A high-pressure water pump pressure compensation method based on real-time data analysis

Abstract

The present invention relates to the field of data processing technology, and in particular to a high-pressure water pump pressure compensation method based on real-time data analysis. The present invention establishes a prediction model for input water pressure changes based on historical water pressure data, which can effectively capture long-term dependencies in time series data, thereby improving the prediction accuracy of future water pressure change trends. This high-precision prediction capability enables the high-pressure water pump control system to make adjustments in advance to achieve more accurate pressure compensation. The drive motor speed and mechanical valve opening can be dynamically adjusted according to the error signal between the real-time input water pressure data and the preset pressure target value, which can ensure that the actual output water pressure is always close to the target value, thereby achieving efficient closed-loop control.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a high-pressure water pump pressure compensation method based on real-time data analysis. Background Art

[0002] At present, the pressure compensation method of high-pressure water pump mainly relies on a variety of control technologies and methods to ensure that the system can operate stably and efficiently. Common high-pressure water pump pressure compensation methods include:

[0003] 1.PID controller:

[0004] The proportional-integral-derivative (PID) controller is the most commonly used controller in industrial control. It controls the output by adjusting the proportional term (P), integral term (I), and derivative term (D) to achieve tracking of the target value.

[0005] 2. Mechanical valve control:

[0006] The water flow rate is controlled by adjusting the opening of the mechanical valve, thereby indirectly controlling the output pressure of the pump.

[0007] However, the above existing high-pressure water pump pressure control methods have limitations in real-time response and closed-loop control. Traditional controllers (such as fixed-parameter PID controllers) are difficult to dynamically adjust the drive motor speed and mechanical valve opening based on the error signal between the real-time input water pressure data and the preset pressure target value. This results in a large deviation between the actual output water pressure and the target value.

[0008] Therefore, there is a need for a high-pressure water pump pressure compensation method based on real-time data analysis that can solve the above problems. Summary of the invention

[0009] The present invention provides a high-pressure water pump pressure compensation method based on real-time data analysis. The present invention establishes a prediction model for input water pressure changes based on historical water pressure data, which can effectively capture long-term dependencies in time series data, thereby improving the prediction accuracy of future water pressure change trends. This high-precision prediction capability enables the high-pressure water pump control system to make adjustments in advance to achieve more accurate pressure compensation. The drive motor speed and mechanical valve opening can be dynamically adjusted according to the error signal between the real-time input water pressure data and the preset pressure target value, which can ensure that the actual output water pressure is always close to the target value, thereby achieving efficient closed-loop control.

[0010] The technical solution adopted by the present invention to solve the above technical problems is: a high-pressure water pump pressure compensation method based on real-time data analysis, comprising the following steps:

[0011] Step S1: installing at least one pressure sensor at the inlet of the high-pressure water pump for real-time detection and recording of input water pressure data;

[0012] Step S2: performing digital signal processing on the input water pressure data in a data processing unit of the high-pressure water pump control system, and using a Kalman filter method to eliminate noise interference in the data;

[0013] Step S3: using a long short-term memory network algorithm to perform pattern recognition on the processed input water pressure data, and establishing a prediction model for input water pressure changes based on historical water pressure data;

[0014] Step S4: using an adaptive PID controller and combining it with a fuzzy logic control method to adjust PID control parameters to respond to changes in input water pressure in real time;

[0015] Step S5: installing a frequency converter on the driving motor of the high-pressure water pump, and installing an adjustable mechanical valve at the outlet of the high-pressure water pump;

[0016] Step S6: dynamically adjusting the speed of the driving motor using the artificial intelligence model, and adjusting the opening of the mechanical valve according to the preset pressure target value to form a comprehensive pressure compensation mechanism;

[0017] Step S7: Applying the differential evolution algorithm to optimize the control parameters of the high-pressure water pump pressure control system.

[0018] Furthermore, the step S2 of performing digital signal processing on the input water pressure data in the data processing unit of the high-pressure water pump control system includes:

[0019] Step S2-1: The data processing unit of the high-pressure water pump control system receives real-time input water pressure data from the high-precision pressure sensor;

[0020] Step S2-2: sampling and quantizing the input water pressure data through the data processing unit of the high-pressure water pump control system, and converting the analog signal into a digital signal;

[0021] Step S2-3: preprocessing the digitized input water pressure data, including removing obvious outliers and performing preliminary smoothing;

[0022] Step S2-4: Time synchronization of the preprocessed data to ensure that the timestamps of the data are consistent;

[0023] Step S2-5: standardize or normalize the time-synchronized data so that they are within the same numerical range;

[0024] Step S2-6: The data processing unit of the high-pressure water pump control system stores the processed input water pressure data in the data buffer.

[0025] Furthermore, the step S2 of using the Kalman filter method to eliminate noise interference in the data includes:

[0026] Step S2-7: Initializing the state vector and covariance matrix of the Kalman filter through the data processing unit of the high-pressure water pump control system, and setting the initial estimation value and error covariance;

[0027] Step S2-8: The data processing unit of the high-pressure water pump control system predicts the state vector and error covariance at the current moment according to the system model of the dynamic change of the input water pressure of the high-pressure water pump;

[0028] Step S2-9: obtaining the actual measurement value at the current moment through the data processing unit of the high-pressure water pump control system, and calculating the measurement residual, that is, the difference between the actual measurement value and the predicted value;

[0029] Step S2-10: calculating the Kalman gain by the data processing unit of the high-pressure water pump control system;

[0030] Step S2-11: The data processing unit of the high-pressure water pump control system updates the state vector and the error covariance, and combines the predicted value and the measured value to obtain an estimated value of the water pressure state at the current moment;

[0031] Step S2-12: The data processing unit of the high-pressure water pump control system outputs the input water pressure data after Kalman filtering for subsequent pattern recognition and control operations.

[0032] Furthermore, in step S3, the long short-term memory network algorithm is used to perform pattern recognition on the processed input water pressure data, and a prediction model of input water pressure changes is established based on the historical water pressure data, including:

[0033] Step S3-1: preparing a historical water pressure dataset;

[0034] Step S3-2: preprocessing the historical water pressure data set;

[0035] Step S3-3: Divide the preprocessed historical water pressure data into a training set and a validation set for training and evaluating the LSTM model;

[0036] Step S3-4: Build an LSTM neural network model and set the network structure parameters, including the number of layers, the number of neurons in each layer, and the activation function;

[0037] Step S3-5: Use the training set data to train the LSTM model and optimize the model parameters through the back propagation algorithm and gradient descent method;

[0038] Step S3-6: Use the validation set data to validate the trained LSTM model and evaluate the prediction performance of the model;

[0039] Step S3-7: Adjust the hyperparameters of the LSTM model according to the verification results;

[0040] Step S3-8: Use the trained LSTM model to perform pattern recognition on the processed input water pressure data to extract the time series features of the input water pressure data;

[0041] Step S3-9: Based on the historical water pressure data and the extracted time series features, a prediction model for input water pressure changes is established to predict the future input water pressure change trend;

[0042] Step S3-10: Output the input water pressure data and its prediction results after being processed by the LSTM model for subsequent control operations.

[0043] Furthermore, the historical water pressure data set in step S3-1 includes the continuous input water pressure measurement value of the high-pressure water pump recorded every half hour within one month, the temperature and humidity environmental condition data of the high-pressure water pump, the operation time of the high-pressure water pump and the flow data.

[0044] Furthermore, in step S4, the adaptive PID controller is used, and the PID control parameters are adjusted in combination with the fuzzy logic control method to respond to the change of the input water pressure in real time, including:

[0045] Step S4-1: Initialize the initial parameters of the adaptive PID controller;

[0046] Step S4-2: receiving the processed real-time input water pressure data, comparing it with the preset target water pressure value, and calculating the error signal at the current moment;

[0047] Step S4-3: input the error signal into the fuzzy logic controller, and the fuzzy logic controller performs fuzzy processing on the error signal and its rate of change according to a predefined fuzzy rule base;

[0048] Step S4-4: the fuzzy logic controller performs reasoning based on the fuzzy rule base to generate fuzzy outputs for adjusting PID control parameters;

[0049] Step S4-5: Defuzzify the fuzzy output to obtain specific PID control parameter adjustment values;

[0050] Step S4-6: updating the parameters of the adaptive PID controller according to the adjustment amount;

[0051] Step S4-7: Calculate a control signal using the updated PID control parameters, where the control signal is used to adjust the speed of the driving motor of the high-pressure water pump or the opening of the mechanical valve;

[0052] Step S4-8: sending a control signal to the actuator to make the actual water pressure close to the target water pressure;

[0053] Step S4-9: Continue to monitor the change of input water pressure and repeat steps S4-2 to S4-8.

[0054] Furthermore, in step S6, the driving motor speed is dynamically adjusted using the artificial intelligence model, and the opening of the mechanical valve is adjusted according to the preset pressure target value to form a comprehensive pressure compensation mechanism including:

[0055] Step S6-1: Build and train an artificial intelligence model;

[0056] Step S6-2: The artificial intelligence model receives the real-time input water pressure data after digital signal processing, compares it with the preset pressure target value, and calculates the error signal at the current moment;

[0057] Step S6-3: inputting the error signal into the artificial intelligence model;

[0058] Step S6-4: the artificial intelligence model generates an adjustment amount of the driving motor speed and an adjustment amount of the mechanical valve opening according to the input error signal;

[0059] Step S6-5: updating the speed setting of the drive motor according to the adjustment amount generated by the artificial intelligence model, and adjusting the actual speed of the motor through the frequency converter;

[0060] Step S6-6: updating the opening setting of the mechanical valve according to the adjustment amount generated by the artificial intelligence model, and adjusting the actual opening of the mechanical valve through the actuator;

[0061] Step S6-7: Monitor the changes in the actual output water pressure and feed back the new water pressure data to the artificial intelligence model.

[0062] Furthermore, in step S7, the differential evolution algorithm is used to optimize the control parameters of the high-pressure water pump pressure control system, including:

[0063] Step S7-1: Initialize the parameters of the differential evolution algorithm;

[0064] Step S7-2: Generate an initial population, where each individual represents a set of control parameters of a high-pressure water pump pressure control system;

[0065] Step S7-3: Evaluate each individual by applying the control parameters to the high-pressure water pump pressure control system and running simulation or actual testing to calculate the corresponding performance indicators;

[0066] Step S7-4: Select individuals in the initial population to perform mutation operations and generate mutation vectors;

[0067] Step S7-5: Perform a crossover operation on the mutation vector to generate a test vector;

[0068] Step S7-6: Evaluate the test vector and calculate its performance index;

[0069] Step S7-7: Perform a selection operation. If the performance index of the test vector is better than that of the target vector, the target vector is replaced by the test vector; otherwise, the target vector is retained.

[0070] Step S7-8: Check whether the maximum number of iterations has been reached. If not, return to step S7-4 to continue the next iteration. If reached, end the optimization process.

[0071] Step S7-9: Output the control parameter combination, and apply the control parameter combination to the high-pressure water pump control system.

[0072] Furthermore, the parameters of the differential evolution algorithm in step S7-1 include the initial population size, the mutation factor F, the crossover probability CR and the maximum number of iterations.

[0073] Furthermore, the mutation operation in step S7-4 is performed by randomly selecting three different individuals and performing the mutation according to the formula Perform mutation operation, where V i represents the generated mutation vector, X r1 represents a randomly selected target vector from the initial population; where X r2 With X r3 represents two other randomly selected different vectors in the initial population; where F represents the mutation factor, which is a real number between 0 and 1 and is used to control the amplitude of the mutation.

[0074] The advantages of the present invention are:

[0075] 1. The present invention establishes a prediction model for input water pressure changes based on historical water pressure data, which can effectively capture the long-term dependencies in time series data, thereby improving the prediction accuracy of future water pressure change trends. This high-precision prediction capability enables the high-pressure water pump control system to make adjustments in advance to achieve more accurate pressure compensation. It can dynamically adjust the drive motor speed and mechanical valve opening according to the error signal between the real-time input water pressure data and the preset pressure target value, ensuring that the actual output water pressure is always close to the target value, thereby achieving efficient closed-loop control.

[0076] 2. The present invention forms a comprehensive pressure compensation mechanism through the synergy of digital signal processing, Kalman filtering, LSTM prediction, adaptive PID control, fuzzy logic control and differential evolution optimization, so that the high-pressure water pump control system can achieve efficient and reliable pressure control. DETAILED DESCRIPTION

[0077] The technical solution of the present invention is described clearly and completely below. Obviously, the described embodiments are only 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.

[0078] Embodiment 1: The present invention provides a high-pressure water pump pressure compensation method based on real-time data analysis, comprising the following steps:

[0079] Step S1: at least one pressure sensor is installed at the inlet of the high-pressure water pump for real-time detection and recording of input water pressure data. The pressure sensor is used to detect and record input water pressure data in real-time.

[0080] Step S2: performing digital signal processing on the input water pressure data in the data processing unit of the high-pressure water pump control system, and using a Kalman filter method to eliminate noise interference in the data, specifically, including:

[0081] Step S2-1: The data processing unit of the high-pressure water pump control system receives real-time input water pressure data from the high-precision pressure sensor. Specifically, the data processing unit of the high-pressure water pump control system receives real-time input water pressure data (these data are in the form of analog signals) from the high-precision pressure sensor through a sensor interface or a communication protocol (such as RS485, Modbus).

[0082] Step S2-2: The input water pressure data is sampled and quantized by the data processing unit of the high-pressure water pump control system, and the analog signal is converted into a digital signal. The data processing unit uses an analog-to-digital converter (ADC) to sample and quantize the analog signal and convert it into a digital signal. The sampling frequency is set according to system requirements to ensure the accuracy and timeliness of the data. For example, the sampling frequency can be set to 100 times per second.

[0083] Step S2-3: Preprocess the digitized input water pressure data, including removing obvious outliers and performing preliminary smoothing. The preprocessing includes removing obvious outliers (such as data points that exceed a reasonable range) and performing preliminary smoothing. Simple statistical methods (such as mean filtering or median filtering) can be used to remove outliers, and a low-pass filter can be used for smoothing to reduce the impact of high-frequency noise.

[0084] Step S2-4: Time synchronization of the pre-processed data to ensure that the timestamps of the data are consistent for subsequent processing. The Network Time Protocol (NTP) can be used to calibrate the time of each sensor and data processing unit.

[0085] Step S2-5: Standardize or normalize the time-synchronized data so that they are within the same numerical range. Specifically, the present invention uses a standardization method including Z-score standardization (converting the data to a standard normal distribution with a mean of 0 and a standard deviation of 1) or Min-Max normalization (scaling the data to the interval [0, 1]) to process the data.

[0086] Step S2-6: The data processing unit of the high-pressure water pump control system stores the processed input water pressure data in the data buffer for subsequent processing and analysis.

[0087] Step S2-7: Initialize the state vector and covariance matrix of the Kalman filter through the data processing unit of the high-pressure water pump control system, set the initial estimate and error covariance, where the state vector includes the water pressure estimate at the current moment and related state variables (temperature and humidity environment of the high-pressure water pump), and the covariance matrix represents the uncertainty of the state estimate;

[0088] Step S2-8: The data processing unit of the high-pressure water pump control system predicts the state vector and error covariance at the current moment according to the system model of the dynamic change of the input water pressure of the high-pressure water pump; the system model includes a state transfer equation and an observation equation, which describes the change relationship of the input water pressure over time, wherein the state transfer equation is used to predict the state vector at the next moment, and the observation equation is used to map the state vector to the observation space. Through these equations, the predicted value of the state vector at the current moment and the corresponding error covariance are calculated. Specifically, the state transfer equation describes the evolution process of the system state from one moment to the next. Assume that the state vector of the high-pressure water pump control system is X k Contains the estimated water pressure at the current moment and related state variables (ambient temperature and humidity of the high-pressure water pump). The state transfer equation is expressed as ,in represents the predicted value of the state vector at time k, A represents the state transfer matrix, which describes how the state variables change over time, Indicates time The estimated value of the state vector is , B is the control input matrix, which describes the influence of the control input on the state, Indicates at time The control input is is the process noise. The observation equation describes how to obtain the observation value from the state vector. Assume that the observation vector Zk contains the actual measured water pressure value and sensor data. The observation equation is expressed as , where Z k is the observation vector at time k, and H is the observation matrix, which describes how the state vector is mapped to the observation space. k is the state vector at time k. V kis the observation noise.

[0089] Step S2-9: The actual measurement value at the current moment is obtained through the data processing unit of the high-pressure water pump control system, and the measurement residual, that is, the difference between the actual measurement value and the predicted value, is calculated. The actual measurement value at the current moment is obtained from the sensor, and the measurement residual, that is, the difference between the actual measurement value and the predicted value, is calculated. The measurement residual reflects the deviation between the predicted value and the actual value.

[0090] Step S2-10: The Kalman gain is calculated by the data processing unit of the high-pressure water pump control system. The gain determines the weight of the predicted value and the measured value in the final estimation. Specifically, the Kalman gain formula is Calculate the Kalman gain, where is the prior error covariance, H is the observation matrix, and R is the covariance matrix of the measurement noise.

[0091] Step S2-11: The data processing unit of the high-pressure water pump control system updates the state vector and error covariance, and combines the predicted value and the measured value to obtain the best estimate of the water pressure state at the current moment.

[0092] Step S2-12: The data processing unit of the high-pressure water pump control system outputs the input water pressure data after Kalman filtering for subsequent pattern recognition and control operations. Since these data have been free of noise interference and are optimal estimates, the reliability and accuracy of the data are improved.

[0093] Step S3: Use the long short-term memory network (LSTM) algorithm to perform pattern recognition on the processed input water pressure data, and establish a prediction model for input water pressure changes based on historical water pressure data, specifically including:

[0094] Step S3-1: Prepare a historical water pressure data set, which includes the continuous input water pressure measurement value of the high-pressure water pump recorded every half hour within one month, the temperature and humidity environmental condition data of the high-pressure water pump, the operation time of the high-pressure water pump, and the flow data. These data can be obtained from the historical records of the high-pressure water pump control system and stored in a database or file system.

[0095] Step S3-2: Preprocess the historical water pressure data set, including data cleaning, missing value processing, outlier processing, and data standardization or normalization. Specifically, data cleaning: remove duplicate records and invalid data. Missing value processing: use interpolation methods (such as linear interpolation, forward filling, etc.) to fill missing values. Outlier processing: detect and process outliers through statistical methods (such as Z-score, IQR). Data standardization or normalization: convert the data to the same numerical range to facilitate neural network training. Commonly used standardization methods include Z-score standardization (converting data to a standard normal distribution with a mean of 0 and a standard deviation of 1) or Min-Max normalization (scaling the data to the interval [0, 1]).

[0096] Step S3-3: Divide the preprocessed historical water pressure data into a training set and a validation set for training and evaluating the LSTM model. Specifically, divide the historical water pressure data set into a training set and a validation set at a ratio of 70%:30% or 80%:20%. The training set is used to train the LSTM model, and the validation set is used to evaluate the generalization ability of the model (the chronological order of the data should be ensured during the division to avoid future data appearing in the training set);

[0097] Step S3-4: Build an LSTM neural network model and set the network structure parameters, including the number of layers, the number of neurons in each layer, and the activation function. Number of layers: Select the appropriate number of layers according to the complexity of the problem, such as a single-layer LSTM or a stacked multi-layer LSTM. Number of neurons in each layer: The number of neurons in each layer can be set based on experience, starting from a smaller number and gradually increasing until the performance is no longer significantly improved. Activation function: The LSTM unit uses tanh and sigmoid activation functions internally, and the output layer uses a linear activation function.

[0098] Step S3-5: Use the training set data to train the LSTM model, optimize the model parameters through the back propagation algorithm and the gradient descent method, and minimize the prediction error. The back propagation algorithm updates the model parameters by calculating the gradient of the loss function to the model parameters. The gradient descent method uses an optimization algorithm (such as Adam, RMSprop) to minimize the prediction error. The initial learning rate can be set to 0.001 and adjusted according to the training process. The batch size is selected as 32, 64 or 128 to balance the training speed and memory consumption.

[0099] Step S3-6: Use the validation set data to validate the trained LSTM model and evaluate the prediction performance of the model, such as the mean square error (MSE) and mean absolute error (MAE) indicators.

[0100] Step S3-7: Adjust the hyperparameters of the LSTM model according to the verification results, such as adjusting the learning rate, batch size, and number of iterations, to further optimize the model performance.

[0101] Step S3-8: Use the trained LSTM model to perform pattern recognition on the processed input water pressure data, extract the time series features of the input water pressure data, input the processed input water pressure data into the trained LSTM model, and extract the time series features of the input water pressure data.

[0102] Step S3-9: Based on the historical water pressure data and the extracted time series features, a prediction model for input water pressure changes is established to predict future input water pressure change trends.

[0103] Step S3-10: Output the input water pressure data and its prediction results after being processed by the LSTM model for subsequent control operations. Output the input water pressure data and its prediction results after being processed by the LSTM model for subsequent control operations.

[0104] Step S4: using an adaptive PID controller and combining it with a fuzzy logic control method to adjust the PID control parameters to respond to changes in the input water pressure in real time, specifically including:

[0105] Step S4-1: Initializing the initial parameters of the adaptive PID controller, specifically, including the proportional coefficient (Kp), the integral coefficient (Ki) and the differential coefficient (Kd);

[0106] Step S4-2: receiving the processed real-time input water pressure data, comparing it with the preset target water pressure value, and calculating the error signal at the current moment;

[0107] Step S4-3: input the error signal into the fuzzy logic controller, and the fuzzy logic controller performs fuzzy processing on the error signal and its rate of change according to a predefined fuzzy rule base;

[0108] Step S4-4: the fuzzy logic controller performs reasoning based on the fuzzy rule base to generate fuzzy outputs for adjusting PID control parameters;

[0109] Step S4-5: Defuzzify the fuzzy output to obtain specific PID control parameter adjustment values;

[0110] Step S4-6: updating the parameters (Kp, Ki, Kd) of the adaptive PID controller according to the adjustment amount to adapt to the current system state and input water pressure changes;

[0111] Step S4-7: Calculate a control signal using the updated PID control parameters, where the control signal is used to adjust the speed of the driving motor of the high-pressure water pump or the opening of the mechanical valve;

[0112] Step S4-8: sending a control signal to an actuator (such as a frequency converter or a mechanical valve) to achieve real-time control of the high-pressure water pump so that the actual water pressure approaches the target water pressure;

[0113] Step S4-9: Continuously monitor the changes in the input water pressure and repeat steps S4-2 to S4-8 to ensure the real-time response and stability of the system.

[0114] Step S5: installing a frequency converter on the driving motor of the high-pressure water pump, and installing an adjustable mechanical valve at the outlet of the high-pressure water pump;

[0115] Step S6: Dynamically adjust the speed of the drive motor using the artificial intelligence model trained based on historical data, and adjust the opening of the mechanical valve according to the preset pressure target value to form a comprehensive pressure compensation mechanism to change the water flow resistance, thereby affecting the output pressure, specifically including:

[0116] Step S6-1: Build and train an artificial intelligence model; the artificial intelligence model is trained based on a historical water pressure data set to generate an artificial intelligence model that can dynamically adjust the drive motor speed and the mechanical valve opening.

[0117] Step S6-2: The artificial intelligence model receives the real-time input water pressure data after digital signal processing, compares it with the preset pressure target value, and calculates the error signal at the current moment; specifically, the real-time input water pressure data after digital signal processing is obtained from the high-pressure water pump control system, a desired pressure target value is preset, the real-time input water pressure data is compared with the preset pressure target value, and the error signal at the current moment is calculated. Where ptarget represents the preset pressure target value, and pactual,k is the actual input water pressure value at the current moment.

[0118] Step S6-3: The calculated error signal ek is used as input and passed to the trained artificial intelligence model;

[0119] Step S6-4: the artificial intelligence model generates an adjustment amount of the driving motor speed and an adjustment amount of the mechanical valve opening according to the input error signal;

[0120] Step S6-5: updating the speed setting of the drive motor according to the adjustment amount generated by the artificial intelligence model, and adjusting the actual speed of the motor through the frequency converter;

[0121] Step S6-6: updating the opening setting of the mechanical valve according to the adjustment amount generated by the artificial intelligence model, and adjusting the actual opening of the mechanical valve through an actuator (such as an electric or pneumatic actuator);

[0122] Step S6-7: Monitor the changes in the actual output water pressure and feed back the new water pressure data to the artificial intelligence model to achieve closed-loop control.

[0123] Step S7: Applying the differential evolution algorithm (DE) to optimize the control parameters of the high-pressure water pump pressure control system, specifically, including:

[0124] Step S7-1: Initialize the parameters of the differential evolution algorithm, where the parameters of the differential evolution algorithm include population size, mutation factor F, crossover probability CR and maximum number of iterations;

[0125] Step S7-2: Generate an initial population, where each individual represents a set of control parameters of the high-pressure water pump pressure control system, such as a proportional coefficient Kp, an integral coefficient Ki, and a differential coefficient Kd of a PID controller;

[0126] Step S7-3: Evaluate each individual by applying the set of control parameters to the high-pressure water pump pressure control system and running simulation or actual test to calculate the corresponding performance indicators (including mean square error MSE, integrated absolute error IAE);

[0127] Step S7-4: Select individuals in the population for mutation operation to generate mutation vectors. The mutation operation is performed by randomly selecting three different individuals and performing mutation operations according to the formula Perform mutation operation, where V i represents the generated mutation vector, X r1 represents a randomly selected target vector in the population; where X r2 With X r3 represents two other randomly selected different vectors in the population; where F represents the mutation factor, which is a real number between 0 and 1 and is used to control the amplitude of the mutation.

[0128] Step S7-5: Perform a crossover operation on the mutation vector to generate a test vector; specifically, the crossover operation compares the target vector element by element and determines whether to adopt the element in the mutation vector based on the crossover probability CR.

[0129] Step S7-6: Evaluate the test vector and calculate its performance index;

[0130] Step S7-7: Perform a selection operation. If the performance index of the test vector is better than that of the target vector, the target vector is replaced by the test vector; otherwise, the target vector is retained.

[0131] Step S7-8: Check whether the maximum number of iterations has been reached. If not, return to step S7-4 to continue the next iteration. If reached, end the optimization process.

[0132] Step S7-9: Output the control parameter combination, and apply the control parameter combination to the high-pressure water pump control system to improve the performance of the water pump pressure control system.

[0133] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A high-pressure water pump pressure compensation method based on real-time data analysis, characterized in that: The following steps are involved: Step S1: installing at least one pressure sensor at the inlet of the high-pressure water pump for real-time detection and recording of input water pressure data; Step S2: performing digital signal processing on the input water pressure data in a data processing unit of the high-pressure water pump control system, and using a Kalman filter method to eliminate noise interference in the data; Step S3: using a long short-term memory network algorithm to perform pattern recognition on the processed input water pressure data, and establishing a prediction model for input water pressure changes based on historical water pressure data; Step S3-1: prepare a historical water pressure data set, which includes the continuous input water pressure measurement value of the high-pressure water pump recorded every half hour within one month, the temperature and humidity environmental condition data of the high-pressure water pump, the operation time of the high-pressure water pump, and the flow data; Step S3-2: preprocessing the historical water pressure data set; Step S3-3: Divide the preprocessed historical water pressure data into a training set and a validation set for training and evaluating the LSTM model; Step S3-4: Build an LSTM neural network model and set the network structure parameters, including the number of layers, the number of neurons in each layer, and the activation function; Step S3-5: Use the training set data to train the LSTM model and optimize the model parameters through the back propagation algorithm and gradient descent method; Step S3-6: Use the validation set data to validate the trained LSTM model and evaluate the prediction performance of the model; Step S3-7: Adjust the hyperparameters of the LSTM model according to the verification results; Step S3-8: Use the trained LSTM model to perform pattern recognition on the processed input water pressure data to extract the time series features of the input water pressure data; Step S3-9: Based on the historical water pressure data and the extracted time series features, a prediction model for input water pressure changes is established to predict the future input water pressure change trend; Step S3-10: Output the input water pressure data and its prediction results after being processed by the LSTM model for subsequent control operations; Step S4: using an adaptive PID controller and combining it with a fuzzy logic control method to adjust PID control parameters to respond to changes in input water pressure in real time; Step S5: installing a frequency converter on the driving motor of the high-pressure water pump, and installing an adjustable mechanical valve at the outlet of the high-pressure water pump; Step S6: dynamically adjusting the speed of the driving motor using the artificial intelligence model, and adjusting the opening of the mechanical valve according to the preset pressure target value to form a comprehensive pressure compensation mechanism; Step S7: Applying the differential evolution algorithm to optimize the control parameters of the high-pressure water pump pressure control system.

2. A high-pressure water pump pressure compensation method based on real-time data analysis according to claim 1, characterized in that: The step S2 of performing digital signal processing on the input water pressure data in the data processing unit of the high-pressure water pump control system comprises: Step S2-1: The data processing unit of the high-pressure water pump control system receives real-time input water pressure data from the high-precision pressure sensor; Step S2-2: sampling and quantizing the input water pressure data through the data processing unit of the high-pressure water pump control system, and converting the analog signal into a digital signal; Step S2-3: preprocessing the digitized input water pressure data, including removing obvious outliers and performing preliminary smoothing; Step S2-4: Time synchronization of the preprocessed data to ensure that the timestamps of the data are consistent; Step S2-5: standardize or normalize the time-synchronized data so that they are within the same numerical range; Step S2-6: The data processing unit of the high-pressure water pump control system stores the processed input water pressure data in the data buffer.

3. A high-pressure water pump pressure compensation method based on real-time data analysis according to claim 1, characterized in that: The step S2 of using the Kalman filter method to eliminate noise interference in the data includes: Step S2-7: Initializing the state vector and covariance matrix of the Kalman filter through the data processing unit of the high-pressure water pump control system, and setting the initial estimation value and error covariance; Step S2-8: The data processing unit of the high-pressure water pump control system predicts the state vector and error covariance at the current moment according to the system model of the dynamic change of the input water pressure of the high-pressure water pump; Step S2-9: obtaining the actual measurement value at the current moment through the data processing unit of the high-pressure water pump control system, and calculating the measurement residual, that is, the difference between the actual measurement value and the predicted value; Step S2-10: calculating the Kalman gain by the data processing unit of the high-pressure water pump control system; Step S2-11: The data processing unit of the high-pressure water pump control system updates the state vector and the error covariance, and combines the predicted value and the measured value to obtain an estimated value of the water pressure state at the current moment; Step S2-12: The data processing unit of the high-pressure water pump control system outputs the input water pressure data after Kalman filtering for subsequent pattern recognition and control operations.

4. A high-pressure water pump pressure compensation method based on real-time data analysis according to claim 1, characterized in that: In step S4, the adaptive PID controller is used, and the PID control parameters are adjusted in combination with the fuzzy logic control method to respond to the change of the input water pressure in real time, including: Step S4-1: Initialize the initial parameters of the adaptive PID controller; Step S4-2: receiving the processed real-time input water pressure data, comparing it with the preset target water pressure value, and calculating the error signal at the current moment; Step S4-3: input the error signal into the fuzzy logic controller, and the fuzzy logic controller performs fuzzy processing on the error signal and its rate of change according to a predefined fuzzy rule base; Step S4-4: the fuzzy logic controller performs reasoning based on the fuzzy rule base to generate fuzzy outputs for adjusting PID control parameters; Step S4-5: Defuzzify the fuzzy output to obtain specific PID control parameter adjustment values; Step S4-6: updating the parameters of the adaptive PID controller according to the adjustment amount; Step S4-7: Calculate a control signal using the updated PID control parameters, where the control signal is used to adjust the speed of the driving motor of the high-pressure water pump or the opening of the mechanical valve; Step S4-8: sending a control signal to the actuator to make the actual water pressure close to the target water pressure; Step S4-9: Continue to monitor the change of input water pressure and repeat steps S4-2 to S4-8.

5. A high-pressure water pump pressure compensation method based on real-time data analysis according to claim 1, characterized in that: In step S6, the artificial intelligence model is used to dynamically adjust the speed of the driving motor, and the opening of the mechanical valve is adjusted according to the preset pressure target value to form a comprehensive pressure compensation mechanism, including: Step S6-1: Build and train an artificial intelligence model; Step S6-2: The artificial intelligence model receives the real-time input water pressure data after digital signal processing, compares it with the preset pressure target value, and calculates the error signal at the current moment; Step S6-3: inputting the error signal into the artificial intelligence model; Step S6-4: the artificial intelligence model generates an adjustment amount of the driving motor speed and an adjustment amount of the mechanical valve opening according to the input error signal; Step S6-5: updating the speed setting of the drive motor according to the adjustment amount generated by the artificial intelligence model, and adjusting the actual speed of the motor through the frequency converter; Step S6-6: updating the opening setting of the mechanical valve according to the adjustment amount generated by the artificial intelligence model, and adjusting the actual opening of the mechanical valve through the actuator; Step S6-7: Monitor the changes in the actual output water pressure and feed back the new water pressure data to the artificial intelligence model.

6. A high-pressure water pump pressure compensation method based on real-time data analysis according to claim 1, characterized in that: The step S7 of applying the differential evolution algorithm to optimize the control parameters of the high-pressure water pump pressure control system includes: Step S7-1: Initialize the parameters of the differential evolution algorithm; Step S7-2: Generate an initial population, where each individual represents a set of control parameters of a high-pressure water pump pressure control system; Step S7-3: Evaluate each individual by applying the control parameters to the high-pressure water pump pressure control system and running simulation or actual testing to calculate the corresponding performance indicators; Step S7-4: Select individuals in the initial population to perform mutation operations and generate mutation vectors; Step S7-5: Perform a crossover operation on the mutation vector to generate a test vector; Step S7-6: Evaluate the test vector and calculate its performance index; Step S7-7: Perform a selection operation. If the performance index of the test vector is better than that of the target vector, the target vector is replaced by the test vector; otherwise, the target vector is retained. Step S7-8: Check whether the maximum number of iterations has been reached. If not, return to step S7-4 to continue the next iteration. If reached, end the optimization process. Step S7-9: Output the control parameter combination, and apply the control parameter combination to the high-pressure water pump control system.

7. A high-pressure water pump pressure compensation method based on real-time data analysis according to claim 6, characterized in that: The parameters of the differential evolution algorithm in step S7-1 include the initial population size, the mutation factor F, the crossover probability CR and the maximum number of iterations.

8. A high-pressure water pump pressure compensation method based on real-time data analysis according to claim 6, characterized in that: The mutation operation in step S7-4 is performed by randomly selecting three different individuals and performing the mutation according to the formula Perform mutation operation, where V i represents the generated mutation vector, X r1 represents a randomly selected target vector from the initial population; where X r2 With X r3 represents two other randomly selected different vectors in the initial population; where F represents the mutation factor, which is a real number between 0 and 1 and is used to control the amplitude of the mutation.