CN117249940B

CN117249940B - Gas pressure detection method and system

Info

Publication number: CN117249940B
Application number: CN202311544785.4A
Authority: CN
Inventors: 葛汉霄; 徐其勇; 李凯迪
Original assignee: Shenzhen Gaofa Gas Co ltd
Current assignee: Shenzhen Gaofa Gas Co ltd
Priority date: 2023-11-20
Filing date: 2023-11-20
Publication date: 2024-10-22
Anticipated expiration: 2043-11-20
Also published as: CN117249940A

Abstract

The invention relates to the field of data processing, and discloses a gas pressure detection method and a gas pressure detection system, which are used for improving the control accuracy of gas output pressure. The method comprises the following steps: detecting gas output pressure based on an immersion boundary method to obtain oxygen pressure data and carbon dioxide pressure data; acquiring temperature monitoring data, and constructing a first temperature-pressure relation model of oxygen pressure data and temperature monitoring data, carbon dioxide pressure data and a second temperature-pressure relation model of oxygen pressure data and temperature monitoring data; performing pressure spectrum conversion to generate an oxygen pressure spectrum and a carbon dioxide pressure spectrum; extracting features to obtain a first oxygen pressure feature set and a first carbon dioxide pressure feature set; performing feature compensation to obtain a second oxygen pressure feature set and a second carbon dioxide pressure feature set; and carrying out gas output pressure control analysis through a gas output pressure control analysis model to generate a target pressure control parameter strategy.

Description

Gas pressure detection method and system

Technical Field

The present invention relates to the field of data processing, and in particular, to a method and a system for detecting gas pressure.

Background

In today's industrial and scientific applications, accurate detection and control of gas pressure is of paramount importance. The gas system is accurately monitored in the oxygen storage field to ensure the safety, efficiency and reliability of operation.

The conventional gas pressure detection method generally has some limitations, such as insufficient accuracy and stability under complex working conditions, sensitivity to temperature variation, and the like, so that the control accuracy of the gas output pressure of the conventional scheme is low.

Disclosure of Invention

The invention provides a gas pressure detection method and a gas pressure detection system, which are used for improving the control accuracy of gas output pressure.

The first aspect of the present invention provides a gas pressure detection method, comprising:

setting a sensor array of a target oxygen storage device based on an immersion boundary method, and detecting the gas output pressure of the target oxygen storage device through the sensor array to obtain oxygen pressure data and carbon dioxide pressure data;

Acquiring temperature monitoring data of the target oxygen storage device through the sensor array, constructing a first temperature-pressure relation model of the oxygen pressure data and the temperature monitoring data, and constructing a second temperature-pressure relation model of the carbon dioxide pressure data and the temperature monitoring data;

Performing pressure spectrum conversion on the oxygen pressure data to generate an oxygen pressure spectrum, and performing pressure spectrum conversion on the carbon dioxide pressure data to generate a carbon dioxide pressure spectrum;

performing feature extraction on the oxygen pressure spectrum to obtain a first oxygen pressure feature set, and performing feature extraction on the carbon dioxide pressure spectrum to obtain a first carbon dioxide pressure feature set;

Performing feature compensation on the first oxygen pressure feature set through the first temperature-pressure relation model to obtain a second oxygen pressure feature set, and performing feature compensation on the first carbon dioxide pressure feature set through the second temperature-pressure relation model to obtain a second carbon dioxide pressure feature set;

And inputting the second oxygen pressure characteristic set and the second carbon dioxide pressure characteristic set into a preset gas output pressure control analysis model to perform gas output pressure control analysis, so as to generate a target pressure control parameter strategy.

With reference to the first aspect, in a first implementation manner of the first aspect of the present invention, the setting a sensor array of a target oxygen storage device based on an immersion boundary method, and performing gas output pressure detection on the target oxygen storage device through the sensor array, to obtain oxygen pressure data and carbon dioxide pressure data, includes:

Setting a sensor array at a gas output end of a target oxygen storage device based on an immersion boundary method, and connecting the sensor array with a data acquisition unit;

Detecting the gas output pressure of the target oxygen storage device through the sensor array to obtain original oxygen signal data and original carbon dioxide signal data;

Performing signal processing on the original oxygen signal data to obtain initial oxygen data, and performing signal processing on the original carbon dioxide signal data to obtain initial carbon dioxide data;

and carrying out data filtering and data calibration on the initial oxygen data to obtain oxygen pressure data, and carrying out data filtering and data calibration on the initial carbon dioxide data to obtain carbon dioxide pressure data.

With reference to the first aspect, in a second implementation manner of the first aspect of the present invention, the acquiring, by the sensor array, temperature monitoring data of the target oxygen storage device, and constructing a first temperature-pressure relationship model of the oxygen pressure data and the temperature monitoring data, and constructing a second temperature-pressure relationship model of the carbon dioxide pressure data and the temperature monitoring data, includes:

The temperature of the gas output end of the target oxygen storage device is monitored through the sensor array, and temperature monitoring data are obtained;

performing kernel matrix calculation on the temperature monitoring data through a preset Gaussian kernel function to obtain a Gaussian kernel matrix;

Taking the oxygen pressure data as a first target value, taking the Gaussian kernel matrix as a first characteristic matrix, constructing a first input-output pair according to the first target value and the first characteristic matrix, taking the carbon dioxide pressure data as a second target value, taking the Gaussian kernel matrix as a second characteristic matrix, and constructing a second input-output pair according to the second target value and the second characteristic matrix;

Carrying out Gaussian linear regression parameter calculation on the first input-output pair through a preset first Gaussian linear regression algorithm to obtain a first Gaussian linear regression parameter, and carrying out Gaussian linear regression parameter calculation on the second input-output pair through a preset second Gaussian linear regression algorithm to obtain a second Gaussian linear regression parameter;

and constructing a first temperature-pressure relation model of the oxygen pressure data and the temperature monitoring data according to the first Gaussian linear regression parameters, and constructing a second temperature-pressure relation model of the carbon dioxide pressure data and the temperature monitoring data according to the second Gaussian linear regression parameters.

With reference to the first aspect, in a third implementation manner of the first aspect of the present invention, the performing pressure spectrum conversion on the oxygen pressure data to generate an oxygen pressure spectrum, and performing pressure spectrum conversion on the carbon dioxide pressure data to generate a carbon dioxide pressure spectrum includes:

Performing Fourier transform on the oxygen pressure data to obtain a first Fourier transform result, and performing Fourier transform on the carbon dioxide pressure data to obtain a second Fourier transform result;

square processing is carried out on the first Fourier transform result to obtain a first power spectrum density, and square processing is carried out on the second Fourier transform result to obtain a second power spectrum density;

generating a corresponding oxygen pressure spectrum according to the first power spectral density, and generating a corresponding carbon dioxide pressure spectrum according to the second power spectral density.

With reference to the first aspect, in a fourth implementation manner of the first aspect of the present invention, the performing feature extraction on the oxygen pressure spectrum to obtain a first oxygen pressure feature set, and performing feature extraction on the carbon dioxide pressure spectrum to obtain a first carbon dioxide pressure feature set, includes:

Performing spectrum segmentation on the oxygen pressure spectrum to obtain a plurality of first pressure spectrum segments, and performing spectrum segmentation on the carbon dioxide pressure spectrum to obtain a plurality of second pressure spectrum segments;

Respectively carrying out frequency spectrum feature extraction on each first pressure spectrum segment to obtain a plurality of first initial pressure feature sets, and respectively carrying out frequency spectrum feature extraction on each second pressure spectrum segment to obtain a plurality of second initial pressure feature sets;

and carrying out feature normalization on the plurality of first initial pressure feature sets to obtain a first oxygen pressure feature set, and carrying out feature normalization on the plurality of second initial pressure feature sets to obtain a first carbon dioxide pressure feature set.

With reference to the first aspect, in a fifth implementation manner of the first aspect of the present invention, the performing feature compensation on the first oxygen pressure feature set through the first temperature-pressure relation model to obtain a second oxygen pressure feature set, and performing feature compensation on the first carbon dioxide pressure feature set through the second temperature-pressure relation model to obtain a second carbon dioxide pressure feature set, where the performing step includes:

calculating a first characteristic compensation parameter of the first oxygen pressure characteristic set through the first temperature-pressure relation model, and calculating a second characteristic compensation parameter of the first carbon dioxide pressure characteristic set through the second temperature-pressure relation model;

Performing feature weight analysis on the first feature compensation parameters to obtain first pressure feature weights, and performing feature weight analysis on the second feature compensation parameters to obtain second pressure feature weights;

and carrying out feature weighting treatment on the first oxygen pressure feature set according to the first pressure feature weight to obtain a second oxygen pressure feature set, and carrying out feature weighting treatment on the first carbon dioxide pressure feature set according to the second pressure feature weight to obtain a second carbon dioxide pressure feature set.

With reference to the first aspect, in a sixth implementation manner of the first aspect of the present invention, inputting the second oxygen pressure feature set and the second carbon dioxide pressure feature set into a preset gas output pressure control analysis model to perform gas output pressure control analysis, and generating a target pressure control parameter policy includes:

Vector encoding is carried out on the second oxygen pressure characteristic set to obtain a first pressure characteristic encoding vector, and vector encoding is carried out on the second carbon dioxide pressure characteristic set to obtain a second pressure characteristic encoding vector;

vector fusion is carried out on the first pressure characteristic coding vector and the second pressure characteristic coding vector to obtain a target fusion characteristic coding vector;

Inputting the target fusion characteristic coding vector into a preset gas output pressure control analysis model, wherein the gas output pressure control analysis model comprises a coding network, a decoding network and a strategy optimization network, the coding network comprises a bidirectional threshold circulation unit, the decoding network comprises a unidirectional threshold circulation unit and a full connection layer, and the strategy optimization network comprises a genetic algorithm;

Extracting the characteristics of the target fusion characteristic coding vector through a bidirectional threshold circulation unit in the coding network to obtain a target dimensional pressure characteristic vector;

performing feature mapping on the target high-dimensional pressure feature vector through a unidirectional threshold circulation unit in the decoding network to obtain a target pressure feature mapping vector, and performing gas output pressure control parameter prediction on the target pressure feature mapping vector through a full connection layer in the decoding network to obtain a plurality of initial pressure control parameters;

Carrying out parameter strategy initialization on the plurality of initial pressure control parameters through a genetic algorithm in the strategy optimization network to generate a plurality of initial pressure control parameter strategies;

Respectively calculating the adaptation data of each initial pressure control parameter strategy, and carrying out strategy group division on the plurality of initial pressure control parameter strategies according to the adaptation data to obtain strategy group division results;

And generating a plurality of corresponding candidate pressure control parameter strategies according to the strategy group division result, and carrying out strategy optimization solution on the plurality of candidate pressure control parameter strategies to generate a target pressure control parameter strategy.

A second aspect of the present invention provides a gas pressure detection system comprising:

The detection module is used for setting a sensor array of the target oxygen storage device based on an immersion boundary method, and detecting the gas output pressure of the target oxygen storage device through the sensor array to obtain oxygen pressure data and carbon dioxide pressure data;

The construction module is used for acquiring temperature monitoring data of the target oxygen storage device through the sensor array, constructing a first temperature-pressure relation model of the oxygen pressure data and the temperature monitoring data, and constructing a second temperature-pressure relation model of the carbon dioxide pressure data and the temperature monitoring data;

the conversion module is used for performing pressure spectrum conversion on the oxygen pressure data to generate an oxygen pressure spectrum, and performing pressure spectrum conversion on the carbon dioxide pressure data to generate a carbon dioxide pressure spectrum;

The feature extraction module is used for carrying out feature extraction on the oxygen pressure spectrum to obtain a first oxygen pressure feature set, and carrying out feature extraction on the carbon dioxide pressure spectrum to obtain a first carbon dioxide pressure feature set;

the characteristic compensation module is used for carrying out characteristic compensation on the first oxygen pressure characteristic set through the first temperature-pressure relation model to obtain a second oxygen pressure characteristic set, and carrying out characteristic compensation on the first carbon dioxide pressure characteristic set through the second temperature-pressure relation model to obtain a second carbon dioxide pressure characteristic set;

And the analysis module is used for inputting the second oxygen pressure characteristic set and the second carbon dioxide pressure characteristic set into a preset gas output pressure control analysis model to carry out gas output pressure control analysis and generate a target pressure control parameter strategy.

A third aspect of the present invention provides a gas pressure detection apparatus comprising: a memory and at least one processor, the memory having instructions stored therein; the at least one processor invokes the instructions in the memory to cause the gas pressure detection apparatus to perform the gas pressure detection method described above.

A fourth aspect of the present invention provides a computer-readable storage medium having instructions stored therein that, when run on a computer, cause the computer to perform the gas pressure detection method described above.

In the technical scheme provided by the invention, gas output pressure detection is performed based on an immersion boundary method to obtain oxygen pressure data and carbon dioxide pressure data; acquiring temperature monitoring data, and constructing a first temperature-pressure relation model of oxygen pressure data and temperature monitoring data, carbon dioxide pressure data and a second temperature-pressure relation model of oxygen pressure data and temperature monitoring data; performing pressure spectrum conversion to generate an oxygen pressure spectrum and a carbon dioxide pressure spectrum; extracting features to obtain a first oxygen pressure feature set and a first carbon dioxide pressure feature set; performing feature compensation to obtain a second oxygen pressure feature set and a second carbon dioxide pressure feature set; the invention can carry out high-precision detection on the oxygen and carbon dioxide pressure of the target oxygen storage device through an immersion boundary method and a sensor array, and provides accurate gas pressure data. The temperature and pressure relation model is utilized to carry out temperature compensation on the gas pressure characteristic set, so that the stability and the reliability of the system are improved, and accurate pressure detection results can be still maintained under different temperature conditions. Through Fourier transformation and power spectral density analysis, the optimized extraction of oxygen and carbon dioxide pressure spectrums is realized, more detailed information of gas pressure is captured, and the characteristic representation capability of the characteristics is improved. By adopting feature extraction and temperature feature compensation, key features are effectively extracted from the pressure spectrum, and meanwhile, the temperature relation model is used for compensation, so that the stability and consistency of the features are improved. The intelligent control of the output pressure of the target gas is realized through the coding, decoding and strategy optimization network, and the target pressure control parameter strategy adapting to different working conditions is generated. By comprehensively analyzing a plurality of feature sets and adopting a vector fusion technology, the comprehensive analysis of oxygen and carbon dioxide is realized, the adaptability of the system to the complex gas pressure condition is improved, and the control accuracy of the gas output pressure is further improved.

Drawings

FIG. 1 is a schematic diagram of an embodiment of a method for detecting gas pressure according to an embodiment of the present invention;

FIG. 2 is a flow chart of constructing a temperature-pressure relationship model in an embodiment of the invention;

FIG. 3 is a flow chart of generating a pressure spectrum in an embodiment of the invention;

FIG. 4 is a flow chart of feature extraction in an embodiment of the invention;

FIG. 5 is a schematic diagram of an embodiment of a gas pressure detection system according to an embodiment of the present invention;

fig. 6 is a schematic view of an embodiment of a gas pressure detection apparatus according to an embodiment of the present invention.

Detailed Description

The embodiment of the invention provides a gas pressure detection method and a gas pressure detection system, which are used for improving the control accuracy of gas output pressure. The terms "first," "second," "third," "fourth" and the like in the description and in the claims and in the above drawings, if any, are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments described herein may be implemented in other sequences than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed or inherent to such process, method, article, or apparatus.

For ease of understanding, a specific flow of an embodiment of the present invention is described below with reference to fig. 1, and an embodiment of a method for detecting gas pressure in an embodiment of the present invention includes:

s101, setting a sensor array of a target oxygen storage device based on an immersion boundary method, and detecting the gas output pressure of the target oxygen storage device through the sensor array to obtain oxygen pressure data and carbon dioxide pressure data;

it is to be understood that the execution body of the present invention may be a gas pressure detection system, and may also be a terminal or a server, which is not limited herein. The embodiment of the invention is described by taking a server as an execution main body as an example.

Specifically, a sensor array is arranged at the gas output end of the target oxygen storage device based on the immersion boundary method, and the sensor array is connected with a data acquisition unit. The sensor array is responsible for monitoring the pressure conditions at the gas output. These sensors may be pressure sensors or other types of sensors for measuring the pressure of oxygen and carbon dioxide. The data acquisition unit will be used to capture raw signal data generated by the sensor. The sensor array monitors the gas output pressure of the target oxygen storage device and generates raw oxygen signal data as well as raw carbon dioxide signal data. These raw signal data will contain information about the gas pressure, but usually require further processing to obtain useful data. The raw oxygen signal data and raw carbon dioxide signal data will be processed to remove noise, eliminate unnecessary interference, and extract information related to gas pressure. Signal processing may include digital filtering, filter design, time and frequency domain analysis, etc. techniques to ensure that high quality initial oxygen data and initial carbon dioxide data are obtained. Data filtering and data calibration are performed. The data filtering is to further remove noise and fluctuations in the signal to obtain more stable oxygen pressure data and carbon dioxide pressure data. The data calibration rule is to ensure the accuracy and reliability of the data, and through the calibration process, the data can be compared with the known standard, so that the gas pressure condition of the target oxygen storage device can be known more accurately.

S102, acquiring temperature monitoring data of a target oxygen storage device through a sensor array, constructing a first temperature-pressure relation model of oxygen pressure data and the temperature monitoring data, and constructing a second temperature-pressure relation model of carbon dioxide pressure data and the temperature monitoring data;

Specifically, the temperature of the gas output end of the target oxygen storage device is monitored through the sensor array. This process generates temperature monitoring data to reflect the temperature changes of the oxygen storage device. The temperature monitoring data is subjected to a kernel matrix calculation by using a preset gaussian kernel function. Gaussian kernel functions can be used to map raw temperature data into gaussian kernel matrices, which helps to handle nonlinear relationships and prepare the data for modeling. And taking the oxygen pressure data as a first target value, and taking a Gaussian kernel matrix as a first feature matrix to construct a first input-output pair. And similarly, taking the carbon dioxide pressure data as a second target value, and taking a Gaussian kernel matrix as a second feature matrix to construct a second input-output pair. This step is to prepare the data for model training. And carrying out Gaussian linear regression parameter calculation on the first input-output pair through a preset first Gaussian linear regression algorithm, so as to obtain a first Gaussian linear regression parameter. Similarly, a second Gaussian linear regression algorithm is used for carrying out Gaussian linear regression parameter calculation on the second input-output pair to obtain a second Gaussian linear regression parameter. These parameters will be used to construct a temperature pressure relationship model. And constructing a first temperature-pressure relation model of oxygen pressure data and temperature monitoring data according to the first Gaussian linear regression parameters, and constructing a second temperature-pressure relation model of carbon dioxide pressure data and temperature monitoring data according to the second Gaussian linear regression parameters. These models will be used to predict the relationship between gas pressure and temperature to better understand the performance of the gas storage device.

S103, performing pressure spectrum conversion on the oxygen pressure data to generate an oxygen pressure spectrum, and performing pressure spectrum conversion on the carbon dioxide pressure data to generate a carbon dioxide pressure spectrum;

the oxygen pressure data was fourier transformed. Fourier transform is a mathematical tool that converts a time domain signal (in this case, the change in oxygen pressure over time) into a frequency domain signal to reveal the spectral content of the signal. This will generate a first fourier transform result, which contains information about the oxygen pressure spectrum. Likewise, the carbon dioxide pressure data is fourier transformed to produce a second fourier transform result. This result contains information about the carbon dioxide pressure spectrum. Fourier transforms transform time domain data into frequency domain data, thereby enabling the server to better understand the spectral characteristics of pressure changes. The first fourier transform result is squared to obtain a first power spectral density. The squaring helps to highlight the energy distribution in the spectrum, enabling the server to better analyze the energy distribution of the pressure spectrum. Also, the second fourier transform result is squared to obtain a second power spectral density. This power spectral density contains information about the energy distribution of the carbon dioxide pressure spectrum. A corresponding oxygen pressure spectrum is generated based on the first power spectral density and a corresponding carbon dioxide pressure spectrum is generated based on the second power spectral density. These spectra will reflect the distribution of oxygen and carbon dioxide pressures over the different frequency components, helping the server to learn more about the gas pressure characteristics. For example, by fourier transforming the oxygen and carbon dioxide pressure data, the server may convert the pressure data into the spectral domain, revealing specific frequency components in the spectrum. Squaring the first and second fourier transform results will enable the server to better understand the energy distribution of the spectrum, i.e. the contributions of the different frequency components. The resulting power spectral density will show the energy distribution of the oxygen and carbon dioxide pressure spectra. This helps to monitor fluctuations in gas pressure over different frequency ranges, for example, whether a particular frequency component is present in relation to plant operation or process parameters.

S104, carrying out feature extraction on the oxygen pressure spectrum to obtain a first oxygen pressure feature set, and carrying out feature extraction on the carbon dioxide pressure spectrum to obtain a first carbon dioxide pressure feature set;

Specifically, the server performs spectrum segmentation on the oxygen pressure spectrum. The spectral data is divided into a plurality of different spectral segments. Similarly, the carbon dioxide pressure spectrum is also subjected to spectrum segmentation to obtain a plurality of different spectrum segments. This step helps to divide the spectral data into smaller portions in order to analyze the characteristics of each portion in more detail. The server performs spectral feature extraction on each first pressure spectrum segment. The server analyzes each spectrum segment to extract its spectral features. These features include the main frequency content in the spectrum, the amplitude information of the spectrum, the phase information of the spectrum, etc. These feature extraction processes will generate a plurality of first initial pressure feature sets. And likewise, the server performs spectrum characteristic extraction on each second pressure spectrum segment to obtain a plurality of second initial pressure characteristic sets. These feature sets will reflect the characteristics of the different spectrum segments, helping the server to more fully understand the characteristics of the oxygen and carbon dioxide pressure spectrum. The server performs feature normalization on the first plurality of initial pressure feature sets. Feature normalization is to ensure that the numerical ranges of the different features are consistent for comparison and analysis. This will result in a first set of oxygen pressure characteristics. Similarly, the server performs feature normalization on the plurality of second initial pressure feature sets to obtain a first carbon dioxide pressure feature set.

S105, performing feature compensation on the first oxygen pressure feature set through a first temperature-pressure relation model to obtain a second oxygen pressure feature set, and performing feature compensation on the first carbon dioxide pressure feature set through a second temperature-pressure relation model to obtain a second carbon dioxide pressure feature set;

Specifically, the server calculates a first feature compensation parameter for the first set of oxygen pressure features using the first temperature-pressure relationship model, and calculates a second feature compensation parameter for the first set of carbon dioxide pressure features using the second temperature-pressure relationship model. These feature compensation parameters are calculated based on known relationships between temperature and pressure to correct the data in the feature set. The server performs feature weight analysis on the first feature compensation parameter to obtain a first pressure feature weight. These weights reflect the extent to which different features in the first set of oxygen pressure features affect pressure. Similarly, the server performs a feature weight analysis on the second feature compensation parameter to obtain second pressure feature weights reflecting the degree of impact of different features in the first set of carbon dioxide pressure features on pressure. And the server performs characteristic weighting processing on the first oxygen pressure characteristic set according to the first pressure characteristic weight to obtain a second oxygen pressure characteristic set. This weighting helps to highlight features that have the greatest impact on pressure, thereby improving the information quality of the feature set. Similarly, the server performs feature weighting processing on the first carbon dioxide pressure feature set according to the second pressure feature weight to obtain a second carbon dioxide pressure feature set. For example, the server calculates a first feature compensation parameter for the first set of oxygen pressure features using the first temperature-pressure relationship model to correct for the effects of temperature changes on the oxygen pressure features. And simultaneously, calculating a second characteristic compensation parameter of the first carbon dioxide pressure characteristic set by using a second temperature-pressure relation model so as to correct the influence of temperature change on the carbon dioxide pressure characteristic. The server performs a feature weight analysis to determine which features have the greatest impact on gas pressure at different temperature conditions. Based on these weights, the server weights the oxygen and carbon dioxide pressure feature sets to obtain a more accurate second oxygen pressure feature set and second carbon dioxide pressure feature set.

S106, inputting the second oxygen pressure characteristic set and the second carbon dioxide pressure characteristic set into a preset gas output pressure control analysis model to perform gas output pressure control analysis, and generating a target pressure control parameter strategy.

Specifically, the server performs vector coding on the second oxygen pressure feature set to obtain a first pressure feature coding vector, and performs vector coding on the second carbon dioxide pressure feature set to obtain a second pressure feature coding vector. This encoding process facilitates the representation of the information of the feature set in vector form for subsequent analysis and processing. And the server performs vector fusion on the first pressure characteristic coding vector and the second pressure characteristic coding vector to obtain a target fusion characteristic coding vector. This fusion process helps to combine the characteristic information of oxygen and carbon dioxide together to more fully represent the state of the gas system. The server inputs the target fusion feature coding vector into a preset gas output pressure control analysis model. This model includes an encoding network, a decoding network, and a policy optimization network. The coding network comprises a bidirectional threshold circulation unit, the decoding network comprises a unidirectional threshold circulation unit and a full connection layer, and the strategy optimization network comprises a genetic algorithm. And extracting the characteristics of the target fusion characteristic coding vector by a server through a bidirectional threshold circulation unit in the coding network to obtain a target dimensional pressure characteristic vector. This high-dimensional feature vector contains critical information about the gas system and can be used for subsequent analysis and control. And performing feature mapping on the target high-dimensional pressure feature vector by the server through a unidirectional threshold circulation unit in the decoding network to obtain a target pressure feature mapping vector. This mapping process helps to transform the high-dimensional feature vectors into lower-dimensional vectors that are more suitable for control parameter prediction. And predicting the gas output pressure control parameters by the server according to the target pressure characteristic mapping vector through a full connection layer in the decoding network to obtain a plurality of initial pressure control parameters. These parameters are predicted from the input characteristic information and training of the model for controlling the gas output pressure. And carrying out parameter strategy initialization on the plurality of initial pressure control parameters by the server through a genetic algorithm in the strategy optimization network, and generating a plurality of initial pressure control parameter strategies. Genetic algorithms are an optimization method for finding the best combination of parameters to achieve the desired control effect. And the server calculates the fitness data of each initial pressure control parameter strategy, and performs strategy group division on a plurality of initial pressure control parameter strategies according to the data to obtain a strategy group division result. This step helps determine which parameter policies are more suitable for practical use, searching through a wider policy space. And generating a plurality of corresponding candidate pressure control parameter strategies according to the strategy group division result, and carrying out strategy optimization solution on the strategies to generate target pressure control parameter strategies. This strategy will be used for control of the actual gas output pressure, ensuring stable operation of the system under different conditions.

According to the embodiment of the invention, the oxygen and carbon dioxide pressure of the target oxygen storage device can be detected with high precision through the immersion boundary method and the sensor array, and accurate gas pressure data can be provided. The temperature and pressure relation model is utilized to carry out temperature compensation on the gas pressure characteristic set, so that the stability and the reliability of the system are improved, and accurate pressure detection results can be still maintained under different temperature conditions. Through Fourier transformation and power spectral density analysis, the optimized extraction of oxygen and carbon dioxide pressure spectrums is realized, more detailed information of gas pressure is captured, and the characteristic representation capability of the characteristics is improved. By adopting feature extraction and temperature feature compensation, key features are effectively extracted from the pressure spectrum, and meanwhile, the temperature relation model is used for compensation, so that the stability and consistency of the features are improved. The intelligent control of the output pressure of the target gas is realized through the coding, decoding and strategy optimization network, and the target pressure control parameter strategy adapting to different working conditions is generated. By comprehensively analyzing a plurality of feature sets and adopting a vector fusion technology, the comprehensive analysis of oxygen and carbon dioxide is realized, the adaptability of the system to the complex gas pressure condition is improved, and the control accuracy of the gas output pressure is further improved.

In a specific embodiment, the process of executing step S101 may specifically include the following steps:

(1) Setting a sensor array at a gas output end of the target oxygen storage device based on an immersion boundary method, and connecting the sensor array with a data acquisition unit;

(2) Detecting the gas output pressure of the target oxygen storage device through the sensor array to obtain original oxygen signal data and original carbon dioxide signal data;

(3) Performing signal processing on the original oxygen signal data to obtain initial oxygen data, and performing signal processing on the original carbon dioxide signal data to obtain initial carbon dioxide data;

(4) And performing data filtering and data calibration on the initial oxygen data to obtain oxygen pressure data, and performing data filtering and data calibration on the initial carbon dioxide data to obtain carbon dioxide pressure data.

In particular, the server is provided with an array of sensors arranged at the gas output of the target oxygen storage device to capture the pressure conditions of the gas output. These sensors may employ different technologies, such as piezoelectric sensors, piezoresistive sensors, or other types of sensors, to meet the requirements of a particular application. This sensor array would be a key tool for the server to monitor the gas pressure. The server connects the sensor array with the data acquisition unit. The data acquisition unit is an important component and is used for receiving signal data acquired by the sensor array, processing the signal and storing the data. The data acquisition unit may be a dedicated hardware device or an embedded system, depending on the application requirements. By connecting the sensor array and the data acquisition unit, the server establishes a complete data acquisition system, and can acquire gas pressure information in real time. Once the sensor array and data acquisition unit setup is complete and connected, the server may begin gas output pressure detection. The sensor array continuously monitors the gas output pressure of the target oxygen storage device and transmits the raw oxygen signal data and the raw carbon dioxide signal data to the data acquisition unit. The raw oxygen signal data and the raw carbon dioxide signal data contain noise and interference from the sensor, and thus require signal processing. The signal processing includes steps such as noise removal, filtering and calibration to extract useful information and accurately reflect the pressure change of the gas. For raw oxygen signal data, the signal processing includes removing high frequency noise, filtering to smooth the data, and then calibrating to convert the signal to a true oxygen pressure value. Similarly, for raw carbon dioxide signal data, similar signal processing steps are also required to obtain accurate carbon dioxide pressure data. These processed data will be the basis for gas output pressure detection and can be used for further analysis and control. Through the process, the server can realize the gas output pressure detection of the target oxygen storage device, and acquire accurate oxygen and carbon dioxide pressure data so as to meet the requirements of various application fields. For example, a sensor array is mounted on the oxygen output conduit to monitor the output pressure of oxygen. Raw oxygen signal data collected by the sensor array is subjected to a signal processing step that includes removing noise from pipe vibrations and other environmental disturbances, and then converting the signal to an actual oxygen pressure value by calibration. These data may be used to ensure the stability of the oxygen supply system to meet oxygen demand.

In a specific embodiment, as shown in fig. 2, the process of executing step S102 may specifically include the following steps:

S201, performing temperature monitoring on a gas output end of a target oxygen storage device through a sensor array to obtain temperature monitoring data;

S202, performing kernel matrix calculation on the temperature monitoring data through a preset Gaussian kernel function to obtain a Gaussian kernel matrix;

S203, taking oxygen pressure data as a first target value, taking a Gaussian kernel matrix as a first characteristic matrix, constructing a first input-output pair according to the first target value and the first characteristic matrix, simultaneously taking carbon dioxide pressure data as a second target value, taking the Gaussian kernel matrix as a second characteristic matrix, and constructing a second input-output pair according to the second target value and the second characteristic matrix;

S204, carrying out Gaussian linear regression parameter calculation on the first input-output pair through a preset first Gaussian linear regression algorithm to obtain a first Gaussian linear regression parameter, and carrying out Gaussian linear regression parameter calculation on the second input-output pair through a preset second Gaussian linear regression algorithm to obtain a second Gaussian linear regression parameter;

S205, constructing a first temperature-pressure relation model of oxygen pressure data and temperature monitoring data according to the first Gaussian linear regression parameters, and constructing carbon dioxide pressure data and a second temperature-pressure relation model of the oxygen pressure data and the temperature monitoring data according to the second Gaussian linear regression parameters.

Specifically, the temperature of the gas output end of the target oxygen storage device is monitored through the sensor array. The sensors monitor the temperature of the gas output end in real time and record temperature monitoring data. These temperature monitoring data are critical for subsequent analysis, as temperature changes can affect the pressure of the gas. And performing kernel matrix calculation on the temperature monitoring data by using a preset Gaussian kernel function. Gaussian kernel functions are a common type of kernel function used for feature mapping in machine learning and statistical analysis. The purpose of this step is to convert the temperature monitoring data into a gaussian kernel matrix for subsequent analysis. Taking the oxygen pressure data as a first target value, taking a Gaussian kernel matrix as a first feature matrix, and constructing a first input-output pair according to the first target value and the first feature matrix. And simultaneously taking the carbon dioxide pressure data as a second target value, taking the Gaussian kernel matrix as a second characteristic matrix, and constructing a second input-output pair according to the second target value and the second characteristic matrix. These steps are to prepare the data for regression analysis. And carrying out Gaussian linear regression parameter calculation on the first input-output pair by using a preset first Gaussian linear regression algorithm to obtain a first Gaussian linear regression parameter. And meanwhile, carrying out Gaussian linear regression parameter calculation on the second input-output pair through a preset second Gaussian linear regression algorithm to obtain a second Gaussian linear regression parameter. These regression parameters will be used to model the relationship between temperature and oxygen pressure, carbon dioxide pressure. And constructing a first temperature-pressure relation model of oxygen pressure data and temperature monitoring data according to the first Gaussian linear regression parameters, and constructing a second temperature-pressure relation model of carbon dioxide pressure data and temperature monitoring data according to the second Gaussian linear regression parameters. These relational models will be used to analyze and predict pressure changes of oxygen and carbon dioxide for better control and management of the gas output pressure of the oxygen storage device.

In a specific embodiment, as shown in fig. 3, the process of executing step S103 may specifically include the following steps:

s301, carrying out Fourier transform on oxygen pressure data to obtain a first Fourier transform result, and carrying out Fourier transform on carbon dioxide pressure data to obtain a second Fourier transform result;

s302, squaring a first Fourier transform result to obtain a first power spectrum density, and squaring a second Fourier transform result to obtain a second power spectrum density;

S303, generating a corresponding oxygen pressure spectrum according to the first power spectrum density, and generating a corresponding carbon dioxide pressure spectrum according to the second power spectrum density.

Specifically, the server needs to perform fourier transform on the oxygen pressure data and the carbon dioxide pressure data, respectively. Fourier transforms are mathematical tools for converting time domain data (data over time) into frequency domain data (data over frequency). For example, assume that there is oxygen pressure data recorded over a period of time. By performing a fourier transform on these data, the server can convert them into a frequency domain representation, which contains component information of different frequencies. This helps the server to identify periodic vibrations or fluctuations in the signal. The power spectral density is calculated. The power spectral density represents the intensity of the different frequency components of the signal in the frequency domain. The server calculates a power spectral density of the oxygen pressure data and a power spectral density of the carbon dioxide pressure data. In order to calculate the power spectral density, the server needs to square the fourier transform result, which can be seen as the strength of the signal frequency domain components. This will result in a first power spectral density and a second power spectral density corresponding to the spectra of the oxygen and carbon dioxide signals, respectively. From the first and second power spectral densities, the server may generate a spectrum of oxygen and carbon dioxide. The spectrum is a graphical representation of the frequency content with respect to its intensity. This may help the server visualize the frequency domain characteristics of the gas pressure data. For example, assume that after the server performs a fourier transform on the oxygen data, the server sees distinct peaks at certain frequencies, indicating the presence of particular vibrations or fluctuations. These vibrations are associated with particular aspects of the industrial process. Also, after performing fourier transform on the carbon dioxide data and calculating the power spectral density, the server may identify frequency components related to the carbon dioxide pressure variation. Such analysis may help the server better understand the behavior of the gas in the industrial process and take appropriate action when needed. For example, if the server finds that vibrations of a particular frequency are associated with equipment failure, the server may perform repair and maintenance early to improve production efficiency.

In a specific embodiment, as shown in fig. 4, the process of executing step S104 may specifically include the following steps:

s401, performing spectrum segmentation on an oxygen pressure spectrum to obtain a plurality of first pressure spectrum segments, and performing spectrum segmentation on a carbon dioxide pressure spectrum to obtain a plurality of second pressure spectrum segments;

S402, respectively carrying out frequency spectrum feature extraction on each first pressure spectrum segment to obtain a plurality of first initial pressure feature sets, and respectively carrying out frequency spectrum feature extraction on each second pressure spectrum segment to obtain a plurality of second initial pressure feature sets;

S403, performing feature normalization on the first initial pressure feature sets to obtain a first oxygen pressure feature set, and performing feature normalization on the second initial pressure feature sets to obtain a first carbon dioxide pressure feature set.

Specifically, the oxygen pressure spectrum and the carbon dioxide pressure spectrum are subjected to spectrum division. The entire spectrum is divided into a plurality of smaller spectral segments, which may be accomplished by means of a window function or the like. This will generate a plurality of first pressure bands of spectrum and a plurality of second pressure bands of spectrum. And respectively extracting spectral features of each first pressure spectrum segment. This involves extracting features from each spectrum segment that are related to oxygen pressure. Common spectral features include center frequency, spectral width, peak amplitude, etc. These feature extraction methods typically use signal processing techniques such as fourier transforms and the like. And simultaneously, extracting the frequency spectrum characteristics of each second pressure frequency spectrum section. These feature extraction methods are similar to the first pressure spectrum segment, but should be designed for features related to carbon dioxide pressure. Feature normalization is performed on the first plurality of pressure feature sets. The extracted features are normalized to ensure that they are on the same scale. This may be achieved using a normalization method, such as Z-score normalization. Likewise, a plurality of second pressure feature sets are feature normalized. This is to ensure that the second set of oxygen pressure features are on the same scale. Through these steps, the server will obtain a first set of oxygen pressure characteristics and a first set of carbon dioxide pressure characteristics. These feature sets contain useful information extracted from the oxygen pressure and carbon dioxide pressure spectra that can be used for subsequent analysis, modeling, or control purposes. For example, assume that a server divides a spectrum into different frequency bands, such as a low band and a high band. For the oxygen spectrum in the low frequency range, the server extracts the dominant low frequency vibration frequency and corresponding amplitude. For the carbon dioxide spectrum in the high frequency band, the server finds the frequency and amplitude of the dither. The server will normalize these features to ensure that they have similar scope for better comparison and analysis. This enables the server to compare the characteristics of the gas pressure over different spectrum segments, identifying the impact of different frequency components on the industrial process. For example, the server may find that high-band carbon dioxide pressure changes are related to equipment vibration, while low-band oxygen pressure changes are related to gas supply system adjustments.

In a specific embodiment, the process of executing step S105 may specifically include the following steps:

(1) Calculating a first characteristic compensation parameter of a first oxygen pressure characteristic set through a first temperature-pressure relation model, and calculating a second characteristic compensation parameter of a first carbon dioxide pressure characteristic set through a second temperature-pressure relation model;

(2) Performing feature weight analysis on the first feature compensation parameters to obtain first pressure feature weights, and performing feature weight analysis on the second feature compensation parameters to obtain second pressure feature weights;

(3) And carrying out feature weighting treatment on the first oxygen pressure feature set according to the first pressure feature weight to obtain a second oxygen pressure feature set, and carrying out feature weighting treatment on the first carbon dioxide pressure feature set according to the second pressure feature weight to obtain a second carbon dioxide pressure feature set.

Specifically, a first characteristic compensation parameter of a first set of oxygen pressure characteristics is calculated using a first temperature-pressure relationship model. This parameter will be used to correct or compensate the first set of oxygen pressure characteristics to take into account the effect of temperature on oxygen pressure. The calculation of the first characteristic compensation parameter is typically based on an established temperature-pressure relationship model. Simultaneously, a second characteristic compensation parameter of the first carbon dioxide pressure characteristic set is calculated by using a second temperature-pressure relation model. This parameter will be used to correct or compensate the first set of carbon dioxide pressure characteristics to take into account the effect of temperature on the carbon dioxide pressure. The calculation of the second characteristic compensation parameter is also based on an established temperature-pressure relationship model. And carrying out characteristic weight analysis on the first characteristic compensation parameters. The first characteristic compensation parameter is analyzed to determine the importance of the different parameters to the oxygen pressure characteristic. These weights may be assigned according to the degree of influence of the parameters. And simultaneously, carrying out characteristic weight analysis on the second characteristic compensation parameters. In this way the importance of the different parameters to the carbon dioxide pressure characteristics can be determined. And carrying out feature weighting processing on the first oxygen pressure feature set according to the first pressure feature weight. This will take into account the weight of each feature to generate a second set of oxygen pressure features. This weighting process can be implemented by a simple linear combination in which each feature is multiplied by its corresponding weight, and then added together. And similarly, carrying out characteristic weighting treatment on the first carbon dioxide pressure characteristic set according to the second pressure characteristic weight. This will generate a second set of secondary oxygen pressure characteristics, wherein the weight of each characteristic is taken into account.

In a specific embodiment, the process of executing step S106 may specifically include the following steps:

(1) Vector encoding is carried out on the second oxygen pressure characteristic set to obtain a first pressure characteristic encoding vector, and vector encoding is carried out on the second carbon dioxide pressure characteristic set to obtain a second pressure characteristic encoding vector;

(2) Vector fusion is carried out on the first pressure characteristic coding vector and the second pressure characteristic coding vector to obtain a target fusion characteristic coding vector;

(3) Inputting the target fusion characteristic coding vector into a preset gas output pressure control analysis model, wherein the gas output pressure control analysis model comprises a coding network, a decoding network and a strategy optimization network, the coding network comprises a bidirectional threshold circulation unit, the decoding network comprises a unidirectional threshold circulation unit and a full connection layer, and the strategy optimization network comprises a genetic algorithm;

(4) Extracting features of the target fusion feature coding vector through a bidirectional threshold circulation unit in the coding network to obtain a target dimensional pressure feature vector;

(5) Performing feature mapping on the target high-dimensional pressure feature vector through a unidirectional threshold circulation unit in the decoding network to obtain a target pressure feature mapping vector, and performing gas output pressure control parameter prediction on the target pressure feature mapping vector through a full-connection layer in the decoding network to obtain a plurality of initial pressure control parameters;

(6) Carrying out parameter strategy initialization on a plurality of initial pressure control parameters through a genetic algorithm in a strategy optimization network to generate a plurality of initial pressure control parameter strategies;

(7) Respectively calculating the adaptability data of each initial pressure control parameter strategy, and carrying out strategy group division on a plurality of initial pressure control parameter strategies according to the adaptability data to obtain strategy group division results;

(8) And generating a plurality of corresponding candidate pressure control parameter strategies according to the strategy group division result, and carrying out strategy optimization solution on the plurality of candidate pressure control parameter strategies to generate a target pressure control parameter strategy.

Specifically, the second set of oxygen pressure features is vector coded. The data in the second set of oxygen pressure features is converted into a vector representation. This typically involves mapping the eigenvalues into a multidimensional vector space for computation and processing. Likewise, the second set of carbon dioxide pressure features is vector coded. And vector fusion is carried out on the first pressure characteristic coding vector and the second pressure characteristic coding vector. This is the process of combining two feature encoded vectors into one target fusion feature encoded vector. Fusion may take various forms, such as simply joining them together or applying a matrix operation to ensure that they are effectively bonded together. And inputting the target fusion feature coding vector into a preset gas output pressure control analysis model. This model includes an encoding network, a decoding network, and a policy optimization network. The coding network typically includes a bi-directional threshold cycling unit for extracting features. The decoding network typically includes a one-way threshold cycling unit and a fully connected layer for mapping features into pressure control parameters. Policy optimization networks typically include genetic algorithms for optimizing pressure control policies. And extracting the characteristics of the target fusion characteristic coding vector through a bidirectional threshold circulation unit in the coding network to obtain a target dimensional pressure characteristic vector. This stage is used to extract valid features from the code for further processing in subsequent steps. And performing feature mapping on the target high-dimensional pressure feature vector through a unidirectional threshold circulation unit in the decoding network to obtain a target pressure feature mapping vector. This is a representation that maps the high dimensional features to lower dimensions in order to better predict the gas output pressure control parameters. And then, predicting the gas output pressure control parameters through the full connection layer in the decoding network to the target pressure characteristic mapping vector to obtain a plurality of initial pressure control parameters. This step is used to generate an initial estimate of the control parameters from the mapped features. And initializing parameter strategies for a plurality of initial pressure control parameters through a genetic algorithm in the strategy optimization network, and generating a plurality of initial pressure control parameter strategies. Genetic algorithms are typically used to optimize parameter policies to better meet specific performance metrics or objectives. Fitness data for each initial pressure control parameter strategy is calculated. This is to evaluate the performance of each policy for comparison and selection in subsequent policy optimizations. And carrying out strategy group division on a plurality of initial pressure control parameter strategies according to the fitness data to obtain strategy group division results. This helps to separate the strategies into different populations in order to better select candidate strategies. And generating a plurality of candidate pressure control parameter strategies according to the strategy group division result, and carrying out strategy optimization solving on the plurality of candidate strategies. This step will ultimately generate a target pressure control parameter strategy for control of the gas output pressure.

The method for detecting gas pressure in the embodiment of the present invention is described above, and the gas pressure detecting system in the embodiment of the present invention is described below, referring to fig. 5, where an embodiment of the gas pressure detecting system in the embodiment of the present invention includes:

the detection module 501 is configured to set a sensor array of a target oxygen storage device based on an immersion boundary method, and perform gas output pressure detection on the target oxygen storage device through the sensor array to obtain oxygen pressure data and carbon dioxide pressure data;

A construction module 502, configured to acquire temperature monitoring data of the target oxygen storage device through the sensor array, and construct a first temperature-pressure relationship model of the oxygen pressure data and the temperature monitoring data, and construct a second temperature-pressure relationship model of the carbon dioxide pressure data and the temperature monitoring data;

the conversion module 503 is configured to perform pressure spectrum conversion on the oxygen pressure data to generate an oxygen pressure spectrum, and perform pressure spectrum conversion on the carbon dioxide pressure data to generate a carbon dioxide pressure spectrum;

the feature extraction module 504 is configured to perform feature extraction on the oxygen pressure spectrum to obtain a first oxygen pressure feature set, and perform feature extraction on the carbon dioxide pressure spectrum to obtain a first carbon dioxide pressure feature set;

the feature compensation module 505 is configured to perform feature compensation on the first oxygen pressure feature set through the first temperature-pressure relationship model to obtain a second oxygen pressure feature set, and perform feature compensation on the first carbon dioxide pressure feature set through the second temperature-pressure relationship model to obtain a second carbon dioxide pressure feature set;

the analysis module 506 is configured to input the second oxygen pressure feature set and the second carbon dioxide pressure feature set into a preset gas output pressure control analysis model to perform gas output pressure control analysis, and generate a target pressure control parameter policy.

Through the cooperation of the components, the oxygen pressure and the carbon dioxide pressure of the target oxygen storage device can be detected with high precision through an immersion boundary method and a sensor array, and accurate gas pressure data can be provided. The temperature and pressure relation model is utilized to carry out temperature compensation on the gas pressure characteristic set, so that the stability and the reliability of the system are improved, and accurate pressure detection results can be still maintained under different temperature conditions. Through Fourier transformation and power spectral density analysis, the optimized extraction of oxygen and carbon dioxide pressure spectrums is realized, more detailed information of gas pressure is captured, and the characteristic representation capability of the characteristics is improved. By adopting feature extraction and temperature feature compensation, key features are effectively extracted from the pressure spectrum, and meanwhile, the temperature relation model is used for compensation, so that the stability and consistency of the features are improved. The intelligent control of the output pressure of the target gas is realized through the coding, decoding and strategy optimization network, and the target pressure control parameter strategy adapting to different working conditions is generated. By comprehensively analyzing a plurality of feature sets and adopting a vector fusion technology, the comprehensive analysis of oxygen and carbon dioxide is realized, the adaptability of the system to the complex gas pressure condition is improved, and the control accuracy of the gas output pressure is further improved.

The gas pressure detection system in the embodiment of the present invention is described in detail from the point of view of the modularized functional entity in fig. 5 above, and the gas pressure detection apparatus in the embodiment of the present invention is described in detail from the point of view of hardware processing below.

Fig. 6 is a schematic diagram of a gas pressure detection apparatus according to an embodiment of the present invention, where the gas pressure detection apparatus 600 may have a relatively large difference due to different configurations or performances, and may include one or more processors (central processing units, CPU) 610 (e.g., one or more processors) and a memory 620, one or more storage mediums 630 (e.g., one or more mass storage devices) storing application programs 633 or data 632. Wherein the memory 620 and the storage medium 630 may be transitory or persistent storage. The program stored in the storage medium 630 may include one or more modules (not shown), each of which may include a series of instruction operations on the gas pressure detection device 600. Still further, the processor 610 may be configured to communicate with the storage medium 630 and execute a series of instruction operations in the storage medium 630 on the gas pressure detection device 600.

The gas pressure detection apparatus 600 may also include one or more power supplies 640, one or more wired or wireless network interfaces 650, one or more input/output interfaces 660, and/or one or more operating systems 631, such as Windows Server, mac OS X, unix, linux, freeBSD, and the like. It will be appreciated by those skilled in the art that the gas pressure detection apparatus structure shown in fig. 6 is not limiting of the gas pressure detection apparatus and may include more or fewer components than shown, or certain components may be combined, or a different arrangement of components.

The present invention also provides a gas pressure detection apparatus comprising a memory and a processor, the memory storing computer readable instructions which, when executed by the processor, cause the processor to perform the steps of the gas pressure detection method in the above embodiments.

The present invention also provides a computer readable storage medium, which may be a non-volatile computer readable storage medium, or a volatile computer readable storage medium, having stored therein instructions that, when executed on a computer, cause the computer to perform the steps of the gas pressure detection method.

It will be clear to those skilled in the art that, for convenience and brevity of description, specific working procedures of the above-described systems, apparatuses and units may refer to corresponding procedures in the foregoing method embodiments, which are not repeated herein.

The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied essentially or in part or all of the technical solution or in part in the form of a software product stored in a storage medium, including instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a read-only memory (ROM), a random access memory (random access memory, RAM), a magnetic disk, or an optical disk, or other various media capable of storing program codes.

The above embodiments are only for illustrating the technical solution of the present invention, and not for limiting the same; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A gas pressure detection method, characterized in that the gas pressure detection method comprises:

Acquiring temperature monitoring data of the target oxygen storage device through the sensor array, constructing a first temperature-pressure relation model of the oxygen pressure data and the temperature monitoring data, and constructing a second temperature-pressure relation model of the carbon dioxide pressure data and the temperature monitoring data; the method specifically comprises the following steps: the temperature of the gas output end of the target oxygen storage device is monitored through the sensor array, and temperature monitoring data are obtained; performing kernel matrix calculation on the temperature monitoring data through a preset Gaussian kernel function to obtain a Gaussian kernel matrix; taking the oxygen pressure data as a first target value, taking the Gaussian kernel matrix as a first characteristic matrix, constructing a first input-output pair according to the first target value and the first characteristic matrix, taking the carbon dioxide pressure data as a second target value, taking the Gaussian kernel matrix as a second characteristic matrix, and constructing a second input-output pair according to the second target value and the second characteristic matrix; carrying out Gaussian linear regression parameter calculation on the first input-output pair through a preset first Gaussian linear regression algorithm to obtain a first Gaussian linear regression parameter, and carrying out Gaussian linear regression parameter calculation on the second input-output pair through a preset second Gaussian linear regression algorithm to obtain a second Gaussian linear regression parameter; constructing a first temperature-pressure relation model of the oxygen pressure data and the temperature monitoring data according to the first Gaussian linear regression parameters, and constructing a second temperature-pressure relation model of the carbon dioxide pressure data and the temperature monitoring data according to the second Gaussian linear regression parameters;

Inputting the second oxygen pressure characteristic set and the second carbon dioxide pressure characteristic set into a preset gas output pressure control analysis model to perform gas output pressure control analysis, and generating a target pressure control parameter strategy; the method specifically comprises the following steps: vector encoding is carried out on the second oxygen pressure characteristic set to obtain a first pressure characteristic encoding vector, and vector encoding is carried out on the second carbon dioxide pressure characteristic set to obtain a second pressure characteristic encoding vector; vector fusion is carried out on the first pressure characteristic coding vector and the second pressure characteristic coding vector to obtain a target fusion characteristic coding vector; inputting the target fusion characteristic coding vector into a preset gas output pressure control analysis model, wherein the gas output pressure control analysis model comprises a coding network, a decoding network and a strategy optimization network, the coding network comprises a bidirectional threshold circulation unit, the decoding network comprises a unidirectional threshold circulation unit and a full connection layer, and the strategy optimization network comprises a genetic algorithm; extracting the characteristics of the target fusion characteristic coding vector through a bidirectional threshold circulation unit in the coding network to obtain a target dimensional pressure characteristic vector; performing feature mapping on the target high-dimensional pressure feature vector through a unidirectional threshold circulation unit in the decoding network to obtain a target pressure feature mapping vector, and performing gas output pressure control parameter prediction on the target pressure feature mapping vector through a full connection layer in the decoding network to obtain a plurality of initial pressure control parameters; carrying out parameter strategy initialization on the plurality of initial pressure control parameters through a genetic algorithm in the strategy optimization network to generate a plurality of initial pressure control parameter strategies; respectively calculating the adaptation data of each initial pressure control parameter strategy, and carrying out strategy group division on the plurality of initial pressure control parameter strategies according to the adaptation data to obtain strategy group division results; and generating a plurality of corresponding candidate pressure control parameter strategies according to the strategy group division result, and carrying out strategy optimization solution on the plurality of candidate pressure control parameter strategies to generate a target pressure control parameter strategy.

2. The method for detecting gas pressure according to claim 1, wherein the step of setting a sensor array of a target oxygen storage device based on an immersion boundary method, and detecting gas output pressure of the target oxygen storage device by the sensor array to obtain oxygen pressure data and carbon dioxide pressure data comprises the steps of:

3. The gas pressure detection method according to claim 1, wherein the performing pressure spectrum conversion on the oxygen pressure data to generate an oxygen pressure spectrum, and performing pressure spectrum conversion on the carbon dioxide pressure data to generate a carbon dioxide pressure spectrum, comprises:

4. The method for detecting gas pressure according to claim 1, wherein the performing feature extraction on the oxygen pressure spectrum to obtain a first oxygen pressure feature set, and performing feature extraction on the carbon dioxide pressure spectrum to obtain a first carbon dioxide pressure feature set, includes:

5. The method for detecting gas pressure according to claim 1, wherein the performing feature compensation on the first oxygen pressure feature set through the first temperature-pressure relation model to obtain a second oxygen pressure feature set, and performing feature compensation on the first carbon dioxide pressure feature set through the second temperature-pressure relation model to obtain a second carbon dioxide pressure feature set, includes:

6. A gas pressure detection system, the gas pressure detection system comprising:

The construction module is used for acquiring temperature monitoring data of the target oxygen storage device through the sensor array, constructing a first temperature-pressure relation model of the oxygen pressure data and the temperature monitoring data, and constructing a second temperature-pressure relation model of the carbon dioxide pressure data and the temperature monitoring data; the method specifically comprises the following steps: the temperature of the gas output end of the target oxygen storage device is monitored through the sensor array, and temperature monitoring data are obtained; performing kernel matrix calculation on the temperature monitoring data through a preset Gaussian kernel function to obtain a Gaussian kernel matrix; taking the oxygen pressure data as a first target value, taking the Gaussian kernel matrix as a first characteristic matrix, constructing a first input-output pair according to the first target value and the first characteristic matrix, taking the carbon dioxide pressure data as a second target value, taking the Gaussian kernel matrix as a second characteristic matrix, and constructing a second input-output pair according to the second target value and the second characteristic matrix; carrying out Gaussian linear regression parameter calculation on the first input-output pair through a preset first Gaussian linear regression algorithm to obtain a first Gaussian linear regression parameter, and carrying out Gaussian linear regression parameter calculation on the second input-output pair through a preset second Gaussian linear regression algorithm to obtain a second Gaussian linear regression parameter; constructing a first temperature-pressure relation model of the oxygen pressure data and the temperature monitoring data according to the first Gaussian linear regression parameters, and constructing a second temperature-pressure relation model of the carbon dioxide pressure data and the temperature monitoring data according to the second Gaussian linear regression parameters;

The analysis module is used for inputting the second oxygen pressure characteristic set and the second carbon dioxide pressure characteristic set into a preset gas output pressure control analysis model to carry out gas output pressure control analysis and generate a target pressure control parameter strategy; the method specifically comprises the following steps: vector encoding is carried out on the second oxygen pressure characteristic set to obtain a first pressure characteristic encoding vector, and vector encoding is carried out on the second carbon dioxide pressure characteristic set to obtain a second pressure characteristic encoding vector; vector fusion is carried out on the first pressure characteristic coding vector and the second pressure characteristic coding vector to obtain a target fusion characteristic coding vector; inputting the target fusion characteristic coding vector into a preset gas output pressure control analysis model, wherein the gas output pressure control analysis model comprises a coding network, a decoding network and a strategy optimization network, the coding network comprises a bidirectional threshold circulation unit, the decoding network comprises a unidirectional threshold circulation unit and a full connection layer, and the strategy optimization network comprises a genetic algorithm; extracting the characteristics of the target fusion characteristic coding vector through a bidirectional threshold circulation unit in the coding network to obtain a target dimensional pressure characteristic vector; performing feature mapping on the target high-dimensional pressure feature vector through a unidirectional threshold circulation unit in the decoding network to obtain a target pressure feature mapping vector, and performing gas output pressure control parameter prediction on the target pressure feature mapping vector through a full connection layer in the decoding network to obtain a plurality of initial pressure control parameters; carrying out parameter strategy initialization on the plurality of initial pressure control parameters through a genetic algorithm in the strategy optimization network to generate a plurality of initial pressure control parameter strategies; respectively calculating the adaptation data of each initial pressure control parameter strategy, and carrying out strategy group division on the plurality of initial pressure control parameter strategies according to the adaptation data to obtain strategy group division results; and generating a plurality of corresponding candidate pressure control parameter strategies according to the strategy group division result, and carrying out strategy optimization solution on the plurality of candidate pressure control parameter strategies to generate a target pressure control parameter strategy.

7. A gas pressure detection apparatus, characterized by comprising: a memory and at least one processor, the memory having instructions stored therein;

the at least one processor invokes the instructions in the memory to cause the gas pressure detection apparatus to perform the gas pressure detection method of any one of claims 1-5.

8. A computer readable storage medium having instructions stored thereon, which when executed by a processor, implement the gas pressure detection method of any of claims 1-5.