Adaptive slip surface control method, device, equipment, storage medium and product

Abstract

The application discloses a kind of self-adapting sliding surface control method, device, equipment, storage medium and product, involve equipment control technical field, the self-adapting sliding surface control method includes: obtaining the displacement tracking state information of piezoelectric displacement platform;Displacement tracking state information is input to integrated with pre-trained fourier neural operator self-adapting sliding surface controller, and output dynamic adjustment self-adapting control parameter;Based on displacement tracking state information and self-adapting control parameter, control voltage is generated by the self-adapting sliding surface controller;Control voltage is applied to piezoelectric displacement platform, to drive piezoelectric displacement platform to track target displacement.The application has integrated fourier neural operator self-adapting sliding surface control architecture, when piezoelectric platform needs to be compensated, real-time processing is carried out by fourier neural operator, and the adaptive control parameter adjustment amount matched with system state is output, parameter adaptive adjustment is realized, and finally high-precision displacement tracking control is completed.

Description

Technical Field

[0001] This application relates to the field of equipment control technology, and in particular to adaptive sliding surface control methods, devices, equipment, storage media, and products. Background Technology

[0002] In high-end industrial fields such as precision manufacturing, semiconductor testing, and biomedical engineering, piezoelectric ceramic actuators are widely used as core actuation components for precision motion platforms due to their extremely high resolution, fast response speed, and huge output. However, the inherent nonlinear characteristics of piezoelectric materials, such as hysteresis and creep, as well as problems such as parameter perturbations and external disturbances during system operation, can severely degrade their positioning accuracy and tracking performance, becoming a bottleneck restricting their application in high-precision scenarios.

[0003] Existing technologies typically employ a strategy combining feedforward compensation and feedback control. For example, feedforward compensation is achieved by establishing an accurate inverse hysteresis model, combined with traditional proportional-integral-derivative (PID) feedback control. However, hysteresis exhibits complex characteristics such as strong nonlinearity, multiple loops, and rate dependence, making accurate modeling difficult. Furthermore, the inverse model is often difficult to obtain analytically, resulting in limited compensation effectiveness and ultimately low control accuracy. Summary of the Invention

[0004] The main objective of this application is to provide an adaptive sliding surface control method, device, equipment, storage medium, and product, which aims to solve the technical problem of low control accuracy when compensating for piezoelectric platforms.

[0005] To achieve the above objectives, this application proposes an adaptive sliding surface control method, the method comprising:

[0006] Obtain the displacement tracking status information of the piezoelectric displacement platform; The displacement tracking state information is input into an adaptive sliding surface controller that integrates a pre-trained Fourier neural operator, and the dynamically adjusted adaptive control parameters are output. Based on the displacement tracking state information and the adaptive control parameters, the adaptive sliding surface controller generates a control voltage; The control voltage is applied to the piezoelectric displacement platform to drive the piezoelectric displacement platform to track the target displacement.

[0007] In one embodiment, the step of acquiring the displacement tracking status information of the piezoelectric displacement platform includes: Obtain the actual and target displacements of the piezoelectric displacement platform within the current control cycle; The difference between the actual displacement and the target displacement is calculated to obtain the displacement tracking status information.

[0008] In one embodiment, the step of inputting the displacement tracking state information to an adaptive sliding surface controller integrated with a pre-trained Fourier neural operator and outputting dynamically adjusted adaptive control parameters includes: The input voltage and output displacement data of the piezoelectric displacement platform under various working conditions were collected to construct a training dataset. The initial model containing Fourier neural operators is trained based on the training dataset to obtain a pre-trained model that can characterize the hysteresis dynamics of the piezoelectric displacement platform. The pre-trained model is validated, and when the preset accuracy condition is met, the operators in it are extracted as the pre-trained Fourier neural operators.

[0009] In one embodiment, the step of training the initial model containing Fourier neural operators based on the training dataset includes: The input voltage sequence in the training data is combined with the corresponding time information to generate the model input data; The model input data is input into the Fourier neural operator and processed through multiple cascaded Fourier layers. The operations performed by each Fourier layer include: frequency domain transformation of the input, linear processing of selected frequency domain components, inverse transformation back to the time domain and fusion with bypass branches, and application of nonlinear activation. The parameters of the Fourier neural operator are adjusted by an optimization algorithm to minimize the difference between the model's predicted displacement and the actual displacement, and the trained initial model containing the Fourier neural operator is output.

[0010] In one embodiment, the step of validating the pre-trained model includes: The test data that was not used in the training is input into the pre-trained model to obtain the predicted displacement command; The predicted displacement command is applied to the piezoelectric displacement platform, and the actual displacement response is measured. The verification error is calculated based on the actual displacement response and the true displacement corresponding to the test data. If the verification error does not meet the preset accuracy threshold, the process returns to the step of training the initial model containing Fourier neural operators based on the training dataset until the verification error meets the preset accuracy threshold.

[0011] In one embodiment, the step of generating a control voltage by the adaptive slip surface controller includes: Based on preset initial control parameters, an initial control voltage is generated by the adaptive sliding surface controller and applied to the piezoelectric displacement platform to obtain the initial actual displacement; Based on the difference between the initial actual displacement and the target displacement, the displacement tracking error and the sliding surface error are obtained; The displacement tracking error and sliding surface error are input into the Fourier neural operator, and the output is used to adjust the parameters of the controller. The control parameters of the adaptive slip surface controller are updated based on the parameter adjustment amount, and an updated control voltage for the adaptive slip surface controller is generated based on the updated control parameters. The process involves verifying whether the updated adaptive sliding surface controller is suitable for the current operating conditions. If it is not suitable, the process returns to the preset initial control parameters, generates an initial control voltage through the adaptive sliding surface controller, applies it to the piezoelectric displacement platform, and obtains the initial actual displacement. This process continues until the updated adaptive sliding surface controller is suitable for the current operating conditions.

[0012] Furthermore, to achieve the above objectives, this application also proposes an adaptive sliding surface control device, which includes: The acquisition module is used to acquire the displacement tracking status information of the piezoelectric displacement platform; The output module is used to input the displacement tracking state information to an adaptive sliding surface controller integrated with a pre-trained Fourier neural operator and output dynamically adjusted adaptive control parameters. The generation module is used to generate a control voltage from the adaptive sliding surface controller based on the displacement tracking state information and the adaptive control parameters. A drive module is used to apply the control voltage to the piezoelectric displacement platform to drive the piezoelectric displacement platform to track the target displacement.

[0013] In addition, to achieve the above objectives, this application also proposes an adaptive sliding surface control device, the device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the adaptive sliding surface control method as described above.

[0014] In addition, to achieve the above objectives, this application also proposes a storage medium, which is a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the steps of the adaptive sliding surface control method as described above.

[0015] In addition, to achieve the above objectives, this application also provides a computer program product, which includes a computer program that, when executed by a processor, implements the steps of the adaptive sliding surface control method as described above.

[0016] One or more technical solutions proposed in this application have at least the following technical effects: In related technologies, a strategy combining feedforward compensation and feedback control is employed. For example, feedforward compensation is performed by establishing an accurate hysteresis inverse model, combined with traditional proportional-integral-derivative (PID) feedback control. However, hysteresis exhibits complex characteristics such as strong nonlinearity, multiple loops, and rate correlation, making accurate modeling difficult, and the inverse model is often difficult to obtain analytically, resulting in limited compensation effects and low final control accuracy. In contrast, this application obtains the displacement tracking state information of a piezoelectric displacement platform; inputs the displacement tracking state information into an adaptive sliding surface controller integrated with a pre-trained Fourier neural operator, and outputs dynamically adjusted adaptive control parameters; based on the displacement tracking state information and the adaptive control parameters, the adaptive sliding surface controller generates a control voltage; and applies the control voltage to the piezoelectric displacement platform to drive it to track the target displacement. Understandably, this application employs an adaptive sliding surface control architecture integrated with pre-trained Fourier neural operators. When compensation is required for the complex hysteresis nonlinearity and time-varying disturbances of the piezoelectric platform, the Fourier neural operators process the displacement tracking state information in real time, dynamically generating and outputting adaptive control parameter adjustments that match the current system state. This achieves accurate online characterization and adaptive parameter adjustment of the piezoelectric platform's hysteresis dynamics. Therefore, the control voltage generated based on the real-time updated adaptive control parameters and displacement tracking state information can more accurately counteract the hysteresis effect and suppress disturbances. Consequently, the optimized control signal required to drive the piezoelectric platform to track the target displacement is determined, ultimately achieving high-precision and highly robust displacement tracking control. Attached Figure Description

[0017] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating an embodiment of the adaptive sliding surface control method of this application. Figure 2 This is a schematic diagram of the control communication system of an embodiment of the adaptive sliding surface control method of this application; Figure 3 This is a conceptual diagram of the control system of an embodiment of the adaptive sliding surface control method of this application; Figure 4 This is a schematic diagram of the module structure of the adaptive slip surface control device according to an embodiment of this application; Figure 5 This is a schematic diagram of the device structure of the hardware operating environment involved in the adaptive sliding surface control method in the embodiments of this application.

[0020] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0021] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0022] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0023] The main solution in this application embodiment is: Obtain the displacement tracking status information of the piezoelectric displacement platform; The displacement tracking state information is input into an adaptive sliding surface controller that integrates a pre-trained Fourier neural operator, and the dynamically adjusted adaptive control parameters are output. Based on the displacement tracking state information and the adaptive control parameters, the adaptive sliding surface controller generates a control voltage; The control voltage is applied to the piezoelectric displacement platform to drive the piezoelectric displacement platform to track the target displacement.

[0024] In this embodiment, the adaptive sliding surface control device is used as the execution subject. For ease of description, it will be referred to as "device" in detail below.

[0025] Existing technologies employ a strategy combining feedforward compensation and feedback control. For example, feedforward compensation is achieved by establishing an accurate hysteresis inverse model, combined with traditional proportional-integral-derivative (PID) feedback control. However, hysteresis exhibits complex characteristics such as strong nonlinearity, multiple loops, and rate dependence, making accurate modeling difficult. Furthermore, the inverse model is often difficult to obtain analytically, resulting in limited compensation effectiveness and ultimately low control accuracy.

[0026] This application provides a solution that employs an adaptive sliding surface control architecture integrated with a pre-trained Fourier neural operator. When compensation is needed for the complex hysteresis nonlinearity and time-varying disturbances of the piezoelectric platform, the Fourier neural operator processes the displacement tracking state information in real time, dynamically generating and outputting adaptive control parameter adjustments that match the current system state. This achieves accurate online characterization and adaptive parameter adjustment of the piezoelectric platform's hysteresis dynamics. Therefore, the control voltage generated based on the real-time updated adaptive control parameters and displacement tracking state information can more accurately counteract the hysteresis effect and suppress disturbances. This further determines the optimized control signal required to drive the piezoelectric platform to track the target displacement, ultimately achieving high-precision and highly robust displacement tracking control.

[0027] Based on this, embodiments of this application provide an adaptive sliding surface control method, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the adaptive sliding surface control method of this application.

[0028] In this embodiment, the adaptive sliding surface control method includes steps S10~S40: Step S10: Obtain the displacement tracking status information of the piezoelectric displacement platform; It should be noted that a piezoelectric displacement platform refers to a precision motion device that includes a piezoelectric ceramic actuator and its mechanical structure, and its displacement is controlled by an applied voltage. Displacement tracking state information is a state quantity used to characterize the deviation between the actual motion of the platform and the desired target, including but not limited to the instantaneous error between the actual displacement and the target displacement, and the differential or integral quantity of the error.

[0029] Understandably, this step provides the necessary feedback information for the control closed loop, which forms the basis for subsequent accurate error compensation and control decisions.

[0030] Step S20: Input the displacement tracking state information into an adaptive sliding surface controller integrated with a pre-trained Fourier neural operator, and output dynamically adjusted adaptive control parameters. It should be noted that the adaptive sliding surface controller is a composite controller combining sliding mode variable structure control and adaptive control law. The Fourier neural operator is a neural network-based operator learning method that learns the dynamic mapping of complex systems through global convolution operations in the Fourier frequency domain. Pre-training refers to training the operator in advance using historical input voltage and output displacement datasets, enabling it to learn and memorize the dynamic characteristics of the piezoelectric platform, such as hysteresis and nonlinearity. Integration refers to embedding the trained Fourier neural operator as part of the adaptive law into the controller structure.

[0031] Understandably, this step utilizes the powerful nonlinear fitting and generalization capabilities of the Fourier neural operator, which can calculate the most suitable adaptive control parameters (such as switching gain, parameter estimation update rate, etc.) online and dynamically based on the real-time system state, thereby replacing the traditional fixed or empirically set parameters and enabling the controller to have the intelligent adjustment capability to cope with hysteresis, time-varying and unknown disturbances.

[0032] Step S30: Based on the displacement tracking state information and the adaptive control parameters, the adaptive sliding surface controller generates a control voltage; It should be noted that the control voltage is the electrical signal ultimately applied to the piezoelectric displacement platform drive circuit. The generation process specifically includes: defining the sliding surface function based on the displacement tracking state information; calculating or updating the switching term gain and equivalent control term in the sliding mode control law based on the dynamically adjusted adaptive control parameters; and synthesizing the real-time control voltage value by combining this information through the sliding mode control algorithm.

[0033] Understandably, this step is the execution phase of the control strategy. It utilizes the optimized parameters obtained in the previous step, combined with the classic sliding mode control framework, to calculate the control command that forces the system state to converge along a predetermined sliding surface. Since the control parameters have been dynamically optimized by Fourier neural operators, the generated control voltage, while ensuring strong robustness, effectively reduces the high-frequency chattering inherent in traditional sliding mode control, achieving a smoother and more precise control output.

[0034] Step S40: Apply the control voltage to the piezoelectric displacement platform to drive the piezoelectric displacement platform to track the target displacement.

[0035] It should be noted that "applying" refers to converting the digital control signal into an analog voltage through a high-precision digital-to-analog converter and inputting it to the power amplifier of the piezoelectric actuator, ultimately acting on the piezoelectric ceramic. "Driving" refers to the voltage signal causing deformation of the piezoelectric material, which in turn drives the load to produce mechanical displacement.

[0036] Understandably, this step completes the final physical implementation of the control action. The control voltage obtained through the aforementioned intelligent decision-making and optimization calculations is directly applied to the controlled object, aiming to accurately compensate for hysteresis and overcome disturbances, enabling the platform's actual displacement to quickly and accurately follow changes in the target displacement, thereby achieving nanometer-level high-precision trajectory tracking control of the target.

[0037] This embodiment provides an adaptive sliding surface control method, which adopts an adaptive sliding surface control architecture integrating a pre-trained Fourier neural operator. When it is necessary to compensate for the complex hysteresis nonlinearity and time-varying disturbances of the piezoelectric platform, the Fourier neural operator processes the displacement tracking state information in real time, dynamically generates and outputs adaptive control parameter adjustment amounts that match the current system state. This achieves accurate online characterization and adaptive parameter adjustment of the piezoelectric platform's hysteresis dynamics. Therefore, the control voltage generated based on the real-time updated adaptive control parameters and displacement tracking state information can more accurately counteract the hysteresis effect and suppress disturbances. This further determines the optimized control signal required to drive the piezoelectric platform to track the target displacement, ultimately achieving high-precision and highly robust displacement tracking control.

[0038] For example, Figure 2 This diagram illustrates the control and communication system provided in this application embodiment. The system architecture mainly includes a computer (host computer), a high-precision digital-to-analog converter, and a piezoelectric displacement platform (piezoelectric system). The computer, as the control core, is responsible for running a predictive model containing Fourier neural operators and an adaptive sliding surface control algorithm. The high-precision digital-to-analog converter, as the key signal interface, connects the computer and the piezoelectric displacement platform. Specifically, it includes a D / A converter and an A / D converter, and works in conjunction with a voltage amplifier and a position sensor: In the downlink path of the control signal, the control commands (digital quantities) generated by the computer are converted into analog voltage signals by the D / A converter, then amplified by the voltage amplifier, and applied to the piezoelectric ceramic actuator, thereby driving the piezoelectric displacement platform to generate mechanical displacement; in the uplink path of the feedback signal, the position sensor installed on the displacement platform detects the actual displacement in real time and outputs an analog feedback signal. This signal is converted into a digital quantity by the A / D converter and then transmitted back to the computer, thus forming a complete closed-loop control circuit. This schematic diagram clearly illustrates the flow and interaction of signals (including input voltage, output displacement, and control commands) during the data acquisition and online control phases, demonstrating the hardware foundation for the combination of data-driven modeling and real-time control in this application.

[0039] In one feasible implementation, the step of acquiring the displacement tracking status information of the piezoelectric displacement platform includes: Obtain the actual and target displacements of the piezoelectric displacement platform within the current control cycle; The difference between the actual displacement and the target displacement is calculated to obtain the displacement tracking status information.

[0040] It should be noted that the control cycle refers to the fixed or variable time interval during which the digital controller samples, calculates, and updates its output. The actual displacement is the physical position measured in real time by a high-precision displacement sensor (including but not limited to a grating ruler, laser interferometer, or capacitive sensor) mounted on a piezoelectric displacement platform. The target displacement is the desired position command generated based on the upper-level task planning (e.g., the focusing trajectory in wafer inspection). The difference is the algebraic difference between the target displacement and the actual displacement, which directly quantifies the tracking error at the current moment.

[0041] Understandably, this implementation method forms the foundation of the feedback control closed loop. By accurately acquiring and calculating the displacement error, the actual output of the system is compared with the desired target, thus providing the controller with a clear adjustment direction and quantitative basis. This error-based control is a prerequisite for eliminating steady-state deviations and achieving precise tracking.

[0042] In one feasible implementation, the step of inputting the displacement tracking state information to an adaptive sliding surface controller integrated with a pre-trained Fourier neural operator and outputting dynamically adjusted adaptive control parameters includes: The input voltage and output displacement data of the piezoelectric displacement platform under various working conditions were collected to construct a training dataset. The initial model containing Fourier neural operators is trained based on the training dataset to obtain a pre-trained model that can characterize the hysteresis dynamics of the piezoelectric displacement platform. The pre-trained model is validated, and when the preset accuracy condition is met, the operators in it are extracted as the pre-trained Fourier neural operators.

[0043] It should be noted that various operating conditions, including but not limited to platform operating states under different frequencies, amplitudes, temperatures, or loads, are included to ensure the dataset covers the system's main nonlinear and dynamic characteristics. Input voltage is the sequence of control signals driving the platform's motion. Output displacement data is the platform's position response sequence, acquired synchronously with the input voltage and measured by sensors. The training dataset is a collection of numerous pairs of input voltage and output displacement data samples used for supervised learning of the model. The initial model containing Fourier neural operators refers to a neural network model with a pre-defined network structure (e.g., number of Fourier layers, number of channels) but whose parameters have not been trained. Hysteresis dynamics specifically refers to the complex nonlinear dynamic relationship, exhibiting memory and path dependence, in relation to the piezoelectric platform's output displacement relative to the input voltage. A pre-trained model refers to a neural network model whose internal parameters have been optimized after training, enabling it to map from input voltage to predicted displacement with high accuracy. Pre-defined accuracy conditions are pre-set performance metric thresholds used to determine the model's suitability, such as the root mean square error of prediction on the test set.

[0044] Understandably, this implementation method represents the offline preparation phase for constructing the core module of the intelligent controller. By systematically collecting experimental data covering the operating range, sufficient learning material can be provided for the Fourier neural operator, enabling it to accurately extract and internalize the complex hysteresis characteristics of the piezoelectric platform from the data. Training and rigorous validation of the model ensure that the obtained operator possesses good generalization ability and reliability, providing accurate dynamic characteristic references for subsequent online control. This process transforms the hysteresis nonlinearity problem, which is difficult to model precisely analytically, into a data-driven, learnable function approximation problem, laying a solid foundation for achieving high-precision online adaptive compensation.

[0045] In one feasible implementation, the step of training the initial model containing Fourier neural operators based on the training dataset includes: The input voltage sequence in the training data is combined with the corresponding time information to generate the model input data; The model input data is input into the Fourier neural operator and processed through multiple cascaded Fourier layers. The operations performed by each Fourier layer include: frequency domain transformation of the input, linear processing of selected frequency domain components, inverse transformation back to the time domain and fusion with bypass branches, and application of nonlinear activation. The parameters of the Fourier neural operator are adjusted by an optimization algorithm to minimize the difference between the model's predicted displacement and the actual displacement, and the trained initial model containing the Fourier neural operator is output.

[0046] It should be noted that the input voltage sequence is a set of voltage values ​​arranged in chronological order. Time information refers to the time data or time step identifier corresponding to each voltage sampling point. The model input data is a multidimensional tensor constructed by concatenating, stacking, or otherwise fusing the voltage sequence and time information. A Fourier neural operator is a neural network architecture specifically designed for learning mappings between infinite-dimensional function spaces. Multiple cascaded Fourier layers refer to a multi-layered structure where the output of the previous layer serves as the input of the next layer. Frequency domain transformation is a mathematical operation that converts an input signal from the time or spatial domain to the frequency domain, specifically including but not limited to the Fast Fourier Transform. Selected frequency domain components refer to components within a specific frequency range chosen for further processing in the frequency domain representation, such as low-frequency components. Linear processing refers to applying a learnable linear transformation to the selected frequency domain components, such as multiplying with a weight matrix. Inverse transformation is the operation of converting the processed frequency domain signal back to the original domain (time domain). Bypass branch refers to a parallel path that performs another learnable linear transformation on the input directly without frequency domain processing. Fusion refers to operations such as element-wise addition or concatenation of the result after inverse transformation with the output of the bypass branch. Nonlinear activation refers to applying a nonlinear function, such as ReLU or GELU, to the fused result to introduce nonlinear expressive power. Optimization algorithms are computational methods used to iteratively update network parameters to minimize the loss function, including but not limited to stochastic gradient descent and its variants (such as Adam). Model-predicted displacement is the displacement estimate calculated by the initial model based on the input data. True displacement is the displacement value actually measured by sensors in the training dataset. Measured differences include, but are not limited to, mean squared error or mean absolute error.

[0047] Understandably, this implementation method represents the core training process for constructing a high-precision hysteresis prediction model. By combining time information with voltage sequences, the model can learn the rate-dependent characteristics of hysteresis. The unique frequency domain processing mechanism of the Fourier neural operator enables it to efficiently capture and model long-range dependencies and global patterns in signals, which is crucial for characterizing hysteresis loops. Cascaded Fourier layers progressively abstract, gradually constructing a nonlinear mapping from input voltage to complex displacement output. The optimization process systematically adjusts all learnable parameters, making the model's predicted output continuously approximate the real physical response, ultimately resulting in a neural network model that accurately characterizes the hysteresis dynamics of the piezoelectric platform. This data-driven modeling method overcomes the difficulties of traditional analytical modeling, providing a reliable dynamic characteristic descriptor for subsequent precise online control.

[0048] In one feasible implementation, the step of validating the pre-trained model includes: The test data that was not used in the training is input into the pre-trained model to obtain the predicted displacement command; The predicted displacement command is applied to the piezoelectric displacement platform, and the actual displacement response is measured. The verification error is calculated based on the actual displacement response and the true displacement corresponding to the test data. If the verification error does not meet the preset accuracy threshold, the process returns to the step of training the initial model containing Fourier neural operators based on the training dataset until the verification error meets the preset accuracy threshold.

[0049] It should be noted that the test data is a subset of data reserved from the overall dataset that has not been used for model parameter updates, used to evaluate the model's generalization performance. The predicted displacement command is the displacement trajectory expected to be executed by the platform, calculated by the pre-trained model based on the input voltage sequence in the test data. Application refers to the process of converting the digital predicted displacement command into an analog voltage signal and driving the platform through a digital-to-analog converter and control circuit. The actual displacement response is the actual motion trajectory measured by the displacement sensor after the platform receives the predicted displacement command. The validation error is an indicator used to quantify the consistency between the model's predictive performance and the platform's actual response; its specific forms include, but are not limited to, root mean square error or maximum absolute error. The preset accuracy threshold is a numerical standard pre-set according to the accuracy requirements of the specific application scenario.

[0050] It is understood that this implementation constitutes a complete closed loop of model performance evaluation and iterative optimization. By directly applying the model's predicted output on unseen data to the physical system and comparing the resulting actual response with real data, the accuracy of the model's representation of system dynamics and its usability in practical control can be most realistically verified. This process ensures that the final Fourier neural operator can not only fit hysteresis characteristics at the data level but also reliably be embedded into subsequent real-time control loops. If the verification fails, the process returns to the training step, optimizing by adjusting the model structure, hyperparameters, or supplementing training data until the model performance meets the accuracy requirements of practical applications, thereby ensuring the reliability and effectiveness of subsequent adaptive control strategies.

[0051] In one feasible implementation, the step of generating the control voltage by the adaptive slip surface controller includes: Based on preset initial control parameters, an initial control voltage is generated by the adaptive sliding surface controller and applied to the piezoelectric displacement platform to obtain the initial actual displacement; Based on the difference between the initial actual displacement and the target displacement, the displacement tracking error and the sliding surface error are obtained; The displacement tracking error and sliding surface error are input into the Fourier neural operator, and the output is used to adjust the parameters of the controller. The control parameters of the adaptive slip surface controller are updated based on the parameter adjustment amount, and an updated control voltage for the adaptive slip surface controller is generated based on the updated control parameters. The process involves verifying whether the updated adaptive sliding surface controller is suitable for the current operating conditions. If it is not suitable, the process returns to the preset initial control parameters, generates an initial control voltage through the adaptive sliding surface controller, applies it to the piezoelectric displacement platform, and obtains the initial actual displacement. This process continues until the updated adaptive sliding surface controller is suitable for the current operating conditions.

[0052] It should be noted that the initial control parameters are a set of controller parameters pre-set according to the target control effect, including but not limited to proportional, integral, and derivative coefficients, as well as the initial gain value of the sliding mode control, used to initiate the control process. The initial control voltage is the driving voltage within the first control cycle calculated by the adaptive sliding surface controller based on the initial control parameters and the current target displacement. The initial actual displacement is the actual position of the piezoelectric displacement platform first measured by the displacement sensor after the initial control voltage is applied. The displacement tracking error is the instantaneous difference between the initial actual displacement and the target displacement. The sliding surface error is a comprehensive error metric calculated based on the displacement tracking error and its higher-order derivatives (such as the error change rate) according to a preset sliding surface function, used to characterize the degree to which the system state deviates from the desired dynamic trajectory. The parameter adjustment amount is the numerical change amount of the controller's internal parameters calculated in real time by the Fourier neural operator based on the input error signal, used to correct the error. For example, it is the correction term for the estimated parameters in the adaptive law or the adjustment amount for the sliding mode gain. The current working condition refers to the working state of the piezoelectric displacement platform at a specific moment, which is defined by a variety of factors, including but not limited to the platform's current speed, acceleration, external disturbance intensity, temperature environment, and the hysteresis state of the piezoelectric material itself.

[0053] Understandably, this implementation describes a complete, iterative online self-tuning and optimization process for the controller. This process initiates control with a safe set of initial parameters and evaluates the initial error by acquiring the response after the first control. Utilizing the intelligent processing capabilities of Fourier neural operators, the system can analyze error patterns and generate targeted parameter adjustment instructions, thereby achieving fine-tuning of the controller parameters online. By cyclically executing a series of steps—control, measurement, analysis, and parameter updates—the system can gradually match the controller's performance to the current complex dynamic operating conditions. This method achieves a smooth transition from "open-loop preset" to "closed-loop adaptive," effectively avoiding control instability or performance degradation caused by improper initial parameter settings. It ensures that the system can, under various unknown or time-varying conditions, learn and adjust itself, ultimately converging to a stable control state that meets high-precision tracking requirements.

[0054] For example, refer to Figure 3 The implementation process of generating control voltage by the adaptive sliding surface controller can be specifically described as follows: A reference command (i.e., the target position value) is input to the control system. First, the SMC control module (sliding mode control module), which integrates preset initial parameters, calculates a preliminary control signal based on this input and the actual position feedback from the position sensor. This control signal is converted into a driving voltage by the driving device and applied to the controlled object PEAS (piezoelectric displacement platform). The actual displacement output by the platform is measured by the position sensor, and the obtained position information is fed back to the error judgment module and the error measurement module, thereby obtaining the displacement tracking error and the sliding surface error. Subsequently, this error information is input to the FNO adaptive rate (Fourier neural operator adaptive law module). The FNO adaptive rate, as the core of parameter adjustment, analyzes the real-time error and outputs a dynamic adjustment amount for the controller parameters. This adjustment amount acts on the controller's automatic parameter update mechanism, adjusting the relevant gains and parameters in the SMC control module in real time. The updated controller parameters are used for calculation in the next control cycle, thereby generating an updated control voltage, which is then applied to the controlled object again through the driving device, forming a closed-loop iterative optimization process. The above process is repeated until the SMC control module, under the continuous adjustment of the FNO adaptive rate, outputs a control signal that can stabilize the system tracking error within the allowable range, indicating that the controller has adapted to the current operating conditions.

[0055] It should be noted that the above examples are only for understanding this application and do not constitute a limitation on the adaptive sliding surface control method of this application. Any simple modifications based on this technical concept are within the protection scope of this application.

[0056] This application also provides an adaptive sliding surface control device, please refer to... Figure 4 The adaptive slip surface control device includes: The acquisition module 10 is used to acquire the displacement tracking status information of the piezoelectric displacement platform; Output module 20 is used to input the displacement tracking state information to an adaptive sliding surface controller integrated with a pre-trained Fourier neural operator and output dynamically adjusted adaptive control parameters. The generation module 30 is used to generate a control voltage from the adaptive sliding surface controller based on the displacement tracking state information and the adaptive control parameters. The drive module 40 is used to apply the control voltage to the piezoelectric displacement platform to drive the piezoelectric displacement platform to track the target displacement.

[0057] And / or, the adaptive slip surface control device includes: The first acquisition module is used to acquire the actual displacement and target displacement of the piezoelectric displacement platform within the current control cycle. The first calculation module is used to calculate the difference between the actual displacement and the target displacement to obtain the displacement tracking status information.

[0058] And / or, the adaptive slip surface control device includes: The first acquisition module is used to acquire the input voltage and output displacement data of the piezoelectric displacement platform under various working conditions and construct a training dataset. The first training module is used to train an initial model containing Fourier neural operators based on the training dataset to obtain a pre-trained model that can characterize the hysteresis dynamics of the piezoelectric displacement platform. The first extraction module is used to verify the pre-trained model. When the preset accuracy condition is met, the operators in the model are extracted as the pre-trained Fourier neural operators.

[0059] And / or, the adaptive slip surface control device includes: The first combination module is used to combine the input voltage sequence in the training data with the corresponding time information to generate model input data; The first input module is used to input the model input data into the Fourier neural operator, which processes the data through multiple cascaded Fourier layers. The operations performed by each Fourier layer include: frequency domain transformation of the input, linear processing of selected frequency domain components, inverse transformation back to the time domain and fusion with bypass branches, and application of nonlinear activation. The first output module is used to adjust the parameters of the Fourier neural operator through an optimization algorithm to minimize the difference between the model's predicted displacement and the actual displacement, and output the trained initial model containing the Fourier neural operator.

[0060] And / or, the adaptive slip surface control device includes: The second input module is used to input test data that has not been trained into the pre-trained model to obtain the predicted displacement command; The first measurement module is used to apply the predicted displacement command to the piezoelectric displacement platform and measure the actual displacement response; The second calculation module is used to calculate the verification error based on the actual displacement response and the true displacement corresponding to the test data; The first verification module is used to return to the step of training the initial model containing Fourier neural operators based on the training dataset if the verification error does not meet the preset accuracy threshold, until the verification error meets the preset accuracy threshold.

[0061] And / or, the adaptive slip surface control device includes: The first difference module is used to obtain the displacement tracking error and the sliding surface error based on the difference between the initial actual displacement and the target displacement. The third input module is used to input the displacement tracking error and the sliding surface error into the Fourier neural operator and output the parameter adjustment amount for adjusting the controller. The first update module is used to update the control parameters of the adaptive slip surface controller based on the parameter adjustment amount, and to generate an updated control voltage for the adaptive slip surface controller based on the updated control parameters. The first verification module is used to verify whether the updated adaptive sliding surface controller is suitable for the current working conditions. If it is not suitable, it returns the steps of generating an initial control voltage based on the preset initial control parameters, applying it to the piezoelectric displacement platform, and obtaining the initial actual displacement, until the updated adaptive sliding surface controller is suitable for the current working conditions.

[0062] The adaptive sliding surface control device provided in this application, employing the adaptive sliding surface control method in the above embodiments, can solve the technical problem of low control accuracy during piezoelectric platform compensation. Compared with the prior art, the beneficial effects of the adaptive sliding surface control device provided in this application are the same as those of the adaptive sliding surface control method provided in the above embodiments, and other technical features in the adaptive sliding surface control device are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0063] This application provides an adaptive sliding surface control device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the adaptive sliding surface control method in the first embodiment described above.

[0064] The following is for reference. Figure 5 The diagram illustrates a structural schematic suitable for implementing the adaptive smooth surface control device of the embodiments of this application. The adaptive smooth surface control device in the embodiments of this application may include, but is not limited to, mobile terminals such as mobile phones, tablets, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PMPs (Portable Media Players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital televisions and desktop computers. Figure 5 The adaptive slip surface control device shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of this application.

[0065] like Figure 5As shown, the adaptive smooth surface control device may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.) that can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage device 1003 into a random access memory (RAM) 1004. The RAM 1004 also stores various programs and data required for the operation of the adaptive smooth surface control device. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to the I / O interface 1006: input devices 1007 including, for example, a touchscreen, touchpad, keyboard, mouse, image sensor, microphone, accelerometer, gyroscope, etc.; output devices 1008 including, for example, a liquid crystal display (LCD), speaker, vibrator, etc.; storage devices 1003 including, for example, magnetic tape, hard disk, etc.; and communication devices 1009. Communication device 1009 allows the adaptive sliding surface control device to communicate wirelessly or wiredly with other devices to exchange data. Although the figures show adaptive sliding surface control devices with various systems, it should be understood that implementation or possession of all the systems shown is not required. More or fewer systems may be implemented alternatively.

[0066] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from ROM 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0067] The adaptive sliding surface control device provided in this application, employing the adaptive sliding surface control method in the above embodiments, can solve the technical problem of low control accuracy during piezoelectric platform compensation. Compared with the prior art, the beneficial effects of the adaptive sliding surface control device provided in this application are the same as those of the adaptive sliding surface control method provided in the above embodiments, and other technical features of this adaptive sliding surface control device are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0068] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0069] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0070] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to execute the adaptive sliding surface control method in the above embodiments.

[0071] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0072] The aforementioned computer-readable storage medium may be included in the adaptive sliding surface control device; or it may exist independently and not assembled into the adaptive sliding surface control device.

[0073] The aforementioned computer-readable storage medium carries one or more programs that, when executed by the adaptive sliding surface control device, cause the adaptive sliding surface control device to: Obtain the displacement tracking status information of the piezoelectric displacement platform; The displacement tracking state information is input into an adaptive sliding surface controller that integrates a pre-trained Fourier neural operator, and the dynamically adjusted adaptive control parameters are output. Based on the displacement tracking state information and the adaptive control parameters, the adaptive sliding surface controller generates a control voltage; The control voltage is applied to the piezoelectric displacement platform to drive the piezoelectric displacement platform to track the target displacement.

[0074] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a Local Area Network (LAN) or a Wide Area Network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0075] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0076] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0077] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the above-described adaptive sliding surface control method, which can solve the technical problem of low control accuracy during piezoelectric platform compensation. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the adaptive sliding surface control method provided in the above embodiments, and will not be repeated here.

[0078] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the adaptive sliding surface control method as described above.

[0079] The computer program product provided in this application can solve the technical problem of low control accuracy when compensating for piezoelectric platforms. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as those of the adaptive sliding surface control method provided in the above embodiments, and will not be repeated here.

[0080] All acquisition of signals, information, or actions in this application are carried out in compliance with the relevant data protection laws and policies of the country where the application is located, and with the authorization of the relevant device owner.

[0081] The above description is only a part of the embodiments of this application and does not limit the scope of protection of this application. All equivalent structural transformations made under the technical concept of this application and using the content of this application specification and drawings, or direct / indirect applications in other related technical fields, are included in the scope of protection of this application.

Claims

1. An adaptive sliding surface control method, characterized in that, The method includes: Obtain the displacement tracking status information of the piezoelectric displacement platform; The displacement tracking state information is input into an adaptive sliding surface controller that integrates a pre-trained Fourier neural operator, and the dynamically adjusted adaptive control parameters are output. Based on the displacement tracking state information and the adaptive control parameters, the adaptive sliding surface controller generates a control voltage; The control voltage is applied to the piezoelectric displacement platform to drive the piezoelectric displacement platform to track the target displacement.

2. The method as described in claim 1, characterized in that, The step of obtaining the displacement tracking status information of the piezoelectric displacement platform includes: Obtain the actual and target displacements of the piezoelectric displacement platform within the current control cycle; The difference between the actual displacement and the target displacement is calculated to obtain the displacement tracking status information.

3. The method as described in claim 1, characterized in that, Before the step of inputting the displacement tracking state information into an adaptive sliding surface controller integrated with a pre-trained Fourier neural operator and outputting dynamically adjusted adaptive control parameters, the following steps are included: The input voltage and output displacement data of the piezoelectric displacement platform under various working conditions were collected to construct a training dataset. The initial model containing Fourier neural operators is trained based on the training dataset to obtain a pre-trained model that can characterize the hysteresis dynamics of the piezoelectric displacement platform. The pre-trained model is validated, and when the preset accuracy condition is met, the operators in it are extracted as the pre-trained Fourier neural operators.

4. The method as described in claim 3, characterized in that, The step of training the initial model containing Fourier neural operators based on the training dataset includes: The input voltage sequence in the training data is combined with the corresponding time information to generate the model input data; The model input data is input into the Fourier neural operator and processed through multiple cascaded Fourier layers. The operations performed by each Fourier layer include: frequency domain transformation of the input, linear processing of selected frequency domain components, inverse transformation back to the time domain and fusion with bypass branches, and application of nonlinear activation. The parameters of the Fourier neural operator are adjusted by an optimization algorithm to minimize the difference between the model's predicted displacement and the actual displacement, and the trained initial model containing the Fourier neural operator is output.

5. The method as described in claim 3, characterized in that, The step of validating the pre-trained model includes: The test data that was not used in the training is input into the pre-trained model to obtain the predicted displacement command; The predicted displacement command is applied to the piezoelectric displacement platform, and the actual displacement response is measured. The verification error is calculated based on the actual displacement response and the true displacement corresponding to the test data. If the verification error does not meet the preset accuracy threshold, the process returns to the step of training the initial model containing Fourier neural operators based on the training dataset until the verification error meets the preset accuracy threshold.

6. The method as described in claim 1, characterized in that, The step of generating the control voltage by the adaptive slip surface controller includes: Based on preset initial control parameters, an initial control voltage is generated by the adaptive sliding surface controller and applied to the piezoelectric displacement platform to obtain the initial actual displacement; Based on the difference between the initial actual displacement and the target displacement, the displacement tracking error and the sliding surface error are obtained; The displacement tracking error and sliding surface error are input into the Fourier neural operator, and the output is used to adjust the parameters of the controller. The control parameters of the adaptive slip surface controller are updated based on the parameter adjustment amount, and an updated control voltage for the adaptive slip surface controller is generated based on the updated control parameters. The process involves verifying whether the updated adaptive sliding surface controller is suitable for the current operating conditions. If it is not suitable, the process returns to the preset initial control parameters, generates an initial control voltage through the adaptive sliding surface controller, applies it to the piezoelectric displacement platform, and obtains the initial actual displacement. This process continues until the updated adaptive sliding surface controller is suitable for the current operating conditions.

7. An adaptive sliding surface control device, characterized in that, The device includes: The acquisition module is used to acquire the displacement tracking status information of the piezoelectric displacement platform; The output module is used to input the displacement tracking state information to an adaptive sliding surface controller integrated with a pre-trained Fourier neural operator and output dynamically adjusted adaptive control parameters. The generation module is used to generate a control voltage from the adaptive sliding surface controller based on the displacement tracking state information and the adaptive control parameters. A drive module is used to apply the control voltage to the piezoelectric displacement platform to drive the piezoelectric displacement platform to track the target displacement.

8. An adaptive sliding surface control device, characterized in that, The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the adaptive slip surface control method as described in any one of claims 1 to 6.

9. A storage medium, characterized in that, The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, it implements the steps of the adaptive sliding surface control method as described in any one of claims 1 to 6.

10. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the steps of the adaptive sliding surface control method as described in any one of claims 1 to 6.

Images