CN114638666A - Method and device for processing Internet of vehicles data - Google Patents

Abstract

The disclosure provides a method and a device for processing Internet of vehicles data. Wherein, the method can comprise the following steps: in the driving process of at least one vehicle to be monitored, the vehicle networking information of the vehicle to be monitored is collected, wherein the vehicle networking information at least comprises: journey information of a vehicle to be monitored; calling a generation driving program identification model deployed at a vehicle-mounted edge end; analyzing the car networking information by adopting a driving program generation identification model, and acquiring the target driving program generation characteristic of the vehicle to be monitored; acquiring a driving demand replacement identification model based on the driving behavior characteristics of the target generation; adopting a designated driving request of a designated driving demand identification model; uploading the designated driving request of the vehicle to be monitored to the cloud end platform, wherein at least one user end is allowed to obtain the designated driving request of the vehicle to be monitored, and the request is pushed by the cloud end platform.

Description

Method and device for processing data of Internet of Vehicles

technical field

The present disclosure relates to the field of data processing, and in particular, to a method and device for processing Internet of Vehicles data.

Background technique

At present, the recommendation of the chauffeur service is mainly based on the user's historical information, obtaining the historical behavior of the platform providing the chauffeur service for them, and inferring the possibility that the user needs a chauffeur according to the frequency of the behavior, and then using the application (Application, referred to as APP for short) ) to send relevant service content pushes or coupons to users. This recommendation method has the technical problems that the user's needs cannot be accurately identified, and the service push efficiency is low.

In view of the above-mentioned technical problems of inability to accurately identify user needs and low efficiency of service push, no effective solution has been proposed yet.

SUMMARY OF THE INVENTION

Embodiments of the present invention provide a method and device for processing Internet of Vehicles data, so as to at least solve the technical problems of inability to accurately identify user needs and low service push efficiency.

According to an aspect of the embodiments of the present invention, a method for processing Internet of Vehicles data is provided, including: collecting the Internet of Vehicles information of the vehicle to be monitored during the driving of at least one vehicle to be monitored, wherein the Internet of Vehicles information at least includes: : Itinerary information of the vehicle to be monitored; call the driver's trip identification model deployed on the edge of the vehicle; use the driver's trip identification model to analyze the Internet of Vehicles information to obtain the target driver's behavior characteristics of the vehicle to be monitored; Driving demand identification model; use the driving demand identification model to determine the driving request of the vehicle to be monitored; upload the driving request of the vehicle to be monitored to the cloud platform, wherein at least one client is allowed to obtain the driving request of the vehicle to be monitored pushed by the cloud platform. driving request.

Optionally, the method further includes: collecting data on the Internet of Vehicles of a plurality of vehicles in a historical time period, wherein the historical time period at least includes the driving time period in which the driving behavior information is generated, and the data on the Internet of Vehicles at least includes: The surrogate driving trip data generated during the driving time period; the machine learning model is trained based on the surrogate driving itinerary data to generate the surrogate driving itinerary identification model; the surrogate driving demand identification model is used to determine the surrogate driving request of the vehicle to be monitored, including: based on the target generation Driving behavior characteristics, and obtain surrogate driver sample data, wherein the surrogate driver sample data includes at least one of the following: the itinerary information in at least one adjacent time period adjacent to the surrogate driver time period, and the surrogate driver generated within the surrogate driver time period. Itinerary data; the machine learning model is trained based on the driver's sample data to generate the driver's demand identification model, wherein the driver's demand identification model is used to identify the corresponding driver's demand characteristics based on the itinerary information of the vehicle to be identified.

Optionally, after uploading the surrogate driving request of the vehicle to be monitored to the cloud platform, the method further includes: detecting whether the cloud platform receives the surrogate driving request; if the surrogate driving request is received, retrieving at least one registered surrogate driving driver. based on the driver information of the driver; determine whether there is an idle driver based on the driver information; determine at least one target driver based on the push rule, and retrieve the device information of the target driver, among which the push rule It is used to determine the push priority of multiple chauffeurs; based on the device information of the target chauffeur, push the chauffeur request to the device held by the target chauffeur.

According to another aspect of the embodiments of the present invention, an apparatus for processing Internet of Vehicles data is provided, including: a collection module configured to collect the Internet of Vehicles information of the vehicle to be monitored during the driving process of at least one vehicle to be monitored, wherein , the Internet of Vehicles information includes at least: the itinerary information of the vehicle to be monitored; the calling module is used to call the driver's trip identification model deployed on the edge of the vehicle; the first acquisition module is used to use the driver's trip identification model to analyze the Internet of Vehicles information, and obtain the target chauffeur-driven behavior characteristics of the vehicle to be monitored; the second acquisition module is used to obtain a surrogate-driving demand identification model based on the target surrogate-driving behavior characteristics; the determining module is used to determine the surrogate-driving request of the to-be-monitored vehicle by using the surrogate-driving demand identification model; The uploading module is used to upload the driving request of the vehicle to be monitored to the cloud platform, wherein at least one user terminal is allowed to obtain the driving request of the vehicle to be monitored pushed by the cloud platform.

Optionally, the device further includes: a collection module for collecting data on the Internet of Vehicles of a plurality of vehicles in a historical time period, wherein the historical time period includes at least the driving time period in which the driving behavior information is generated, and the Internet of Vehicles The data includes at least: the chauffeured itinerary data generated within the chauffeured time period; the first training module is used to train the machine learning model based on the chauffeured itinerary data to generate the chauffeured itinerary identification model; the acquisition module is used to generate the chauffeured itinerary Driving behavior characteristics, and obtain surrogate driver sample data, wherein the surrogate driver sample data includes at least one of the following: the itinerary information in at least one adjacent time period adjacent to the surrogate driver time period, and the surrogate driver generated within the surrogate driver time period. Itinerary data; the second training module is used to train the machine learning model based on the surrogate driver sample data, and generate a chauffeur demand identification model, wherein the surrogate driver demand identification model is used to identify the corresponding vehicle based on the itinerary information of the vehicle to be identified. chauffeur demand characteristics.

Optionally, the device further includes: a detection module, used to detect whether the cloud platform receives a driver's request; a retrieval module, used to retrieve the driver information of at least one registered driver if the driver's request is received; The determination module is used to determine whether there is an idle driver based on the driver information of the driver; the processing module is used to determine at least one target driver based on the push rule, and retrieve the device of the target driver information, where the push rule is used to determine the push priorities of multiple chauffeurs; the push module is used to push the chauffeur request to the device held by the target chauffeur based on the device information of the target chauffeur.

According to another aspect of the embodiments of the present invention, a computer-readable storage medium is also provided. The computer-readable storage medium includes a stored program, wherein when the program is run by the processor, the device where the computer-readable storage medium is located is controlled to execute the method for processing data of the Internet of Vehicles according to the embodiment of the present invention.

According to another aspect of the embodiment of the present invention, a processor is also provided, and the processor is used for running a program, wherein the method for processing the data of the Internet of Vehicles according to the embodiment of the present invention is executed when the program is running.

According to another aspect of the embodiments of the present invention, a vehicle is also provided, including: implementing the method for processing data of the Internet of Vehicles according to the embodiments of the present disclosure.

In the embodiment of the present invention, by acquiring the historical behavior information that the current platform provides the user with the chauffeur service, and obtaining the data in the corresponding time period of the user as a sample from the Internet of Vehicles data, the behavior characteristics of the chauffeur are obtained, and then, based on the data of the Internet of Vehicles Identify the characteristics of chauffeured driving behavior, obtain a large number of sample data of chauffeured driving behavior, train the user's chauffeured service demand recognition model based on the data, and identify the user's possible surrogate driving needs. Driving service demand identification, and user location data does not need to be uploaded to the cloud, ensuring the security of user privacy data, realizing accurate user identification, greatly improving the efficiency of service push, and ensuring the security of user privacy data, thus solving the problem of inability to accurately identify user needs and service push The technical problem of low efficiency has achieved the technical effect of being able to accurately identify user needs and improve the efficiency of service push.

It should be understood that what is described in this section is not intended to identify key or critical features of embodiments of the disclosure, nor is it intended to limit the scope of the disclosure. Other features of the present disclosure will become readily understood from the following description.

Description of drawings

The accompanying drawings are used for better understanding of the present solution, and do not constitute a limitation to the present disclosure. in:

FIG. 1 is a flowchart of a method for processing Internet of Vehicles data according to an embodiment of the present disclosure;

FIG. 2 is a schematic diagram of a process flow of a driving trip identification solution according to an embodiment of the present disclosure;

3 is a schematic diagram of a process flow of a chauffeur-driven demand identification solution according to an embodiment of the present disclosure;

4 is a schematic diagram of a flow of a solution for implementing a model function at an edge according to an embodiment of the present disclosure;

FIG. 5 is a schematic diagram of an apparatus for processing Internet of Vehicles data according to an embodiment of the present disclosure.

Detailed ways

In order to make those skilled in the art better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only Embodiments are part of the present invention, but not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

It should be noted that the terms "first", "second" and the like in the description and claims of the present invention and the above drawings are used to distinguish similar objects, and are not necessarily used to describe a specific sequence or sequence. It is to be understood that the data so used may be interchanged under appropriate circumstances such that the embodiments of the invention described herein can be practiced in sequences other than those illustrated or described herein. Furthermore, the terms "comprising" and "having" and any variations thereof, are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those expressly listed Rather, those steps or units may include other steps or units not expressly listed or inherent to these processes, methods, products or devices.

Example 1

The following describes the processing method of the Internet of Vehicles data according to the embodiment of the present disclosure.

Internet of Vehicles data: It includes two parts: gray data and color data, of which gray mainly refers to various sensor signals of the vehicle, including vehicle speed, door opening status, trunk door opening status, and Global Positioning System (Global Position System, referred to as GPS) information, etc.; The color data includes various APPs and functional data used by the user in the car, such as navigation information.

Identification scheme: First, combine the platform data to obtain a batch of vehicle IoV data before and after the chauffeur service, first eliminate the relevant user information, and analyze the chauffeured itinerary data. Address) and Point of Information (POI) analysis of the location of the starting point of the trip (there is a business place near the starting point), the characteristics before the start of the chauffeured trip (the number of trunks opened, the number of doors opened, etc.), the driving of the chauffeured trip Behavior (speed, acceleration rate, relatively conservative style), etc.

Based on the characteristics of the above-mentioned chauffeur-driven itineraries, feature construction is carried out, and the obtained chauffeur-driven service information is used as sample data, and the machine learning model is input for training, and then the itinerary information of a large number of vehicles is used as data to identify the itinerary information with the same characteristics, that is, This part of the itinerary is a chauffeured itinerary.

Obtain the information of the previous segment of the chauffeured itinerary. The main features of concern include the travel time period (non-working hours), navigation information and the location of the end of the trip (there is a business place nearby), the driving route (different from the daily driving route, not from the company The corresponding sample itinerary features are input into the machine learning model for training, and the trained model is deployed to the vehicle edge for calculation. When the model recognizes the itinerary information with the same characteristics, the user may have a driver service demand. The identification results are fed back to the cloud, and then the relevant services are accurately pushed to users through the APP.

FIG. 1 is a flowchart of a method for processing Internet of Vehicles data according to an embodiment of the present disclosure. As shown in FIG. 1 , the method may include the following steps:

Step S101 , during the driving process of at least one vehicle to be monitored, collect the Internet of Vehicles information of the vehicle to be monitored, wherein the Internet of Vehicles information at least includes: travel information of the vehicle to be monitored.

In the technical solution provided by the above step S101 of the present disclosure, during the driving process of at least one vehicle to be monitored, the Internet of Vehicles information of the vehicle to be monitored is collected, for example, in combination with platform data, a group of vehicle information is obtained before and after the driving service. Sensor-like signals and various APPs and functional data used by users in the car, such as navigation information.

In this embodiment, the Internet of Vehicles data may include gray data and color data, where gray mainly refers to various sensor signals of the vehicle, including vehicle speed, door open status, trunk door open status, GPS information, etc.; color data is Including various APPs and functional data used by users in the car, such as navigation information.

In this embodiment, the itinerary information of the vehicle to be monitored may be information of the driver's trip of the vehicle to be monitored, and the driver's trip information may be obtained through historical information of the driver's application or some manual identification methods.

In this embodiment, the IoV information includes at least: the itinerary information of the vehicle to be monitored. For example, after obtaining a batch of IoV data before and after the chauffeur service in combination with the platform data, the relevant user information is first eliminated, and the chauffeur trip is analyzed. Data, its characteristics mainly include user navigation information (the destination is the user's frequent place or home address) and the point of interest analysis of the starting point of the trip (there is a business place near the starting point), the characteristics before the start of the driving trip (the trunk is opened, Number of doors opened, etc.), driving behavior (speed, acceleration rate of change, and relatively conservative style) on behalf of the driver.

Step S102, invoking the chauffeur-driven itinerary recognition model deployed on the edge of the vehicle.

In the technical solution provided in the above step S102 of the present disclosure, the driver's trip identification model deployed on the vehicle edge is called. For example, after the training of the driver's trip identification model is completed, it is deployed on the vehicle edge, and the model is called for generation Driving trip identification.

In this embodiment, before invoking the driver's itinerary recognition model deployed on the edge of the vehicle, it may include generating a driver's itinerary recognition model. The data is fed into the machine learning model for training, which in turn generates a chauffeured itinerary recognition model.

Step S103 , analyzing the information of the Internet of Vehicles by using the driving trip identification model to obtain the target driving behavior characteristics of the vehicle to be monitored.

In the technical solution provided in the above step S103 of the present disclosure, the vehicle network information can be analyzed by using the driver's itinerary identification model to obtain the target driver's behavior characteristics of the vehicle to be monitored. The obtained chauffeur service information is used as feature data, and is input into the machine learning model for training. After obtaining the chauffeured itinerary identification model, the itinerary information of a large number of vehicles is used as data to identify the itinerary information with the same characteristics, that is, this part of the itinerary is chauffeured. journey.

In this embodiment, the vehicle network information can be analyzed in the cloud by using the driver's trip identification model to obtain the target driver's behavior characteristics of the vehicle to be monitored.

In this embodiment, the method may include: recognizing the characteristics of the itinerary information in the IoV information inputted into the surrogate driving itinerary recognition model, and judging whether there is the same itinerary information as the features in the sample data used to train the surrogate driving recognition model. .

In this embodiment, characteristic data can be formed by analyzing common features of the driver's trip data, and the driver's trip identification model can be trained by using the feature data.

Step S104 , obtaining a chauffeur-driven demand identification model based on the target chauffeur-driven behavior feature.

In the technical solution provided in the above step S104 of the present disclosure, if it is identified that the vehicle to be monitored has the target driver's behavior feature, the driver's demand recognition model is obtained. Requirement identification model.

In step S105 , the surrogate driving request of the vehicle to be monitored is determined by using the surrogate driving demand identification model.

In the technical solution provided in the above step S105 of the present disclosure, the chauffeur-driven demand identification model can be used to determine the surrogate-driving request of the vehicle to be monitored. Users who need to drive on behalf of others push the chauffeur-driven service.

In this embodiment, the chauffeur-driven demand identification model deployment may be deployed on the vehicle-mounted intelligent terminal, and the user's location information may be obtained from the vehicle-mounted intelligent terminal, so that only one piece of result information is uploaded to the cloud, and the user's private information is not uploaded to the cloud, that is, When the user is currently required by the chauffeur service, the cloud can control the recommendation information of the service to be delivered to the user terminal, effectively protecting the user's privacy while making accurate recommendations.

Step S106, uploading the driving request of the vehicle to be monitored to the cloud platform, wherein at least one user terminal is allowed to obtain the driving request of the vehicle to be monitored pushed by the cloud platform.

In the technical solution provided in the above step S106 of the present disclosure, the chauffeur request of the vehicle to be monitored is uploaded to the cloud platform. For example, after the model recognizes the itinerary information with the same characteristics, the identification result that the user may have a chauffeur service demand Feedback to the cloud, and then accurately push related services to users through the APP.

In this embodiment, before generating the driving request for the vehicle to be monitored, the method may include: uploading the identification result to the cloud.

In this embodiment, after generating a chauffeur request for the vehicle to be monitored, the method may include: performing a service push of the chauffeur request via an APP.

Through the above steps S101 to S106, by obtaining the historical behavior information of the current platform providing the user with the chauffeur service, and obtaining the data in the corresponding time period of the user as a sample from the Internet of Vehicles data, the behavior characteristics of the chauffeur are obtained, and then, based on the Internet of Vehicles The data identifies the chauffeur behavior characteristics, obtains a large number of chauffeured behavior sample data, trains the user chauffeur service demand identification model based on the data, and identifies the user's possible chauffeur demand. Finally, the model recognition calculation is performed on the vehicle edge to achieve real-time At the same time, user location data does not need to be uploaded to the cloud, ensuring the security of user privacy data, realizing accurate user identification, greatly improving the efficiency of service push, and ensuring the security of user privacy data, thus solving the problem of inability to accurately identify user needs. The technical problem of low push efficiency has achieved the technical effect of accurately identifying user needs and improving service push efficiency.

The above method of this embodiment will be further described in detail below.

As an optional implementation manner, in step S102, before invoking the driver identification model deployed at the edge of the vehicle, the method further includes: collecting data on the Internet of Vehicles of multiple vehicles in a historical time period, wherein the historical time period is at least It includes the driving time period in which the driving behavior information is generated. The Internet of Vehicles data includes at least: driving itinerary data generated in the driving time period; training the machine learning model based on the driving driving data to generate the driving itinerary recognition model. ; Using the chauffeur demand identification model to determine the chauffeur request of the vehicle to be monitored, including: obtaining chauffeur sample data based on the target chauffeur behavior characteristics, wherein the chauffeur sample data includes at least one of the following: adjacent to the chauffeur time period. The itinerary information in at least one adjacent time period, and the driving itinerary data generated in the driving time period; the machine learning model is trained based on the driving sample data to generate a driving demand identification model, wherein the driving demand identification model is It is used to identify the corresponding chauffeured itinerary feature based on the itinerary information of the vehicle to be identified.

In this embodiment, the IoV data of multiple vehicles in a historical time period is collected, for example, a batch of vehicle IoV data before and after the chauffeur service is acquired, the relevant user information is first eliminated, and the chauffeured itinerary data is analyzed. It mainly includes user navigation information (the destination is the user's frequent place or home address) and the analysis of the point of interest at the starting point of the trip (there is a business place near the starting point), and the characteristics before the start of the chauffeured trip (the number of trunks opened, the number of doors opened, etc. ), driving behavior (speed, acceleration rate of change, relatively conservative style), etc.

In this embodiment, the machine learning model can be trained based on the driver's itinerary data to generate the driver's trip identification model. For example, after analyzing the common features of the driver's trip data, feature data is formed, and then the feature data is used for training. The driver's trip identification model is used to filter out a large amount of driver's trip data from the historical data of the vehicle to be monitored, and the driver's trip data can be the trip data representing the driver's driving.

For example, a small amount of historical chauffeur-driven itinerary information can be obtained through the chauffeur service APP, and the characteristics of the trip can be analyzed, such as whether it is a multi-person ride by the open state of the car door (single-person rides can be excluded from chauffeur-driven), and through backup The switch state of the box can distinguish the storage of the driver's vehicle, and calculate the speed, acceleration rate of change and other indicators through a series of signals such as speed and acceleration, which are used to evaluate the overall driving style of the driver's trip, and determine whether the starting point of the driver's trip is at the location of the GPS signal. Commercial areas, etc., based on this part of the historical data, complete the relevant feature construction, and train the cost formation recognition model. After the training of the model is completed, the model selects a large amount of driving trip data from the massive Internet of Vehicles data in the cloud. Based on the corresponding The time period obtains the travel period before the chauffeured itinerary, that is, the travel period for the user to go to the business place, which can be characterized in that the arrival at the destination (business place, etc.) in a fixed time period is different from the arrival at the destination (home) on normal working days.

In this embodiment, based on the characteristics of the target driver's behavior, the sample data of the driver is obtained. For example, after analyzing the driver's itinerary data, feature construction is performed based on the characteristics of the driver's trip, and the obtained driver's service information is used as a sample. data, wherein the driver sample data includes at least one of the following: itinerary information in at least one adjacent time period adjacent to the driver time period, and driver itinerary data generated within the driver time period.

In this embodiment, the main features of the driver's sample data may include: travel time period (non-working hours), navigation information and the location of the end of the trip (there is a business place nearby), driving route (different from the daily driving route, not from the company home route) etc.

In this embodiment, the machine learning model can be trained based on the driver's sample data to generate the driver's demand identification model. For example, the sample data is used to train the driver's requirement identification model, and the sample data can be used for training the driver's driver. The itinerary data of the user driving to the restaurant or entertainment place before the driver's trip data of the trip recognition model, and the corresponding sample trip characteristics are input into the machine learning model for training to generate the driver's demand identification model.

As an optional implementation, extract at least one type of surrogate driver feature information from surrogate driver itinerary data generated within the surrogate driver time period, and perform feature construction based on the surrogate driver feature information to generate some samples in the surrogate driver sample data. , wherein the driver's trip data includes at least one of the following: navigation information, point-of-interest analysis data of the location of the driver's trip starting point, vehicle equipment usage characteristics before the driver's trip starts, and driving behavior information during the driver's trip.

In this embodiment, the navigation information may be that the destination is a frequent place of the user or a home address.

In this embodiment, the point of interest analysis data of the starting point of the chauffeured trip may be that there is a business place near the starting point.

In this embodiment, the usage characteristics of the vehicle equipment before the start of the chauffeur trip may be the opening of the trunk, the number of openings of the vehicle doors, and the like.

In this embodiment, the driving behavior information in the driving trip may be speed, acceleration change rate, relatively conservative style, and the like.

As an optional implementation, extracting at least one type of surrogate driving feature information from the itinerary information in at least one adjacent time period adjacent to the surrogate driving time period, and performing feature construction based on the surrogate driving feature information to generate a surrogate driving sample Some samples in the data, wherein the travel information includes at least one of the following: travel time period, navigation information, the end position of the driver's trip, and the driver's route.

In this embodiment, the travel time period may be non-working time.

In this embodiment, the navigation information and the end location of the chauffeured trip may be the route and location of a nearby business place.

In this embodiment, the chauffeured driving route may be different from the daily driving route, and may be a route from a company to a business and leisure place, rather than a route home from the company.

As an optional implementation manner, the historical chauffeur service information is obtained by accessing the chauffeur service platform, and based on the historical chauffeur service information, the IoV data of each vehicle in the historical time period is obtained, wherein the IoV data also includes It includes at least one of the following: vehicle door opening and closing state, trunk opening and closing state, vehicle speed, vehicle acceleration, vehicle steering wheel angle, steering wheel angle speed, accelerator pedal opening, brake pedal opening and vehicle navigation data.

In this embodiment, the door switch state of the vehicle can be used to distinguish whether there are multiple people riding in the vehicle (single-person riding can exclude chauffeurs).

In this embodiment, the trunk switch state can be used to identify the vehicle storage of the chauffeur driver.

In this embodiment, vehicle speed, vehicle acceleration, vehicle steering wheel angle, steering wheel angle speed, accelerator pedal opening, and brake pedal opening may be used to evaluate the overall driving style of the chauffeur trip.

In this embodiment, the vehicle navigation data can be used to determine whether the starting point of the chauffeured trip is located in a commercial area.

As an optional implementation manner, in step S105, after uploading the driving request of the vehicle to be monitored to the cloud platform, the method further includes: detecting whether the cloud platform receives the driving request; if the driving request is received, adjusting Obtain the driver information of at least one registered driver; based on the driver information of the driver, determine whether there is an idle driver; based on the push rules, determine at least one target driver, and retrieve the target driver The push rule is used to determine the push priority of multiple chauffeurs; based on the device information of the target chauffeur, the chauffeur request is pushed to the device held by the target chauffeur.

In this embodiment, it is detected whether the cloud platform receives a driving request. For example, after uploading the driving request of the vehicle to be monitored to the cloud platform, it is detected whether the cloud platform receives the driving request.

In this embodiment, if a surrogate driving request is received, the driver information of at least one registered surrogate driver is retrieved. For example, after it is detected that the cloud platform receives the surrogate driver request, the driver information registered on the surrogate driving service platform is retrieved. The driver information (such as name, ID number, work status and current location, etc.) of at least one chauffeur.

In this embodiment, it is determined whether there is an idle driver based on the driver information of the driver. For example, it is determined whether there is an idle driver according to the retrieved working status in the driver information of at least one driver. chauffeur driver.

In this embodiment, based on the push rule, at least one target driver is determined, and the device information of the target driver is retrieved, for example, the driver's favorable rate is ranked from high to low, and the favorable rate is based on the push rule. The highest one is pushed as the priority, at least one driver with the highest favorable rating is determined, and the device information of such driver is retrieved.

In this embodiment, based on the device information of the target chauffeur, the chauffeur request is pushed to the device held by the target chauffeur. The registration information of the chauffeur service APP in the mobile phone held by the target chauffeur will push the chauffeur request to the mobile phone held by the target chauffeur through the mobile phone text message and/or the voice call interface of the chauffeur service APP.

For example, in step 1, to identify "driver behavior characteristics" is to manually select a part of the data that is likely to be the driver's trip (historical information of the driver application or some manual methods), and analyze the common data in this part of the data. feature; step 2, use the feature data obtained by step 1 to train the "driver's trip identification model", the purpose of this model is to filter out a large amount of driver's trip data (that is, the driver's trip) from the historical data; step Third, use the large amount of data obtained in the previous step (that is, the itinerary of the chauffeur driver) to obtain the data of a previous trip (the user's own trip to the restaurant or entertainment place), and use this trip data to train the "driver's chauffeur". Demand recognition model", this model is to identify whether the user has a chauffeur demand when the user arrives near the destination, and push chauffeur services to users who may need a chauffeur.

In the above-mentioned embodiments of the present disclosure, by obtaining the historical behavior information of the current platform providing the user with the chauffeur service, and obtaining the data in the corresponding time period of the user as a sample from the Internet of Vehicles data, the behavior characteristics of the chauffeur are obtained, and then based on the data of the Internet of Vehicles Identify the characteristics of chauffeured driving behavior, obtain a large number of sample data of chauffeured driving behavior, train the user's chauffeured service demand identification model based on the data, and identify the user's possible chauffeured demand, realize the accurate identification of users, greatly improve the efficiency of service push, and ensure The user privacy data is secure, which solves the technical problems of inability to accurately identify user needs and low service delivery efficiency, and achieves the technical effect of accurately identifying user needs and improving service delivery efficiency.

As a preferred implementation manner, the user's acquisition of the chauffeur service can be divided into two stages, which are respectively introduced below.

Stage 1, the travel segment of the user when going to the business place, which is characterized in that the arrival at the destination (business place, etc.) in a fixed time period is different from the arrival at the destination (home) on a normal working day, etc.;

Stage 2: After the driver arrives at the designated location, the driver will start the vehicle and send the user to the destination. During the journey, there will be features such as opening the trunk door at the beginning of the journey, and there will be driving behavior characteristics in the journey that are the same as those of the car owner under normal circumstances. There are differences in driving behavior and other characteristics.

The recommendation of the chauffeur service needs to be carried out after the above-mentioned itinerary segment of the first stage is identified and before the second stage. The final purpose of this application is to identify the first stage.

The solution of the present application can be divided into three parts, which are respectively introduced below.

The first part, first of all, is the chauffeur-driven itinerary identification scheme. This part is carried out in the cloud, and its ultimate purpose is to obtain a large number of chauffeured itineraries (the itinerary segment driven by the chauffeur driver, that is, the above-mentioned stage 2), which is used for the next sample. Screening (screening the travel segment that the user drives to the destination, that is, stage 1).

The specific method can be to first obtain a small amount of historical chauffeur-driven itinerary information (stage 2) through channels such as the chauffeur service APP, and analyze its itinerary characteristics (distinguish whether it is a multi-person ride by the open state of the door (single-person rides can be excluded). driving); the trunk switch status is used to distinguish the storage of the driver's vehicle; a series of signals such as speed and acceleration are used to calculate the speed, acceleration rate of change and other indicators, which are used to evaluate the overall driving style of the driver's trip; GPS signals are used to judge Whether the starting point of the driver's trip is located in a commercial area, etc.), complete the relevant feature construction based on this part of the historical data, and train the driver's trip identification model. Based on the corresponding time period, the trip segment before the chauffeured itinerary, that is, the sample data of the trip segment in Phase 1, is obtained.

The second part is the chauffeur-driven demand identification scheme, which is also carried out in the cloud. Its ultimate purpose is to construct the characteristics of one trip segment in the stage, and through the training of a large number of sample data, output the chauffeur-driven demand identification model for subsequent vehicles. deploy.

The third part is the deployment of the vehicle and the overall recommendation strategy. The trained driver demand recognition model is deployed to the vehicle intelligent terminal, and some operations involving the acquisition of user location information are performed on the vehicle and will not be uploaded. The user's private information is sent to the cloud, and only one piece of result information is uploaded to the cloud, that is, the user currently has a demand for chauffeur service, and the cloud can control the delivery of service recommendation information to the user, effectively protecting user privacy while making accurate recommendations.

It should be noted that when training the driver's itinerary recognition model based on the driver's itinerary information, since the method of directly obtaining a large amount of driver's itinerary information involves user privacy data, in this application, a small amount is obtained through the driver's service APP and other channels. Based on the historical data of the driver's itinerary, analyze its trip characteristics, complete the relevant feature construction based on this part of the historical data, and train the driver's itinerary recognition model.

Example 2

The processing method of the Internet of Vehicles data of the present disclosure will be further introduced below with reference to the preferred embodiments.

FIG. 2 is a schematic diagram of a process flow of a surrogate driving itinerary identification solution according to an embodiment of the present disclosure. As shown in FIG. 2 , the solution may include the following steps:

S201, obtain historical chauffeur service information;

S202, obtain the data of the Internet of Vehicles within the driving time period;

S203, constructing the surrogate driving itinerary feature;

S204, the algorithm model training for driving trip identification;

S205, obtain a large number of chauffeured itineraries;

In this embodiment, the user's historical driving service information is obtained from a driving service providing platform such as an APP, and the data of the Internet of Vehicles in the corresponding period is matched to obtain sample data. The main initial field signals include door switch status, trunk switch status, speed, Acceleration, steering wheel angle, steering wheel angular velocity, accelerator pedal opening, brake pedal opening, GPS, etc., the main purpose of selecting each signal is to construct the following types of features: distinguish whether it is a multi-person vehicle (single passenger) through the open state of the door. The car can be excluded from driving); the switch status of the trunk can distinguish the storage of the driver's vehicle; a series of signals such as speed and acceleration are used to calculate the speed, acceleration rate of change and other indicators, which are used to evaluate the overall driving style of the driver's trip; GPS The signal is used to determine whether the starting point of the chauffeured trip is located in a commercial area.

Complete the corresponding feature structure (including but not limited to the above four categories) on the data to form the sample data of the driving trip, and use it as the input of the machine learning model to train the model, obtain the trip data of all vehicles, construct the same features, and input them into the trained model. , the output results indicate all the trip segments identified as chauffeur trips, which are used as input conditions for the subsequent model.

Through steps S201 to S205 in this embodiment, the historical chauffeur service information is obtained; the Internet of Vehicles data within the chauffeur travel time period is obtained; the chauffeur trip characteristics are constructed; the algorithm model training of the chauffeur trip identification algorithm; That is to say, obtain the user's historical chauffeur service information from the APP and other chauffeur service providers, match the vehicle-to-vehicle data in the corresponding period, obtain sample data, select multiple signals to construct multiple types of features, train the machine learning model, and The trip data of all vehicles is input into the trained model, and all the trip segments of the chauffeured trip are obtained, which solves the technical problem that the user's needs cannot be accurately identified, and achieves the technical effect of accurately identifying the user's needs.

FIG. 3 is a schematic diagram of a process flow of a chauffeur-driven demand identification solution according to an embodiment of the present disclosure. As shown in FIG. 3 , the solution may include the following steps:

Step S301, obtaining the Internet of Vehicles data within a period of time before the driving trip;

Step S302, constructing a corresponding travel feature;

Step S303, the chauffeur-driven demand identification algorithm model training.

In this embodiment, according to the obtained corresponding times of a large number of chauffeured itineraries, obtain the information of its adjacent previous itinerary, and obtain the relevant field signals in the itinerary, including time, GPS signal, navigation information, etc. The main purpose of the item signal is to construct the following types of features: time is used to judge whether it is working time; GPS signal and navigation information are used to judge whether the user is driving on the home route or the common route (the route is set by the user or the algorithm recognition result, stored At the vehicle end); GPS signals and navigation information are also used to determine whether the end of the journey or the vicinity of the navigation address is located in a commercial area, etc.

After completing the relevant feature construction for all the data (including but not limited to the above three categories), the sample data of the itinerary chauffeur demand identification is used as the input of the machine learning model for model training.

Through steps S301 to S303 in this embodiment, the Internet of Vehicles data for a period of time before the driving trip is obtained; the corresponding itinerary characteristics are constructed; Corresponding time, obtain its adjacent previous itinerary information, and obtain the relevant field signals in the itinerary, select multiple signals to construct multi-type features, and complete the relevant feature construction for all data (including but not limited to the above three categories) itinerary The sample data of chauffeur demand identification is used as the input of the machine learning model for model training, which realizes the accurate identification of the chauffeur demand, solves the technical problem that the user's demand cannot be accurately identified, and achieves the technical effect of being able to accurately identify the user's demand.

FIG. 4 is a schematic diagram of a flow of a solution for implementing a model function at the edge according to an embodiment of the present disclosure. As shown in FIG. 4 , the solution may include the following steps:

Step S401, model edge deployment;

Step S402, the model edge is calculated in real time;

Step S403, the model calculation at the edge end is uploaded to the cloud;

Step S404, the APP sends precise recommendation information.

In this embodiment, after the above model training is completed, the model is deployed to the vehicle-mounted intelligent terminal (including but not limited to the intelligent cockpit, gateway, and vehicle-mounted connected terminal), and the gray and color data of the Internet of Vehicles are acquired in real time, and obtained on the vehicle-side The stored information such as the user's home address, common routes, etc., is calculated and stored on the car side, and only the results are uploaded to the cloud.

Through the above-mentioned steps S401 to S404 in the embodiment of the present disclosure, the model is deployed at the edge; the model is calculated in real time at the edge; the model calculation at the edge is uploaded to the cloud; Obtain the gray and color data of the Internet of Vehicles, and obtain the user's home address, common routes and other information stored on the car end, complete the data calculation and storage on the car end, only upload the results to the cloud, and the cloud will send the judgment results to the APP, It realizes accurate user identification, greatly improves the efficiency of service push, and at the same time ensures the security of user privacy data, which solves the technical problems of inability to accurately identify user needs and low service push efficiency, and achieves a technology that can accurately identify user needs and improve service push efficiency. Effect.

Example 3

An embodiment of the present disclosure further provides a device for processing connected vehicle data for executing the method for processing connected vehicle data in the embodiment shown in FIG. 1 .

FIG. 5 is a schematic diagram of a processing device for Internet of Vehicles data according to an embodiment of the present disclosure. As shown in FIG. 5 , the processing device 50 for Internet of Vehicles data may include: a collection module 51 , a calling module 52 , an identification module 53 , a generation module 54 and upload module 55.

The collection module 51 is configured to collect the Internet of Vehicles information of the vehicle to be monitored during the driving process of at least one vehicle to be monitored, wherein the Internet of Vehicles information includes at least: travel information of the vehicle to be monitored;

The calling module 52 is used to call the driver identification model deployed on the edge of the vehicle;

The first acquisition module 53 is used to analyze the information of the Internet of Vehicles by adopting the driver identification model, and obtain the target driver behavior characteristics of the vehicle to be monitored;

The second obtaining module 53 is used to obtain the identification model of the driving demand based on the target driving behavior;

A determination module 54, used for determining the chauffeur request of the vehicle to be monitored by adopting the chauffeur demand identification model;

The uploading module 55 is configured to upload the driving request of the vehicle to be monitored to the cloud platform, wherein at least one user terminal is allowed to obtain the driving request of the vehicle to be monitored pushed by the cloud platform.

In the above-mentioned processing device of the Internet of Vehicles data of the present disclosure, the device further comprises: a collection module, a first training module, an acquisition module and a second training module. Among them, the collection module is used to collect the Internet of Vehicles data of multiple vehicles in a historical time period, wherein the historical time period at least includes the driving time period in which the driving behavior information is generated, and the Internet of Vehicles data includes at least: driving time The driver's itinerary data generated in the segment; the first training module is used to train the machine learning model based on the driver's trip data to generate the driver's trip identification model; the acquisition module is used to obtain the driver's driver based on the characteristics of the target driver's behavior. Sample data, wherein the surrogate driving sample data includes at least one of the following: itinerary information in at least one adjacent time period adjacent to the surrogate driving time period, and the surrogate driving itinerary data generated within the surrogate driving time period; the second training module , which is used to train the machine learning model based on the surrogate driving sample data to generate a surrogate driving demand identification model, wherein the surrogate driving demand identification model is used to identify the corresponding chauffeured driving itinerary characteristics based on the itinerary information of the vehicle to be identified.

Optionally, the obtaining module further includes a first obtaining unit and a second obtaining unit. Wherein, the first acquisition unit is used to extract at least one type of surrogate driver feature information from surrogate driver itinerary data generated within the surrogate driver time period, and perform feature construction based on the surrogate driver feature information to generate some samples in the surrogate driver sample data. , wherein the driver's itinerary data includes at least one of the following: navigation information, analysis data of points of interest at the starting point of the driver's trip, vehicle equipment usage characteristics before the driver's trip starts, and driving behavior information during the driver's trip; The second acquisition unit is used to extract at least one type of surrogate driving feature information from the itinerary information in at least one adjacent time period adjacent to the surrogate driving time period, and perform feature construction based on the surrogate driving feature information to generate the surrogate driving sample data. For some samples, the travel information includes at least one of the following: travel time period, navigation information, the end position of the chauffeured trip, and the chauffeured travel route.

Optionally, the collection module further includes a first collection unit. Among them, the first collection unit is used to obtain historical chauffeur service information by accessing the chauffeur service platform, and based on matching the historical chauffeur service information to obtain the IoV data of each vehicle in the historical time period, wherein the IoV data also includes It includes at least one of the following: vehicle door opening and closing state, trunk opening and closing state, vehicle speed, vehicle acceleration, vehicle steering wheel angle, steering wheel angle speed, accelerator pedal opening, brake pedal opening and vehicle navigation data.

In the above-mentioned processing device of the Internet of Vehicles data of the present disclosure, the device further comprises: a detection module, a retrieval module, a determination module, a processing module and a push module. Among them, the detection module is used to detect whether the cloud platform has received a chauffeur request; the retrieval module is used to retrieve the driver information of at least one registered driver if the chauffeur request is received; the determination module is used to retrieve the driver information based on the proxy The driver information of the driver, to determine whether there is an idle driver; the processing module is used to determine at least one target driver based on the push rule, and retrieve the device information of the target driver, where the push rule uses It is used to determine the push priority of multiple chauffeurs; the push module is used to push the chauffeur request to the device held by the target chauffeur based on the device information of the target chauffeur.

In the above-mentioned embodiment of the present disclosure, by obtaining the historical behavior information of the current platform providing the user with the chauffeur service, and obtaining the data in the corresponding time period of the user as a sample from the Internet of Vehicles data, the chauffeur behavior characteristics are obtained, and then, based on the vehicle Networked data identifies the characteristics of chauffeured driving behavior, obtains a large number of sample data of chauffeured driving behavior, trains the user's chauffeured service demand identification model based on the data, and identifies the user's possible chauffeured driving needs. Real-time chauffeur service demand identification, and user location data does not need to be uploaded to the cloud, ensuring the security of user privacy data, realizing accurate user identification, greatly improving the efficiency of service push, and ensuring the security of user privacy data, thus solving the inability to accurately identify user needs. The technical problem of low service push efficiency achieves the technical effect of accurately identifying user needs and improving service push efficiency.

Example 4

According to an embodiment of the present invention, a computer-readable storage medium is also provided. The computer-readable storage medium includes a stored program, wherein when the program is run by the processor, the device where the computer-readable storage medium is located is controlled to execute the method for processing data of the Internet of Vehicles according to the embodiment of the present invention.

Example 5

According to an embodiment of the present invention, a processor is also provided, and the processor is used for running a program, wherein, when the program runs, the method for processing data of the Internet of Vehicles according to the embodiment of the present invention is executed.

Example 6

According to an embodiment of the present disclosure, the present disclosure also provides a vehicle, including the method for processing data of the Internet of Vehicles according to the embodiment of the present disclosure.

The above-mentioned serial numbers of the embodiments of the present invention are only for description, and do not represent the advantages or disadvantages of the embodiments.

In the above-mentioned embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference may be made to related descriptions of other embodiments.

In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are only illustrative, for example, the division of the units may be a logical function division, and there may be other division methods in actual implementation, for example, multiple units or components may be combined or Integration into another system, or some features can be ignored, or not implemented. On the other hand, the shown or discussed mutual coupling or direct coupling or communication connection may be indirect coupling or communication connection through some interfaces, units or models, and may be in electrical or other forms.

The units described as separate components may or may not be physically separated, and components shown as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution in this embodiment.

In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically alone, or two or more units may be integrated into one unit. The above-mentioned integrated units may be implemented in the form of hardware, or may be implemented in the form of software functional units.

The integrated unit, if implemented in the form of a software functional unit and sold or used as an independent product, may be stored in a computer-readable storage medium. Based on such understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium , including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program codes .

The above are only the preferred embodiments of the present invention. It should be pointed out that for those skilled in the art, without departing from the principles of the present invention, several improvements and modifications can be made. It should be regarded as the protection scope of the present invention.

Claims (10)

1. a processing method of Internet of Vehicles data, is characterized in that, comprises: During the driving process of at least one vehicle to be monitored, the Internet of Vehicles information of the vehicle to be monitored is collected, wherein the Internet of Vehicles information includes at least: travel information of the vehicle to be monitored; Invoke the driving trip recognition model deployed on the edge of the vehicle; Analyzing the Internet of Vehicles information by using the driving trip identification model to obtain the target driving behavior characteristics of the vehicle to be monitored; Obtaining a chauffeur-driven demand identification model based on the target chauffeur-driven behavior characteristics; Determine the chauffeur request of the to-be-monitored vehicle by using the chauffeur demand identification model; Upload the driving request of the vehicle to be monitored to the cloud platform, wherein at least one user terminal is allowed to obtain the driving request of the vehicle to be monitored pushed by the cloud platform. 2. The method according to claim 1, wherein The method further includes: collecting data on the Internet of Vehicles of a plurality of vehicles in a historical time period, wherein the historical time period at least includes the driving time period in which the driving behavior information is generated, and the data on the Internet of Vehicles at least includes: The driver's itinerary data generated within the driver's time period; the machine learning model is trained based on the driver's trip data to generate the driver's trip identification model; the driver's demand identification is obtained based on the target driver's behavior characteristics The model includes: obtaining driver sample data based on the characteristics of the target driver's behavior, wherein the driver sample data includes at least one of the following: at least one adjacent time period adjacent to the driver time period. Itinerary information, and the chauffeured itinerary data generated within the chauffeured time period; The machine learning model is trained based on the chauffeured sample data to generate the chauffeured demand identification model, wherein the chauffeured demand identification model is used to identify the corresponding chauffeured demand based on the itinerary information of the vehicle to be identified feature. 3. The method according to claim 2, characterized in that, extracting at least one type of surrogate driving feature information from surrogate driving itinerary data generated within the surrogate driving time period, and carrying out feature construction based on the surrogate driving feature information , and generate some samples in the chauffeured sample data, wherein the chauffeured itinerary data includes at least one of the following: navigation information, analysis data of points of interest at the starting point of the chauffeured itinerary, and vehicles before the chauffeured itinerary begins. Device usage characteristics, driving behavior information in chauffeured itineraries. 4. The method according to claim 2, characterized in that, extracting at least one type of surrogate driving feature information from the itinerary information in at least one adjacent time period adjacent to the surrogate driving time period, and based on the surrogate driving time period The feature information is characterized to generate part of the samples in the driver sample data, wherein the itinerary information includes at least one of the following: travel time period, navigation information, the end position of the driver's trip, and the driver's route. 5. The method according to claim 2, characterized in that, obtaining historical chauffeur service information by visiting a chauffeur service platform, and obtaining the information of each vehicle in the historical time period based on matching in the historical chauffeur service information. Internet of Vehicles data, wherein the Internet of Vehicles data also includes at least one of the following: door switch status of the vehicle, trunk switch status, vehicle speed, vehicle acceleration, vehicle steering wheel angle, steering wheel angle speed, accelerator pedal opening, and brake pedal Opening and vehicle navigation data. 6. The method according to any one of claims 1-5, wherein after uploading the chauffeur request of the vehicle to be monitored to the cloud platform, the method further comprises: Detecting whether the cloud platform receives the chauffeur request; If receiving the chauffeur request, retrieve the driver information of at least one registered chauffeur; determining, based on the driver information of the chauffeur driver, whether there is an idle chauffeur driver; Based on the push rule, at least one target driver is determined, and the device information of the target driver is retrieved, wherein the push rule is used to determine the push priorities of multiple drivers; Based on the device information of the target chauffeur, the chauffeur request is pushed to the device held by the target chauffeur. 7. A processing device for Internet of Vehicles data, comprising: a collection module, configured to collect the Internet of Vehicles information of the vehicle to be monitored during the driving process of at least one vehicle to be monitored, wherein the Internet of Vehicles information includes at least: travel information of the vehicle to be monitored; The calling module is used to call the driver recognition model deployed on the edge of the vehicle; a first acquisition module, configured to use the driver identification model to analyze the Internet of Vehicles information, and acquire the target driver behavior characteristics of the vehicle to be monitored; a second obtaining module, configured to obtain a model for identifying a driver's demand based on the target driver's behavior feature; a determination module, used for determining the driver's request of the vehicle to be monitored by using the driver-request identification model; The uploading module is configured to upload the driving request of the vehicle to be monitored to the cloud platform, wherein at least one user terminal is allowed to obtain the driving request of the vehicle to be monitored pushed by the cloud platform. 8. The apparatus according to claim 7, wherein the apparatus further comprises: The collection module is used to collect the data of the Internet of Vehicles of multiple vehicles in a historical time period, wherein the historical time period at least includes the driving time period in which the driving behavior information is generated, and the data of the Internet of Vehicles at least includes: Describe the driver's itinerary data generated during the driver's time period; a first training module, configured to train a machine learning model based on the driving itinerary data to generate the driving itinerary identification model; The acquisition module is used to obtain the driver sample data based on the characteristics of the target driver's behavior, wherein the driver sample data includes at least one of the following: in at least one adjacent time period adjacent to the driver time period The itinerary information of the driver, and the driver's trip data generated within the driver's time period; The second training module is used to train the machine learning model based on the surrogate driving sample data, and generate a chauffeur-driven demand recognition model, wherein the surrogate driving demand recognition model is used to identify the vehicle based on the itinerary information of the vehicle to be identified. Corresponding chauffeur demand characteristics. 9. The apparatus of claim 7, wherein the apparatus further comprises: a detection module, configured to detect whether the cloud platform receives the chauffeur request; a retrieval module, configured to retrieve the driver information of at least one registered driver if the driver request is received; a determination module, configured to determine whether there is an idle driver based on the driver information of the driver; a processing module, configured to determine at least one target chauffeur based on the push rule, and retrieve the device information of the target chauffeur, wherein the push rule is used to determine the push priorities of multiple chauffeurs; A push module, configured to push the chauffeur request to the device held by the target chauffeur based on the device information of the target chauffeur. 10 . A vehicle, characterized in that it comprises the method for processing data of the Internet of Vehicles according to any one of claims 1 to 6 .

Images