abnormal behavior.

Preferably, step S63 includes the steps of:

abnormal data node positioning is carried out on the multi-mode community association data matrix according to unidentified personnel data corresponding to the behavior abnormal data, and multi-mode abnormal data nodes are generated;

performing target early warning equipment marking on optimized community equipment according to the multi-mode abnormal data nodes and community equipment relevance data to generate target early warning equipment;

and executing an early warning control task on the target early warning equipment.

According to the method and the system for locating the abnormal data nodes of the multi-mode community association data matrix, the abnormal data nodes of the multi-mode community association data matrix are located according to the unidentified personnel data corresponding to the behavior abnormal data, the nodes with unidentified personnel behavior abnormality of the monitoring data in the multi-mode community association data matrix can be accurately located, and the accurate location of abnormal events is achieved. And marking the optimized community equipment by target early warning equipment according to the multi-mode abnormal data nodes and the community equipment relevance data, and determining equipment closely related to the abnormal event by combining the multi-mode abnormal data nodes and the community equipment relevance data, wherein the equipment is marked as the target early warning equipment, so that the timely attention and the treatment of the potential problem are realized. The target early warning equipment is subjected to early warning control tasks, abnormal events can be responded rapidly through control of the target early warning equipment, and necessary measures are taken to reduce the influence range of potential risk early warning. Through the accurate positioning of the abnormal event and the marking and control of the target early warning equipment, the intelligent early warning and response to the community equipment are realized, the sensing and processing capacity of the system to the abnormal condition is improved, the pressure of manual intervention is relieved, and the automation and intelligent management of the community digital information control system are realized.

In the embodiment of the invention, monitoring nodes are positioned on the multi-mode community association data matrix according to unidentified personnel data corresponding to the behavior abnormal data to obtain multi-mode abnormal data nodes corresponding to the behavior abnormal data, the multi-mode data nodes comprise monitoring equipment information and the like corresponding to the monitoring data, and early warning equipment most relevant to the monitoring equipment is found according to equipment association relations of the multi-mode abnormal data nodes and the multi-mode community association data matrix to mark the early warning equipment as target early warning start to write. The target early warning device executes corresponding early warning control tasks, such as triggering an alarm, adjusting monitoring parameters, feeding back the position information of the early warning device by a relevant manager, and the like, so as to cope with the detected abnormal behavior.

The system comprises an unidentified personnel behavior data analysis module, a behavior anomaly analysis model, a behavior anomaly analysis module and a behavior analysis module, wherein the unidentified personnel behavior data analysis module is used for acquiring a historical personnel behavior anomaly training sample;

The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein.

The foregoing is only a specific embodiment of the invention to enable those skilled in the art to understand or practice the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for intelligently controlling community digital information, comprising the following steps:

Step S1: Acquire historical community device data of various community devices; design the data interface of the device based on the historical community device data to generate the device data interface;

Wherein, step S1 comprises:

Step S11: Obtain historical community device data of various community devices;

Step S12: Performing format analysis of device data based on historical community device data to generate device format data;

Step S13: Designing the data interface of the device based on the device format data and the data transmission protocol, and generating the device data interface;

Step S2: Design the device identity of the device data interface according to the historical community device data to generate device identity data; optimize the data interface identity of the device data interface based on the device identity data to generate a device identification data interface;

Wherein, step S2 comprises:

Step S21: extracting device data source information from historical community device data to generate device data source information;

Step S22: Design the device identity of the device data interface according to the device data source information to generate device identity data;

Step S23: Optimizing the data interface identity of the device data interface based on the device identity data to generate a device identification data interface;

Step S3: When the community device performs the data collection task, the device signal of the community device is collected to generate a device signal; the community device is divided into community devices with abnormal signals according to the device signal, and valid community devices and invalid community devices are obtained respectively; the invalid community devices are fed back to the terminal to perform the invalid community device repair control task;

Step S31: When the community device performs a data collection task, the device signal of the community device is collected to generate a device signal;

Step S32: using Fourier transform technology to convert the device signal into a spectrum diagram of the device signal to generate a spectrum diagram of the device signal;

Step S33: using the device signal spectrum anomaly detection algorithm to perform spectrum anomaly detection calculation on the device signal spectrum to generate spectrum anomaly detection data;

Step S34: classify the community devices according to the spectrum anomaly detection data. When the spectrum anomaly detection data is less than the preset abnormal spectrum detection threshold, mark the community device corresponding to the spectrum anomaly detection data as a valid community device. When the spectrum anomaly detection data is not less than the preset abnormal spectrum detection threshold, mark the community device corresponding to the spectrum anomaly detection data as an invalid community device.

Step S35: Feedback the invalid community device to the terminal to perform the invalid community device repair control task;

Step S4: using the device identification data interface to receive data of the effective community device for the data collection task to obtain the effective community device data; performing real-time collection of monitoring data of the community device on the effective community device data to generate multimodal monitoring data; extracting unidentified facial image data from the multimodal monitoring data to generate unidentified facial image data; extracting unidentified personnel behavior data from the multimodal monitoring data according to the unidentified facial image data to generate unidentified personnel behavior data;

Wherein, step S4 comprises:

Step S41: using the device identification data interface to receive data from a valid community device for a data collection task, so as to obtain valid community device data;

Step S42: Perform community device relevance analysis based on valid community devices to generate community device relevance data;

Step S43: establishing a multimodal community association matrix node according to the community device association data, and transmitting the valid community device data to the multimodal community association matrix node for data filling, thereby generating a multimodal community association data matrix;

Step S44: collecting the monitoring data of community devices in real time on the multimodal community association data matrix to generate multimodal monitoring data;

Step S45: extracting facial image data of the monitoring data from the multimodal monitoring data to generate monitoring facial image data;

Step S46: Acquire owner identification facial image data from the community database;

Step S47: extracting unidentified facial image data from the monitored facial image data according to the owner's identified facial image data to generate unidentified facial image data;

Step S48: extracting unidentified person behavior data from the multimodal monitoring data according to the unidentified facial image data to generate unidentified person behavior data;

Step S5: Obtain historical personnel behavior abnormality training samples; use the convolutional neural network algorithm and the historical personnel behavior abnormality training samples to establish a relationship model for analyzing the abnormal behavior of unidentified personnel, and generate a behavior abnormality analysis model; transmit the unidentified personnel behavior data to the behavior abnormality analysis model to perform the unidentified personnel behavior abnormality analysis, and generate unidentified personnel behavior abnormality analysis data;

Step S6: Establish an automated early warning engine based on the unidentified personnel behavior abnormality analysis data; transmit the unidentified personnel behavior abnormality analysis data to the automated early warning engine to execute the automated early warning control task.

2. The community digital information intelligent control method according to claim 1 is characterized in that the device signal spectrum anomaly detection algorithm in step S33 is as follows:

Where D(f) represents the spectrum anomaly detection data, f represents the frequency parameter of the device signal spectrum, σ represents the standard deviation of the conventional frequency parameter, μ represents the conventional frequency parameter, α represents the correction adjustment value of the external interference frequency parameter of the device signal spectrum, γ represents the frequency attenuation rate of the device signal spectrum, δ represents the offset of the attenuation frequency parameter of the device signal spectrum, and η represents the attenuation frequency parameter of the device signal spectrum.

3. The method for intelligently controlling community digital information according to claim 1, wherein step S5 comprises the following steps:

Step S51: using a convolutional neural network algorithm to establish a mapping relationship for abnormal behavior analysis of unidentified personnel, and generating an initial abnormal behavior analysis model;

Step S52: Obtain historical personnel behavior abnormal training samples;

Step S53: using historical personnel behavior abnormality training samples to perform model training on the initial behavior abnormality analysis model to generate a behavior abnormality analysis model;

Step S54: transmitting the unidentified person behavior data to the behavior anomaly analysis model to perform behavior anomaly analysis on the unidentified person, and generating unidentified person behavior anomaly analysis data, wherein the unidentified person behavior anomaly analysis data includes behavior anomaly data or normal behavior data.

4. The community digital information intelligent control method according to claim 3 is characterized in that the execution of the automated early warning control task includes executing the early warning control task on the target early warning device or transmitting the unidentified personnel behavior data corresponding to the normal behavior data to the terminal for feedback on the personnel's normal behavior, and step S6 includes the following steps:

Step S61: Establishing an automated early warning engine based on the abnormal behavior analysis data of unidentified personnel;

Step S62: transmitting the unidentified personnel behavior abnormality analysis data to the automated early warning engine for automated early warning control. When the automated engine receives the behavior abnormality data corresponding to the unidentified personnel behavior abnormality analysis data, step S63 is executed. Alternatively, when the automated engine receives the behavior normality data corresponding to the unidentified personnel behavior abnormality analysis data, the unidentified personnel behavior data corresponding to the behavior normality data is transmitted to the terminal for feedback on the personnel having no abnormal behavior.

Step S63: Mark the target early warning device on the optimized community device according to the unidentified personnel data corresponding to the abnormal behavior data to obtain the target early warning device, and perform the early warning control task on the target early warning device.

5. The method for intelligently controlling community digital information according to claim 4, wherein step S63 comprises the following steps:

According to the unidentified personnel data corresponding to the abnormal behavior data, the abnormal data nodes of the multimodal community association data matrix are located to generate multimodal abnormal data nodes;

According to the multi-modal abnormal data nodes and community device correlation data, the optimized community devices are marked as target warning devices to generate target warning devices;

Perform early warning control tasks on target early warning equipment.

6. A community digital information intelligent control system, characterized in that it is used to execute the community digital information intelligent control method according to claim 1, and the community digital information intelligent control system comprises:

The device data interface design module is used to obtain historical community device data; the device data interface is designed based on the historical community device data to generate the device data interface;

The device data interface optimization module is used to design the device identity of the device data interface according to the historical community device data and generate the device identity data; optimize the data interface identity of the device data interface based on the device identity data and generate the device identity data interface;

The community device abnormality analysis module is used to collect device signals from community devices and generate device signals when community devices perform data collection tasks; classify community devices with abnormal signals according to device signals to obtain valid community devices and invalid community devices respectively; and feed back invalid community devices to the terminal to perform invalid community device repair and control tasks;

The unidentified personnel behavior data collection module uses the device identification data interface to optimize the data reception of the community equipment for the data collection task to obtain effective community equipment data; collects the monitoring data of the community equipment in real time for the effective community equipment data to generate multimodal monitoring data; extracts the unidentified facial image data from the multimodal monitoring data to generate the unidentified facial image data; extracts the unidentified personnel behavior data from the multimodal monitoring data based on the unidentified facial image data to generate the unidentified personnel behavior data;

The unidentified personnel behavior data analysis module obtains historical personnel behavior abnormality training samples; uses the convolutional neural network algorithm and historical personnel behavior abnormality training samples to establish a relationship model for unidentified personnel behavior abnormality analysis, and generates a behavior abnormality analysis model; transmits the unidentified personnel behavior data to the behavior abnormality analysis model to perform unidentified personnel behavior abnormality analysis, and generates unidentified personnel behavior abnormality analysis data;

The automated early warning and control module establishes an automated early warning engine based on the abnormal behavior analysis data of unidentified personnel; transmits the abnormal behavior analysis data of unidentified personnel to the automated early warning engine to execute the automated early warning and control tasks.