Pressure state detection method and system and storage medium

Abstract

The invention provides a pressure state detection method and system and a storage medium, and relates to the technical field of pressure detection.The detection method comprises the steps that a detection environment is selected, preset detection time is determined, and pressure signals in the detection environment within the preset detection time are collected based on the multi-sensor fusion technology; receiving the pressure signal, preprocessing the pressure signal according to a signal processing algorithm, and generating corresponding pressure data; obtaining a target pressure model, and determining a pressure model index of the target pressure model according to the pressure data; and inputting the pressure data into the target pressure model to obtain the pressure state in the detection environment within the preset detection time, so that efficient and real-time detection of the pressure state can be realized, the detection precision is relatively high, the accuracy and reliability of pressure detection are ensured, and meanwhile, the model is relatively high in applicability and flexibility, can adapt to complex and changeable application scenes, and is suitable for popularization and application. The change of the detection environment is reflected in time, and the algorithm complexity and the calculation cost are low.

Description

Technical Field

[0001] The present invention relates to the field of pressure detection technology, and in particular to a pressure state detection method, system and storage medium. Background Art

[0002] In key areas such as industrial automation, environmental monitoring, aerospace, etc., accurate detection of pressure status plays a vital role in ensuring system safety and improving operational efficiency.

[0003] In the prior art, with the increasing requirements for pressure state detection accuracy and reliability in the fields of industrial automation, environmental monitoring, aerospace, etc., the limitations of traditional detection methods are becoming increasingly prominent. Traditional pressure detection methods mainly rely on a single sensor for data collection. However, the detection accuracy of a single sensor is greatly affected by environmental factors, such as temperature changes, humidity changes, and aging and wear of the sensor itself, which will lead to deviations in the detection results. Secondly, the data of a single sensor lacks redundancy. Once the sensor fails, the entire detection system may fail, seriously affecting the stability and reliability of the system. In addition, when processing signals in complex environments, traditional methods are often difficult to effectively filter noise, resulting in limited accuracy and real-time performance of signal processing. At the same time, the existing pressure detection method cannot establish a pressure model that meets specific application conditions, and thus adapt to complex and changeable detection environments. Therefore, it cannot achieve efficient and accurate detection of pressure states, has poor applicability and flexibility, has large limitations in applicable scenarios, cannot reflect changes in the detection environment in a timely manner, and has high algorithm complexity and computational cost.

[0004] Therefore, it is necessary to provide a pressure state detection method, system and storage medium to solve the above technical problems. Summary of the invention

[0005] In order to solve the above technical problems, the present invention provides a pressure state detection method, system and storage medium for solving the problems that the existing pressure detection model has poor detection accuracy, cannot achieve efficient and real-time detection of pressure state, has poor applicability and flexibility, has large limitations on applicable scenarios, cannot timely reflect changes in the detection environment, and has high algorithm complexity and computational cost.

[0006] The present invention provides a pressure state detection method, comprising: Selecting a detection environment and determining a preset detection time, and collecting a pressure signal in the detection environment within the preset detection time based on a multi-sensor fusion technology; receiving the pressure signal, preprocessing the pressure signal according to a signal processing algorithm, and generating corresponding pressure data; Acquire a target pressure model, and determine a pressure model index of the target pressure model according to the pressure data; The pressure data is input into the target pressure model to obtain the pressure state in the detection environment within the preset detection time.

[0007] Preferably, the collecting of the pressure signal in the detection environment within the preset detection time based on the multi-sensor fusion technology specifically includes: Based on the preset detection time, obtaining initial pressure signals collected by pressure sensors at different positions in the detection environment; Obtaining weights corresponding to the plurality of pressure sensors and generating a corresponding weight matrix, and performing weighted processing on the plurality of initial pressure signals according to the weight matrix to obtain a weighted pressure signal; Performing fusion processing on the weighted pressure signal to generate a fused pressure signal; The fused pressure signal is time-series calibrated to obtain the pressure signal.

[0008] Preferably, there are differences in the initial pressure signals collected by pressure sensors at different positions in the detection environment, and it is necessary to adjust the pressure signal collection frequency of the pressure sensor according to the position of the pressure sensor in the detection environment and in combination with the type of the pressure sensor.

[0009] Preferably, the preprocessing of the pressure signal according to the signal processing algorithm to generate corresponding pressure data specifically includes: Analyzing the frequency domain characteristics corresponding to the pressure signal to obtain a pressure frequency domain signal; determining a high-pressure fluctuation portion of the pressure frequency domain signal as signal noise; Based on a sliding window averaging algorithm, the signal noise is removed by a noise filter to generate a denoised pressure signal; The denoised pressure signal is subjected to noise verification based on a noise verifier. If the signal noise does not exist in the denoised pressure signal, the pressure data is generated. Otherwise, the denoising operation is continued.

[0010] Preferably, determining the pressure model index of the target pressure model according to the pressure data specifically includes: Acquiring the frequency-time characteristics of the pressure data; determining a target recognition mode of the target pressure model according to the frequency-time characteristic; determining target fitting parameters of the target pressure model based on a target recognition mode of the target pressure model mode; The pressure model indicators include target recognition mode and target fitting parameters.

[0011] Preferably, the step of obtaining the frequency-time characteristics of the pressure data specifically includes: Obtaining the frequency component ratio and frequency component intensity corresponding to the pressure data; Constructing a frequency component distribution corresponding to the pressure data according to the frequency component proportion and the frequency component intensity; The frequency-time characteristic of the pressure data is generated based on the frequency component distribution and a preset prediction time.

[0012] Preferably, determining the target recognition mode of the target pressure model according to the frequency-time characteristic specifically includes: Acquire multiple preset recognition modes corresponding to the target pressure model, and determine the standard characteristics corresponding to each of the preset recognition modes; Comparing the frequency-time characteristic with a plurality of standard characteristics respectively to obtain a plurality of characteristic similarities; A maximum characteristic similarity among the multiple characteristic similarities is determined, and a preset recognition mode corresponding to the maximum characteristic similarity is determined as the target recognition mode.

[0013] Preferably, the target identification mode based on the target pressure model mode determines the target fitting parameters of the target pressure model, and the determination is performed by the following fitting effect judgment algorithm: Acquire historical pressure data and pressure state corresponding to the detection environment, determine multiple sets of preset fitting parameters corresponding to the target pressure model, input the historical pressure data and pressure state into the target pressure model, and determine the target fitting parameters with the best fitting effect; Determine the historical pressure feature corresponding to the historical pressure data, input the historical pressure feature into the target pressure model, and obtain the historical pressure feature output. The expression of the historical pressure feature output is: ; In the formula, Indicates Output of historical pressure characteristics; Indicates An exponential function with base ; represents the total number of historical pressure features; Indicates Historical pressure characteristics to The weight of the historical pressure feature output; Indicates Historical stress characteristics; Indicates The bias of the historical pressure characteristic output; Σ represents the summation sign; After obtaining the historical pressure feature output, the historical pressure feature output is processed by a linear activation function to obtain a predicted pressure state, and the expression of the predicted pressure state is: ; In the formula, Indicates predicted stress state; Represents the total number of historical pressure feature outputs; Indicates The weight of the historical pressure feature output to the predicted pressure state; Indicates Output of historical pressure characteristics; represents the bias of the predicted pressure state; Σ represents the summation sign; After obtaining the predicted pressure state, the actual pressure state corresponding to the historical pressure data is obtained to determine the degree of difference between the predicted pressure state and the actual pressure state: ; In the formula, Indicates the degree of difference between the predicted stress state and the actual stress state; Indicates predicted stress state; Indicates the actual pressure status; According to the difference between the predicted pressure state and the actual pressure state, a set of target fitting parameters with the best fitting effect is determined.

[0014] The present invention provides a pressure state detection system, comprising: An acquisition module, used to select a detection environment and determine a preset detection time, and to acquire a pressure signal in the detection environment within the preset detection time based on a multi-sensor fusion technology; A preprocessing module, used for receiving the pressure signal, preprocessing the pressure signal according to a signal processing algorithm, and generating corresponding pressure data; A determination module, used to obtain a target pressure model and determine a pressure model index of the target pressure model according to the pressure data; The output module is used to input the pressure data into the target pressure model to obtain the pressure state in the detection environment within the preset detection time.

[0015] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, any one of the method steps in the pressure state detection method is implemented.

[0016] Compared with the related art, the pressure state detection method, system and storage medium provided by the present invention have the following beneficial effects: The present invention selects a detection environment and determines a preset detection time, and collects pressure signals in the detection environment within the preset detection time based on multi-sensor fusion technology, thereby improving the pressure detection accuracy and enhancing the robustness and fault tolerance of the pressure detection system by integrating data from multiple sensors. By receiving the pressure signal, the pressure signal is pre-processed according to the signal processing algorithm to generate the corresponding pressure data, so that the noise in the pressure signal can be effectively filtered with the help of advanced signal processing technology, the signal-to-noise ratio of the pressure signal can be improved, and a more accurate pressure state judgment can be achieved.

[0017] The present invention obtains a target pressure model and determines the pressure model index of the target pressure model according to pressure data, thereby effectively fusing multi-sensor data, establishing a pressure model that meets specific application conditions, and realizing efficient and real-time detection of pressure states, thereby ensuring the accuracy and reliability of pressure detection. By inputting pressure data into the target pressure model, the pressure state in the detection environment within a preset detection time is obtained, thereby realizing efficient and accurate detection of pressure states, and the applicability and flexibility are strong, the limitations of applicable scenarios are small, and the changes in the detection environment can be reflected in a timely manner, and the algorithm complexity and calculation cost are low. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 is a flow chart of a pressure state detection method of the present invention; Figure 2 is a schematic diagram of a pressure sensor of the present invention; Figure 3 The system block diagram of a pressure state detection system of the present invention. DETAILED DESCRIPTION

[0019] The present invention will be further described below in conjunction with the accompanying drawings and implementation modes.

[0020] Embodiment 1 like Figure 1 As shown, a pressure state detection method includes: S1, selecting a detection environment and determining a preset detection time, and collecting a pressure signal in the detection environment within the preset detection time based on a multi-sensor fusion technology; Among them, first of all, a representative detection environment can be selected. For this detection environment, it is necessary to ensure that it can truly reflect the pressure conditions of the area to be tested, and set a preset detection time according to actual needs. Generally speaking, a suitable time period can capture the complete change process of the pressure signal, while avoiding the impact of too long a time period on the detection results, saving detection resources, and improving detection efficiency.

[0021] Based on advanced multi-sensor fusion technology, multiple sensors of different types and locations can be deployed to collect pressure signals in the detection environment within the preset detection time from all directions and angles.

[0022] Through the above method, the accuracy and reliability of data acquisition can be significantly improved, which facilitates the subsequent processing and analysis of pressure signals.

[0023] S2, receiving the pressure signal, preprocessing the pressure signal according to a signal processing algorithm, and generating corresponding pressure data; It is understandable that after receiving the pressure signals from multiple sensors, since these pressure signals may contain various noises and interference factors, in order to extract effective pressure data, advanced signal processing algorithms can be used to pre-process the pressure signals.

[0024] Through the above method, the noise components in the pressure signal, such as mechanical vibration, electromagnetic interference, etc., can be effectively filtered out, while retaining the key information reflecting the pressure state, generating clearer and more accurate pressure data, and facilitating the subsequent analysis and modeling of the pressure data.

[0025] S3, obtaining a target pressure model, and determining a pressure model index of the target pressure model according to the pressure data; Among them, a specific target pressure model can be obtained, and the model can reflect the pressure change law under different environmental conditions. According to the preprocessed pressure data, the pressure model index of the target pressure model can be calculated and determined. In the present invention, the pressure model index specifically includes the target recognition mode and the target fitting parameter of the target pressure model.

[0026] Through the above method, multi-sensor data can be effectively integrated to establish a pressure model that meets specific application conditions, realize efficient and real-time detection of pressure status, and ensure the accuracy and reliability of pressure detection. At the same time, the model has strong applicability and flexibility, less limitations on applicable scenarios, can promptly reflect changes in the detection environment, and has low algorithm complexity and computational cost.

[0027] S4, inputting the pressure data into the target pressure model to obtain the pressure state in the detection environment within the preset detection time.

[0028] It should be noted that the pressure data can be input into a pre-built and verified target pressure model. Through calculation and analysis of the target pressure model, the pressure state information corresponding to the pressure data can be obtained, and then the pressure state in the detection environment within the preset detection time can be determined.

[0029] Through the above methods, the real-time changes of pressure signals can be revealed, and the future pressure trends can be predicted through the model, thereby providing important reference information on the pressure status.

[0030] In a specific implementation process, the pressure signal in the detection environment within the preset detection time is collected based on the multi-sensor fusion technology, specifically including: Based on the preset detection time, obtaining initial pressure signals collected by pressure sensors at different positions in the detection environment; Obtaining weights corresponding to the plurality of pressure sensors and generating a corresponding weight matrix, and performing weighted processing on the plurality of initial pressure signals according to the weight matrix to obtain a weighted pressure signal; Performing fusion processing on the weighted pressure signal to generate a fused pressure signal; The fused pressure signal is time-series calibrated to obtain the pressure signal.

[0031] Among them, based on the preset detection time, the pressure sensor data from different positions in the detection environment can be synchronously collected first. It should be noted that the selection of these positions is based on the spatial distribution characteristics of the detection environment and the possible pressure change trend, so as to ensure the comprehensiveness and representativeness of the collected pressure signals. And each pressure sensor can capture and output the initial pressure signal at its location according to a specific data collection frequency.

[0032] In this embodiment, Figure 2 As shown, there are three sensors in a certain detection environment, namely: sensor 1, sensor 2 and sensor 3. These sensors have different positions and sensor types in the detection environment, and each pressure sensor collects data at a specific data collection frequency. Therefore, based on the detection time set in advance, the terminal device can synchronously collect pressure sensor data from different positions in the detection environment.

[0033] In order to reduce data deviations caused by individual differences in sensors and environmental interference, the weight values ​​corresponding to each pressure sensor in the detection environment can be automatically calculated and obtained. It should be noted that these weight values ​​are obtained based on a comprehensive evaluation of the sensor's historical performance, accuracy, stability and current environmental conditions. The obtained weight values ​​are combined to obtain the corresponding weight matrix, through which multiple initial pressure signals can be weighted, thereby effectively balancing the contribution of each sensor data and generating the corresponding weighted pressure signal.

[0034] The Kalman filtering algorithm can be used to perform deep fusion processing on weighted pressure signals, thereby eliminating redundant and conflicting information in single sensor data. It can also greatly improve the reliability and accuracy of data through mutual verification and complementarity of multi-source data, and generate high-quality fused pressure signals.

[0035] In practical applications, in order to ensure the consistency and continuity of the fused pressure signal in the time series, a corresponding time series calibration can be performed. Specifically, by comparing the data changes at each time point, the time series analysis algorithm can adjust the possible time offset or jump in the fused pressure signal, and then output an accurate, continuous pressure signal that meets the preset detection time requirements.

[0036] In the above manner, through a series of operations including data collection, weighted processing, deep fusion and time series calibration, the pressure signal in the detection environment within the preset detection time can be collected based on the multi-sensor fusion technology, thereby making full use of the advantages of multi-source data and ensuring the comprehensiveness, accuracy and reliability of the output pressure signal through the multi-sensor fusion technology.

[0037] There are differences in the initial pressure signals collected by the pressure sensors at different positions in the detection environment. It is necessary to adjust the pressure signal collection frequency of the pressure sensor according to the position of the pressure sensor in the detection environment and the type of the pressure sensor.

[0038] It is understandable that due to various complex factors in the detection environment, such as temperature gradient, fluid flow velocity differences, and the impact of different materials on pressure conduction, the initial pressure signals collected by pressure sensors at different locations are significantly different. In order to ensure the consistency and accuracy of the data, the pressure signal collection frequency of each pressure sensor can be finely adjusted according to the specific location of each pressure sensor in the detection environment and in combination with its sensor type.

[0039] Specifically, the environmental characteristics of each pressure sensor location, such as temperature stability, fluid dynamics, and possible interference sources, can be identified and evaluated first. Then, based on the type of sensor, such as piezoresistive, piezoelectric, and capacitive, its response speed, sensitivity, and measurement range can be determined, and the most appropriate acquisition frequency can be selected.

[0040] For example, for sensors in high-speed fluid flow areas, a higher acquisition frequency may need to be set to capture transient pressure changes; while for sensors in static or low-speed fluid areas, the acquisition frequency can be appropriately lowered to reduce the data processing burden.

[0041] In addition, in order to ensure the stability and reliability of the collected pressure signal so that it can truly reflect the dynamic changes of the detection environment, the pressure sensor can be calibrated and verified regularly to eliminate measurement errors caused by sensor aging, drift or external interference.

[0042] Through the above method, according to the differences in the position and type of pressure sensors in the detection environment, the pressure signal acquisition frequency can be finely adjusted, thereby ensuring the accuracy and reliability of pressure state detection, effectively improving the overall performance of the detection system, and facilitating subsequent pressure signal processing and analysis.

[0043] Preprocessing the pressure signal according to the signal processing algorithm to generate corresponding pressure data specifically includes: Analyzing the frequency domain characteristics corresponding to the pressure signal to obtain a pressure frequency domain signal; determining a high-pressure fluctuation portion of the pressure frequency domain signal as signal noise; Based on a sliding window averaging algorithm, the signal noise is removed by a noise filter to generate a denoised pressure signal; The denoised pressure signal is subjected to noise verification based on a noise verifier. If the signal noise does not exist in the denoised pressure signal, the pressure data is generated. Otherwise, the denoising operation is continued.

[0044] It should be noted that the received pressure signal can first be converted into the frequency domain through frequency domain analysis tools such as fast Fourier transform (FFT), and then the corresponding frequency domain characteristics can be analyzed and generated to obtain the pressure frequency domain signal, so that different frequency components in the pressure signal can be identified, which is convenient for subsequent noise separation. In the pressure frequency domain signal, the high-pressure fluctuation part is often regarded as signal noise, which may come from environmental interference, sensor error or distortion in the signal transmission process, and may interfere with the accurate judgment of the pressure state.

[0045] In order to effectively remove these signal noises, the signal noise can be processed accordingly through the noise filter based on the sliding window averaging algorithm. Specifically, the sliding window averaging algorithm can effectively smooth the random fluctuations in the signal by calculating the average value of the signal within a certain window, thereby accurately suppressing high-frequency noise. The noise filter can further enhance the denoising ability, ensure that the noise is removed more thoroughly, while retaining the useful information in the signal, generating a denoised pressure signal, and significantly improving the signal-to-noise ratio of the denoised pressure signal.

[0046] In order to ensure the accuracy of the pressure data finally generated, a noise detector can be equipped to re-test the denoised pressure signal to confirm whether there is still unidentified noise. If the test result shows that there is no significant noise in the signal, it indicates that the denoising operation is completed, and the corresponding pressure data can be generated at this time; if the test result shows that there is still noise, a new round of denoising operation can be automatically triggered until the denoising effect meets the preset standard.

[0047] In the above manner, through frequency domain analysis, sliding window averaging algorithm, noise filter and noise detector and other technologies, efficient preprocessing of the original pressure signal can be achieved to generate high-quality pressure data, thereby improving the accuracy and reliability of pressure detection and facilitating subsequent pressure state analysis.

[0048] Determining the pressure model index of the target pressure model according to the pressure data specifically includes: Acquiring the frequency-time characteristics of the pressure data; determining a target recognition mode of the target pressure model according to the frequency-time characteristic; determining target fitting parameters of the target pressure model based on a target recognition mode of the target pressure model mode; The pressure model indicators include target recognition mode and target fitting parameters.

[0049] It is understandable that the frequency-time characteristics corresponding to the pressure data can be accurately obtained first. Specifically, the performance of the pressure data in the frequency domain can be deeply analyzed through algorithms such as the Fast Fourier Transform (FFT), and then the key frequency components and their time distribution characteristics can be extracted. It should be noted that these characteristics can not only reflect the basic structure of the pressure data, but also imply the dynamic change information of the detection environment.

[0050] Based on the frequency-time characteristics of the acquired pressure data, the target recognition mode of the target pressure model can be determined. After the target recognition mode is determined, the target fitting parameters of the target pressure model can be further determined based on the mode. It should be noted that the target fitting parameters can be quantitative indicators of the similarity and difference between the pressure data and the target pressure model. The target fitting parameters can be accurately calculated by mathematical methods such as least squares method and maximum likelihood estimation, thereby achieving the best fit between the pressure data and the target pressure model. Among them, the pressure model indicators include target recognition mode and target fitting parameters, and the target recognition mode and target fitting parameters are interrelated and supportive of each other.

[0051] Through the above method, the target pressure model that meets the actual pressure state of the detection environment can be accurately determined according to the pressure data, which is convenient for subsequent pressure monitoring, early warning and control. The target pressure model has strong applicability and flexibility, less limitation in applicable scenarios, can timely reflect changes in the detection environment, and has low algorithm complexity and computational cost.

[0052] The obtaining of the frequency-time characteristics of the pressure data specifically includes: Obtaining the frequency component ratio and frequency component intensity corresponding to the pressure data; Constructing a frequency component distribution corresponding to the pressure data according to the frequency component proportion and the frequency component intensity; The frequency-time characteristic of the pressure data is generated based on the frequency component distribution and a preset prediction time.

[0053] It is understandable that the pressure data can first be deeply analyzed through signal processing technology, such as fast Fourier transform (FFT), etc. Among them, FFT, as an efficient frequency domain analysis tool, can accurately convert the pressure signal in the time domain into a frequency domain representation, and then clearly reveal the proportion and intensity of each frequency component in the pressure signal, thereby ensuring the accuracy of the pressure data and facilitating the subsequent analysis and processing of the pressure data.

[0054] Based on the obtained frequency component proportion and intensity information, the frequency component distribution corresponding to the pressure data can be constructed. This frequency component distribution can intuitively display the distribution of each frequency component in the pressure data, including the main frequency component, the secondary frequency component and the relative intensity relationship between them, so as to comprehensively and deeply analyze the spectral characteristics corresponding to the pressure data.

[0055] Furthermore, based on the constructed frequency component distribution, the frequency-time characteristics corresponding to the pressure data can be generated in combination with the preset prediction time. Under the premise of fully considering the influence of time factors on frequency characteristics, combined with the preset prediction time, the change trend of pressure data in the future can be more accurately determined. At the same time, potential problems can be discovered in advance based on the change trend, and corresponding measures can be taken to intervene.

[0056] Through the above method, detailed information of pressure data in the frequency domain can be obtained, and a comprehensive and in-depth analysis can be performed on it in combination with time factors.

[0057] The step of determining the target recognition mode of the target pressure model according to the frequency-time characteristic specifically includes: Acquire multiple preset recognition modes corresponding to the target pressure model, and determine the standard characteristics corresponding to each of the preset recognition modes; Comparing the frequency-time characteristic with a plurality of standard characteristics respectively to obtain a plurality of characteristic similarities; A maximum characteristic similarity among the multiple characteristic similarities is determined, and a preset recognition mode corresponding to the maximum characteristic similarity is determined as the target recognition mode.

[0058] Among them, firstly, multiple preset recognition modes corresponding to the target pressure model can be extracted, and each preset recognition mode has specific standard characteristics, which can be set based on historical data and experimental verification to characterize typical behaviors under different pressure states.

[0059] Furthermore, the frequency-time characteristics acquired in real time can be compared one by one with the standard characteristics of the above-mentioned multiple preset recognition modes. It should be noted that this comparison process needs to take into account various potential interference factors, such as environmental noise, sensor errors, etc. And through algorithms such as correlation coefficient method and Euclidean distance, the characteristic similarity between the frequency-time characteristics and each standard characteristic can be calculated to quantify the similarity between the two. After obtaining multiple characteristic similarities, the maximum characteristic similarity can be identified and determined. Once the maximum characteristic similarity is determined, the corresponding preset recognition mode is the target recognition mode.

[0060] Through the above method, it is possible to ensure accurate matching of the target recognition mode, achieve accurate recognition and rapid response of the pressure state of the detection environment, and improve the accuracy and reliability of pressure state detection.

[0061] The target recognition mode based on the target pressure model mode determines the target fitting parameters of the target pressure model, which is determined by the following fitting effect judgment algorithm: Acquire historical pressure data and pressure state corresponding to the detection environment, determine multiple sets of preset fitting parameters corresponding to the target pressure model, input the historical pressure data and pressure state into the target pressure model, and determine the target fitting parameters with the best fitting effect; Determine the historical pressure feature corresponding to the historical pressure data, input the historical pressure feature into the target pressure model, and obtain the historical pressure feature output. The expression of the historical pressure feature output is: ; In the formula, Indicates Output of historical pressure characteristics; Indicates An exponential function with base ; represents the total number of historical pressure features; Indicates Historical pressure characteristics to The weight of the historical pressure feature output; Indicates Historical stress characteristics; Indicates The bias of the historical pressure characteristic output; Σ represents the summation sign; After obtaining the historical pressure feature output, the historical pressure feature output is processed by a linear activation function to obtain a predicted pressure state, and the expression of the predicted pressure state is: ; In the formula, Indicates predicted stress state; Represents the total number of historical pressure feature outputs; Indicates The weight of the historical pressure feature output to the predicted pressure state; Indicates Output of historical pressure characteristics; represents the bias of the predicted pressure state; Σ represents the summation sign; After obtaining the predicted pressure state, the actual pressure state corresponding to the historical pressure data is obtained to determine the degree of difference between the predicted pressure state and the actual pressure state: ; In the formula, Indicates the degree of difference between the predicted stress state and the actual stress state; Indicates predicted stress state; Indicates the actual pressure status; According to the difference between the predicted pressure state and the actual pressure state, a set of target fitting parameters with the best fitting effect is determined.

[0062] For example, five historical pressure characteristics are set to k1=100, k2=150, k3=120, k4=130, and k5=110, and a set of preset fitting parameters is set, namely, w 11 =0.2, w 21 =0.3, w 31 =0.2, w 41 =0.15, w 51 =0.15, b i =0,v j =1, c=0, at this time, the corresponding historical pressure characteristic output can be obtained , and the predicted stress state s=e 107 ,Since the actual pressure state corresponding to the historical pressure data can be obtained through measurement, ,thus, by quantifying the difference between the predicted ,pressure state and the actual pressure state, a set of target fitting ,parameters with the best fitting effect can be determined.

[0063] Embodiment 2 like Figure 3 As shown, a pressure state detection system comprises: An acquisition module, used to select a detection environment and determine a preset detection time, and to acquire a pressure signal in the detection environment within the preset detection time based on a multi-sensor fusion technology; A preprocessing module, used for receiving the pressure signal, preprocessing the pressure signal according to a signal processing algorithm, and generating corresponding pressure data; A determination module, used to obtain a target pressure model and determine a pressure model index of the target pressure model according to the pressure data; The output module is used to input the pressure data into the target pressure model to obtain the pressure state in the detection environment within the preset detection time.

[0064] Figure 3 The apparatus of the embodiment shown can be used to perform Figure 1 The implementation principles and technical effects of the steps in the method embodiment shown are similar and will not be repeated here.

[0065] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, any one of the method steps in the pressure state detection method is implemented.

[0066] Through the introduction of the above embodiments, the present invention can obtain the target pressure model through the pressure state detection method, system and storage medium, determine the pressure model index of the target pressure model according to the pressure data, so that the multi-sensor data can be effectively integrated, and the pressure model that meets the specific application conditions can be established to achieve efficient and real-time detection of the pressure state, and the accuracy and reliability of the pressure detection can be guaranteed. By inputting the pressure data into the target pressure model, the pressure state in the detection environment within the preset detection time is obtained, so that the pressure state can be efficiently and accurately detected, and the applicability and flexibility are strong, the limitation of the applicable scene is small, the change of the detection environment can be reflected in time, and the algorithm complexity and calculation cost are low. The present invention obtains the target pressure model, determines the pressure model index of the target pressure model according to the pressure data, so that the multi-sensor data can be effectively integrated, and the pressure model that meets the specific application conditions can be established, so that the pressure state can be efficiently and accurately detected, and the accuracy and reliability of the pressure detection can be guaranteed. By inputting the pressure data into the target pressure model, the pressure state in the detection environment within the preset detection time is obtained, so that the pressure state can be efficiently and accurately detected, and the applicability and flexibility are strong, the limitation of the applicable scene is small, the change of the detection environment can be reflected in time, and the algorithm complexity and calculation cost are low.

[0067] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0068] Those skilled in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable rewritable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0069] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

Claims

1. A pressure state detection method, characterized in that: include: Selecting a detection environment and determining a preset detection time, and collecting a pressure signal in the detection environment within the preset detection time based on a multi-sensor fusion technology; receiving the pressure signal, preprocessing the pressure signal according to a signal processing algorithm, and generating corresponding pressure data; Acquire a target pressure model, and determine a pressure model index of the target pressure model according to the pressure data; The pressure data is input into the target pressure model to obtain the pressure state in the detection environment within the preset detection time.

2. A pressure state detection method according to claim 1, characterized in that: The collecting of the pressure signal in the detection environment within the preset detection time based on the multi-sensor fusion technology specifically includes: Based on the preset detection time, obtaining initial pressure signals collected by pressure sensors at different positions in the detection environment; Obtaining weights corresponding to the plurality of pressure sensors and generating a corresponding weight matrix, and performing weighted processing on the plurality of initial pressure signals according to the weight matrix to obtain a weighted pressure signal; Performing fusion processing on the weighted pressure signal to generate a fused pressure signal; The fused pressure signal is time-series calibrated to obtain the pressure signal.

3. A pressure state detection method according to claim 2, characterized in that: There are differences in the initial pressure signals collected by the pressure sensors at different positions in the detection environment. It is necessary to adjust the pressure signal collection frequency of the pressure sensor according to the position of the pressure sensor in the detection environment and the type of the pressure sensor.

4. A pressure state detection method according to claim 1, characterized in that: Preprocessing the pressure signal according to the signal processing algorithm to generate corresponding pressure data specifically includes: Analyzing the frequency domain characteristics corresponding to the pressure signal to obtain a pressure frequency domain signal; determining a high-pressure fluctuation portion of the pressure frequency domain signal as signal noise; Based on a sliding window averaging algorithm, the signal noise is removed by a noise filter to generate a denoised pressure signal; The denoised pressure signal is subjected to noise verification based on a noise verifier. If the signal noise does not exist in the denoised pressure signal, the pressure data is generated. Otherwise, the denoising operation is continued.

5. A pressure state detection method according to claim 1, characterized in that: Determining the pressure model index of the target pressure model according to the pressure data specifically includes: Acquiring the frequency-time characteristics of the pressure data; determining a target recognition mode of the target pressure model according to the frequency-time characteristic; determining target fitting parameters of the target pressure model based on a target recognition mode of the target pressure model mode; The pressure model indicators include target recognition mode and target fitting parameters.

6. A pressure state detection method according to claim 5, characterized in that: The obtaining of the frequency-time characteristics of the pressure data specifically includes: Obtaining the frequency component ratio and frequency component intensity corresponding to the pressure data; Constructing a frequency component distribution corresponding to the pressure data according to the frequency component proportion and the frequency component intensity; The frequency-time characteristic of the pressure data is generated based on the frequency component distribution and a preset prediction time.

7. A pressure state detection method according to claim 5, characterized in that: The step of determining the target recognition mode of the target pressure model according to the frequency-time characteristic specifically includes: Acquire multiple preset recognition modes corresponding to the target pressure model, and determine the standard characteristics corresponding to each of the preset recognition modes; Comparing the frequency-time characteristic with a plurality of standard characteristics respectively to obtain a plurality of characteristic similarities; A maximum characteristic similarity among the multiple characteristic similarities is determined, and a preset recognition mode corresponding to the maximum characteristic similarity is determined as the target recognition mode.

8. A pressure state detection method according to claim 5, characterized in that: The target recognition mode based on the target pressure model mode determines the target fitting parameters of the target pressure model, which is determined by the following fitting effect judgment algorithm: Acquire historical pressure data and pressure state corresponding to the detection environment, determine multiple sets of preset fitting parameters corresponding to the target pressure model, input the historical pressure data and pressure state into the target pressure model, and determine the target fitting parameters with the best fitting effect; Determine the historical pressure feature corresponding to the historical pressure data, input the historical pressure feature into the target pressure model, and obtain the historical pressure feature output. The expression of the historical pressure feature output is: ； In the formula, Indicates Output of historical pressure characteristics; Indicates An exponential function with base ; represents the total number of historical pressure features; Indicates Historical pressure characteristics to The weight of the historical pressure feature output; Indicates Historical stress characteristics; Indicates The bias of the historical pressure characteristic output; Σ represents the summation sign; After obtaining the historical pressure feature output, the historical pressure feature output is processed by a linear activation function to obtain a predicted pressure state, and the expression of the predicted pressure state is: ； In the formula, Indicates predicted stress state; Represents the total number of historical pressure feature outputs; Indicates The weight of the historical pressure feature output to the predicted pressure state; Indicates Output of historical pressure characteristics; represents the bias of the predicted pressure state; Σ represents the summation sign; After obtaining the predicted pressure state, the actual pressure state corresponding to the historical pressure data is obtained to determine the degree of difference between the predicted pressure state and the actual pressure state: ； In the formula, Indicates the degree of difference between the predicted stress state and the actual stress state; Indicates predicted stress state; Indicates the actual pressure status; According to the difference between the predicted pressure state and the actual pressure state, a set of target fitting parameters with the best fitting effect is determined.

9. A pressure state detection system, applied to a pressure state detection method according to any one of claims 1 to 8, the detection system comprising: An acquisition module, used to select a detection environment and determine a preset detection time, and to acquire a pressure signal in the detection environment within the preset detection time based on a multi-sensor fusion technology; A preprocessing module, used for receiving the pressure signal, preprocessing the pressure signal according to a signal processing algorithm, and generating corresponding pressure data; A determination module, used to obtain a target pressure model and determine a pressure model index of the target pressure model according to the pressure data; The output module is used to input the pressure data into the target pressure model to obtain the pressure state in the detection environment within the preset detection time.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method steps described in any one of claims 1 to 8 are implemented.

Images