Pressure control method and device for accurately adjusting air pressure intensity

Abstract

The present invention relates to the technical field of air pressure intensity regulation, and discloses a pressure control method and device for accurately regulating air pressure intensity, wherein the method comprises: performing digital conversion and temperature and humidity compensation on the original pressure signal collected by the pressure sensor array to obtain a pressure data matrix and a pressure gradient matrix; calculating the pressure distribution centroid and the pressure fluctuation spectrum to obtain a pressure spatial distribution feature matrix and a pressure time domain feature matrix; inputting a three-state Markov model for dynamic state recognition to obtain a pressure state transition probability matrix and a pressure response prediction matrix; constructing an objective function including a pressure tracking error term and an organizational safety constraint term, and solving to obtain a PID control parameter matrix and a pressure safety limit matrix; constructing a three-level control system including a basic controller, an interference compensator and a safety protector, and outputting an air pressure regulation drive signal. The method realizes high-precision pressure sensing control.

Description

Technical Field

[0001] The present invention relates to the technical field of air pressure intensity regulation, and in particular to a pressure control method and device for accurately regulating air pressure intensity. Background Art

[0002] The pressure control system and method of the traditional pressure sensor array are difficult to adapt to the complex clinical application environment. At present, the pressure control system faces multiple technical challenges: the elastic properties of the surface of the traditional pressure sensor array vary significantly and change dynamically, which causes the pressure distribution to show nonlinear and time-varying characteristics; uncertain factors such as external interference, environmental factors and sensor failure will affect the control accuracy and reliability of the system; the pressure control method lacks an effective safety protection mechanism, and it is difficult to ensure the safety of pressure control while ensuring control performance. In addition, traditional pressure control systems and methods generally have problems such as slow response speed, poor anti-interference ability, and high energy consumption. Summary of the invention

[0003] The present invention provides a pressure control method and device for accurately adjusting air pressure intensity, which are used to realize high-precision pressure sensing control.

[0004] In a first aspect, the present invention provides a pressure control method for accurately adjusting air pressure intensity, the pressure control method for accurately adjusting air pressure intensity comprising:

[0005] The original pressure signal collected by the pressure sensor array is digitized and temperature and humidity compensated to obtain a pressure data matrix and a pressure gradient matrix;

[0006] Calculate the pressure distribution centroid and the pressure fluctuation spectrum based on the pressure data matrix and the pressure gradient matrix to obtain a pressure spatial distribution feature matrix and a pressure time domain feature matrix;

[0007] Inputting the pressure spatial distribution characteristic matrix and the pressure time domain characteristic matrix into a three-state Markov model for dynamic state recognition to obtain a pressure state transition probability matrix and a pressure response prediction matrix;

[0008] According to the pressure state transition probability matrix and the pressure response prediction matrix, an objective function including a pressure tracking error term and an organizational safety constraint term is constructed, and a PID control parameter matrix and a pressure safety limit matrix are obtained by solving the matrix;

[0009] A three-level control system including a basic controller, a disturbance compensator and a safety protector is constructed based on the PID control parameter matrix and the pressure safety limit matrix to output an air pressure regulating drive signal.

[0010] In a second aspect, the present invention provides a pressure control device for accurately adjusting air pressure intensity, the pressure control device for accurately adjusting air pressure intensity comprising:

[0011] A conversion module is used to perform digital conversion and temperature and humidity compensation on the original pressure signal collected by the pressure sensor array to obtain a pressure data matrix and a pressure gradient matrix;

[0012] A calculation module, used to calculate the pressure distribution centroid and the pressure fluctuation spectrum based on the pressure data matrix and the pressure gradient matrix, and obtain a pressure spatial distribution feature matrix and a pressure time domain feature matrix;

[0013] An identification module, used for inputting the pressure spatial distribution feature matrix and the pressure time domain feature matrix into a three-state Markov model for dynamic state identification, and obtaining a pressure state transition probability matrix and a pressure response prediction matrix;

[0014] A solution module, used for constructing an objective function including a pressure tracking error term and an organizational safety constraint term according to the pressure state transition probability matrix and the pressure response prediction matrix, and solving to obtain a PID control parameter matrix and a pressure safety limit matrix;

[0015] The output module is used to construct a three-level control system including a basic controller, a disturbance compensator and a safety protector based on the PID control parameter matrix and the pressure safety limit matrix, and output an air pressure regulation drive signal.

[0016] The third aspect of the present invention provides a pressure control device for accurately adjusting the intensity of air pressure, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor calls the instructions in the memory so that the pressure control device for accurately adjusting the intensity of air pressure executes the above-mentioned pressure control method for accurately adjusting the intensity of air pressure.

[0017] A fourth aspect of the present invention provides a computer-readable storage medium, in which instructions are stored, which, when executed on a computer, enable the computer to execute the above-mentioned pressure control method for accurately adjusting the air pressure intensity.

[0018] In the technical solution provided by the present invention, by constructing a three-state Markov model for dynamic state recognition, combined with the pressure spatial distribution feature matrix and the pressure time domain feature matrix, the pressure change characteristics can be accurately captured and the pressure control error can be reduced; a multi-level control architecture including a basic controller, an interference compensator and a safety protector is adopted to achieve effective suppression of system disturbances; the introduction of a temperature and humidity compensation mechanism and a 16-bit high-precision digital conversion significantly improves the acquisition accuracy and environmental adaptability of pressure data; an objective function based on a pressure tracking error term and an organizational safety constraint term is designed to achieve dual guarantees of control performance and safety; adaptive adjustment of system parameters is achieved through the coordinated optimization of the pressure state transition probability matrix and the pressure response prediction matrix; dual-channel data acquisition and multi-scale feature extraction technology are adopted to improve the robustness of the system; an adaptive adjustment mechanism based on the PID parameter matrix realizes real-time optimization of control parameters and enhances system stability; and through energy consumption analysis and parameter optimization, the overall energy consumption of the system is optimized and energy utilization efficiency is improved. The present invention not only solves the problems of traditional pressure control systems in terms of accuracy, safety and reliability, but also realizes high-precision pressure sensing control.

[0019] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention are realized and obtained by the structures particularly pointed out in the description, claims and drawings.

[0020] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 A schematic diagram of an embodiment of a pressure control method for accurately adjusting air pressure intensity in an embodiment of the present invention;

[0022] Figure 2 A schematic diagram of an embodiment of a pressure control device for accurately adjusting air pressure intensity in an embodiment of the present invention;

[0023] Figure 3 This is a schematic diagram of an embodiment of a pressure control device for accurately adjusting air pressure intensity in an embodiment of the present invention. DETAILED DESCRIPTION

[0024] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0025] The terms "including" and "having" and any variations thereof mentioned in the embodiments of the present invention are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device end including a series of steps or units is not limited to the listed steps or units, but may optionally include other steps or units that are not listed, or may optionally include other steps or units that are inherent to these processes, methods, products or device ends.

[0026] To facilitate understanding of this embodiment, a pressure control method for accurately adjusting air pressure intensity disclosed in an embodiment of the present invention is first described in detail. Figure 1 As shown, the method comprises the following steps:

[0027] 101. Perform digital conversion and temperature and humidity compensation on the original pressure signal collected by the pressure sensor array to obtain a pressure data matrix and a pressure gradient matrix;

[0028] It is understandable that the execution subject of the present invention may be a pressure control device for accurately adjusting the air pressure intensity, or may be a terminal or a server, which is not limited here. The embodiment of the present invention is described by taking a server as the execution subject as an example.

[0029] It should be noted that the high-precision pressure sensing array used in this application adopts an M×N matrix layout structure, where M and N are both integers greater than or equal to 2, and the array size can be adjusted according to the needs of the actual application scenario. Each pressure sensing unit adopts a capacitive pressure sensor, which measures the change in capacitance between electrodes based on the deformation characteristics of the dielectric material when under pressure to achieve pressure measurement. The spacing Δx and Δy between adjacent sensing units in the array are optimized to avoid crosstalk between adjacent units while ensuring the spatial sampling density. The physical layout of the sensing unit adopts a flexible substrate structure, which improves the fit of the array to the surface of the object being measured and reduces the measurement dead zone. The sampling frequency of each sensing unit is set to 1000Hz to ensure that the system can capture rapidly changing pressure signals and meet the needs of dynamic pressure monitoring. The analog signal of the sensing unit is digitized by a 16-bit analog-to-digital converter to convert the continuous pressure signal into a discrete digital quantity, providing a basis for subsequent digital signal processing. In the design of the array, the influence of environmental factors on the measurement accuracy is taken into account, and temperature sensors and humidity sensors are arranged around the array to collect environmental parameters. These sensors are distributed in key positions of the array to ensure that the spatial distribution characteristics of environmental parameters can be accurately captured. The array adopts a multi-layer shielding structure to effectively reduce the impact of external electromagnetic interference on the measurement signal. Signal transmission adopts a differential method to improve the anti-interference ability of the signal. The array as a whole adopts a modular design, which is easy to maintain and replace. At the same time, an expansion interface is reserved to support the flexible expansion of the array scale. The layout of the pressure sensing unit also comprehensively considers the influence of thermal effects. By optimizing the layout and adding a heat dissipation structure, the influence of temperature drift on measurement accuracy is reduced. At the same time, a moisture-proof structure is used in the packaging design of the sensor unit to reduce the impact of humidity changes on the measurement. This array design that comprehensively considers many factors provides a reliable hardware foundation for achieving high-precision pressure distribution measurement and dynamic response analysis.

[0030] Specifically, the capacitance value of each pressure sensing unit in the pressure sensing array is collected and processed, and the original analog pressure signal is converted into a digital signal that can be further analyzed and processed. The capacitance value of each pressure sensing unit is proportional to the air pressure, and the pressure data of each sensing unit is obtained by accurately collecting these capacitance values.

[0031] The original pressure signal is input into a 16-bit analog-to-digital converter for digital conversion, and the result of the conversion is a digital pressure signal matrix. Based on the digital pressure signal matrix of two adjacent samples, the sampling time difference calculation is performed to obtain the pressure difference matrix, and the change of the pressure signal over time is obtained to analyze the fluctuation characteristics of the air pressure. The calculation of the sampling time difference can reveal the rate of change of pressure at different time points and help to analyze the subsequent pressure change trend.

[0032] The temperature and relative humidity parameters of the surrounding environment are digitally collected to obtain the environmental compensation parameter set. The environmental compensation parameter set is used to perform temperature and humidity compensation operations on the digital pressure signal matrix to eliminate the error caused by environmental changes on the pressure data and obtain a compensated pressure data matrix. Through further processing of the pressure data matrix and combining it with the pressure difference matrix, the pressure change trend is calculated to obtain a dynamic characteristic matrix. This matrix reflects the dynamic change characteristics of air pressure over time.

[0033] The pressure values ​​of adjacent spatial positions in the pressure data matrix are interpolated to obtain the spatial pressure gradient matrix. The pressure gradient matrix reveals the changes in air pressure in space and can provide the spatial characteristics of air pressure distribution. The calculation of this matrix depends on the pressure values ​​of surrounding sensors, and the pressure change trend of the local area is obtained through spatial difference. Combining the dynamic feature matrix and the spatial pressure gradient matrix, the final pressure gradient matrix is ​​obtained through weighted average calculation.

[0034] 102. Calculate the pressure distribution centroid and pressure fluctuation spectrum based on the pressure data matrix and the pressure gradient matrix to obtain the pressure spatial distribution characteristic matrix and the pressure time domain characteristic matrix;

[0035] Specifically, the pressure data matrix is ​​subjected to pressure data segmentation processing, and the data is cut into fixed-length pressure data sequences, so that the pressure change characteristics can be analyzed independently in different time windows, and the pressure data in each time window can reflect the change of air pressure in that time period.

[0036] Based on each fixed-length pressure data sequence, the center of gravity position coordinates and peak position coordinates of the pressure distribution in each time window are calculated. These position coordinates can help determine the spatial characteristics of the pressure distribution. The center of gravity position coordinates reflect the center of the pressure distribution, while the peak position reveals the maximum value position of the pressure distribution, thereby obtaining the pressure distribution positioning matrix. Based on the pressure distribution positioning matrix and the pressure gradient matrix, spatial gradient accumulation calculations are performed to obtain the pressure action range characteristic matrix. This matrix can reflect the expansion range of pressure in space and its action intensity, and provide data support for analyzing the spatial impact of pressure distribution. At the same time, a multi-scale spectral decomposition operation is performed on the fixed-length pressure data sequence to extract the characteristics of pressure fluctuations in different frequency bands and obtain a multi-band characteristic matrix of pressure fluctuations. Multi-scale spectral decomposition can reveal the frequency characteristics of pressure fluctuations, including low-frequency and high-frequency fluctuations, which helps to more comprehensively understand the dynamic changes of air pressure, especially the fluctuation patterns at different time scales.

[0037] Based on the multi-band characteristic matrix of pressure fluctuations, the phase information and amplitude information of pressure changes are calculated to obtain the pressure time domain dynamic characteristic matrix, which reflects the dynamic change process of air pressure in the time domain. The phase information reveals the relative time delay of pressure fluctuations in different frequency bands, while the amplitude information reflects the intensity of fluctuations in each frequency band. By extracting this information, the time characteristics of air pressure changes are obtained. At the same time, the pressure range characteristic matrix and the pressure gradient matrix are fused with spatial features to obtain the pressure spatial distribution characteristic matrix. In the pressure time domain dynamic characteristic matrix, the main frequency extraction and phase correlation analysis are performed on the frequency components to find the main frequency components of pressure fluctuations, and the relationship between them and the phase is analyzed to obtain the initial time domain characteristic matrix, which preliminarily reflects the time domain characteristics of air pressure. Based on the pressure spatial distribution characteristic matrix, the initial time domain characteristic matrix is ​​optimized with spatial constraints to obtain the final pressure time domain characteristic matrix to ensure that the time domain characteristics conform to the actual spatial distribution characteristics of pressure.

[0038] 103. Input the pressure spatial distribution feature matrix and the pressure time domain feature matrix into the three-state Markov model for dynamic state recognition to obtain the pressure state transition probability matrix and the pressure response prediction matrix;

[0039] Specifically, the pressure spatial distribution feature matrix is ​​subjected to binary threshold segmentation processing, and the continuous pressure data is converted into discrete state labels to obtain the pressure distribution state labeling matrix. This matrix can clearly identify whether each area of ​​the pressure field is in a normal state, an abnormal state, or other specific states. At the same time, the pressure time domain feature matrix is ​​subjected to frequency domain transformation processing, and the time characteristics of the pressure are converted into frequency characteristics to obtain the pressure spectrum feature matrix. This matrix reveals the fluctuation characteristics of air pressure under different frequency components, which helps to identify the pattern of pressure fluctuations and their changing laws.

[0040] Based on the pressure distribution state labeling matrix and the pressure spectrum feature matrix, the three-state division operation of the state space is performed to divide the state of the pressure field into three basic states: normal pressure state, pressure mutation state and sensor fault state. The three-state division is based on the spatial distribution of pressure and the frequency domain fluctuation characteristics to identify the working state of the current system. Through the division, the state vector at the current moment is obtained, which describes the pressure state of the current system. According to the state vector at the current moment, the state transfer matrix of the three-state Markov model is updated and calculated to obtain the dynamic state transfer probability matrix. The Markov model predicts the future state through the transfer probability of the current state. The updated dynamic state transfer probability matrix can accurately describe the transfer probability of the system state from one time point to another. Based on the transfer matrix, the maximum likelihood probability calculation is performed to calculate the state prediction vector at the next moment, and the state probability distribution is calculated based on the vector to obtain the complete pressure state transfer probability matrix. The state transfer probability matrix is ​​matched with the historical state sequence to obtain the state evolution feature vector, which describes the evolution law of the pressure state over time. Based on this eigenvector, the state response model is constructed to obtain the state response characteristic matrix, which reveals the time domain characteristics of the system response, including important information such as the amplitude and duration of the response.

[0041] By analyzing the obtained state response feature matrix and pressure spatial distribution feature matrix, the pressure response mode analysis is performed to obtain the pressure response mode matrix, which reflects the manifestation of different pressure response modes under different pressure distributions. On this basis, the response feature extraction is performed to obtain the pressure response feature vector. The pressure response feature vector is subjected to feature fusion operation with the state evolution feature vector to generate a more comprehensive pressure state prediction matrix. This matrix integrates the temporal and spatial characteristics of pressure fluctuations and the law of state changes, and can provide a more accurate basis for state prediction and control. According to the pressure state prediction matrix, the initial response prediction matrix is ​​generated by response feature mapping. The initial response prediction matrix and the pressure time domain feature matrix are subjected to spatiotemporal feature constraint optimization to effectively reduce the prediction error caused by environmental noise or external interference, and obtain the pressure response prediction matrix.

[0042] 104. Construct an objective function including a pressure tracking error term and an organizational safety constraint term according to the pressure state transition probability matrix and the pressure response prediction matrix, and solve to obtain a PID control parameter matrix and a pressure safety limit matrix;

[0043] Specifically, the state weight distribution calculation is performed on the pressure state transition probability matrix to obtain the state weight coefficient matrix, which reflects the importance and influence of each state in the entire system. The time series correlation analysis is performed on the pressure response prediction matrix to obtain the response characteristic weight matrix. The time series correlation analysis can reveal the relationship between pressure fluctuations and time, determine which time periods are more critical to system control, and thus assign different weights to the response characteristics of different time periods. By combining the state weight coefficient matrix and the response characteristic weight matrix, a multi-objective optimization function is constructed, which can simultaneously consider the transition probability of the pressure state and the prediction accuracy of the pressure response. Through this step, a preliminary initial objective function matrix is ​​obtained.

[0044] Add pressure tracking error constraint items to the initial objective function matrix. The constraint items reflect the deviation between the actual pressure of the system and the expected pressure. The goal is to minimize this deviation, thereby improving the control accuracy. After adding the constraint items, the tracking error constraint matrix is ​​obtained, which helps to optimize the objective function to ensure that the response of the air pressure regulation system is more accurate. Add safety constraints to the tracking error constraint matrix to ensure that the operation of the system does not exceed the safety limit during the pressure regulation process to avoid risks to equipment and operators. The organizational safety constraint matrix after adding safety constraints ensures that safety regulations are always followed during the optimization process and avoids over-regulation or unstable pressure fluctuations. Constraint fusion of the organizational safety constraint matrix and the tracking error constraint matrix is ​​obtained to obtain the objective function constraint matrix.

[0045] The objective function constraint matrix is ​​solved by convex optimization. Through the convex optimization method, the optimal control parameter solution set is obtained while satisfying various constraints. Based on the optimized parameter solution set, the proportional, integral, and differential parameters of the PID controller are calculated to obtain the PID parameter initial value matrix. Through the closed-loop response characteristic analysis, the response characteristic matrix of the system is obtained, which describes the dynamic response of the system under given initial control parameters, including the steady-state error, adjustment speed, and oscillation characteristics of the system. According to the response characteristic matrix of the system, parameter optimization calculation is performed to obtain the final PID control parameter matrix. This matrix contains the various parameters of the optimized PID controller, which can ensure that the air pressure regulation system can maintain efficient and stable control performance under different operating environments. At the same time, the pressure limit analysis of the system response characteristic matrix is ​​performed, which can reveal whether the pressure control of the system will exceed the predetermined safety range under various working conditions and obtain the pressure threshold vector.

[0046] Based on the pressure threshold vector, the safety margin calculation is performed to ensure that the system is adjusted within the maximum tolerance range and obtain the safety constraint boundary matrix. The safety constraint boundary matrix and the pressure response prediction matrix are matched with safety features to ensure that the operation of the pressure regulation system is always carried out within the safety range. After safety feature matching, the initial safety limit matrix is ​​obtained. The initial safety limit matrix is ​​dynamically corrected, and the safety limit is corrected through real-time data to obtain the initial safety limit matrix. The initial safety limit matrix is ​​closed-loop verified based on the PID control parameter matrix to ensure that in the actual operation process, the safety limit can correctly reflect the safety boundary of the system, and the final pressure safety limit matrix is ​​obtained to ensure that the air pressure regulation system is always within the safety control range while ensuring precise regulation.

[0047] 105. Based on the PID control parameter matrix and the pressure safety limit matrix, a three-level control system including a basic controller, a disturbance compensator and a safety protector is constructed to output an air pressure regulating drive signal.

[0048] Specifically, a basic controller is constructed based on the PID control parameter matrix. The basic controller can realize proportional, integral, and differential actions. It adjusts the air pressure according to the change of pressure so that the system can be kept in an ideal state as much as possible. Through the basic controller, the basic control action matrix obtained represents the initial response of the system to pressure. The control action amount is calculated based on the basic control action matrix to obtain the basic control amount, which represents the direct influence of the control system on pressure regulation without any interference and constraints.

[0049] The interference component identification and calculation of the basic control quantity are performed. The matching interference component and the mismatching interference component are identified, where the matching interference is the external influence known to the system, and the mismatching interference is the disturbance that cannot be predicted or modeled. Based on these interference components, a finite-time interference observer is constructed, which can accurately estimate the influence of the interference within a finite time and obtain the interference compensation amount. This step can effectively reduce the negative impact of external factors on the control system and improve the robustness and stability of the system. At the same time, a pressure limiting protection unit is constructed based on the pressure safety limit matrix to monitor whether the air pressure exceeds the preset safety threshold and perform protection when the pressure exceeds the limit. Through this protection unit, a safety threshold trigger function is obtained, which can indicate when the system enters the safety protection state. Based on this trigger function, a switching function is designed to ensure that the system quickly switches to the safety mode in a dangerous state and obtain the safety protection amount, that is, when the pressure exceeds the safety limit, the control amount of the system will be adjusted to prevent further danger.

[0050] The weight distribution calculation is performed on the basic control quantity, interference compensation quantity and safety protection quantity to determine the importance and contribution of each quantity in the overall control, and the weighted coefficient matrix is ​​obtained. The matrix is ​​used to balance the role of each control part to ensure that the system can give priority to important control requirements without losing control. Based on the weighted coefficient matrix, the control quantity combination operation is performed on each control quantity to obtain the comprehensive control quantity, which integrates multiple factors of basic control, interference compensation and safety protection to ensure that the final adjustment signal can efficiently and stably control the system. The comprehensive control quantity is input to the controller execution unit for signal modulation operation. Through this process, the control signal is modulated into a drive signal suitable for the actuator, and the signal is converted into an actual actuator drive signal through the step of driving voltage conversion. The pneumatic system response analysis is performed on the actuator drive signal to obtain the air pressure response characteristic matrix, which reflects the response speed, stability and accuracy of the air pressure regulation system under different operating conditions. Based on the response characteristic matrix, dynamic compensation calculation is performed, and the response error caused by system delay or nonlinearity is corrected by the compensation algorithm to obtain the air pressure compensation signal.

[0051] The pressure compensation signal is superimposed on the original actuator drive signal to obtain the initial pressure regulation signal. The initial pressure regulation signal is dynamically corrected and the control signal is adjusted in real time to eliminate external interference and dynamic deviations within the system to obtain the compensated pressure regulation signal. Based on the compensated pressure regulation signal, the drive signal conversion process is used to convert the signal into the final pressure regulation drive signal and send it to the actuator of the pressure regulation system to achieve precise pressure regulation.

[0052] Based on the air pressure regulation drive signal, the system output is subjected to the first pressure response data acquisition and the second pressure response data acquisition. The two sets of pressure response data correspond to the behavior of the system under short-term and long-term pressure responses, respectively. Through these two sets of data, the first real-time pressure response sequence and the second real-time pressure response sequence are obtained, and the two sets of data are subjected to dual-path difference calculation to obtain the first pressure tracking error sequence and the second pressure tracking error sequence. The two sets of error sequences reflect the error fluctuations of the pressure regulation system in actual operation.

[0053] The first pressure tracking error sequence is processed in short time segments according to the first time window to obtain the pressure error matrix of the first time window. Similarly, the second pressure tracking error sequence is processed in long time segments according to the second time window to obtain the pressure error matrix of the second time window. Multi-scale statistical feature extraction is performed on these two pressure error matrices to extract the variation law of errors at different time scales, and they are fused into multi-scale error statistical feature vectors to reveal the fluctuation trend of pressure errors at multiple time scales.

[0054] In the control performance evaluation, based on the first pressure tracking error sequence, the first drive voltage and the first drive current of the actuator are sampled at high frequency to obtain the first power consumption data matrix. At the same time, based on the second pressure tracking error sequence, the second drive voltage and the second drive current of the actuator are sampled at low frequency to obtain the second power consumption data matrix. The dual-frequency energy consumption accumulation calculation is performed on the two power consumption data matrices to obtain the dual-frequency system energy consumption characteristic vector, which reflects the energy consumption of the system at both high-frequency and low-frequency levels. Through the comprehensive analysis of the above-mentioned error statistical characteristic vector and energy consumption characteristic vector, the first control performance evaluation index and the second control performance evaluation index are constructed, and based on these two indicators, a dual performance evaluation matrix is ​​obtained. By performing multi-threshold classification judgment on this matrix, the hierarchical performance evaluation results of the system performance are obtained, providing the performance of the system under different operating conditions.

[0055] Based on the hierarchical performance evaluation results, the state transfer matrix of the three-state Markov model is optimized. The first parameter rough adjustment calculation and the second parameter fine adjustment calculation are performed. These two steps are respectively for the preliminary adjustment and precise adjustment of the model parameters, and the rough adjustment state transfer parameter update amount and the fine adjustment state transfer parameter update amount are obtained. Based on these two update amounts, the state transfer matrix is ​​hierarchically and progressively updated to obtain a multi-level updated state transfer matrix. The update process can optimize the state transfer at multiple levels to adapt to different operating environments and requirements.

[0056] The updated state transfer matrix is ​​input into the three-state Markov model, and the first state prediction optimization calculation and the second state prediction optimization calculation are performed to obtain the short-term optimization state prediction result and the long-term optimization state prediction result. The two prediction results provide the pressure state change trend of the system in the short and long term. Based on these two prediction results, multi-time domain pressure response modeling is performed, and a multi-time domain optimization pressure response matrix is ​​obtained. This matrix can describe the pressure response characteristics at different time scales. On this basis, the multi-time domain optimization pressure response matrix is ​​compared with the actual pressure response, and the first deviation analysis and the second deviation analysis are performed. These two analyses can reveal the deviation characteristics between the optimization model and the actual system, and obtain the short-term response deviation feature vector and the long-term response deviation feature vector. According to these two deviation feature vectors, the hierarchical model parameter correction is performed to obtain the multi-level model correction parameter matrix. The corrected multi-level model correction parameter matrix and the multi-level updated state transfer matrix are subjected to multi-level parameter progressive fusion operation to obtain the target state transfer parameter matrix.

[0057] In the embodiment of the present invention, by constructing a three-state Markov model for dynamic state recognition, combined with the pressure spatial distribution feature matrix and the pressure time domain feature matrix, the pressure change characteristics can be accurately captured and the pressure control error can be reduced; a multi-level control architecture including a basic controller, an interference compensator and a safety protector is adopted to achieve effective suppression of system disturbances; the temperature and humidity compensation mechanism and 16-bit high-precision digital conversion are introduced to significantly improve the acquisition accuracy and environmental adaptability of pressure data; an objective function based on pressure tracking error terms and organizational safety constraint terms is designed to achieve dual guarantees of control performance and safety; adaptive adjustment of system parameters is achieved through the coordinated optimization of the pressure state transition probability matrix and the pressure response prediction matrix; dual-channel data acquisition and multi-scale feature extraction technology are adopted to improve the robustness of the system; an adaptive adjustment mechanism based on the PID parameter matrix is ​​used to achieve real-time optimization of control parameters and enhance system stability; through energy consumption analysis and parameter optimization, the overall energy consumption of the system is optimized and energy utilization efficiency is improved. The present invention not only solves the problems of traditional pressure control systems in terms of accuracy, safety and reliability, but also realizes high-precision pressure sensing control.

[0058] In a specific embodiment, the process of executing step 101 may specifically include the following steps:

[0059] The capacitance value of each pressure sensing unit in the pressure sensing array is collected and processed to obtain an original pressure signal, and the original pressure signal is input into a 16-bit analog-to-digital converter for digital conversion to obtain a digital pressure signal matrix;

[0060] The sampling time difference calculation is performed based on the digital pressure signal matrix of two adjacent samples to obtain the pressure difference matrix, and the temperature parameters and relative humidity parameters of the surrounding environment are digitally collected and processed to obtain the environmental compensation parameter set;

[0061] Perform temperature and humidity compensation operations on the digital pressure signal matrix according to the environmental compensation parameter set to obtain a pressure data matrix, and calculate the pressure change trend based on the pressure difference matrix and the pressure data matrix to obtain a dynamic feature matrix;

[0062] The pressure values ​​of adjacent spatial positions in the pressure data matrix are differenced to obtain a spatial pressure gradient matrix, and a weighted average calculation is performed based on the dynamic feature matrix and the spatial pressure gradient matrix to obtain a pressure gradient matrix.

[0063] Specifically, each unit in the pressure sensor array is measured in correspondence with its capacitance value. Since the capacitance value is usually proportional to the air pressure, the pressure information of each sensor position is determined by accurately measuring the capacitance of the sensing unit. After the capacitance of each sensing unit is collected, a raw pressure signal is obtained. The raw pressure signal is input into a 16-bit analog-to-digital converter (ADC). The accuracy of the 16-bit ADC means that it can subdivide the analog signal into , that is, 65536 discrete values, which can more accurately reflect the subtle fluctuations of pressure changes, and obtain the digital pressure signal matrix, using the symbol Indicates that Representative The digitized pressure value of the sensor at the position. The sampling time difference is calculated based on the digital pressure signal matrix of two adjacent samples to obtain the pressure difference matrix Specifically, assuming that at time and time The digital pressure signal matrices obtained by two samplings are: and , then the pressure difference matrix is ​​calculated by the following formula:

[0064] ;

[0065] in, Represents that between two adjacent sampling moments, The pressure change at the position. By calculating the pressure difference matrix, the pressure fluctuation of the system in a short period of time is revealed. At the same time, environmental factors such as temperature and humidity will have a significant impact on the output of the pressure sensor, and the environmental conditions are compensated. The temperature parameters and relative humidity parameters of the surrounding environment are digitally collected and processed to obtain the environmental compensation parameter set using the symbol Indicates that , Represents the temperature value, Represents the relative humidity value. These environmental parameters are collected by temperature and humidity sensors and are also digitally processed to ensure consistency in the subsequent compensation process. The digital pressure signal matrix is ​​compensated for temperature and humidity using the environmental compensation parameter set to obtain a more accurate pressure data matrix. The calculation of temperature and humidity compensation is completed by the following formula:

[0066] ;

[0067] in, and Respectively represent the influence coefficients of temperature and humidity on the pressure signal, and are the reference temperature and reference humidity. These coefficients need to be determined through experimental calibration in order to accurately compensate for the effects of temperature and humidity. The pressure data matrix obtained after compensation operation Reflects the actual pressure value under current environmental conditions. Based on the obtained pressure difference matrix And the compensated pressure data matrix , calculate the pressure change trend and obtain the dynamic characteristic matrix The calculation of the dynamic characteristic matrix is ​​realized by the rate of change of the time series. For example, the rate of change of the local pressure is expressed by the ratio of the pressure difference matrix to the time difference:

[0068] ;

[0069] in, is the time interval between two adjacent samples. Dynamic feature matrix It can reveal the speed of pressure change over time, which helps to identify the instantaneous fluctuation of pressure and the dynamic behavior in the pressure field. The pressure values ​​of adjacent spatial positions are interpolated to obtain the spatial pressure gradient matrix The spatial pressure gradient reflects the change in pressure distribution in space and is an important indicator for measuring the degree of pressure change between different positions in the pressure field. For a two-dimensional pressure sensor array, the spatial pressure gradient is expressed in the following form:

[0070] ;

[0071] in, and Respectively represent Direction and The pressure change rate in a direction is approximated by calculating the pressure difference between adjacent cells. For example:

[0072] ;

[0073] ;

[0074] in, and They are respectively the sensor units in the pressure sensing array Direction and Spacing in direction. Based on dynamic feature matrix and the spatial pressure gradient matrix , perform weighted average calculation to obtain the final pressure gradient matrix The weighted average of the pressure gradient matrix is ​​calculated using the following formula:

[0075] ;

[0076] in, and They are the weighting coefficients of dynamic features and spatial gradients respectively. These weights are adjusted according to the needs of practical applications to highlight the importance of certain features.

[0077] In a specific embodiment, the process of executing step 102 may specifically include the following steps:

[0078] Performing pressure data segmentation processing on the pressure data matrix to obtain a fixed-length pressure data sequence, and calculating the center of gravity position coordinates and peak position coordinates of the pressure distribution in each time window based on the fixed-length pressure data sequence to obtain a pressure distribution positioning matrix;

[0079] The spatial gradient accumulation calculation is performed according to the pressure distribution positioning matrix and the pressure gradient matrix to obtain the pressure action range characteristic matrix, and the multi-scale spectrum decomposition operation is performed on the fixed-length pressure data sequence to obtain the pressure fluctuation multi-band characteristic matrix;

[0080] The phase information and amplitude information of the pressure change are calculated based on the multi-band characteristic matrix of pressure fluctuations to obtain the pressure time-domain dynamic characteristic matrix, and the pressure action range characteristic matrix and the pressure gradient matrix are subjected to spatial characteristic fusion operation to obtain the pressure spatial distribution characteristic matrix;

[0081] The main frequency extraction and phase correlation analysis are performed on the frequency components in the pressure time-domain dynamic feature matrix to obtain the initial time-domain feature matrix. The initial time-domain feature matrix is ​​then subjected to spatial constraint optimization processing based on the pressure spatial distribution feature matrix to obtain the pressure time-domain feature matrix.

[0082] Specifically, the original pressure data matrix is ​​segmented to obtain several fixed-length pressure data sequences. The pressure data is segmented according to the time dimension, and each sequence contains pressure data within a certain time window, so that the pressure change characteristics in a short period of time can be analyzed. Assume that the pressure data matrix is , whose size is ,in and are the sizes of the pressure sensing array in two spatial directions, is the time step. Divide into fixed-length sequences along the time axis, the length of each sequence is , and obtain multiple fixed-length pressure data sequences, recorded as ,in Indicates Based on the fixed-length pressure data sequence, the coordinates of the center of gravity and peak position of the pressure distribution in each time window are calculated to obtain the pressure distribution positioning matrix The position coordinates of the pressure center of gravity are defined as , represents the center of gravity of pressure in the entire two-dimensional space distribution. The calculation method is:

[0083] ;

[0084] ;

[0085] in, For the The pressure value at the position, is the pressure center coordinate within the time window. At the same time, the peak position coordinate is defined as , which indicates the maximum position of pressure in the array. By finding the matrix The index of the maximum value in determines the coordinate:

[0086] ;

[0087] Through the above calculations, the center of gravity and peak position of each time window are obtained to form the pressure distribution positioning matrix , which contains the pressure center of gravity and peak information in all time windows for subsequent spatial analysis. According to the pressure distribution positioning matrix and the pressure gradient matrix, the spatial gradient accumulation calculation is performed to obtain the pressure range characteristic matrix . Pressure gradient matrix Represents the rate of change of pressure in space, and each component is determined by calculating the change of pressure between two adjacent positions. The pressure range characteristic matrix is ​​accumulated by the following formula:

[0088] ;

[0089] in, For the The pressure positioning matrix of the time window, is the corresponding pressure gradient matrix, For in time The cumulative result in the pressure data is used to describe the cumulative change characteristics of the pressure in the spatial distribution. At the same time, a multi-scale spectrum decomposition operation is performed on the fixed-length pressure data sequence to obtain the multi-band characteristic matrix of the pressure fluctuation. . Use wavelet transform or fast Fourier transform to decompose the pressure data into different frequency components. Assume that for each fixed-length pressure data sequence Perform fast Fourier transform and the obtained frequency components are expressed as:

[0090] ;

[0091] in, Indicates frequency, Represents the pressure characteristics at this frequency. Through spectrum decomposition, the characteristics of pressure fluctuations in different frequency ranges are extracted to form a multi-band feature matrix. Based on the multi-band feature matrix of pressure fluctuations, the phase information and amplitude information of pressure changes are calculated to obtain the pressure time domain dynamic feature matrix The phase information of pressure changes describes the relative time relationship of pressure changes in different frequency bands, while the amplitude information describes the intensity of pressure fluctuations in each frequency band. Phase information and amplitude information Calculated by:

[0092] ;

[0093] ;

[0094] in, Indicates frequency The amplitude below Represents the phase. By combining this information together, the pressure time domain dynamic characteristic matrix is ​​obtained , including the dynamic change characteristics of pressure in time and frequency dimensions. and the pressure gradient matrix Perform spatial feature fusion operations to obtain the pressure spatial distribution feature matrix The fusion operation is implemented by weighted averaging to simultaneously consider the spatial accumulation characteristics and gradient characteristics of pressure. The formula for the fusion operation is expressed as:

[0095] ;

[0096] in, and is a weight parameter used to adjust the influence of cumulative features and gradient features in the fusion results. The frequency components in the pressure time domain dynamic feature matrix are extracted and the phase correlation analysis is performed to obtain the initial time domain feature matrix . The purpose of main frequency extraction is to find out the dominant frequency component in the pressure fluctuation, which is achieved by selecting the frequency with the largest amplitude. Phase correlation analysis is used to analyze the phase relationship between each frequency, so as to understand the coordinated behavior of pressure fluctuations in different frequency bands. Through these analyses, the initial time domain feature matrix is ​​obtained, which contains the main frequency information and its corresponding phase characteristics. Based on the pressure spatial distribution feature matrix, the initial time domain feature matrix is ​​optimized with spatial constraints to obtain the final pressure time domain feature matrix , ensuring that the time domain characteristics conform to the actual spatial distribution characteristics of the pressure, for example, ensuring that the pressure fluctuations in a specific area are consistent with the spatial pressure distribution. This is achieved through the following optimization formula:

[0097] ;

[0098] in, is the constraint coefficient, which is used to control the magnitude of optimization. is the spatial distribution feature matrix, is the initial time domain characteristic matrix. Through optimization, a pressure time domain characteristic matrix that is more consistent with the actual pressure distribution is obtained.

[0099] In a specific embodiment, the process of executing step 103 may specifically include the following steps:

[0100] The pressure spatial distribution feature matrix is ​​subjected to binary threshold segmentation processing to obtain a pressure distribution state labeling matrix, and the pressure time domain feature matrix is ​​subjected to frequency domain transformation processing to obtain a pressure spectrum feature matrix;

[0101] Based on the pressure distribution state labeling matrix and the pressure spectrum feature matrix, a three-state division operation of the state space normal pressure state, pressure mutation state and sensor fault state is performed to obtain the state vector at the current moment;

[0102] According to the current state vector, the state transfer matrix of the three-state Markov model is updated and calculated to obtain a dynamic state transfer probability matrix;

[0103] Perform maximum likelihood probability calculation on the dynamic state transfer probability matrix to obtain the state prediction vector at the next moment, and calculate the state probability distribution based on the state prediction vector to obtain the pressure state transfer probability matrix;

[0104] Perform time series matching operation on the pressure state transition probability matrix and the historical state sequence to obtain the state evolution feature vector, and perform state response modeling based on the state evolution feature vector to obtain the state response feature matrix;

[0105] Performing pressure response pattern analysis according to the state response feature matrix and the pressure space distribution feature matrix to obtain a pressure response pattern matrix, and performing response feature extraction based on the pressure response pattern matrix to obtain a pressure response feature vector;

[0106] Performing feature fusion operation on the pressure response feature vector and the state evolution feature vector to obtain a pressure state prediction matrix, and performing response feature mapping based on the pressure state prediction matrix to obtain an initial response prediction matrix;

[0107] The initial response prediction matrix and the pressure time domain feature matrix are optimized with time and space feature constraints to obtain the pressure response prediction matrix.

[0108] Specifically, the pressure spatial distribution feature matrix is ​​subjected to binary threshold segmentation processing, and the original pressure spatial distribution feature matrix is ​​converted into a pressure distribution state labeling matrix, making the pressure distribution information easier to analyze. Assume that the pressure spatial distribution feature matrix is , set a threshold For binarization:

[0109] ;

[0110] in, represents the pressure distribution state labeling matrix, is the first Status flag of the position. If the pressure value exceeds the threshold , then it is marked as 1, indicating that the pressure state at this position is high, there is pressure abnormality or large fluctuation, otherwise it is marked as 0, indicating that the pressure is relatively normal. Through the binarization process, the complex pressure distribution information is simplified into two states: normal and abnormal. The pressure time domain feature matrix is ​​transformed into the frequency domain to obtain the pressure spectrum feature matrix. Assume that the pressure time domain feature matrix is , by fast Fourier transform, it is converted into frequency domain representation to obtain the characteristic distribution of pressure at different frequencies:

[0111] ;

[0112] in, is the pressure spectrum feature matrix, Indicates frequency. The pressure spectrum feature matrix contains the pressure fluctuation characteristics of the system in the frequency domain, which is used to identify periodic or sudden changes in the pressure signal. Based on the pressure distribution state labeling matrix And the pressure spectrum characteristic matrix , perform the three-state partition operation of the state space and obtain the state vector at the current moment . The system state is divided into three types: normal pressure state, pressure mutation state and sensor failure state. For each position in the pressure distribution state marking matrix, if the mark is 1 and the spectrum feature matrix has a significant peak in the high frequency band, it is considered that there is a pressure mutation state at this position; if the mark is 0 and the spectrum feature is relatively stable, it is considered that the position is in a normal pressure state; if the spectrum feature matrix shows large noise or irregular fluctuations, it is a sensor failure state. Through division, the overall state of the system at the current moment is obtained, which is expressed as the state vector ,in Indicates The state category of the position. According to the state vector at the current moment , perform parameter update calculation on the state transfer matrix of the three-state Markov model and obtain the dynamic state transfer probability matrix Assumptions is the state transfer matrix of the previous moment, then the update at the current moment is adjusted according to the change of the state vector. For example, the Bayesian update method is used to calculate the state transition probability:

[0113] ;

[0114] in, Indicates the status from Transfer to The probability of is the conditional probability, indicating that the state at the previous moment is In the case of, the current moment is the state In this way, the state transfer matrix is ​​gradually updated so that it can dynamically adapt to the changes in the current system. Perform maximum likelihood probability calculation to obtain the state prediction vector for the next moment The goal of maximum likelihood estimation is to find the most likely state transition path and make predictions by maximizing the joint probability of the states:

[0115] ;

[0116] in, Indicates the most likely state category at the next moment. Based on the state prediction vector, calculate the probability distribution of the state in different categories and obtain the pressure state transition probability matrix This matrix is ​​used to describe the possibility of the system transitioning between different states. Perform time series matching operation with the historical state sequence to obtain the state evolution feature vector Compare the current state with the historical state to identify the trend of state change. Assume that the historical state sequence is , then the state evolution feature vector is obtained by calculating the similarity between the historical state and the current state:

[0117] ;

[0118] in, is a weight factor used to adjust the contribution of different time steps in the evolution characteristics. Describes the evolution direction and rate of the system state. Based on the state evolution feature vector , perform state response modeling and obtain the state response characteristic matrix The state response characteristic matrix is ​​used to describe the response behavior of the system to different state changes. The response characteristics are obtained by linear regression modeling of the state evolution characteristics:

[0119] ;

[0120] in, is the weight matrix of the state response, which is fitted through the training data to accurately describe the response characteristics of the system. According to the state response characteristic matrix and the pressure space distribution characteristic matrix, the pressure response mode analysis is performed to obtain the pressure response mode matrix The pressure response mode matrix describes the distribution characteristics of the pressure field under different states and is calculated using the following formula:

[0121] ;

[0122] Through the pressure response pattern matrix, the response feature is extracted to obtain the pressure response feature vector , which contains the main modes and characteristics of the pressure response. and the state evolution feature vector Perform feature fusion operation to obtain the pressure state prediction matrix . Combining information from different feature sources to improve prediction accuracy is achieved through weighted summation:

[0123] ;

[0124] in, and is the weight coefficient of fusion. Based on the pressure state prediction matrix Perform response feature mapping to obtain the initial response prediction matrix The initial response prediction matrix and the pressure time domain feature matrix Perform spatiotemporal feature constraint optimization to obtain the final pressure response prediction matrix . Constrained optimization is achieved through the following formula:

[0125] ;

[0126] in, To optimize the step size, it is used to adjust the gap between the initial prediction and the actual characteristics to obtain a more realistic pressure response prediction.

[0127] In a specific embodiment, the process of executing step 104 may specifically include the following steps:

[0128] The state weight distribution calculation is performed on the pressure state transfer probability matrix to obtain the state weight coefficient matrix, and the time series correlation analysis is performed on the pressure response prediction matrix to obtain the response feature weight matrix;

[0129] A multi-objective optimization function is constructed based on the state weight coefficient matrix and the response characteristic weight matrix to obtain an initial objective function matrix, and a pressure tracking error constraint item is added to the initial objective function matrix to obtain a tracking error constraint matrix;

[0130] Adding safety constraints to the tracking error constraint matrix to obtain the organizational safety constraint matrix, and then fusing the organizational safety constraint matrix with the tracking error constraint matrix to obtain the objective function constraint matrix;

[0131] Perform convex optimization solution operation on the objective function constraint matrix to obtain the optimization parameter solution set, and calculate the proportional, integral and differential parameters of the PID controller based on the optimization parameter solution set to obtain the PID parameter initial value matrix;

[0132] Based on the PID parameter initial value matrix, the closed-loop response characteristic analysis is performed to obtain the system response characteristic matrix, and the parameter optimization calculation is performed based on the system response characteristic matrix to obtain the PID control parameter matrix;

[0133] Perform pressure limit analysis on the system response characteristic matrix to obtain a pressure threshold vector, and perform safety margin calculation based on the pressure threshold vector to obtain a safety constraint boundary matrix;

[0134] The safety constraint boundary matrix and the pressure response prediction matrix are matched with each other for safety characteristics to obtain the safety limit initial value matrix, and the safety limit initial value matrix is ​​dynamically corrected to obtain the initial safety limit matrix;

[0135] Based on the PID control parameter matrix, a closed-loop verification operation is performed on the initial safety limit matrix to obtain the pressure safety limit matrix.

[0136] Specifically, the state weight distribution calculation is performed on the pressure state transition probability matrix to obtain the state weight coefficient matrix. Assume that the pressure state transition probability matrix , each element of which Indicates from the state To status In order to analyze the influence of the state in the entire system, the state weight coefficient matrix is ​​defined :

[0137] ;

[0138] in, Indicates The sum of the total transition probabilities between all possible states of a state is used to measure the importance of the state in the overall system. The calculation of the state weight coefficient can help the system effectively allocate resources under multi-state conditions and ensure that key states receive more adequate responses. Perform time series correlation analysis to obtain the response feature weight matrix The purpose of the response feature weight matrix is ​​to analyze the relationship between the pressure responses at different time points, which is achieved by calculating the autocorrelation function between the responses:

[0139] ;

[0140] in, Indicates at time The correlation between the pressure response at and the subsequent time, It is the timing delay. It can provide information about the degree of mutual influence between pressures at different time points. Based on the obtained state weight coefficient matrix and response characteristic weight matrix, a multi-objective optimization function is constructed, taking into account the characteristics of both state and response. The definition of the multi-objective optimization function is expressed in the following form:

[0141] ;

[0142] in, is the control variable to be optimized, is the state-dependent objective function, To respond to the relevant objective function, and are the weights of the state and response respectively. By weighted combination of the state and response features, the initial objective function matrix is ​​obtained, which is recorded as On the basis of the initial objective function matrix, a pressure tracking error constraint is added to ensure that the actual pressure of the system can track the expected value. This is achieved by defining the pressure tracking error:

[0143] ;

[0144] in, For the expected pressure, is the actual pressure, For in time The tracking error is . Adding this error term to the initial objective function, we get the tracking error constraint matrix

[0145] ;

[0146] in, is the weight coefficient used to adjust the impact of the tracking error term on the objective function. Safety constraints are added to the tracking error constraint matrix to ensure the safety of the system during operation. Ensure that the pressure value does not exceed certain safety limits to avoid equipment damage or safety accidents. By adding safety constraints, the organizational safety constraint matrix is ​​obtained. :

[0147] ;

[0148] in, is the safety constraint weight, is the safety pressure threshold. If the pressure exceeds the safety threshold, this item will increase, thereby punishing unsafe operations. Constraint fusion is performed on the organizational safety constraint matrix and the tracking error constraint matrix to obtain the final objective function constraint matrix. On this basis, a convex optimization solution operation is performed on the objective function constraint matrix to obtain the optimization parameter solution set Convex optimization is an effective solution method that can find the global optimal solution when the objective function is convex. Using the parameter solution set obtained by optimization, the proportional (K_p), integral (K_i) and differential (K_d) parameters of the PID controller are calculated to obtain the PID parameter initial value matrix:

[0149] ;

[0150] Based on the obtained PID parameter initial value matrix, a closed-loop response characteristic analysis is performed to understand the dynamic performance of the system under the current control parameters. Through simulation or experiment, the response characteristic matrix of the system is obtained. , which describes the output response of the system after control is applied, including characteristics such as steady-state error, adjustment time and overshoot. According to the system response characteristic matrix, parameter optimization calculation is performed to obtain the final PID control parameter matrix . By carefully adjusting the proportional, integral and differential parameters, the system dynamic response performance is optimized, such as reducing steady-state errors and oscillations and improving system stability. After obtaining the optimized PID control parameters, the pressure limit of the system is analyzed. By analyzing the pressure limit of the system response characteristic matrix, the pressure threshold vector is obtained. , which describes the maximum allowable pressure of the system under different operating conditions. Based on this pressure threshold vector, the safety margin is calculated to obtain the safety constraint boundary matrix , which is used to ensure that the pressure regulation of the system is always within the safe range. Perform safety feature matching operations to obtain the safety limit initial value matrix According to the predicted response of the system, it is ensured that the safety margin will not be exceeded at any time. The initial safety limit matrix is ​​dynamically corrected to adapt to the changes in the system during operation, and the initial safety limit matrix is ​​obtained. Based on the PID control parameter matrix , perform closed-loop verification operations on the initial safety limit matrix to ensure that the control system can still operate normally after applying these safety limits, and obtain the pressure safety limit matrix This matrix is ​​used in actual control to ensure that each step of pressure regulation is carried out within a safe range.

[0151] In a specific embodiment, the process of executing step 105 may specifically include the following steps:

[0152] Based on the PID control parameter matrix, a basic controller with proportional, integral and differential actions is constructed to obtain a basic control action matrix, and the control action amount is calculated according to the basic control action matrix to obtain a basic control amount;

[0153] Perform interference component identification calculation on the basic control quantity to obtain the matching interference component and the mismatching interference component, and construct a finite time interference observer based on the matching interference component and the mismatching interference component to obtain the interference compensation amount;

[0154] Based on the pressure safety limit matrix, a pressure limiting protection unit is constructed to obtain a safety threshold trigger function, and a switching function is designed according to the safety threshold trigger function to obtain a safety protection amount;

[0155] Perform weight distribution calculation on basic control quantity, interference compensation quantity and safety protection quantity to obtain weight coefficient matrix, and perform control quantity combination operation according to the weight coefficient matrix to obtain comprehensive control quantity;

[0156] The integrated control quantity is input into the controller execution unit for signal modulation operation to obtain a control action signal, and a drive voltage conversion is performed based on the control action signal to obtain an actuator drive signal;

[0157] Perform pneumatic system response analysis on the actuator drive signal to obtain an air pressure response characteristic matrix, and perform dynamic compensation calculation based on the air pressure response characteristic matrix to obtain an air pressure compensation signal;

[0158] The air pressure compensation signal is superimposed on the actuator drive signal to obtain an initial air pressure adjustment signal, and the initial air pressure adjustment signal is dynamically corrected to obtain a compensated air pressure adjustment signal;

[0159] A drive signal conversion process is performed based on the compensated air pressure regulation signal to obtain an air pressure regulation drive signal.

[0160] Specifically, a basic controller with proportional, integral, and differential functions is constructed based on the PID control parameter matrix. ,integral( ) and differential ( ) coefficient. The basic controller adjusts the system output pressure through proportional, integral and differential actions to make it as close to the set value as possible. Assume that the set pressure is The actual pressure is , then at time The pressure error is:

[0161] ;

[0162] Based on this error signal, a basic controller is constructed. Its control action matrix is ​​calculated by the following formula:

[0163] ;

[0164] in, Indicates at time The proportional term is adjusted according to the current error, the integral term accumulates the error to eliminate the steady-state error, and the differential term suppresses rapid changes according to the rate of change of the error to obtain a smooth response of the control system. According to the control action matrix, the basic control quantity is calculated and used It represents the regulation effect of the PID controller on the system at each time point. In order to enhance the anti-interference ability of the control system, the interference component identification calculation is performed on the basic control quantity. By analyzing the different frequency components in the basic control quantity, the matching interference component and the mismatching interference component are distinguished. The matching interference component refers to those known and predictable interferences, while the mismatching interference component represents those unpredictable or random interferences. Assume that the basic control quantity is , decompose it into frequency components by Fourier transform:

[0165] ;

[0166] By analyzing the spectrum characteristics, matching interference components are identified and mismatch interference component Based on these two types of interference components, a finite-time interference observer is constructed, whose purpose is to estimate and compensate for the impact of these interferences in a finite time. The output of the interference observer is the interference compensation , expressed by the following formula:

[0167] ;

[0168] in, The adjustment parameters are used to control the compensation degree of mismatch interference. Through the interference compensation amount, the interference in the basic control quantity is effectively compensated, thereby improving the robustness of the system. At the same time, based on the pressure safety limit matrix, a pressure limit protection unit is constructed to ensure that the pressure value does not exceed the safe threshold range to avoid damage to the system. Through the pressure limit matrix, the safety threshold trigger function is obtained , which is defined as:

[0169] ;

[0170] in, is the safety threshold. If the current pressure exceeds the safety range, the function is triggered will be equal to 1. Based on the trigger function, design the switching function , used to control the system to enter safety protection mode:

[0171] ;

[0172] According to the switching function, the safety protection amount is obtained , which is used to limit the system pressure. The basic control amount, interference compensation amount and safety protection amount are weighted and calculated to obtain the weighted coefficient matrix Assign different weights to different control quantities to ensure that the contribution of each control quantity is fully considered during comprehensive control:

[0173] ;

[0174] in, , , is the weight coefficient of each control quantity. Based on these weights, the control quantity combination operation is performed on each control quantity to obtain the comprehensive control quantity :

[0175] ;

[0176] The comprehensive control Input to the controller execution unit, perform signal modulation operation, and obtain the control action signal . Convert the control signal into a voltage signal suitable for the actuator. Through modulation operation, map the comprehensive control quantity into the required drive voltage range. Based on the control action signal , convert the drive voltage to obtain the actuator drive signal This signal directly acts on the actuator, driving it to adjust the pressure accordingly. After receiving the drive signal, the actuator adjusts the pneumatic system accordingly to change the pressure state of the system. In order to ensure that the adjustment effect of the actuator meets expectations, the pneumatic system response analysis of the actuator drive signal is performed to obtain the air pressure response characteristic matrix This matrix reflects the pressure response characteristics of the system under different driving signals, including information such as the pressure change rate, steady-state value and overshoot. Based on the air pressure response characteristic matrix, dynamic compensation calculation is performed to obtain the air pressure compensation signal , in order to compensate for the deviation in the actual response of the system. With actuator drive signal Perform superposition operation to obtain the initial air pressure adjustment signal The initial air pressure regulation signal includes the combined effect of actuator drive and compensation. The initial air pressure regulation signal is dynamically corrected to correct errors caused by environmental changes or system nonlinearity, and the compensated air pressure regulation signal is obtained. Based on the compensated air pressure regulation signal, the drive signal conversion process is performed to obtain the final air pressure regulation drive signal This signal is used to drive the pneumatic system to ensure that the system pressure is maintained near the set value, achieving precise regulation and control of the pressure.

[0177] In a specific embodiment, the pressure control method for accurately adjusting the air pressure intensity further includes the following steps:

[0178] Based on the air pressure regulating driving signal, pressure response data collection and dual-path difference calculation are performed on the system output to obtain a pressure tracking error sequence;

[0179] Performing time segmentation processing and multi-scale statistical feature extraction based on the pressure tracking error sequence to obtain a multi-scale error statistical feature vector;

[0180] Based on the pressure tracking error sequence, high-frequency sampling is performed on the driving voltage and driving current of the actuator and dual-frequency energy consumption accumulation is calculated to obtain a dual-frequency system energy consumption characteristic vector;

[0181] Constructing a dual performance evaluation matrix according to the multi-scale error statistical feature vector and the dual-frequency system energy consumption feature vector, and performing multi-threshold classification judgment on the dual performance evaluation matrix to obtain a classification performance evaluation result;

[0182] Based on the hierarchical performance evaluation result, the state transfer matrix of the three-state Markov model is fine-tuned and calculated and hierarchically and progressively updated to obtain a multi-level updated state transfer matrix;

[0183] Inputting the multi-stage updated state transfer matrix into the three-state Markov model, performing state prediction optimization calculation and multi-time domain pressure response modeling, and obtaining a multi-time domain optimized pressure response matrix;

[0184] Performing deviation analysis and row-level model parameter correction on the multi-time-domain optimized pressure response matrix and the actual pressure response to obtain a multi-level model correction parameter matrix;

[0185] The multi-level model correction parameter matrix and the multi-level updated state transfer matrix are subjected to a multi-level parameter progressive fusion operation to obtain a target state transfer parameter matrix.

[0186] Furthermore, in this embodiment, based on the air pressure regulation drive signal, the system output is subjected to first pressure response data acquisition and second pressure response data acquisition to obtain a first real-time pressure response sequence and a second real-time pressure response sequence, and a dual-path difference calculation is performed on the first real-time pressure response sequence and the second real-time pressure response sequence to obtain a first pressure tracking error sequence and a second pressure tracking error sequence;

[0187] The first pressure tracking error sequence is processed by short-term segmentation according to the first time window to obtain the first time window pressure error matrix, the second pressure tracking error sequence is processed by long-term segmentation according to the second time window to obtain the second time window pressure error matrix, and multi-scale statistical feature extraction is performed on the first time window pressure error matrix and the second time window pressure error matrix to obtain a multi-scale error statistical feature vector;

[0188] Based on the first pressure tracking error sequence, a first driving voltage and a first driving current of the actuator are sampled at high frequency to obtain a first power consumption data matrix; based on the second pressure tracking error sequence, a second driving voltage and a second driving current of the actuator are sampled at low frequency to obtain a second power consumption data matrix; and dual-frequency energy consumption accumulation calculation is performed on the first power consumption data matrix and the second power consumption data matrix to obtain a dual-frequency system energy consumption characteristic vector;

[0189] Constructing a first control performance evaluation index and a second control performance evaluation index according to the multi-scale error statistical eigenvector and the dual-frequency system energy consumption eigenvector to obtain a dual performance evaluation matrix, and performing multi-threshold classification judgment on the dual performance evaluation matrix to obtain a classification performance evaluation result;

[0190] Based on the hierarchical performance evaluation result, the state transfer matrix of the three-state Markov model is subjected to a first parameter coarse adjustment calculation and a second parameter fine adjustment calculation to obtain a coarse adjustment state transfer parameter update amount and a fine adjustment state transfer parameter update amount, and the state transfer matrix is ​​hierarchically and progressively updated according to the coarse adjustment state transfer parameter update amount and the fine adjustment state transfer parameter update amount to obtain a multi-level updated state transfer matrix;

[0191] The state transfer matrix after multi-level update is input into the three-state Markov model, and the first state prediction optimization calculation and the second state prediction optimization calculation are performed to obtain the short-term optimization state prediction result and the long-term optimization state prediction result, and multi-time domain pressure response modeling is performed based on the short-term optimization state prediction result and the long-term optimization state prediction result to obtain the multi-time domain optimization pressure response matrix;

[0192] Performing a first deviation analysis and a second deviation analysis on the multi-time domain optimized pressure response matrix and the actual pressure response to obtain a short-term response deviation characteristic vector and a long-term response deviation characteristic vector, and performing hierarchical model parameter correction based on the short-term response deviation characteristic vector and the long-term response deviation characteristic vector to obtain a multi-level model correction parameter matrix;

[0193] The multi-level model correction parameter matrix and the multi-level updated state transfer matrix are subjected to multi-level parameter progressive fusion operation to obtain the target state transfer parameter matrix.

[0194] Specifically, based on the air pressure regulation driving signal, the system output is subjected to the first pressure response data collection and the second pressure response data collection. After that, the actuator responds to the signal and monitors the output pressure of the system in real time to obtain the first real-time pressure response sequence and a second real-time pressure response sequence These data correspond to the pressure response of the system at two different time scales: the first real-time pressure response sequence usually represents a faster short-term response, while the second real-time pressure response sequence represents a slower long-term response. The dual-channel acquisition method is to comprehensively evaluate the dynamic performance of the system at different time scales. The dual-channel difference calculation of these two pressure response sequences is performed to obtain the first pressure tracking error sequence and the second pressure tracking error sequence Assume the set pressure is , then the calculation of the error sequence is expressed as:

[0195] ;

[0196] ;

[0197] in, and Respectively, in time The tracking errors of the first and second pressure response sequences are calculated. The accuracy of the system pressure regulation is evaluated through these two sets of error data. The first pressure tracking error sequence is processed in short time segments according to the first time window to obtain the pressure error matrix of the first time window The purpose of short-term segmentation is to analyze the error characteristics of the system at a smaller time scale. Similarly, the second pressure tracking error sequence is processed in long-term segments according to the second time window to obtain the second time window pressure error matrix Long-term segment processing is used to analyze the regulation performance of the system at a longer time scale. Multi-scale statistical feature extraction is performed on the pressure error matrix of the first time window and the pressure error matrix of the second time window to obtain the multi-scale error statistical feature vector The feature extraction process includes calculating the mean, variance, skewness, and kurtosis in each time window to describe the error distribution characteristics of the system at different time scales. Assume that the error matrix is , then its multi-scale statistical characteristics are calculated by the following formula:

[0198] ;

[0199] in, represents the mean of the errors, Represents the standard deviation of the error. Through these statistical features, a multi-scale error statistical feature vector is constructed to describe the overall characteristics of the system error. Based on the first pressure tracking error sequence, the first driving voltage of the actuator and the first drive current Perform high frequency sampling to obtain the first power consumption data matrix The power consumption is calculated as:

[0200] ;

[0201] This data matrix is ​​used to describe the power consumption of the system in a short period of time. Similarly, based on the second pressure tracking error sequence, the second driving voltage of the actuator is and the second drive current Perform low-frequency sampling to obtain the second power consumption data matrix The dual-frequency energy consumption is accumulated by calculating the two power consumption data matrices, and the energy consumption characteristic vector of the dual-frequency system is obtained. , which is expressed as:

[0202] ;

[0203] Represents the total power consumption of the system at two different frequencies, high frequency and low frequency, which helps to evaluate the energy efficiency of the system. and dual-frequency system energy consumption characteristic vector , construct the first control performance evaluation index and the second control performance evaluation index, and obtain the dual performance evaluation matrix The first control performance evaluation index describes the tracking ability of the system through the root mean square error, while the second control performance evaluation index uses energy consumption to measure the energy efficiency of the system. By performing multi-threshold classification judgment on the dual performance evaluation matrix, the hierarchical performance evaluation results of the system are obtained, for example, the system performance is divided into "excellent", "good", "average" and other levels. Based on the hierarchical performance evaluation results, the state transition matrix of the three-state Markov model is First, perform a rough adjustment calculation of the first parameter to obtain the rough adjustment state transfer parameter update amount , and then perform the second parameter fine-tuning calculation to obtain the fine-tuning state transfer parameter update amount The goal of coarse tuning parameter update is to quickly respond to changes in system performance, while fine tuning is used to correct subtle changes in state transfer. By performing hierarchical progressive updates on the state transfer matrix, the multi-level updated state transfer matrix is ​​obtained. :

[0204] ;

[0205] The multi-level updated state transfer matrix is ​​input into the three-state Markov model to perform the first state prediction optimization calculation and the second state prediction optimization calculation to obtain the short-term optimized state prediction result. And the long-term optimization state prediction results The short-term optimization results are used to evaluate the state changes of the system in a short period of time, while the long-term optimization results are used to predict the state trend in a longer period of time. Based on these prediction results, multi-time domain pressure response modeling is performed to obtain the multi-time domain optimization pressure response matrix , which is used to describe the pressure response characteristics of the system at different time scales. In order to improve the accuracy of the model, the multi-time domain optimized pressure response matrix is ​​compared with the actual pressure response, and the first deviation analysis and the second deviation analysis are performed to obtain the short-term response deviation feature vectors. and the long-term response deviation eigenvector These deviation feature vectors describe the difference between the model prediction and the actual system response. Based on these deviation features, hierarchical model parameter correction is performed to obtain the multi-level model correction parameter matrix , which is used to correct the deviation in the model. The multi-level model correction parameter matrix And the state transfer matrix after multi-level update Perform multi-level parameter progressive fusion operations to obtain the target state transfer parameter matrix This matrix is ​​the final state transfer matrix obtained after multiple updates and corrections, and is used to accurately describe the transition rules of the system between different states.

[0206] The pressure control method for accurately adjusting the air pressure intensity in the embodiment of the present invention is described above. The pressure control device for accurately adjusting the air pressure intensity in the embodiment of the present invention is described below. Figure 2 In one embodiment of the present invention, a pressure control device for accurately adjusting the air pressure intensity includes:

[0207] The conversion module 201 is used to perform digital conversion and temperature and humidity compensation on the original pressure signal collected by the pressure sensor array to obtain a pressure data matrix and a pressure gradient matrix;

[0208] A calculation module 202 is used to calculate the pressure distribution centroid and the pressure fluctuation spectrum based on the pressure data matrix and the pressure gradient matrix to obtain a pressure spatial distribution feature matrix and a pressure time domain feature matrix;

[0209] Identification module 203, used for inputting pressure spatial distribution feature matrix and pressure time domain feature matrix into three-state Markov model for dynamic state identification, and obtaining pressure state transition probability matrix and pressure response prediction matrix;

[0210] A solution module 204 is used to construct an objective function including a pressure tracking error term and an organizational safety constraint term according to the pressure state transition probability matrix and the pressure response prediction matrix, and obtain a PID control parameter matrix and a pressure safety limit matrix by solution;

[0211] The output module 205 is used to construct a three-level control system including a basic controller, a disturbance compensator and a safety protector based on the PID control parameter matrix and the pressure safety limit matrix, and output a gas pressure regulating drive signal.

[0212] Through the cooperation of the above components, a three-state Markov model is constructed for dynamic state identification. Combined with the pressure spatial distribution feature matrix and the pressure time domain feature matrix, the pressure change characteristics can be accurately captured and the pressure control error can be reduced. A multi-level control architecture including a basic controller, a disturbance compensator and a safety protector is adopted to achieve effective suppression of system disturbances. The temperature and humidity compensation mechanism and 16-bit high-precision digital conversion are introduced to significantly improve the acquisition accuracy and environmental adaptability of pressure data. An objective function based on pressure tracking error terms and organizational safety constraint terms is designed to achieve dual guarantees of control performance and safety. Adaptive adjustment of system parameters is achieved through the coordinated optimization of the pressure state transition probability matrix and the pressure response prediction matrix. The dual-channel data acquisition and multi-scale feature extraction technology are adopted to improve the robustness of the system. The adaptive adjustment mechanism based on the PID parameter matrix realizes real-time optimization of control parameters and enhances system stability. Through energy consumption analysis and parameter optimization, the overall energy consumption of the system is optimized and energy utilization efficiency is improved. The present invention not only solves the problems of traditional pressure control systems in terms of accuracy, safety and reliability, but also realizes high-precision pressure sensing control.

[0213] above Figure 2 The pressure control device for accurately adjusting the air pressure intensity in the embodiment of the present invention is described in detail from the perspective of modular functional entity. The pressure control device for accurately adjusting the air pressure intensity in the embodiment of the present invention is described in detail from the perspective of hardware processing.

[0214] Figure 3: is a structural schematic diagram of a pressure control device for accurately adjusting the air pressure intensity provided by an embodiment of the present invention. The pressure control device 300 for accurately adjusting the air pressure intensity may have relatively large differences due to different configurations or performances, and may include one or more processors (central processing units, CPU) 310 (for example, one or more processors) and a memory 320, and one or more storage media 330 (for example, one or more mass storage device terminals) storing application programs 333 or data 332. Among them, the memory 320 and the storage medium 330 can be short-term storage or permanent storage. The program stored in the storage medium 330 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations in the pressure control device 300 for accurately adjusting the air pressure intensity. Furthermore, the processor 310 can be configured to communicate with the storage medium 330, and execute a series of instruction operations in the storage medium 330 on the pressure control device 300 for accurately adjusting the air pressure intensity to implement the steps of the pressure control method for accurately adjusting the air pressure intensity.

[0215] The pressure control device 300 for precisely adjusting the air pressure intensity may also include one or more power supplies 340, one or more wired or wireless network interfaces 350, one or more input and output interfaces 360, and / or one or more operating systems 331, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. It will be appreciated by those skilled in the art that Figure 3 The structure of the pressure control device for accurately adjusting the air pressure intensity shown does not constitute a limitation of the pressure control device for accurately adjusting the air pressure intensity provided by the present invention, and may include more or fewer components than shown in the figure, or a combination of certain components, or a different arrangement of components.

[0216] 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. Instructions are stored in the computer-readable storage medium. When the instructions are executed on a computer, the computer executes the steps of the pressure control method for precisely adjusting the air pressure intensity.

[0217] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0218] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or the whole or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program code.

[0219] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A pressure control method for accurately adjusting air pressure intensity, characterized in that: The method comprises: The original pressure signal collected by the pressure sensor array is digitized and temperature and humidity compensated to obtain a pressure data matrix and a pressure gradient matrix; Calculate the pressure distribution centroid and the pressure fluctuation spectrum based on the pressure data matrix and the pressure gradient matrix to obtain a pressure spatial distribution feature matrix and a pressure time domain feature matrix; Inputting the pressure spatial distribution characteristic matrix and the pressure time domain characteristic matrix into a three-state Markov model for dynamic state recognition to obtain a pressure state transition probability matrix and a pressure response prediction matrix; According to the pressure state transfer probability matrix and the pressure response prediction matrix, an objective function including a pressure tracking error term and an organizational safety constraint term is constructed, and a PID control parameter matrix and a pressure safety limit matrix are obtained by solving the problem; specifically, the following steps are performed: a state weight distribution calculation is performed on the pressure state transfer probability matrix to obtain a state weight coefficient matrix, and a time series correlation analysis is performed on the pressure response prediction matrix to obtain a response characteristic weight matrix; a multi-objective optimization function is constructed based on the state weight coefficient matrix and the response characteristic weight matrix to obtain an initial objective function matrix, and a pressure tracking error constraint term is added to the initial objective function matrix to obtain a tracking error constraint matrix; safety constraints are added to the tracking error constraint matrix to obtain an organizational safety constraint matrix, and the organizational safety constraint matrix is ​​constraint-fused with the tracking error constraint matrix to obtain an objective function constraint matrix; the objective function constraint matrix Perform convex optimization solution operation to obtain an optimization parameter solution set, and calculate the proportional, integral, and differential parameters of the PID controller based on the optimization parameter solution set to obtain a PID parameter initial value matrix; perform closed-loop response characteristic analysis based on the PID parameter initial value matrix to obtain a system response characteristic matrix, and perform parameter optimization calculation based on the system response characteristic matrix to obtain a PID control parameter matrix; perform pressure limit analysis on the system response characteristic matrix to obtain a pressure threshold vector, and perform safety margin calculation based on the pressure threshold vector to obtain a safety constraint boundary matrix; perform safety feature matching operation on the safety constraint boundary matrix and the pressure response prediction matrix to obtain a safety limit initial value matrix, and dynamically correct the safety limit initial value matrix to obtain an initial safety limit matrix; perform closed-loop verification operation on the initial safety limit matrix based on the PID control parameter matrix to obtain a pressure safety limit matrix; A three-level control system including a basic controller, a disturbance compensator and a safety protector is constructed based on the PID control parameter matrix and the pressure safety limit matrix to output an air pressure regulating drive signal.

2. The pressure control method for accurately adjusting air pressure intensity according to claim 1, characterized in that: The original pressure signal collected by the pressure sensor array is digitally converted and temperature and humidity compensated to obtain a pressure data matrix and a pressure gradient matrix, including: The capacitance value of each pressure sensing unit in the pressure sensing array is collected and processed to obtain an original pressure signal, and the original pressure signal is input into a 16-bit analog-to-digital converter for digital conversion to obtain a digital pressure signal matrix; The sampling time difference calculation is performed based on the digital pressure signal matrix of two adjacent samples to obtain a pressure difference matrix, and the temperature parameters and relative humidity parameters of the surrounding environment are digitally collected and processed to obtain an environmental compensation parameter set; Performing temperature and humidity compensation operations on the digital pressure signal matrix according to the environmental compensation parameter set to obtain a pressure data matrix, and calculating a pressure change trend based on the pressure difference matrix and the pressure data matrix to obtain a dynamic feature matrix; The pressure values ​​of adjacent spatial positions in the pressure data matrix are subjected to difference calculation to obtain a spatial pressure gradient matrix, and a weighted average calculation is performed based on the dynamic feature matrix and the spatial pressure gradient matrix to obtain a pressure gradient matrix.

3. The pressure control method for accurately adjusting air pressure intensity according to claim 2, characterized in that: The pressure distribution centroid and pressure fluctuation spectrum are calculated based on the pressure data matrix and the pressure gradient matrix to obtain a pressure spatial distribution feature matrix and a pressure time domain feature matrix, including: Performing pressure data segmentation processing on the pressure data matrix to obtain a fixed-length pressure data sequence, and calculating the center of gravity position coordinates and peak position coordinates of the pressure distribution in each time window based on the fixed-length pressure data sequence to obtain a pressure distribution positioning matrix; Performing spatial gradient accumulation calculation according to the pressure distribution positioning matrix and the pressure gradient matrix to obtain a pressure action range characteristic matrix, and performing multi-scale spectrum decomposition operation on the fixed-length pressure data sequence to obtain a pressure fluctuation multi-band characteristic matrix; The phase information and amplitude information of the pressure change are calculated based on the pressure fluctuation multi-band feature matrix to obtain the pressure time domain dynamic feature matrix, and the pressure action range feature matrix and the pressure gradient matrix are subjected to spatial feature fusion operation to obtain the pressure spatial distribution feature matrix; The frequency components in the pressure time-domain dynamic feature matrix are subjected to main frequency extraction and phase correlation analysis to obtain an initial time-domain feature matrix, and the initial time-domain feature matrix is ​​subjected to spatial constraint optimization processing based on the pressure spatial distribution feature matrix to obtain a pressure time-domain feature matrix.

4. The pressure control method for accurately adjusting air pressure intensity according to claim 3, characterized in that: The step of inputting the pressure spatial distribution feature matrix and the pressure time domain feature matrix into a three-state Markov model for dynamic state recognition to obtain a pressure state transition probability matrix and a pressure response prediction matrix includes: Performing a binary threshold segmentation process on the pressure spatial distribution feature matrix to obtain a pressure distribution state labeling matrix, and performing a frequency domain transformation process on the pressure time domain feature matrix to obtain a pressure spectrum feature matrix; Based on the pressure distribution state labeling matrix and the pressure spectrum feature matrix, a three-state division operation of a state space normal pressure state, a pressure mutation state and a sensor failure state is performed to obtain a state vector at the current moment; Perform parameter update calculation on the state transfer matrix of the three-state Markov model according to the state vector at the current moment to obtain a dynamic state transfer probability matrix; Performing maximum likelihood probability calculation on the dynamic state transfer probability matrix to obtain a state prediction vector at the next moment, and calculating the state probability distribution based on the state prediction vector to obtain a pressure state transfer probability matrix; Performing a time series matching operation on the pressure state transition probability matrix and the historical state sequence to obtain a state evolution feature vector, and performing state response modeling based on the state evolution feature vector to obtain a state response feature matrix; Performing pressure response pattern analysis according to the state response feature matrix and the pressure space distribution feature matrix to obtain a pressure response pattern matrix, and performing response feature extraction based on the pressure response pattern matrix to obtain a pressure response feature vector; Performing feature fusion operation on the pressure response feature vector and the state evolution feature vector to obtain a pressure state prediction matrix, and performing response feature mapping based on the pressure state prediction matrix to obtain an initial response prediction matrix; The initial response prediction matrix and the pressure time domain feature matrix are optimized with time and space feature constraints to obtain a pressure response prediction matrix.

5. The pressure control method for accurately adjusting air pressure intensity according to claim 1, characterized in that: The three-level control system including a basic controller, a disturbance compensator and a safety protector is constructed based on the PID control parameter matrix and the pressure safety limit matrix, and outputs a gas pressure regulating drive signal, including: Based on the PID control parameter matrix, a basic controller with proportional, integral and differential actions is constructed to obtain a basic control action matrix, and a control action amount is calculated according to the basic control action matrix to obtain a basic control amount; Performing interference component identification calculation on the basic control amount to obtain a matching interference component and a mismatching interference component, and constructing a finite time interference observer according to the matching interference component and the mismatching interference component to obtain an interference compensation amount; Based on the pressure safety limit matrix, a pressure amplitude limiting protection unit is constructed to obtain a safety threshold trigger function, and a switching function is designed according to the safety threshold trigger function to obtain a safety protection amount; Performing weight distribution calculation on the basic control amount, the interference compensation amount and the safety protection amount to obtain a weight coefficient matrix, and performing control amount combination operation according to the weight coefficient matrix to obtain a comprehensive control amount; Inputting the comprehensive control amount into the controller execution unit for signal modulation operation to obtain a control action signal, and performing drive voltage conversion based on the control action signal to obtain an actuator drive signal; Performing a pneumatic system response analysis on the actuator drive signal to obtain an air pressure response characteristic matrix, and performing a dynamic compensation calculation based on the air pressure response characteristic matrix to obtain an air pressure compensation signal; The air pressure compensation signal is superimposed on the actuator drive signal to obtain an initial air pressure regulation signal, and the initial air pressure regulation signal is dynamically corrected to obtain a compensated air pressure regulation signal; A drive signal conversion process is performed based on the compensated air pressure regulation signal to obtain an air pressure regulation drive signal.

6. The pressure control method for accurately adjusting air pressure intensity according to claim 5, characterized in that: The pressure control method for accurately adjusting the air pressure intensity also includes: Based on the air pressure regulating driving signal, pressure response data collection and dual-path difference calculation are performed on the system output to obtain a pressure tracking error sequence; Performing time segmentation processing and multi-scale statistical feature extraction based on the pressure tracking error sequence to obtain a multi-scale error statistical feature vector; Based on the pressure tracking error sequence, high-frequency sampling is performed on the driving voltage and driving current of the actuator and dual-frequency energy consumption accumulation is calculated to obtain a dual-frequency system energy consumption characteristic vector; Constructing a dual performance evaluation matrix according to the multi-scale error statistical feature vector and the dual-frequency system energy consumption feature vector, and performing multi-threshold classification judgment on the dual performance evaluation matrix to obtain a classification performance evaluation result; Based on the hierarchical performance evaluation result, the state transfer matrix of the three-state Markov model is fine-tuned and calculated and hierarchically and progressively updated to obtain a multi-level updated state transfer matrix; Inputting the multi-stage updated state transfer matrix into the three-state Markov model, performing the first state prediction optimization calculation and multi-time domain pressure response modeling, and obtaining a multi-time domain optimized pressure response matrix; Performing deviation analysis and row-level model parameter correction on the multi-time-domain optimized pressure response matrix and the actual pressure response to obtain a multi-level model correction parameter matrix; The multi-level model correction parameter matrix and the multi-level updated state transfer matrix are subjected to a multi-level parameter progressive fusion operation to obtain a target state transfer parameter matrix.

7. A pressure control device for accurately adjusting air pressure intensity, characterized in that: A pressure control method for accurately adjusting air pressure intensity according to any one of claims 1 to 6, wherein the pressure control device for accurately adjusting air pressure intensity comprises: A conversion module is used to perform digital conversion and temperature and humidity compensation on the original pressure signal collected by the pressure sensor array to obtain a pressure data matrix and a pressure gradient matrix; A calculation module, used to calculate the pressure distribution centroid and the pressure fluctuation spectrum based on the pressure data matrix and the pressure gradient matrix, and obtain a pressure spatial distribution feature matrix and a pressure time domain feature matrix; An identification module, used for inputting the pressure spatial distribution feature matrix and the pressure time domain feature matrix into a three-state Markov model for dynamic state identification, and obtaining a pressure state transition probability matrix and a pressure response prediction matrix; A solution module, used for constructing an objective function including a pressure tracking error term and an organizational safety constraint term according to the pressure state transition probability matrix and the pressure response prediction matrix, and solving to obtain a PID control parameter matrix and a pressure safety limit matrix; The output module is used to construct a three-level control system including a basic controller, a disturbance compensator and a safety protector based on the PID control parameter matrix and the pressure safety limit matrix, and output an air pressure regulation drive signal.

8. A pressure control device for accurately adjusting air pressure intensity, characterized in that: The pressure control device for accurately adjusting the air pressure intensity comprises: a memory and at least one processor, wherein the memory stores instructions; The at least one processor calls the instructions in the memory so that the pressure control device for accurately adjusting the air pressure intensity executes the pressure control method for accurately adjusting the air pressure intensity as described in any one of claims 1-6.

9. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by the processor, the pressure control method for accurately adjusting the air pressure intensity as described in any one of claims 1-6 is implemented.

Images