A precise group control synchronous pushing control method for an extra-large main transformer

Abstract

The present application belongs to the technical field of power transformer, and provides a precision group control synchronous push control method for super large main transformer, which comprises collecting working condition parameters of each pushing point, calculating response deviation and constructing dynamic pressure coupling graph according to branch and main pressure; when the response deviation meets the first trigger condition and / or the pressure coupling graph meets the second trigger condition, a learning trigger instruction of the current point is generated; in response to the instruction, the preset unified dead zone dynamic model parameters are cooperatively identified and updated by applying excitation to the servo valve and collecting response data; in each control cycle, based on the updated model, the current working condition and the pressure coupling graph, a coupling feedforward compensation signal corresponding to the current servo valve is synthesized; finally, the signal is synthesized with the friction feedforward compensation signal and the displacement synchronous closed-loop feedback signal generated in advance to generate the final driving instruction and send it to the corresponding servo valve for execution, which effectively solves the time-varying and coupling interference problems of valve port dead zone compensation.

Description

Technical Field

[0001] This invention relates to the field of power transformer technology, and in particular to a precision group control synchronous jacking control method for extra-large main transformers. Background Technology

[0002] In the precision group-controlled synchronous jacking operation of extra-large transformers (main transformers), modern advanced control methods have formed a relatively fixed processing flow to ensure the smooth and synchronous movement of hundreds of tons of equipment. This method typically includes the following conventional steps: First, synchronously collect multi-source operating parameters (such as valve commands, displacement, and pressure) at each jacking point; second, calculate the feedforward compensation amount based on a preset fixed parameter model (such as a friction model or dead zone model) to offset system nonlinearity; next, perform displacement synchronization closed-loop control through a master-slave or virtual spindle algorithm to eliminate tracking errors; then, synthesize the feedforward compensation signal, the synchronization closed-loop feedback signal, and other compensation signals to generate the final drive command; finally, send the drive command to each servo valve for execution.

[0003] In the aforementioned conventional control process, valve dead zone compensation is one of the key feedforward links to ensure the system's low-speed stability, response consistency, and synchronization accuracy. Currently, this link generally adopts a feedforward compensation strategy based on fixed parameters, that is, a set of dead zone compensation parameters (such as dead zone threshold voltage) are experimentally calibrated during the system commissioning phase and then used consistently in all subsequent operations.

[0004] However, this fixed compensation strategy has significant drawbacks in complex actual jacking operations: First, the dead zone characteristics of the servo valve are not static; they change slowly over time with valve core wear, oil temperature variations, and decreased oil cleanliness. Fixed parameters cannot track these changes, leading to insufficient or over-compensation. Second, in multi-cylinder group control systems, the movement of one jacking point causes system pressure fluctuations, which in turn instantly alter the effective dead zone of other valves. This dynamic coupling effect cannot be predicted or compensated for by the fixed model. These problems cause the dead zone compensation effect to gradually deteriorate during actual jacking operations. This manifests as inconsistent responses from different jacking points during startup, low-speed operation, and direction switching, decreased synchronization accuracy, and may even induce low-frequency oscillations. This becomes a hidden but critical technical bottleneck restricting the entire group control system from achieving and maintaining extremely high synchronization accuracy over a long period.

[0005] Therefore, within the framework of existing precision group control synchronous pushing methods, how to enable the valve dead zone compensation link to have adaptive time-varying characteristics and anti-coupling interference capabilities, thereby fundamentally improving the long-term accuracy and robustness of the entire system under complex working conditions, is a specific technical problem that urgently needs to be solved. Summary of the Invention

[0006] To address the aforementioned technical problems, this invention provides a precision group control synchronous jacking control method for extra-large main transformers. Within the framework of existing precision group control synchronous jacking methods, an intelligent upgrade has been made to the specific link of valve dead zone compensation. Through adaptive learning and coupled feedforward compensation, the problems of time-varying and coupling interference are effectively solved.

[0007] This invention provides a precision group control synchronous jacking control method for extra-large main transformers, the control method comprising the following steps: The response deviation is calculated based on the working condition parameter set of each jacking point collected synchronously, and a dynamic pressure coupling spectrum is constructed based on the branch pressure and main pressure in the working condition parameter set. When the response deviation meets the first triggering condition and / or the dynamic pressure coupling map meets the second triggering condition, a learning triggering command for the current push point is generated. In response to the learning trigger command, based on the response data obtained by applying an excitation signal to the servo valve corresponding to the current push point, the parameter set of the preset unified dead zone dynamic model is collaboratively identified and updated. In each control cycle, based on the updated unified dead zone dynamic model, the current operating condition, and the dynamic pressure coupling spectrum, a coupling feedforward compensation signal for the servo valve corresponding to the current push point is synthesized. The coupling feedforward compensation signal is combined with the friction feedforward compensation signal and displacement synchronization closed-loop feedback signal pre-generated by the corresponding servo valve to generate the final drive command, which is then sent to the corresponding servo valve for execution.

[0008] Preferably, the first triggering condition is that the statistical mean of the response deviation exceeds the adaptive threshold over N consecutive control cycles.

[0009] Preferably, the second triggering condition is that the magnitude of the main pressure mutation exceeds a preset disturbance threshold, and the correlation coefficient between the main pressure mutation and the change in the response characteristics of the current push point exceeds a preset correlation threshold.

[0010] Preferably, the calculation of response deviation based on the working condition parameter set of each jacking point acquired synchronously, and the construction of a dynamic pressure coupling spectrum based on the branch pressure and main pressure in the working condition parameter set, includes: Based on the working condition parameter set of each push point collected synchronously, the servo valve control command and valve core displacement feedback signal are extracted. Based on the servo valve control command and valve core displacement feedback signal, the command-speed response deviation of each pushing point is calculated; Simultaneously, the branch pressure and main system pressure of each jacking point are extracted from the set of operating parameters; Based on the time series of the branch pressure and the main system pressure, the transmission relationship and coupling strength between pressure fluctuations are analyzed and calculated, and the dynamic pressure coupling spectrum is constructed.

[0011] Preferably, the step of generating a learning trigger command for the current push point when the response deviation meets the first trigger condition and / or the dynamic pressure coupling map meets the second trigger condition includes: Receive the calculated response deviation and the constructed dynamic pressure coupling map; Based on the response deviation, determine whether the first triggering condition is met, and output the first determination result; Simultaneously, based on the dynamic pressure coupling spectrum, it is determined whether the second triggering condition is met, and the second determination result is output; When the first judgment result and / or the second judgment result indicate that the triggering condition is met, a learning trigger instruction for the current push point is generated.

[0012] Preferably, the response to the learning trigger command, based on the response data obtained by applying an excitation signal to the servo valve corresponding to the current push point, involves collaboratively identifying and updating the parameter set of a preset unified dead-zone dynamic model, including: Receive the generated learning trigger command, and determine the target servo valve based on the learning trigger command; Based on the determined target servo valve, a corresponding excitation signal is generated and applied to the target servo valve; Collect the response data generated by the target servo valve in response to the excitation signal; Based on the collected response data, the parameter set of the preset unified dead zone dynamic model is collaboratively identified using a parameter identification algorithm, and the parameter set of the unified dead zone dynamic model is updated according to the identification results.

[0013] Preferably, in each control cycle, based on the updated unified dead-zone dynamic model, the current operating condition, and the dynamic pressure coupling spectrum, the synthesis of a coupling feedforward compensation signal for the servo valve corresponding to the current push point includes: In each control cycle, the updated unified dead zone dynamic model, current operating condition data, and dynamic pressure coupling map are acquired; Based on the unified dead zone dynamic model and the current operating condition data, the basic feedforward compensation amount of the servo valve corresponding to the current push point is calculated. Simultaneously, based on the dynamic pressure coupling map, the pressure disturbance caused by the action of adjacent jacking points is predicted, and the coupling correction amount corresponding to the pressure disturbance is calculated. The basic feedforward compensation amount and the coupling correction amount are combined to obtain the coupled feedforward compensation signal.

[0014] Preferably, the step of synthesizing the coupled feedforward compensation signal with the friction feedforward compensation signal and the displacement synchronization closed-loop feedback signal pre-generated by the corresponding servo valve to generate the final drive command and sending it to the corresponding servo valve for execution includes: Obtain the coupling feedforward compensation signal of the servo valve for the current push point; Simultaneously, the friction feedforward compensation signal and displacement synchronization closed-loop feedback signal pre-generated for the servo valve are acquired; The coupling feedforward compensation signal, the friction feedforward compensation signal, and the displacement synchronous closed-loop feedback signal are superimposed and synthesized to generate the final drive command. The final drive command is sent to the corresponding servo valve to drive it to perform the corresponding control action.

[0015] Compared with related technologies, the precision group control synchronous jacking control method for ultra-large main transformers provided by this invention has the following beneficial effects: This invention constructs a unified dead zone dynamic model and introduces an intelligent triggering online identification mechanism, enabling the dead zone compensation parameters to be automatically updated in accordance with changes in the valve's own characteristics (such as wear and oil temperature), thus maintaining the accuracy of the compensation amount and effectively solving the problem of fixed parameters failing due to time-varying effects.

[0016] This invention creates a dynamic pressure coupling spectrum and proactively predicts and counteracts the pressure disturbance caused by the action of adjacent jacking points in the compensation calculation, enabling the dead zone compensation of this valve to have the ability to be "immune" to system coupling interference, thereby ensuring the overall synchronization accuracy of the group control system under complex interactive working conditions.

[0017] This invention employs a dual-condition triggering strategy based on the performance deviation and pressure coupling spectrum, which enables the system to initiate simplified learning only when necessary (i.e., when performance degrades or is subjected to explicit interference), avoiding unnecessary frequent testing, balancing learning effectiveness and process efficiency, and enhancing the system's practicality and robustness. Attached Figure Description

[0018] Figure 1 The flowchart illustrates a precision group control synchronous jacking control method for an extra-large main transformer provided by this invention. Detailed Implementation

[0019] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it. Furthermore, it should be noted that, for ease of description, only the parts relevant to the invention are shown in the drawings, not all structures. Moreover, unless otherwise specified, the embodiments and features described herein can be combined with each other.

[0020] It should also be noted that, for ease of description, the accompanying drawings show only the parts relevant to the invention and not all of them. Before discussing exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although the flowcharts describe the operations (or steps) as sequential processes, many of the operations can be performed in parallel, concurrently, or simultaneously. Furthermore, the order of the operations can be rearranged. The process can be terminated when its operation is completed, but it may also have additional steps not included in the drawings. The process may correspond to a method, function, procedure, subroutine, subroutine, etc.

[0021] Before detailing the specific implementation methods of each step of this invention, the conventional process of the industry-standard precision group control synchronous pushing control method upon which this invention is based is first explained. This conventional process constitutes the basic framework for achieving synchronous control and typically includes the following five steps in sequence: The first step is to synchronously acquire multi-source operating parameters. That is, under a unified time base, the system acquires the control commands of the servo valve at each jacking point, the displacement feedback of the valve core or cylinder, the branch pressure and the main system pressure in real time, thereby forming a set of operating parameters for control.

[0022] The second step is to calculate the feedforward compensation based on a fixed model. The system calls a pre-calibrated nonlinear model with fixed parameters (such as a dead zone model or a friction model) during the debugging phase, and calculates the corresponding feedforward compensation signal according to the current motion command, in order to preemptively offset the deterministic nonlinear interference in the system.

[0023] The third step is to execute displacement synchronization closed-loop feedback control. Using algorithms such as master-slave synchronization or virtual master axis, and taking a certain reference trajectory as a benchmark, the displacement tracking deviation of each slave point is calculated in real time, and feedback control quantities for correcting the deviation are generated by controllers such as PID controllers.

[0024] The fourth step is to synthesize multi-source control signals. The fixed feedforward compensation signal obtained in the second step, the displacement synchronization feedback signal obtained in the third step, and other compensation signals are superimposed to generate preliminary drive commands for each servo valve.

[0025] The fifth step is the sending and execution of drive commands. The synthesized commands are sent to each servo valve, driving the hydraulic cylinders to perform the jacking action. This conventional process constitutes the basic closed-loop control for achieving synchronous jacking.

[0026] However, this conventional method, especially the "fixed parameter feedforward compensation" strategy relied upon in its second step, has inherent flaws in complex engineering applications. The dead zone characteristics of servo valves slowly change over time with equipment wear, oil temperature variations, and oil contamination, causing the pre-calibrated fixed compensation parameters to gradually become ineffective, resulting in undercompensation or overcompensation. Furthermore, in multi-cylinder group-controlled hydraulic systems, the movement of a certain jacking point causes system pressure fluctuations. This pressure disturbance instantly alters the effective dead zone of other valves; this dynamic coupling effect is completely unpredictable and uncompensable by static fixed models. These flaws lead to inconsistent responses and decreased synchronization accuracy during startup, low-speed operation, and reversing phases after long-term operation, becoming key technical bottlenecks restricting the entire system from achieving and maintaining long-term high-precision and robust operation.

[0027] To address the core problem of insufficient adaptability caused by the fixed dead zone compensation model, this invention proposes a precision group control synchronous push control method that deeply integrates intelligent sensing, dynamic learning, and coupled feedforward.

[0028] This method, while inheriting the reasonable architecture of the aforementioned conventional process, fundamentally innovates its core bottleneck—the feedforward compensation stage. This invention abandons the fixed model and instead constructs an intelligent closed-loop subsystem with autonomous "monitoring-diagnosis-learning-compensation" capabilities. This subsystem aims to track the time-varying characteristics of dead zone in real time and actively cancel multi-cylinder coupling interference, thereby ensuring the continuous accuracy of feedforward compensation and ultimately overcoming the technical problem of maintaining ultra-high synchronization accuracy over a long period. The specific embodiments of this invention will be described in detail below with reference to the accompanying drawings.

[0029] refer to Figure 1 As shown, the control method includes the following steps: S1: Calculate the response deviation based on the set of operating parameters of each jacking point collected synchronously, and construct a dynamic pressure coupling spectrum based on the branch pressure and main pressure in the set of operating parameters.

[0030] Specifically, step S1 includes the following steps: S11: Based on the operating parameter set of each push point acquired synchronously, extract the servo valve control command and valve core displacement feedback signal.

[0031] In this embodiment, to achieve high-precision synchronous control, it is necessary to first complete the real-time and synchronous acquisition of multi-source data.

[0032] Specifically, each jacking point acts as an independent control unit, equipped with corresponding servo valves, hydraulic cylinders, and sensors. The system uses a central controller (such as a high-performance PLC or industrial PC) as its core, with a built-in or external multi-channel synchronous data acquisition card. All input channels of this acquisition card are driven by a highly stable hardware clock, ensuring that the sampling action of all jacking points is triggered at the same microsecond level, fundamentally eliminating comparison errors caused by asynchronous sampling times.

[0033] The synchronous acquisition process of the operating parameter set is as follows: For servo valve control commands, the acquisition card directly reads the analog voltage command signals (typically ranging from -10V to +10V) sent by the controller's digital output (DA) module to the amplifiers of each servo valve. For valve spool displacement feedback, the high-precision LVDT (linear variable differential transformer) displacement sensor integrated into each servo valve or its valve spool position feedback signal is connected to the analog input channel of the acquisition card. The controller uses a pre-configured sampling period (e.g., At the start of each cycle, the command acquisition card synchronously captures both types of signals from all jacking points and packages them into a structured "operating condition parameter set" for subsequent steps. This step ensures that all control and feedback data have a unified timestamp, which is the basis for accurate deviation calculation and coupling analysis.

[0034] S12: Calculate the command-speed response deviation of each push point based on the servo valve control command and valve core displacement feedback signal.

[0035] In this embodiment, the command-velocity response deviation is a key indicator for evaluating the dynamic tracking performance of a single push point and for subsequent triggering of self-learning. Its calculation is performed in real-time by a software algorithm in the central controller. First, the valve core displacement feedback signal is processed: the valve core displacement value collected in the current cycle is... The displacement value of the previous cycle Perform a difference operation, then divide by the sampling period. This allows for the estimation of the valve core movement speed in the current cycle. :

[0036] To prevent measurement noise from amplifying the differential calculation, a first-order low-pass digital filter can be applied to the displacement signal before differential calculation. Secondly, the servo valve control commands are processed: the control command voltage value... This inherently represents the desired direction and speed of valve spool movement (within the linear operating range of the servo valve). To obtain a "command speed" that can be directly compared to the actual speed, a calibrated speed gain coefficient is required. (unit: The command voltage is converted using a coefficient obtained during initial system calibration, representing the valve spool speed corresponding to a unit command voltage under ideal conditions. The formula for calculating the command speed is:

[0037] Finally, calculate the response deviation: in each control cycle, convert the actual speed obtained. With command speed By subtracting, we obtain the instantaneous velocity deviation. :

[0038] this This refers to the command-speed response deviation for the current cycle. This deviation not only reflects the instantaneous tracking error caused by nonlinear factors such as dead zone and friction, but its statistical characteristics over a period of time (such as the continuous average value described in step S2) can also characterize the slow decline trend of valve performance.

[0039] S13: At the same time, extract the branch pressure and main system pressure of each jacking point from the set of operating parameters.

[0040] In this embodiment, pressure data is the core input for analyzing system coupling relationships and constructing a dynamic pressure coupling map. Similar to displacement and command signals, pressure data acquisition is also accomplished by the aforementioned synchronous data acquisition card. High-response-frequency piezoresistive or piezoelectric pressure sensors are installed at the hydraulic cylinder inlet or servo valve outlet of each jacking point to measure the oil pressure in that branch, i.e., the "branch pressure". (in (Numbering of the jacking point).

[0041] Install another pressure sensor of the same type near the system's common hydraulic source (such as the main pump station outlet or near the accumulator) to measure the "system main pressure". All the analog signal outputs (typically 4-20mA or 0-10V) from these pressure sensors are connected to a synchronous data acquisition card. At the same synchronous sampling moment described in step S11, these pressure signals, along with valve commands and valve displacement signals, are acquired together and stored in the operating parameter set at the same timestamp.

[0042] Therefore, in subsequent processing, each push point has a set of fully time-synchronized data at any time t: This synchronization is crucial, as it ensures that when analyzing the phase and amplitude relationship between branch pressure fluctuations and system main pressure fluctuations, erroneous causal relationship judgments will not be made due to data asynchrony.

[0043] in, The valve core displacement feedback signal (referred to as in step S12) This signal is directly measured by a high-precision LVDT displacement sensor integrated into each servo valve, which characterizes the actual physical position of the valve core in real time and is used to estimate the actual speed. Calculate command-speed response deviation This is the fundamental basis and also provides core feedback for the valve's own status monitoring and closed-loop control.

[0044] S14: Based on the time series of the branch pressure and the main system pressure, analyze and calculate the transmission relationship and coupling strength between pressure fluctuations, and construct the dynamic pressure coupling spectrum.

[0045] In this embodiment, the dynamic pressure coupling map is a quantitative model used to describe the dynamic pressure influence relationship between any two jacking points in the system, and between any point and the main system pressure. Its construction is a continuous online analysis process. The central controller maintains a sliding data window (e.g., storing the most recent data). Time series data of all jacking points and main system pressure.

[0046] The core algorithm for constructing the map includes the following steps: First, calculate the cross-correlation function. For the first... The first apex push point and the first Calculate the branch pressure signals of each jacking point (or system main pressure) and its corresponding jacking points. and Cross-correlation function within the sliding window ,in For time lag:

[0047] By finding the peak value of the cross-correlation function and its corresponding time delay This allows us to determine the pressure fluctuation from a point. Pass to point The main time delay characteristic is that the magnitude of the peak reflects the coupling strength. Secondly, frequency domain or transfer function analysis is performed. In more complex implementations, a Fast Fourier Transform (FFT) can be performed on the pressure signal pair to analyze it at specific frequencies (especially the frequency band related to the valve actuation frequency). right The coherence coefficients and frequency responses are analyzed to more finely characterize the dynamic properties of the coupling. Finally, a graphical representation and update are performed. The analysis results (coupling strength coefficients, dominant time delays, dominant coupling frequencies, etc.) are stored in memory as a matrix or graph structure, called a "dynamic pressure coupling graph". For example, an adjacency matrix of a weighted directed graph can be used. To represent, where matrix elements Represents the source node To the target node The stress coupling strength, the properties of the edges can include time delay information. .

[0048] This graph is not static but dynamically updated as the system operates (e.g., changes in oil temperature and load). The system periodically (e.g., every second) or when a significant change in pressure fluctuation pattern is detected, the above analysis is re-executed using new sliding window data to update the graph parameters, thus ensuring that it always reflects the current true hydraulic coupling state of the system and provides accurate prediction basis for coupling feedforward compensation in step S4.

[0049] S2: When the response deviation meets the first triggering condition and / or the dynamic pressure coupling map meets the second triggering condition, a learning triggering command for the current push point is generated; Specifically, step S2 includes the following steps: S21: Receive the calculated response deviation and the constructed dynamic pressure coupling map.

[0050] In this embodiment, this step is the data preparation stage for triggering the judgment logic. The central controller maintains two real-time storage areas for key data: one is the data continuously calculated and updated by the S12 sub-step of step S1 for each push point. Command-speed response deviation time series ,in The first part refers to the window length used for statistics; the second part refers to the dynamic pressure coupling map, which is periodically updated (e.g., once per second) by the S14 sub-step of step S1. This map is typically stored in memory as a data structure of a weighted adjacency matrix C and a corresponding time delay matrix. Step S21 is executed by the "intelligent trigger judgment" module actively reading the latest data from these storage areas during each judgment period (this period can be the same as the main control period, such as 1ms, or a longer period, such as 10ms). Specifically, it reads the response deviation sequence of the current push point i in the most recent M periods, as well as the dynamic pressure coupling map of the entire system. This step ensures that the data used for subsequent conditional judgments is up-to-date and complete, forming the input basis for trigger judgments.

[0051] S22: Based on the response deviation, determine whether the first triggering condition is met, and output the first judgment result.

[0052] In this embodiment, this step performs quantitative detection of "performance degradation". The first trigger condition is defined as: the statistical mean of the response deviation over N consecutive control cycles exceeds an adaptive threshold. .

[0053] The specific judgment process is as follows: First, calculate the statistical mean. For the response deviation sequence of the current push point i obtained from step S21, calculate its arithmetic mean over the most recent N periods (N is a preset value, such as 100, corresponding to 100ms). .

[0054] Secondly, determine or update the adaptive threshold. Adaptive threshold It is not completely fixed; its initial value can be set as the average deviation measured during the stable operation phase after system calibration. times ( >1, such as 1.5). More importantly, It can slowly adapt to the "background noise" level of the system over a long period of time.

[0055] For example, it can be recorded during periods without triggering. long-term moving average and order ( >1, such as 2.0), which allows the threshold to adapt to the slow drift of the entire system. Finally, a comparison is made. If it meets the requirements... > If the condition is met, the first triggering condition is satisfied, and the first judgment result Flag1 is set to "True"; otherwise, it is set to "False". This judgment result Flag1 will be output to the arbitration logic.

[0056] S23: Simultaneously, based on the dynamic pressure coupling spectrum, determine whether the second triggering condition is met, and output the second determination result.

[0057] In this embodiment, this step performs the detection of "pressure coupling interference". The second triggering condition includes two sub-conditions, which must be satisfied simultaneously: (a) System main pressure The mutation magnitude exceeds the preset perturbation threshold. ; (b) The correlation coefficient between the mutation and the change in response characteristics at the current push point i exceeds a preset correlation threshold. .

[0058] The specific implementation is as follows: First, detect sudden changes in the main pressure. Calculate in real time the rate of change of the system's main pressure signal or the range (maximum value minus minimum value) within a short time window (e.g., 5 control cycles). If the absolute value of this change... Greater than the preset disturbance threshold (Based on the system's rated pressure setting, such as 5% of the rated pressure), a significant pressure surge is considered to have occurred, and sub-condition (a) is satisfied. Next, the correlation coefficient is calculated. Within the time window where the pressure surge is detected, the response deviation of the current push point i is extracted. The synchronous change sequence. Calculate the main pressure change sequence. Response deviation change sequence The Pearson correlation coefficient in this window If the calculated correlation coefficient Greater than the preset relevant threshold If the result is 0.7, it indicates that the changes of the two conditions are highly correlated, and sub-condition b is satisfied. Finally, a comprehensive judgment is made. The second triggering condition is determined to be satisfied only if sub-conditions a and b are satisfied simultaneously, and the result of the second judgment, Flag2, is set to "True"; otherwise, it is set to "False".

[0059] This judgment utilizes the coupling relationship revealed by the dynamic pressure coupling map to quickly locate the correlation between the action of a specific adjacent point (whose pressure fluctuations are transmitted to the main system) and the performance of the current point.

[0060] S24: When the first judgment result and / or the second judgment result indicate that the triggering condition is met, a learning trigger instruction for the current push point is generated.

[0061] In this embodiment, this step is the decision and output stage that triggers the judgment. It receives two Boolean judgment results, Flag1 and Flag2, from S22 and S23. The triggering logic adopts an "OR" relationship, that is, if Flag1=True or Flag2=True, it means that the performance of the current push point i has been significantly degraded due to its own decay or external coupling interference, and the model parameters need to be relearned.

[0062] Once the triggering conditions are met, the system immediately generates a structured "learning trigger instruction". This instruction contains at least the following information fields: target push point ID (i.e., i), trigger timestamp, trigger type ("performance degradation trigger", "coupled interference trigger", or "composite trigger"), and optional associated stress disturbance characteristics (if triggered by the second triggering condition).

[0063] After generating the instruction, the system places it in a priority "learning task queue" or sends it directly to the model parameter update module in step S3. Simultaneously, to prevent repeated and frequent triggering during the duration of the disturbance or before learning is complete, the system sets a "learning inhibition period" (e.g., 2 seconds) for the push point i. During this period, trigger judgments for that point will be temporarily suspended to ensure the integrity and stability of the learning process. This step ultimately completes the transition from anomaly perception to learning decision-making.

[0064] S3: In response to the learning trigger command, based on the response data obtained by applying an excitation signal to the servo valve corresponding to the current push point, collaboratively identify and update the parameter set of the preset unified dead zone dynamic model.

[0065] Specifically, step S3 includes the following steps: S31: Receive the generated learning trigger command and determine the target servo valve based on the learning trigger command.

[0066] In this embodiment, the "model parameter update module" of the central controller continuously monitors the "learning task queue" generated in step S24. When there is an instruction to be processed in the queue, the module immediately reads and parses the instruction. The instruction explicitly includes a "target push point ID" field, which has been uniquely mapped to physical hardware (specific servo valves, sensors) during system initialization. By querying this mapping table, the module can determine the target servo valve for which parameter learning is required and its associated valve core displacement sensor, pressure sensor, etc. Simultaneously, the module reads auxiliary information such as "trigger type" and "associated pressure disturbance characteristics" from the instruction. This information will be used to guide the design of excitation signals and the focus of model updates in subsequent steps. For example, if the trigger type is "coupled interference trigger," special attention will be paid to the impact of pressure changes on dead zone parameters during the learning process. After the target is determined, the system can temporarily switch the valve from normal synchronous push closed-loop control to "parameter identification mode," or assign it an independent test control thread to prepare for applying a dedicated excitation signal.

[0067] S32: Based on the determined target servo valve, generate a corresponding excitation signal and apply the excitation signal to the target servo valve.

[0068] In this embodiment, to accurately identify the dynamic characteristics of the dead zone, a carefully designed and safely executable excitation signal needs to be applied to the target servo valve. This signal must meet two core requirements: First, the amplitude is small enough to ensure that the movement of the hydraulic cylinder and transformer is negligible and does not affect the safety and synchronization of the overall jacking operation; Second, the spectral components can effectively excite nonlinear dynamics related to the dead zone.

[0069] In practice, the system generates an excitation signal consisting of a stepped wave or a low-frequency swept sine wave sequence with increasing amplitude, encompassing both positive and negative directions. For example, it first generates five positive stepped voltages from +0.1V to +1.0V (0.5 to 1.5 times the estimated dead zone voltage), each step lasting 50ms; then returns to zero for 100ms; followed by five negative stepped voltages from -0.1V to -1.0V. The entire sequence is short (approximately 1.1s) and its amplitude is much lower than the normal push command. After generating this excitation signal, the system uses the controller's DA output module to superimpose it onto the current control command for the valve, or directly replaces its normal command, applying it to the driver of the target servo valve. During the application process, it is crucial to ensure that other push points in the system remain under normal synchronous closed-loop control to maintain overall stability.

[0070] S33: Collect the response data generated by the target servo valve in response to the excitation signal.

[0071] In this embodiment, throughout the entire process of applying the excitation signal, the system acquires data at a high rate (e.g., 1 ms) synchronized with the control cycle. The goal of the acquisition is to obtain the precise input-output correspondence of the servo valve under minute commands. The acquired data includes: 1. The applied excitation voltage sequence; 2. Valve spool displacement feedback signal of the target valve (used to calculate the actual speed); 3. The branch pressure of the branch where the target valve is located; 4. System main pressure.

[0072] All data acquisition must maintain high precision and synchronization, and the implementation method is exactly the same as the synchronous acquisition mechanism described in steps S11 and S13. The acquired data is temporarily stored as a dedicated "test dataset" whose time span covers the entire excitation signal application process. This dataset fully records the dynamic response of the servo valve when crossing the positive and negative dead zones and making micro-movements near the dead zones under specific operating conditions (current system pressure, oil temperature), and serves as the direct basis for subsequent parameter identification.

[0073] S34: Based on the collected response data, the parameter set of the preset unified dead zone dynamic model is collaboratively identified using a parameter identification algorithm, and the parameter set of the unified dead zone dynamic model is updated according to the identification results.

[0074] In this embodiment, this step is the core computational step in model self-learning. The unified dead-zone dynamic model can be described using a simplified dynamic model that includes the dead zone and nonlinear gain, for example: When the absolute value of the valid command is less than or equal to the dead zone threshold, the actual speed of the valve core is zero.

[0075] When the absolute value of the valid command is greater than the dead zone threshold, the actual speed of the valve core is described by the following dynamic equation:

[0076] in It is the equivalent command after taking into account the effects of pressure coupling. Dead zone threshold, For gain, The damping coefficient is... Represents unmodeled dynamics and measurement noise. It is a function that describes the effect (coupling) of system pressure on dead zone characteristics.

[0077] The parameter set that needs to be identified is and functions The coefficients in.

[0078] The identification process is as follows: First, data preprocessing and segmentation. Using the test dataset, the data is divided into positive and negative excitation segments based on the polarity of the excitation signal, since the dead zones for positive and negative directions may differ. Within each segment, the precise moment when the valve core begins to produce a measurable displacement (e.g., 5 μm) is determined; the command voltage value corresponding to this moment is the dead zone threshold for that direction. or Preliminary estimate.

[0079] Secondly, dynamic parameter identification. For data segments outside the dead zone, the model is linearized to construct a linear regression equation. Recursive least squares with a forgetting factor (FFRLS) is used to estimate parameters K and B online. The algorithm iterates through the test data points, continuously updating the parameter estimates, and finally obtains a set of convergent parameters.

[0080] At the same time, through analysis Changes and estimates The relationship between instantaneous fluctuations can update the coupling function. The coefficients in.

[0081] Finally, the model is smoothly updated. The new parameter set obtained in this identification is then applied. , with the old parameters stored in the model Perform weighted fusion:

[0082] in, A small fusion factor (e.g., 0.1) ensures that the model parameters are updated smoothly and gradually, avoiding sudden model changes due to noise or accidental disturbances in single test data, thus guaranteeing the stability of the control system. The updated model will be immediately used for the feedforward compensation calculation in step S4, completing a full "learning-application" cycle.

[0083] S4: In each control cycle, based on the updated unified dead zone dynamic model, the current operating condition, and the dynamic pressure coupling spectrum, synthesize the coupling feedforward compensation signal of the servo valve corresponding to the current push point.

[0084] Specifically, step S4 includes the following steps: S41: In each control cycle, acquire the updated unified dead zone dynamic model, current operating condition data, and dynamic pressure coupling map.

[0085] In this embodiment, this step is the data synchronization and preparation stage before performing dynamic feedforward compensation.

[0086] At the start of each control cycle (i.e., sampling cycle, e.g., 1 millisecond), the central controller's "dynamic feedforward compensation module" needs to synchronously acquire three sets of key data.

[0087] First, obtain the latest unified dead-zone dynamic model. This model is updated online in step S34, and its parameter set (including dead-zone threshold, gain, damping coefficient, and coefficients in the coupling function, etc.) is stored in shared memory.

[0088] The compensation module reads the latest parameters of the model via pointers or message passing to ensure that the compensation calculation is based on the most accurate knowledge. Next, it acquires the current operating condition data. This includes the servo valve control commands for the target jacking point i under the current control cycle, as well as the real-time measured system main pressure and oil temperature (if an oil temperature sensor is installed). This data is directly read from the operating condition parameter set continuously updated in step S1. Finally, it acquires the dynamic pressure coupling map. This map exists as a data structure of adjacency matrix and time delay matrix, maintained by step S14. The compensation module reads this map to determine the pressure coupling strength i from any jacking point j to the target point i and the estimated time delay of disturbance propagation in the current system. This step ensures that the model, real-time state, and system coupling relationship information upon which the compensation calculation depends are complete, consistent, and up-to-date.

[0089] S42: Based on the unified dead zone dynamic model and the current operating condition data, calculate the basic feedforward compensation amount of the servo valve corresponding to the current push point.

[0090] In this embodiment, the core of this step is to use the updated dead-zone model to calculate a basic compensation amount to overcome the dead zone of the valve for the control command to be issued. The calculation process is an "inverse operation" or "feedforward query" for the nonlinear dynamic model.

[0091] Specifically, the process is as follows: First, construct the equivalent command. Considering the coupling effect of pressure on the dead zone, add the current control command to the pressure correction term calculated from the current system pressure using a coupling function to obtain the equivalent command.

[0092] Secondly, the basic dead-zone compensation is queried or calculated. The purpose of the basic compensation is to ensure that the total command after compensation exactly offsets the dead zone and produces the desired dynamic response. Its calculation logic is as follows: determine whether the absolute value of the current equivalent command is less than or equal to the dead-zone threshold, and whether the equivalent command is close to zero (in a command reversal or micro-command state). If so, it indicates that the command may be "stuck" in the dead zone and requires compensation. The compensation amount should make the equivalent command cross the dead-zone threshold. A typical calculation method is: the basic compensation amount equals (the sign function value of the control command) multiplied by the dead-zone threshold, and then subtracts the pressure coupling term. After this compensation amount is superimposed on the original command, it ensures that the equivalent command at least reaches the dead-zone threshold, ensuring valve spool activation. Furthermore, the dynamic parameters (gain, damping coefficients) in the model can be used to fine-tune the shape of the compensation signal (such as pulse width) to optimize the dynamic crossover process. This basic compensation amount is the basic feedforward compensation amount directly derived from the model for the current command and operating condition.

[0093] S43: Simultaneously, based on the dynamic pressure coupling map, the pressure disturbance caused by the action of adjacent jacking points is predicted, and the coupling correction amount corresponding to the pressure disturbance is calculated.

[0094] In this embodiment, this step aims to proactively counteract pressure coupling interference from other push points, and is the core manifestation of "coupling feedforward". Its implementation relies on the predictive capabilities provided by the dynamic pressure coupling map.

[0095] The specific process is as follows: First, predict the impending pressure disturbance. The compensation module monitors or receives in real time the planned control command sequence of each adjacent jacking point in the near future (the following few control cycles) from the "synchronization control module". Based on the coupling strength (coupling from adjacent point j to the current point i) and time delay in the dynamic pressure coupling spectrum, it can predict the branch pressure disturbance that will be transmitted to the current point i due to the action of adjacent point j in the near future (e.g., the current time plus the time delay).

[0096] The prediction can be simplified to: the predicted disturbance is approximately equal to the coupling strength multiplied by the change in control command at the adjacent point j at time (current time minus time delay).

[0097] Secondly, the coupling correction is calculated. This predicted pressure disturbance will cause an instantaneous change in the current valve's effective dead zone. Based on the functional relationship between pressure and dead zone in the unified dead zone dynamic model, the change in the dead zone threshold caused by the pressure disturbance can be calculated. It is equal to the value of function f at (current system pressure plus predicted disturbance) minus the value of function f at the current system pressure. The coupling correction is the additional compensation needed to offset this future change, and can be calculated as: the coupling correction equals the negative (signed function value of the control command) multiplied by the change in the dead zone threshold. Its negative sign indicates that a reverse correction is applied in advance to "hedge" the upcoming coupling effect. This correction is an intelligent predictive compensation based on system interaction relationships.

[0098] S44: Combine the basic feedforward compensation amount with the coupling correction amount to obtain the coupled feedforward compensation signal.

[0099] In this embodiment, this step fuses the compensation amounts for its own dead zone and unexpected coupling interference into a single, comprehensive intelligent feedforward signal. The synthesis is numerically a simple algebraic sum: The coupled feedforward compensation signal equals the basic feedforward compensation amount plus the coupling correction amount.

[0100] Among them, the coupled feedforward compensation signal is the final coupled feedforward compensation signal.

[0101] In practical implementation, it should be noted that the timing of the two compensation quantities may differ slightly. The basic compensation quantity acts on the current control cycle to respond to the current command. The coupled correction quantity, on the other hand, is calculated based on predicted future disturbances and should theoretically be injected at an appropriate time before the disturbance occurs.

[0102] Therefore, a simple time-shift (lead) control may be needed during the synthesis, or the coupled feedforward compensation signal may be smoothly injected within a short time window. The synthesized signal coupled with the feedforward compensation signal is a dynamically changing voltage signal that includes both a "precision guidance" component to overcome the current valve's own time-varying dead zone and an "active defense" component to mitigate crosstalk within the immune system. This signal is then fed into step S5 for final synthesis with other control components, thereby driving the servo valve to achieve precise and robust motion control.

[0103] S5: Combine the coupling feedforward compensation signal with the friction feedforward compensation signal and displacement synchronization closed-loop feedback signal pre-generated by the corresponding servo valve to generate the final drive command, and send it to the corresponding servo valve for execution.

[0104] Specifically, step S5 includes the following steps: S51: Obtain the coupling feedforward compensation signal of the servo valve for the current push point.

[0105] In this embodiment, this step is the interface step for obtaining the calculation results from the aforementioned dynamic compensation module. Within each control cycle, when the instruction synthesis module of the central controller begins execution, it first synchronously reads the calculated coupling feedforward compensation signal for the servo valve at the current push point from the output buffer of the dynamic feedforward compensation module (which executes step four) via a predefined data interface. This signal is a numerical value with physical dimensions, representing the amount of voltage compensation injected to overcome the valve's dead zone and cancel predicted coupling interference.

[0106] To ensure data consistency and real-time performance, this read operation occurs at a fixed timing point in the control cycle, typically after sensor data acquisition and core control algorithm calculations are completed, but before final instruction synthesis. The module verifies the timeliness of the signal and may perform anti-pulse interference filtering on it before temporarily storing it, awaiting synthesis with other control components.

[0107] S52: Simultaneously, acquire the friction feedforward compensation signal and displacement synchronization closed-loop feedback signal pre-generated for the servo valve.

[0108] In this embodiment, this step acquires two additional key control components for the final synthesis in parallel.

[0109] First, the friction feedforward compensation signal is acquired. This signal is typically generated by a model-based friction observer or friction feedforward calculator. This module pre-calculates a voltage compensation amount to counteract the expected frictional force based on the target velocity at the current jacking point and a preset friction model.

[0110] The instruction synthesis module reads this signal directly from the output of the friction compensation module via the internal data bus.

[0111] Secondly, the displacement synchronization closed-loop feedback signal is acquired. This signal is the output of the system's displacement synchronization closed-loop controller. The synchronization closed-loop controller uses the displacement of the master point or other pushing points as a reference, calculates the displacement tracking deviation of the current point in real time, and generates a feedback control voltage to correct this deviation through proportional-integral-derivative operations.

[0112] The instruction synthesis module reads this real-time calculation result from the output register of the controller. The acquisition of these two signals must be strictly time-synchronized with the acquisition of the coupling feedforward compensation signal in step fifty-one to ensure that all signals used for synthesis correspond to the same state of the system at the same moment, and to avoid introducing additional control errors or jitter due to signal asynchrony.

[0113] S53: The coupling feedforward compensation signal, the friction feedforward compensation signal, and the displacement synchronization closed-loop feedback signal are superimposed and synthesized to generate the final drive command.

[0114] In this embodiment, this step is the fusion point of the core control commands. The synthesis operation is numerically an algebraic superposition of three signals, that is, directly adding the coupled feedforward compensation signal, the friction feedforward compensation signal, and the displacement synchronization closed-loop feedback signal.

[0115] In actual software implementation, this is usually a simple floating-point addition operation. However, to ensure system safety and stability, the following processing steps are often included before and after synthesis: 1. Limiting Processing: Immediately after synthesis, the final drive command is output-limited to ensure its absolute value does not exceed the maximum input voltage allowed by the servo valve amplifier. If it does, it is limited to the maximum or minimum allowed value.

[0116] 2. Rate of change limit: To prevent sudden changes in commands from impacting the servo valve, the rate of change of the final drive command in adjacent cycles can be limited to make it change smoothly.

[0117] 3. Enable logic: In abnormal system conditions, the synthesized logic can be overridden, and the final drive instruction is forced to be zeroed or set to a safe value.

[0118] The final drive command generated is the total control voltage that is intended to be applied to the target servo valve within the current control cycle, which integrates feedforward compensation and feedback correction. It directly determines the desired valve core opening and cylinder movement.

[0119] S54: Send the final drive command to the corresponding servo valve to drive it to perform the corresponding control action.

[0120] In this embodiment, this step is the final output stage from the digital control quantity to the physical actuator.

[0121] The synthesized final drive instruction is a digital quantity stored in the controller's memory. The system converts the final drive instruction into a corresponding analog voltage signal through its digital-to-analog converter output module.

[0122] The specific process is as follows: The controller's central processing unit writes the value of the final drive instruction into the register of the digital-to-analog converter output channel corresponding to the specified push point. The digital-to-analog converter chip then converts this digital quantity into a precise analog voltage.

[0123] This analog voltage signal is transmitted via a shielded cable to the servo amplifier of the corresponding servo valve. The servo amplifier amplifies the voltage command and converts it into a control current that drives the proportional electromagnet of the servo valve. Under the action of electromagnetic force, the valve core is displaced, thereby regulating the direction and flow of hydraulic oil, driving the push cylinder to generate precise force and displacement. The timing of the output command is strictly triggered by the controller's hardware timer interrupt to ensure the timeliness of periodic execution.

[0124] Meanwhile, the system can be configured with a readback mechanism, which samples the actual output analog voltage through the analog-to-digital conversion module and compares it with the command value to verify the normal operation of the output channel, thus completing a complete control cycle.

[0125] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0126] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, including read-only memory (ROM), random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), one-time programmable read-only memory (OTPROM), electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, disk storage, magnetic tape storage, or any other computer-readable medium capable of carrying or storing data.

[0127] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

Claims

1. A precision group control synchronous jacking control method for an ultra-large main transformer, characterized in that, The control method includes the following steps: The response deviation is calculated based on the working condition parameter set of each jacking point collected synchronously, and a dynamic pressure coupling spectrum is constructed based on the branch pressure and main pressure in the working condition parameter set. When the response deviation meets the first triggering condition and / or the dynamic pressure coupling map meets the second triggering condition, a learning triggering command for the current push point is generated. In response to the learning trigger command, based on the response data obtained by applying an excitation signal to the servo valve corresponding to the current push point, the parameter set of the preset unified dead zone dynamic model is collaboratively identified and updated. In each control cycle, based on the updated unified dead zone dynamic model, the current operating condition, and the dynamic pressure coupling spectrum, a coupling feedforward compensation signal for the servo valve corresponding to the current push point is synthesized. The coupling feedforward compensation signal is combined with the friction feedforward compensation signal and displacement synchronization closed-loop feedback signal pre-generated by the corresponding servo valve to generate the final drive command, which is then sent to the corresponding servo valve for execution.

2. The precision group control synchronous jacking control method for an extra-large main transformer according to claim 1, characterized in that, The first triggering condition is that the statistical mean of the response deviation exceeds the adaptive threshold over N consecutive control cycles.

3. The precision group control synchronous jacking control method for an extra-large main transformer according to claim 1, characterized in that, The second triggering condition is that the magnitude of the main pressure mutation exceeds a preset disturbance threshold, and the correlation coefficient between the main pressure mutation and the change in the response characteristics of the current push point exceeds a preset correlation threshold.

4. The precision group control synchronous jacking control method for an extra-large main transformer according to claim 1, characterized in that, The response deviation is calculated based on the operating condition parameter set of each jacking point acquired synchronously, and a dynamic pressure coupling spectrum is constructed based on the branch pressure and main pressure in the operating condition parameter set, including: Based on the working condition parameter set of each push point collected synchronously, the servo valve control command and valve core displacement feedback signal are extracted. Based on the servo valve control command and valve core displacement feedback signal, the command-speed response deviation of each pushing point is calculated; Simultaneously, the branch pressure and main system pressure of each jacking point are extracted from the set of operating parameters; Based on the time series of the branch pressure and the main system pressure, the transmission relationship and coupling strength between pressure fluctuations are analyzed and calculated, and the dynamic pressure coupling spectrum is constructed.

5. The precision group control synchronous jacking control method for an extra-large main transformer according to claim 4, characterized in that, The step of generating a learning trigger instruction for the current push point when the response deviation meets the first trigger condition and / or the dynamic pressure coupling map meets the second trigger condition includes: Receive the calculated response deviation and the constructed dynamic pressure coupling map; Based on the response deviation, determine whether the first triggering condition is met, and output the first determination result; Simultaneously, based on the dynamic pressure coupling spectrum, it is determined whether the second triggering condition is met, and the second determination result is output; When the first judgment result and / or the second judgment result indicate that the triggering condition is met, a learning trigger instruction for the current push point is generated.

6. The precision group control synchronous jacking control method for an extra-large main transformer according to claim 5, characterized in that, The response to the learning trigger command, based on the response data obtained by applying an excitation signal to the servo valve corresponding to the current push point, involves collaborative identification and updating of the parameter set of a preset unified dead-zone dynamic model, including: Receive the generated learning trigger command, and determine the target servo valve based on the learning trigger command; Based on the determined target servo valve, a corresponding excitation signal is generated and applied to the target servo valve; Collect the response data generated by the target servo valve in response to the excitation signal; Based on the collected response data, the parameter set of the preset unified dead zone dynamic model is collaboratively identified using a parameter identification algorithm, and the parameter set of the unified dead zone dynamic model is updated according to the identification results.

7. The precision group control synchronous jacking control method for an extra-large main transformer according to claim 6, characterized in that, In each control cycle, based on the updated unified dead-zone dynamic model, the current operating condition, and the dynamic pressure coupling spectrum, a coupling feedforward compensation signal for the servo valve corresponding to the current push point is synthesized, including: In each control cycle, the updated unified dead zone dynamic model, current operating condition data, and dynamic pressure coupling map are acquired; Based on the unified dead zone dynamic model and the current operating condition data, the basic feedforward compensation amount of the servo valve corresponding to the current push point is calculated. Simultaneously, based on the dynamic pressure coupling map, the pressure disturbance caused by the action of adjacent jacking points is predicted, and the coupling correction amount corresponding to the pressure disturbance is calculated. The basic feedforward compensation amount and the coupling correction amount are combined to obtain the coupled feedforward compensation signal.

8. The precision group control synchronous jacking control method for an extra-large main transformer according to claim 7, characterized in that, The step of synthesizing the coupled feedforward compensation signal with the friction feedforward compensation signal and the displacement synchronization closed-loop feedback signal pre-generated by the corresponding servo valve to generate the final drive command and sending it to the corresponding servo valve for execution includes: Obtain the coupling feedforward compensation signal of the servo valve for the current push point; Simultaneously, the friction feedforward compensation signal and displacement synchronization closed-loop feedback signal pre-generated for the servo valve are acquired; The coupling feedforward compensation signal, the friction feedforward compensation signal, and the displacement synchronous closed-loop feedback signal are superimposed and synthesized to generate the final drive command. The final drive command is sent to the corresponding servo valve to drive it to perform the corresponding control action.

Images