Energy-saving operation method of hydraulic motor based on pressure detection energy compensation

Abstract

This invention discloses an energy-saving operation method for hydraulic motors based on pressure detection and energy replenishment, specifically relating to the field of energy-saving control. The method includes: firstly, collecting, filtering, and preprocessing multiple operating parameters of the hydraulic system and drive motor using time synchronization; then constructing a dimensionless characteristic function from multiple dimensions and achieving nonlinear fusion through adaptive weight allocation to obtain a comprehensive state coupling value; constructing a pressure advance prediction model based on historical data, and performing consistency verification and confidence analysis on the comprehensive state and predicted state; using the comprehensive state coupling value, pressure prediction value, and confidence value as inputs, employing fuzzy logic reasoning to output motor speed and accumulator valve group control commands, ultimately driving the actuator and transmitting the operating parameters back to form a closed loop; this invention improves the system's dynamic response and anti-interference capability, optimizes the motor operating range while ensuring pressure stability, and improves energy utilization, making it suitable for efficient and energy-saving operation of hydraulic systems under complex working conditions.

Description

Technical Field

[0001] This invention relates to the field of energy-saving control technology, and more specifically, to a method for energy-saving operation of a hydraulic motor based on pressure detection and energy replenishment. Background Technology

[0002] Hydraulic transmission systems, due to their high power density, fast response speed, and stable operation, are widely used in industrial fields such as engineering machinery, equipment manufacturing, and automated production lines. With the continuous advancement of industrial energy conservation and intelligent control technologies, hydraulic systems powered by electric motors are gradually replacing traditional power structures, becoming the mainstream development direction in the industry. Pressure detection and accumulator replenishment technologies, as important means of energy-saving regulation of hydraulic systems, achieve dynamic matching between the power source and load demand by monitoring operating parameters such as system pressure and flow.

[0003] Currently, hydraulic motor drive systems have formed a technical system combining multi-sensor information acquisition, closed-loop actuator regulation, energy recovery, and auxiliary power replenishment. Related control strategies are continuously iterating around optimizing system efficiency, suppressing pressure fluctuations, and improving operational reliability. In practical engineering applications, hydraulic systems often exhibit nonlinear, strongly coupled, and time-varying operational characteristics. Related control technologies are continuously developing towards multi-source information fusion, advanced prediction, and adaptive decision-making, providing sustained technical support and an application foundation for energy-saving operation of hydraulic motors.

[0004] However, it still has some drawbacks in practical use, such as: 1. Traditional hydraulic motor control often uses single pressure feedback regulation, without integrating multiple operating parameters, making it difficult to fully reflect the true operating status of the system. Under complex variable load conditions, it is prone to control lag or over-regulation, and cannot achieve precise matching of pressure and flow. The overall control accuracy and dynamic response performance are significantly limited. 2. Existing energy-saving control methods mostly rely on fixed parameters and empirical thresholds, without adaptive adjustment based on changes in operating conditions. They lack quantitative characterization of the coordinated operation of motors and accumulators, which can easily cause motors to idle or frequently start and stop under low flow and intermittent operating conditions, resulting in limited improvement in energy utilization. 3. Most hydraulic control schemes do not introduce data validity verification mechanisms. Interference such as sensor noise and sudden changes in operating conditions can easily lead to incorrect output of control commands. The system has insufficient robustness and is prone to problems such as large pressure fluctuations and low reliability in harsh industrial environments, making it difficult to meet the requirements for long-term stable operation. 4. Existing technologies mostly adopt post-compensation feedback control, which lacks the ability to predict system pressure trends in advance and cannot intervene in situations such as sudden load changes and energy fluctuations in advance. The control process has obvious lag, making it difficult to balance the dynamic response speed of the system with the goal of energy-saving operation, and the overall optimization space is limited. Summary of the Invention

[0005] In order to overcome the above-mentioned defects of the prior art, the present invention provides a hydraulic motor energy-saving operation method based on pressure detection and energy replenishment, which solves the problems mentioned in the background art through the following scheme.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a hydraulic motor energy-saving operation method based on pressure detection and energy replenishment, comprising: S1: Multi-source data acquisition and preprocessing: Synchronously acquire multiple operating parameters of the hydraulic system and drive motor, and sequentially complete filtering, time synchronization and buffer storage, and output preprocessed multi-source data; S2: Construct dimensionless characteristic functions: Based on the preprocessed multi-source data, construct dimensionless characteristic functions for each dimension from the perspectives of pressure-flow coupling, motor-accumulator coordination, energy conversion efficiency, accumulator response, temperature-leakage coupling, and power-flow hysteresis. S3: Adaptive Feature Coupling: Based on the dimensionless feature functions constructed in S2, multi-dimensional features are nonlinearly fused through statistical analysis and adaptive weight allocation to complete adaptive feature coupling and obtain a comprehensive state coupling value. S4: Predictive Model Construction: Combining historical operating data with system dynamic characteristics, construct a system pressure advance prediction model to obtain the pressure prediction state value; S5: Coupling value validity and confidence analysis: The consistency between the integrated state coupling value and the pressure prediction state value is verified, the validity and confidence analysis of the coupling value is completed, and the state confidence is obtained; S6: Adaptive control decision based on fuzzy logic: Taking the comprehensive state coupling value, pressure prediction state value and state confidence as input, the adaptive control decision is completed through fuzzy logic reasoning, and the control commands for motor speed and accumulator valve group are output. S7: Control command execution and closed-loop operation: Executes control commands to drive the corresponding actuators and collects operational feedback parameters in real time and sends them back to S1 to realize control command execution and closed-loop operation.

[0007] The technical effects and advantages of this invention are as follows: 1. This solution achieves standardized quantification of multiple dimensions such as pressure-flow coupling, electromechanical coordination, efficiency, and leakage by acquiring multi-source data and constructing dimensionless features. It comprehensively characterizes the system's operating characteristics and, compared with traditional single-parameter control, can more accurately reflect the system's true operating conditions, providing a reliable basis for subsequent decision-making. 2. The scheme adopts adaptive feature coupling and advanced pressure prediction to realize the transformation from passive feedback to active predictive control. It can identify load changes and pressure trends in advance, and combine fuzzy logic to adaptively output control commands, effectively avoiding adjustment lag and shock, and improving the dynamic response performance and operational stability of the system. 3. This solution adds a coupling value validity and confidence analysis step, and ensures data reliability through residual verification and multi-feature consistency assessment. It automatically switches to a conservative strategy in low confidence state, which greatly improves the system's anti-interference ability and robustness, and can adapt to long-term stable operation in complex industrial environments. 4. The solution uses a coordinated control of the motor and accumulator and a closed-loop iterative optimization strategy to accurately match load demand, reduce inefficient motor operation and waste energy consumption, and achieve reasonable energy recovery and utilization. Under the premise of ensuring stable system pressure, it improves energy utilization and achieves the goal of high-efficiency and energy-saving operation. Attached Figure Description

[0008] Figure 1 This is a schematic diagram of the overall structure of the present invention.

[0009] Figure 2 This is a schematic diagram of the S2 process of the present invention.

[0010] Figure 3 This is a schematic diagram of the S3 process of the present invention.

[0011] Figure 4 This is a schematic diagram of the S4 process of the present invention.

[0012] Figure 5 This is a schematic diagram of the S5 process of the present invention.

[0013] Figure 6 This is a schematic diagram of the S6 process of the present invention. Detailed Implementation

[0014] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0015] refer to Figure 1 - Figure 6 The energy-saving operation method of hydraulic motor based on pressure detection and energy replenishment shown includes: S1: Multi-source data acquisition and preprocessing: By synchronously acquiring multiple key operating parameters from the hydraulic system and drive motor, hardware filtering, software filtering, time synchronization, and cache storage are sequentially performed to construct a standardized dataset with high time alignment, noise suppression, and reliability. This provides stable and reliable data input for subsequent feature construction, state coupling, pressure prediction, and fuzzy decision-making, eliminating the impact of noise, delay, and outliers on the entire control process at its source. The specific steps are as follows: S101: Multi-source data acquisition: based on a fixed system sampling period The following physical quantities are collected synchronously through corresponding sensors, specifically including: The real-time pressure of the main working oil circuit of the hydraulic system is collected by a pressure sensor and recorded as follows: Unit: Pa; The real-time pressure of the working chamber inside the accumulator is collected by a pressure sensor and recorded as follows: Unit: Pa; The instantaneous flow rate of the main hydraulic circuit is collected by a flow sensor and recorded as follows: Unit: m³ / s; The real-time speed of the drive motor is acquired by a speed sensor or encoder and recorded as follows: Unit: r / min; The effective value of the three-phase current of the motor is collected by a current transformer and recorded as follows: Unit: A; The real-time temperature of the hydraulic oil is collected by a temperature sensor and recorded as follows: , Unit: ℃.

[0016] S102: Data Preprocessing: The raw sensor signals contain high-frequency electromagnetic interference, pressure pulsation, transmission delay, and abnormal abrupt changes, and the following standardized preprocessing steps are required: Hardware filtering: By simulating a low-pass filter circuit, high-frequency electromagnetic interference on the sensor transmission line, hydraulic pump pulsation noise, and high-frequency impact caused by valve movement are suppressed. Software filtering: A combination of moving average filtering and median filtering is used to remove outliers, spikes, and abnormal jumps in the data, ensuring a smooth and continuous signal. Time synchronization: Using the controller's internal system clock as a unified reference, timestamps are aligned for data from multiple sources to eliminate phase offsets and delays caused by inconsistencies in acquisition modules, transmission paths, and AD conversion speeds; Cache storage: The filtered and synchronized data is stored in a fixed-length circular cache queue, providing continuous and reliable data support for real-time differential calculation, sliding window statistics, online model updates, and historical state comparison.

[0017] S2: Constructing Characteristic Functions: Based on the multi-source operating data collected and preprocessed in S1, a set of dimensionless characteristic functions with clear physical meaning, strictly normalized dimensions, and physically constrained value ranges are constructed from six dimensions: pressure-flow coupling characteristics, motor-accumulator synergy, energy conversion efficiency, accumulator response capability, temperature-leakage coupling effect, and power-flow hysteresis characteristics. These functions transform heterogeneous physical quantities into standardized features that can be directly used for state fusion and decision analysis, achieving accurate quantification of system operating status and energy consumption characteristics. The specific steps are as follows: S201: Characteristic Function 1: Pressure-Flow Coupling Inertia Coefficient: The product of the flow rate and pressure change rate characterizes the transient energy change of the system. Multiplying this by the sampling period yields the transient energy increment, which is then compared to the product of the system's rated pressure and rated flow rate to achieve dimensionless normalization. Simultaneously, an exponential saturation term is introduced to suppress low-flow-rate conditions and highlight the characteristic response under high-flow-rate and strong transient conditions, ultimately forming a characteristic that reflects the dynamic inertia strength of the system. Its specific mathematical function is as follows: in, Indicates the rated working pressure of the hydraulic system; Indicates the rated flow rate of the hydraulic system; This represents the reference traffic threshold, with a value of [value to be filled in]. / 10 is used to suppress the impact of small flow disturbances on the characteristics; This feature is used to characterize the inertial strength of the system pressure response during flow rate changes. The larger the value, the stronger the transient change in system flow rate and pressure, and the higher the requirement for the energy storage device's replenishment response speed.

[0018] S202: Characteristic Function 2: Dynamic Coupling Degree Between Motor and Accumulator Multiplying the accumulator pressure, the motor speed change rate, and the sampling period yields the electromechanical coupling dynamic term, which is then normalized using the motor's rated electrical parameters to eliminate the influence of differences in equipment parameters. Simultaneously, the relative difference between the accumulator and system pressures is introduced to characterize the energy flow direction. Multiplying these two terms forms a dimensionless feature that simultaneously reflects the coupling strength and energy flow direction. Its specific mathematical function is as follows: in, The rated operating current of the drive motor; This refers to the rated speed of the drive motor; This is the rated operating voltage for the drive motor.

[0019] This feature is used to quantify the dynamic coordination between motor speed regulation and accumulator energy charging and discharging. A positive function value indicates that the accumulator pressure is higher than the system pressure, and it can provide auxiliary energy replenishment. A negative function value indicates that the system pressure is higher than the accumulator pressure, and the motor needs to bear the main load output.

[0020] S203: Characteristic Function 3: Energy Conversion Efficiency Coefficient Using hydraulic output power (flow rate × pressure) as the numerator and the motor input electrical power multiplied by the motor efficiency and pump efficiency as the denominator, a real-time efficiency expression is directly constructed using the power ratio. Efficiency curves are used to compensate for differences in motor and pump operating conditions, ensuring the calculation results accurately reflect the electrical and hydraulic energy conversion efficiency at the current speed and pressure. The specific mathematical function is as follows: in, The operating efficiency of the motor at the current speed is obtained by fitting the motor's factory characteristic curve. This indicates the working efficiency of the hydraulic pump at the current speed and system pressure, which is obtained from bench test data.

[0021] This feature reflects the real-time efficiency of converting electrical energy into hydraulic energy under the current operating conditions. The closer the value is to 1, the higher the system energy efficiency. When the value is below 0.6, the system is judged to be in an inefficient operating range.

[0022] S204: Characteristic Function 4: Pressure Gradient-Accumulator Response Matching Coefficient Multiplying the pressure change rate by the accumulator volume yields the system's transient flow demand on the accumulator. This demand is then compared to a baseline quantity comprised of the system's rated flow rate, rated pressure, and the accumulator's time constant to achieve dimensionlessness. Further multiplying this by the ratio of the accumulator's current pressure to its maximum pressure reflects the accumulator's available energy level. Finally, a matching characteristic is obtained that determines whether the accumulator can independently support pressure changes. The specific mathematical function is as follows: in, Indicates the effective working volume of the accumulator; This represents the charging and discharging time constant of the accumulator; This indicates the maximum allowable operating pressure of the accumulator; This feature is used to determine whether the current system pressure change rate exceeds the accumulator's independent compensation capability. When the value is greater than 1, it indicates that the system pressure cannot be stabilized by the accumulator alone, and the drive motor needs to quickly increase its speed to supplement the energy.

[0023] S205 Characteristic Function 5: Temperature-Leakage Coupling Compensation Factor The relative deviation between the current oil temperature and the optimal oil temperature is used as the leakage trend term to reflect the influence of temperature on leakage. Simultaneously, the pressure change rate is integrated and averaged over a certain time window to obtain the pressure fluctuation intensity term. These two terms are multiplied and normalized to form a coupled compensation feature that simultaneously reflects the influence of temperature and pressure stability. The specific mathematical function is as follows: in, Indicates the optimal operating oil temperature of the hydraulic system; Indicates the maximum allowable operating oil temperature of the system; This indicates the minimum allowable operating oil temperature of the system; The time window for integrating pressure changes is 5 to 10 seconds; u is a temporary time variable within the integration interval, iterated through... All time points up to t are used to calculate the average rate of pressure change during this period; This feature is used to quantify the impact of oil temperature changes on system leakage and pressure maintenance capabilities. A positive function with a larger value indicates that higher oil temperature leads to increased leakage and stronger pressure fluctuations, requiring enhanced energy replenishment control strategies.

[0024] S206 Characteristic Function 6: Power and Flow Inertial Hysteresis Coefficients: The response speed at the drive end is characterized by the rate of change of motor power, and the demand speed at the load end is characterized by the rate of change of flow rate. The ratio of the two yields the degree of response lag. A pressure margin factor is then introduced to mitigate operating conditions close to the rated pressure, preventing false triggering of emergency energy replenishment in high-pressure areas. This ultimately results in a lag characteristic that can predict the risk of pressure drop. The specific mathematical function is as follows: in, These are extremely small positive numbers, used to avoid denominators of 0 and to ensure the stability of the calculation process; This feature characterizes the dynamic lag of the motor's power response relative to changes in flow demand. A larger value indicates a more significant power response lag, making the system more prone to insufficient pressure and requiring predictive energy replenishment in advance.

[0025] S3: Adaptive Feature Coupling Based on six sets of dimensionless features constructed using S2, statistical analysis is performed using historical data from a sliding window. Principal component analysis is used to adaptively allocate feature weights. Combined with standardization and nonlinear coupling operations, multi-dimensional, strongly correlated, and dispersed features are fused into a single comprehensive state coupling value. This achieves unified quantification of system energy consumption, energy replenishment requirements, and operational trends, providing highly refined and adaptable input indicators for subsequent control decisions. The specific steps are as follows: S301: Constructing Feature Vectors: Integrating the six sets of dimensionless features at the same time into a standard column vector form, realizing the structured aggregation of multiple features, providing a regular data carrier for subsequent statistical analysis and matrix operations, and eliminating the logical fragmentation problem of scattered calculation of multiple features.

[0026] The superscript T indicates matrix transpose, which converts a row vector into a column vector to adapt to subsequent linear algebra operations.

[0027] S302: Sliding window statistical calculation: Select a fixed-duration sliding time window (Typical values ​​are 30-60 seconds). Extract historical feature data cached within the window to perform statistical calculations, capture the long-term distribution patterns of features, eliminate instantaneous fluctuations, and provide a reliable statistical benchmark for weight calculation.

[0028] Feature mean calculation: Calculate the average level of the i-th feature within the window, representing the normal operating state of that feature. The formula is: Where N is the total number of sampling points within the time window, This represents the k-th sampling time within the window.

[0029] Feature standard deviation calculation: Calculate the fluctuation range of the i-th feature within the window, characterizing the stability of the feature. The formula is: ; Correlation coefficient matrix construction: quantifies the degree of linear correlation between any two features, filters redundant features, and highlights core correlation characteristics. The formula is: The matrix R is a symmetric square matrix that reflects the coupling correlation between features and is the core input of principal component analysis.

[0030] S303: Principal component analysis to determine coupling weights: Principal component decomposition is performed based on the correlation coefficient matrix. Core features are selected and adaptive weights are assigned through data-driven methods, retaining more than 85% of the effective information of the system, thereby achieving feature dimensionality reduction and information condensation.

[0031] Perform eigenvalue decomposition on the correlation coefficient matrix R to obtain the eigenvalues ​​arranged in descending order. and the corresponding unit eigenvector ; Select the first p principal components and ensure that their cumulative contribution rate is ≥0.85, which satisfies the following formula: ; Construct the principal component loading matrix to map the correlation between the principal components and the original features. The specific mathematical formula is as follows: ; Calculate the global coupling weight of each individual feature, and after normalization, obtain the final contribution percentage of each feature. The formula is as follows: The weight values ​​change dynamically with operating conditions and historical data, automatically strengthening the weight of features with high information content and weakening the weight of redundant features.

[0032] S304: Feature Normalization Processing To address the issue of large differences in amplitude among different features, a standardization transformation is performed on the original features to map all features to the same dimension range, eliminating the interference of amplitude magnitude on coupled calculations and ensuring fairness when each feature participates in the calculation.

[0033] in: Use extremely small positive numbers to avoid denominators of 0 and ensure calculation stability; These are the standardized dimensionless features, with a mean of 0 and a variance approaching 1.

[0034] S305: Nonlinear Coupling and Synthetic State Calculation: By introducing a quadratic interaction term between features, the nonlinear interaction of features is captured. With the total energy consumption per unit time of the system as the optimization objective, a partial least squares regression model is constructed. After solving the interaction weights, the comprehensive coupling value that characterizes the overall state of the system is finally obtained.

[0035] An energy consumption regression model is constructed, taking into account both linear characteristic terms and nonlinear interaction terms, to accurately fit the energy consumption change pattern. Its specific mathematical function is as follows: in The feature interaction weights are solved using the least squares method to minimize the energy consumption prediction error. This represents the model residuals.

[0036] By removing residual terms, the combined state coupling value is obtained, achieving single numerical quantization of the system state: This value is dimensionless. A positive value indicates that the system is in a state of high energy consumption and high energy replenishment demand; a negative value indicates that the system is in a state of low energy consumption and energy-saving standby. The magnitude of the absolute value corresponds to the state intensity.

[0037] S4: Predictive Model Construction: Based on historical operational data collected by S1 and combined with the system's dynamic characteristics, a multivariate autoregressive prediction model is built. By iteratively updating the model coefficients online, the model accurately predicts the system's main pressure at future moments. The predicted pressure is then normalized into standardized state values, overcoming the lag of traditional feedback control and enabling proactive energy replenishment decisions. This provides advanced trend information for subsequent control processes, avoiding system fluctuations and energy waste caused by sudden pressure changes. The specific steps are as follows: S401: Multivariate Autoregressive Stress Prediction: A multivariate prediction model is constructed based on historical time-series data, integrating historical information on three key parameters: pressure, flow rate, and motor current. It takes into account both the system's hydraulic inertia and electromechanical coupling characteristics to achieve quantitative prediction of the main pressure at future moments. Simultaneously, an online update mechanism is employed to adapt to operating condition drift and equipment aging. The specific mathematical functions are as follows: ; in: This represents the system principal pressure predicted at time τ from the current time t, in Pa. This indicates the pressure prediction step size, which is calibrated based on the length of the system's hydraulic lines and the response speed of the actuators to ensure that the prediction time matches the dynamic response rhythm of the system. m is the autoregressive order of the model, which is optimally determined by the Bayesian Information Criterion (BIC) to balance prediction accuracy and computational efficiency and avoid overfitting. This is a model constant bias term used to correct the steady-state deviation between the theoretical values ​​and actual operating conditions, and to offset fixed errors such as sensor zero-point drift and hydraulic pipeline losses. It has no fixed value and its value is obtained through online iterative solution using the recursive least squares (RLS) method, dynamically and adaptively updated throughout the process according to the system's operating conditions. The specific rules are as follows: Initial value: During the model power-on initialization phase, the default value is 0 as the initial bias to adapt to the system's cold start steady-state benchmark; Online solution: During operation, the residuals between historical measured and predicted pressure values ​​are combined with the regression coefficients. , , Synchronous iterative calculations are performed to match characteristics such as steady-state pressure deviation, oil temperature drift, and equipment wear in the system. , , The regression coefficients corresponding to historical pressure, historical flow, and historical current are updated online in real time using the recursive least squares method to adapt to fluctuations in operating conditions such as sudden load changes and oil temperature variations; their value selection rules are as follows: During model initialization and system cold start phases, all regression coefficients are initialized to 0 by default, and the constant bias term is set to 0. The initial value is 0 to establish a blank prediction model baseline, avoiding prediction distortion caused by initial parameter deviations. After the system enters steady-state operation, the recursive least squares (RLS) method is used for real-time iterative solution. The core process is as follows: Each sampling cycle is completed Collect the latest measured data of pressure, flow rate and current, and construct a time-series observation vector; Calculate the residual between the current predicted pressure and the measured pressure, and use the residual inverse correction coefficient to reduce the prediction error; By combining a forgetting factor (typically 0.95~0.99), the weight of long-term historical data is weakened, while the influence of recent data is strengthened, adapting to the drift of operating conditions such as sudden load changes, oil temperature changes, and equipment aging. The coefficients have no fixed upper or lower limits, but are automatically constrained through iterative algorithms to avoid abnormal extreme values. Every N sampling periods (N is the number of sliding window points), a coefficient normalization calibration is performed to prevent long-term cumulative errors. When the system undergoes major repairs, sensor replacements, or hydraulic pipeline modifications, the initialization process is re-executed, and the coefficient iteration process is restarted. The coefficient order j strictly corresponds to the model autoregressive order m (j=1, 2, ..., m). After m is optimally determined by the Bayesian Information Criterion (BIC), the number of coefficients is fixed synchronously. After the order is adjusted, the coefficient matrix is ​​re-initialized and iterated. The system has a fixed sampling period, which is consistent with the S1 acquisition period to ensure timing continuity.

[0038] S402: Predicted State Normalization: The predicted absolute pressure value is mapped to the standard interval [0, 1] to eliminate decision interference caused by differences in the system's rated parameters. The predicted pressure is transformed into an intuitive state level, which facilitates the identification of pressure trends in subsequent fuzzy decision-making processes. The specific mathematical function is as follows: in, The normalized predicted state value is strictly limited to the range [0, 1]. This is the minimum safe operating pressure allowed by the hydraulic system. Below this value, the actuator will move slowly and the system will fail. The rated working pressure of the hydraulic system is the upper limit of the pressure for normal operation of the system. Explanation of normalization logic: The closer it is to 0, the lower the main pressure of the system will be in the future, and the urgent need for energy replenishment will be. The closer it is to 1, the more sufficient the system pressure will be in the future, allowing for a reduction in motor output to achieve energy savings; When the value is in the middle of the range, it indicates that the system pressure is stable, and the existing operating mode can be maintained.

[0039] S5: Analysis of the validity and confidence of coupling values: The consistency between the integrated state coupling value obtained from S3 and the pressure prediction state value obtained from S4 is verified. Through residual analysis, multi-feature joint distribution testing, and confidence function calculation, the reliability of the current system state information is quantitatively assessed. The aim is to identify abnormal scenarios such as sensor failures, transient disturbances, and sudden changes in operating conditions, providing confidence constraints for subsequent fuzzy decision-making, avoiding erroneous control strategies due to distorted state information, and improving the robustness and safety of the entire control process. The specific steps are as follows: S501: Calculation of the residual between the coupled value and the predicted value: By calculating the deviation between the integrated state coupling value and the predicted pressure state value, the degree of deviation between the current system state and the predicted trend is measured, which is used to preliminarily determine whether there are abnormal fluctuations in the data; the specific mathematical function is: ; in: The combined state coupling value obtained in S3; The normalized pressure prediction state value obtained in S4; The residuals are dimensionless. The smaller the absolute value of the residuals, the more consistent the current system state is with the predicted trend, and the higher the data reliability. The larger the absolute value of the residuals, the more significant the deviation of the state information, which may indicate disturbances or faults, and the confidence level should be reduced.

[0040] S502: Multi-source feature consistency check: Based on historical statistical patterns within the sliding window, a multi-feature joint consistency function is constructed to determine whether the current six feature values ​​are within a normal distribution range, thus verifying reliability through the statistical characteristics of the data itself. Its specific mathematical function is as follows: in: Let be the calculated value of the i-th dimensionless characteristic function at the current time. Let be the statistical mean of the i-th feature within the sliding time window; Let be the statistical standard deviation of the i-th feature within the sliding time window.

[0041] The principle of consistency testing is as follows: The current value of each feature is compared with its historical mean and standard deviation to construct a Gaussian consistency index; the consistency indices of the six features are multiplied together to obtain the overall joint confidence index. ; The closer the value is to 1, the more it indicates that all current features conform to historical operating patterns, and the data is stable and reliable; The closer the value is to 0, the more likely it is that at least one feature has deviated abnormally, thus reducing the reliability of the data.

[0042] S503: Coupling value confidence calculation: By combining residual bias and multi-feature consistency, a final confidence function is constructed, normalizing the reliability assessment results to continuous values ​​within the interval [0, 1], which are directly used as confidence constraints for subsequent fuzzy decision-making. Its specific mathematical function is as follows: The state confidence level has a value range of [0, 1]. This is the residual penalty coefficient, calibrated through offline perturbation tests, and is generally taken as 2-5; The consistency deviation penalty coefficient is calibrated through offline fault tests and is generally set to 3-6. , It is a fixed constant and will not be changed after the system debugging is completed.

[0043] Confidence level judgment rules: ≥0.7: Determined as highly reliable, the optimal control strategy can be executed normally; <0.7: Determined as low reliability, forced to switch to a conservative security control strategy, prohibiting aggressive energy-saving actions, and ensuring that the system pressure remains stable and does not drop.

[0044] S6: Adaptive control decision based on fuzzy logic: Using the comprehensive state coupling value, pressure prediction state value, and confidence level as input, and through a complete process of fuzzification, fuzzy inference, and declarative analysis, intelligent decision-making under multivariable, strongly coupled, and nonlinear operating conditions is achieved. The output is the motor speed setpoint and the accumulator valve group control command. Under the premise of ensuring system pressure stability and smooth response, the optimal energy-saving operation strategy is obtained, which is the core control output link of this method. The specific steps are as follows: S601 Fuzzy Input Variable Definition: The precise values ​​obtained from S3, S4, and S5 are converted into fuzzy linguistic variables to construct the input space of the fuzzy controller: Integrated state coupling value The fuzzy subset is {low, medium, high}, used to characterize the current energy consumption level and the intensity of energy replenishment demand of the system; Predicted state values The fuzzy subset is {decreasing, stable, increasing}, used to characterize the changing trend of system pressure over a future period of time; Confidence The fuzzy subset is {low confidence, high confidence}, with a threshold of 0.7. Data with a value greater than or equal to 0.7 is considered high confidence, and data with a value less than 0.7 is considered low confidence. This is used to characterize the reliability of the current state data.

[0045] The comprehensive state coupling value The fuzzy subset partitioning method is as follows: Low: ;middle: ;high: ; The predicted state value The fuzzy subset partitioning method is as follows: decline: ;smooth: ;rise: ; S602: Definition of fuzzy output variables: This step determines the fuzzy control quantity output by the controller, corresponding to the operating mode of the actuator: Motor operating mode The fuzzy subset {stop, low-speed intermittent, speed following, full speed} is used to determine the motor's operating state; its output range is divided as follows: stop: [0, 0.2], low-speed intermittent: (0.2, 0.45], speed following: (0.45, 0.75], full speed: (0.75, 1]; after defuzzification, the values ​​in this range are linearly mapped to the actual motor speed range. ;in, Normalized values ​​for fuzzy output (∈[0,1]); This is the minimum operating speed of the motor; This refers to the motor's maximum rated speed. Energy replenishment valve group mode The fuzzy subset is {energy storage priority, collaborative power supply, energy recovery, forced power replenishment}, which is used to determine the energy storage charging and discharging strategy; its normalized output range is: energy storage priority: [0, 0.25], collaborative power supply: (0.25, 0.5], energy recovery: (0.5, 0.75], forced power replenishment: (0.75, 1]; S603: Fuzzy Rules and Reasoning This step uses the Mamdani fuzzy inference method to establish a logical mapping relationship between input and output, forming a complete fuzzy rule base: When the coupling value is low, the predicted pressure is stable, and the confidence level is high, the generator should be shut down and the energy storage device should be replenished first. When the coupling value is low, the predicted pressure decreases, and the confidence level is high, the motor is executed at low speed intermittently and in a coordinated manner to provide energy. When the coupling value is in the middle, the predicted pressure rises, and the confidence level is high, the motor speed follows and the energy is supplied in a coordinated manner. When the coupling value is high, the predicted pressure rises, and the confidence level is highly reliable, the motor runs at full speed and forced energy replenishment is executed. When the coupling value is in the middle, the predicted pressure decrease is high and the confidence level is reliable, motor speed tracking and energy recovery are implemented. When the confidence level is low, a conservative safety strategy of motor speed following and coordinated power supply is uniformly adopted to ensure reliable system operation.

[0046] S604: Defuzzy output command: The result obtained from fuzzy inference is a fuzzy set, which cannot directly drive the actuator. This step uses the centroid method for defuzzification, transforming the fuzzy quantities into continuous, precise, and executable control commands. For the membership function curve of the inference output, calculate the geometric center of the area enclosed by the curve and the horizontal axis. The horizontal axis of this center is the exact value of the output. Map this precise value to the motor speed setpoint It is continuously adjustable to ensure smooth changes in motor speed without impact or vibration; Based on the valve group mode, output the corresponding switching state, opening ratio or PWM duty cycle command to achieve precise control of accumulator charging, discharging, pressure holding and energy recovery; Physically limit and restrict the speed and rate of valve group commands to ensure that the output does not exceed the rated range of the equipment and avoid overshoot and equipment damage; The final output can be directly sent to the frequency converter and valve group drive module to complete the complete conversion from fuzzy decision-making to actual control.

[0047] S7: Control command execution and closed-loop operation: The motor speed command and accumulator replenishment valve group command output from the S6 fuzzy decision are sent to the corresponding actuators for drive control. Simultaneously, real-time operational feedback parameters from the hydraulic system and drive motor are collected and transmitted back to the S1 data acquisition stage, forming a complete closed-loop control process of "data acquisition—feature construction—state coupling—pressure prediction—confidence assessment—fuzzy decision—command execution—data feedback." This enables the entire system to achieve pressure stability, smooth response, and energy-saving operation through continuous iteration. Specifically: S701: Actuator drive control: The precise control commands obtained after defuzzification of S6 are output to the frequency converter and hydraulic valve group respectively, completing the conversion of control commands into physical actions: Set the motor speed to a set value The data is transmitted to the frequency converter, which adjusts the actual output speed of the drive motor according to the target speed, thereby changing the output flow of the hydraulic pump and achieving dynamic matching of system flow and pressure. The control command of the energy replenishment valve group is output to the corresponding solenoid valve, proportional valve or servo valve. By controlling the opening and closing status, opening degree or PWM duty cycle of the valve port, the working modes of the accumulator such as liquid filling, liquid release, pressure holding and energy recovery can be switched. The motor speed and valve group action are synchronized and coordinated to avoid pressure shocks, sudden flow changes or system oscillations caused by asynchronous actions, thus ensuring the smooth and reliable operation of the actuator.

[0048] S702: Closed-loop feedback and iterative operation: The system collects feedback parameters defined in S1 in real time, such as main system pressure, accumulator pressure, system flow rate, motor speed, motor three-phase current, and hydraulic oil temperature. These parameters are then fed back into the S1 multi-source data acquisition and preprocessing stage as input data for the next control cycle. This enables the entire control process to form a continuously iterative closed-loop operation mechanism, ensuring that the system can keep up with changes in operating conditions, continuously correct control strategies, and maintain optimal operating status.

[0049] S703: System Stability and Energy Efficiency Optimization During closed-loop operation, with the premise of dynamic stability of system pressure and the goal of reducing motor ineffective energy consumption and improving energy utilization, the motor operation mode and accumulator energy replenishment strategy are adaptively adjusted based on the comprehensive state coupling value, pressure prediction results and confidence level. This minimizes the motor's operating time in the low load and low efficiency range, and realizes the adaptive, highly reliable and efficient energy-saving operation of the hydraulic motor based on pressure detection energy replenishment.

[0050] Secondly: The accompanying drawings of the embodiments disclosed in this invention only involve the structures involved in the embodiments disclosed in this invention. Other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of this invention can be combined with each other. In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A hydraulic motor energy-saving operation method based on pressure detection and energy replenishment, characterized in that, include: S1: Multi-source data acquisition and preprocessing: Synchronously acquire multiple operating parameters of the hydraulic system and drive motor, and sequentially complete filtering, time synchronization and buffer storage, and output preprocessed multi-source data; S2: Construct dimensionless characteristic functions: Based on the preprocessed multi-source data, construct dimensionless characteristic functions for each dimension from the perspectives of pressure-flow coupling, motor-accumulator coordination, energy conversion efficiency, accumulator response, temperature-leakage coupling, and power-flow hysteresis. S3: Adaptive Feature Coupling: Based on the dimensionless feature functions constructed in S2, multi-dimensional features are nonlinearly fused through statistical analysis and adaptive weight allocation to complete adaptive feature coupling and obtain a comprehensive state coupling value. S4: Predictive Model Construction: Combining historical operating data with system dynamic characteristics, construct a system pressure advance prediction model to obtain the pressure prediction state value; S5: Coupling value validity and confidence analysis: The consistency between the integrated state coupling value and the pressure prediction state value is verified, the validity and confidence analysis of the coupling value is completed, and the state confidence is obtained; S6: Adaptive control decision based on fuzzy logic: Taking the comprehensive state coupling value, pressure prediction state value and state confidence as input, the adaptive control decision is completed through fuzzy logic reasoning, and the control commands for motor speed and accumulator valve group are output. S7: Control command execution and closed-loop operation: Executes control commands to drive the corresponding actuators and collects operational feedback parameters in real time and sends them back to S1 to realize control command execution and closed-loop operation.

2. The energy-saving operation method of hydraulic motor based on pressure detection and energy replenishment according to claim 1, characterized in that: The preprocessed multi-source data includes: The system synchronously collects hydraulic system main oil circuit pressure, accumulator pressure, hydraulic main oil circuit flow, drive motor speed, motor three-phase current and hydraulic oil temperature through sensors. After filtering, multi-channel time synchronization and buffer storage processing, the system obtains standardized data with time regularization and noise suppression.

3. The energy-saving operation method of hydraulic motor based on pressure detection and energy replenishment according to claim 1, characterized in that: The construction of dimensionless characteristic functions for each dimension includes: Based on the preprocessed multi-source data, six types of dimensionless characteristic functions are constructed respectively: pressure-flow coupling inertia coefficient, dynamic coupling degree between motor and accumulator, energy conversion efficiency coefficient, pressure gradient and accumulator response matching coefficient, temperature and leakage coupling compensation factor, and power and flow inertia hysteresis coefficient. The pressure-flow coupling inertia coefficient characterizes the transient energy change of the system by multiplying the flow rate and the pressure change rate. It is then multiplied by the sampling period to obtain the transient energy increment. The value is then compared with the product of the system's rated pressure and rated flow rate to achieve dimensionless normalization. At the same time, an exponential saturation term is introduced to suppress low flow conditions, forming a characteristic that reflects the dynamic inertia strength of the system. The dynamic coupling degree between the motor and the accumulator is formed by the product of the accumulator pressure, the rate of change of motor speed and the sampling period to constitute the electromechanical coupling dynamic term. It is normalized by combining the rated electrical parameters of the motor and the relative difference between the accumulator and the system pressure to characterize the energy flow direction, thus forming a feature that reflects the dynamic coordination degree between the motor and the accumulator. The energy conversion efficiency coefficient is calculated with hydraulic output power as the numerator and motor input power combined with motor efficiency and pump efficiency as the denominator. It is formed by power ratio and working condition efficiency compensation to reflect the real-time conversion efficiency of electrical energy and hydraulic energy. The pressure gradient and accumulator response matching coefficient characterizes the transient flow demand by multiplying the pressure change rate by the accumulator volume. It is compared with the benchmark quantity composed of the system benchmark flow rate and the accumulator time constant to achieve dimensionlessness. It is combined with the ratio of the current pressure of the accumulator to the maximum working pressure to reflect the available energy level, forming a characteristic for judging the accumulator's compensation capability. The temperature and leakage coupling compensation factor characterizes the leakage trend by the relative deviation between the current oil temperature and the optimal oil temperature. It is combined with the pressure fluctuation intensity for integral averaging and normalization to form a feature that reflects the influence of oil temperature on leakage and pressure stability. The power and flow inertial hysteresis coefficients characterize the degree of response hysteresis by the ratio of the rate of change of motor power to the rate of change of flow rate, and introduce a pressure margin factor to weaken the influence of high pressure range, forming a hysteresis feature for predicting the risk of pressure drop.

4. The energy-saving operation method of hydraulic motor based on pressure detection and energy replenishment according to claim 1, characterized in that: The integrated state coupling value includes: Based on the dimensionless feature functions constructed by S2, historical feature data are extracted and statistical analysis is carried out by selecting a fixed-duration sliding time window. The feature mean and standard deviation are calculated and the correlation coefficient matrix is ​​constructed. The correlation coefficient matrix is ​​decomposed by principal component analysis to determine the adaptive coupling weight of each feature. After standardizing the dimensionless features, nonlinear cross terms between features are introduced for fusion operation. Finally, the comprehensive state coupling value of the overall operating state of the quantified system is obtained.

5. The energy-saving operation method of hydraulic motor based on pressure detection and energy replenishment according to claim 1, characterized in that: The construction of the system pressure advance prediction model to obtain pressure prediction state values ​​includes: By integrating historical pressure, historical flow, and motor current time-series data, a multivariate autoregressive prediction model is built. The model coefficients are updated online using the recursive least squares method to predict the future system main pressure value. The predicted pressure value is then normalized and mapped to the system's rated pressure range to obtain the pressure prediction state value.

6. The energy-saving operation method of hydraulic motor based on pressure detection and energy replenishment according to claim 1, characterized in that: The validity and confidence analysis of the coupling values ​​includes: The residual between the integrated state coupling value and the pressure prediction state value is calculated to measure the degree of deviation between the two. Based on the historical mean and standard deviation of each dimensionless feature within the sliding window, a Gaussian multi-feature joint consistency function is constructed, and the data stability is determined by synchronous verification of multi-dimensional features. Combining the two indicators of residual deviation and multi-feature consistency, a normalized confidence function with a penalty coefficient is constructed to normalize the evaluation results to a specified interval, thereby obtaining the state confidence. The multi-feature joint consistency function obtains the overall data consistency index by performing Gaussian normalization on the deviation between the current value and the historical statistical value of a single feature, and then multiplying the normalization results of each feature dimension together. The normalized confidence function introduces residual penalty coefficients and consistency deviation penalty coefficients, and performs a weighted operation by combining the absolute value of the residuals and the consistency deviation value, and finally outputs the state confidence.

7. The energy-saving operation method of hydraulic motor based on pressure detection and energy replenishment according to claim 1, characterized in that: The adaptive control decision-making process, which uses fuzzy logic reasoning to output control commands for motor speed and accumulator valve group, includes: The integrated state coupling value, pressure prediction state value, and state confidence are fuzzified and transformed into corresponding fuzzy linguistic variables. Based on the preset fuzzy rule base, the fuzzy inference method is used to complete the logical inference of multiple input variables and obtain the fuzzy control output. The fuzzy control output is defuzzified by the centroid method and transformed into a continuous control quantity. After physical amplitude and rate limiting, the motor speed control command and accumulator valve group control command adapted to the actuator are output.

8. The energy-saving operation method of hydraulic motor based on pressure detection and energy replenishment according to claim 7, characterized in that: The execution of control commands and closed-loop operation include: The defuzzified motor speed command is sent to the frequency converter, and the accumulator valve group command is sent to the corresponding hydraulic valves, synchronously driving the actuator to achieve dynamic pressure matching. Real-time feedback data such as system main pressure, accumulator pressure, flow rate, motor operating parameters and oil temperature are collected and synchronously transmitted back to the S1 multi-source data acquisition and preprocessing stage as the input data source for the next control cycle, forming a closed-loop control process of data acquisition, analysis and decision-making, command execution and feedback iteration.

Images