Precise macro-micro composite driving system control method based on self-learning nonlinear PID algorithm

Abstract

The invention relates to the technical field of automatic control and intelligent driving, and discloses a precise macro-micro composite driving system control method based on a self-learning nonlinear PID algorithm. The method comprises the steps of collecting real-time operation data to judge nonlinear dynamic characteristics, obtaining a deviation value, extracting nonlinear characteristics and adjusting control parameters, switching to a self-learning mode to update a model, compensating interference and the like. By monitoring and compensating the nonlinear characteristics and external interference of the system in real time, the response speed and control precision of the system are improved, the dependence on external experience parameters is reduced, and the system adapts to complex and changeable working environment requirements.

Description

Technical Field

[0001] The present invention relates to the field of automation control and intelligent drive technology, and in particular to a control method for a precision macro-micro composite drive system based on a self-learning nonlinear PID algorithm. Background Art

[0002] In the fields of industrial automation and precision manufacturing, macro-micro hybrid drive systems, with their ability to combine large-scale motion with high-precision positioning, are widely used in high-end manufacturing scenarios such as semiconductor packaging, optical component processing, and precision measurement. The core requirement of these systems is to achieve precise control across scales from nanometers to millimeters. However, their operation is susceptible to multiple factors, resulting in significant challenges for traditional control methods.

[0003] Although the traditional PID control algorithm has become the mainstream solution due to its simple structure and easy implementation, it has obvious limitations under the complex working conditions of the macro-micro hybrid drive system: on the one hand, the system has inherent nonlinear dynamic characteristics, such as the coupling effect between the macro-motion mechanism and the micro-motion mechanism, the friction nonlinearity of the transmission components, and the gain drift caused by load changes, which lead to a strong nonlinear input-output relationship and make it difficult for traditional linear PID parameters to be adaptively adjusted; on the other hand, the superposition of external environmental interference (such as temperature fluctuations, vibration, and electromagnetic noise) and the time-varying characteristics of internal parameters further aggravates the uncertainty of the system's dynamic response, leading to problems such as delayed response speed, increased overshoot, and decreased steady-state accuracy.

[0004] Existing control methods for nonlinear systems often rely on empirical parameter tuning or pre-set model compensation, resulting in poor adaptability to complex and changing operating conditions. When the system's operating environment undergoes sudden changes (such as sudden load changes or interference source changes), manual parameter re-adjustment is required, which not only increases operational complexity but also makes it difficult to meet the stringent real-time and stability requirements of high-precision manufacturing. Therefore, developing an intelligent control method that can autonomously sense the system's nonlinear characteristics, compensate for external interference in real time, and reduce reliance on empirical parameters is key to improving the control performance of macro-micro hybrid drive systems and represents a pressing technical challenge in this field. Summary of the Invention

[0005] The present invention provides a precision macro-micro composite drive system control method based on a self-learning nonlinear PID algorithm, which is used to solve the problems of slow response speed, low control accuracy and dependence on experience in parameter adjustment of the macro-micro composite drive system in the prior art when dealing with complex nonlinear dynamic characteristics.

[0006] The present invention provides a control method for a precision macro-micro hybrid drive system based on a self-learning nonlinear PID algorithm. The control method includes: collecting real-time operating data of the macro-micro hybrid drive system, and judging whether the system has nonlinear dynamic characteristics based on the real-time operating data; when it is detected that the system has nonlinear dynamic characteristics and the characteristics affect the operating state of the system, obtaining a first deviation value between the current system operating state and a target state; when the first deviation value is greater than or equal to a preset deviation threshold, extracting nonlinear characteristics of the system based on the real-time operating data, and judging whether the nonlinear characteristics meet preset nonlinear conditions, the nonlinear characteristics including input and output characteristics and dynamic response characteristics of the system; if the nonlinear characteristics meet the preset nonlinear conditions, adjusting control parameters of the system and switching to a self-learning mode; in the self-learning mode, updating a control model of the system based on the real-time operating data and compensating for external interference of the system; after the external interference compensation is completed, judging whether the system meets optimization conditions based on the control model; if the system meets the optimization conditions, adjusting the control parameters of the system to optimal values, and shutting down the redundant control modules of the system.

[0007] According to the control method for a precision macro-micro composite drive system based on a self-learning nonlinear PID algorithm provided by the present invention, real-time operating data of the macro-micro composite drive system is collected, and based on the real-time operating data, it is determined whether the system has nonlinear dynamic characteristics, including: when a variable to be analyzed is detected in the real-time operating data, key indicators in the variable to be analyzed are extracted, and the key indicators are compared with corresponding indicators in a preset nonlinear feature database; the similarity between the key indicators and the corresponding indicators in the preset nonlinear feature database is evaluated through a calculation formula; when the similarity is greater than or equal to a preset similarity threshold, it is determined that the system has nonlinear dynamic characteristics.

[0008] According to the control method of a precision macro-micro composite drive system based on a self-learning nonlinear PID algorithm provided by the present invention, when it is detected that the system has nonlinear dynamic characteristics and the characteristics affect the operating state of the system, a first deviation value between the current system operating state and the target state is obtained, including: determining the input signal direction of the system based on the real-time operating data, and obtaining the angle between the input signal direction and the target position direction with the current position of the system as the starting point; continuously collecting a plurality of the real-time operating data, and judging whether the deviation between the system and the target position increases based on the multiple real-time operating data; if the angle is less than or equal to the preset angle and the deviation between the system and the target position increases, it is determined that the nonlinear dynamic characteristics have an impact on the operating state of the system; when the nonlinear dynamic characteristics have an impact on the operating state of the system, the first deviation value is obtained based on the real-time operating data.

[0009] According to the precision macro-micro composite drive system control method based on the self-learning nonlinear PID algorithm provided by the present invention, when the first deviation value is greater than or equal to a preset deviation threshold, the nonlinear characteristics of the system are extracted based on the real-time operation data, and it is judged whether the nonlinear characteristics meet the preset nonlinear conditions, and the nonlinear characteristics include the input and output characteristics and dynamic response characteristics of the system, including: obtaining the input and output characteristics and dynamic response characteristics of the system based on the real-time operation data; analyzing whether the input and output characteristics meet the input and output conditions, and analyzing whether the dynamic response characteristics meet the dynamic response conditions; when the input and output characteristics meet the input and output conditions and the dynamic response characteristics meet the dynamic response conditions, the nonlinear characteristics meet the preset nonlinear conditions.

[0010] According to the control method for a precision macro-micro composite drive system based on a self-learning nonlinear PID algorithm provided by the present invention, whether the input-output characteristics meet the input-output conditions is analyzed, including: determining the gain variation range of the system through the input-output characteristics, and comparing the gain variation range with the standard gain range in a preset nonlinear characteristic database; when the difference between the gain variation range and the standard gain range is greater than or equal to a first preset difference value, it is determined that the input-output characteristics meet the input-output conditions.

[0011] According to the precision macro-micro composite drive system control method based on the self-learning nonlinear PID algorithm provided by the present invention, the dynamic response characteristic is analyzed to see whether it meets the dynamic response condition, including: obtaining the response time, overshoot, oscillation frequency and steady-state error of the system based on the dynamic response characteristic; judging whether the response speed of the system is lower than the preset response speed based on the response time, judging whether the stability of the system is lower than the preset stability threshold based on the overshoot, judging whether the vibration amplitude of the system is higher than the preset amplitude based on the oscillation frequency, and judging whether the accuracy of the system is lower than the preset accuracy threshold based on the steady-state error; when the response speed of the system is lower than the preset response speed and the stability of the system is lower than the preset stability threshold and the vibration amplitude of the system is higher than the preset amplitude and the accuracy of the system is lower than the preset accuracy threshold, the dynamic response characteristic meets the dynamic response condition.

[0012] According to the precision macro-micro composite drive system control method based on the self-learning nonlinear PID algorithm provided by the present invention, if the nonlinear characteristics meet the preset nonlinear conditions, the control parameters of the system are adjusted and switched to the self-learning mode, including: when the nonlinear characteristics meet the preset nonlinear conditions, the proportional coefficient of the control system is increased by 10%, the integral coefficient is reduced by 5%, the differential coefficient is increased by 8%, and the system is switched to the self-learning mode.

[0013] According to the control method of a precision macro-micro composite drive system based on a self-learning nonlinear PID algorithm provided by the present invention, in a self-learning mode, the control model of the system is updated based on the real-time operating data and the external interference of the system is compensated, including: in the self-learning mode, the real-time interference intensity is compared with the interference intensity at the last moment before entering the self-learning mode based on the external interference characteristics of the system; when the real-time interference intensity is greater than the interference intensity at the last moment before entering the self-learning mode, it is determined whether the position of the interference source is in the input signal direction of the system; if the position of the interference source is in the input signal direction of the system, the control system maintains the current operating state and operating parameters; if the position of the interference source is not in the input signal direction of the system, the system meets the optimization conditions.

[0014] According to a control method for a precision macro-micro composite drive system based on a self-learning nonlinear PID algorithm provided by the present invention, the method also includes: when the deviation between the system and the target state is less than a preset deviation threshold, the operating state of the control system is restored to the initial operating state and initial operating parameters.

[0015] According to the precision macro-micro composite drive system control method based on the self-learning nonlinear PID algorithm provided by the present invention, in the self-learning mode, the real-time interference intensity is compared with the interference intensity at the last moment before entering the self-learning mode based on the external interference characteristics of the system, including: based on the external interference characteristics of the system, the average value of the interference intensity in the first time period after the system switches to the self-learning mode is used as the real-time interference intensity; based on the external interference characteristics of the system, the average value of the interference intensity in the last first time period before the system switches to the self-learning mode is used as the interference intensity at the last moment before entering the self-learning mode; and the real-time interference intensity is compared with the interference intensity at the last moment before entering the self-learning mode.

[0016] The proposed control method for a precision macro-micro hybrid drive system based on a self-learning nonlinear PID algorithm collects system operating data in real time, accurately extracts input-output characteristics and dynamic response characteristics, and compares and analyzes these characteristics with a pre-set nonlinear feature database. This method can specifically compensate for precision losses caused by friction nonlinearity, mechanism coupling, load fluctuations, and other factors. Combined with dynamic updates of the control model in self-learning mode, the system's steady-state error can be controlled to the nanometer level, effectively addressing the accuracy degradation of traditional PID algorithms in nonlinear scenarios.

[0017] In order to address the problems of response lag and large overshoot in traditional control methods, the present invention can significantly improve the system's following speed to the input signal and shorten the response time by dynamically adjusting the PID parameters. At the same time, by suppressing the oscillation frequency and overshoot, the vibration amplitude of the system in high-speed motion is reduced, ensuring a smoother coordinated response of the macro-motion and micro-motion mechanisms, thus meeting the dual requirements of "high speed and high precision" in precision manufacturing.

[0018] The system is capable of autonomously identifying nonlinear dynamic characteristics: when nonlinear features such as gain drift or abnormal response characteristics are detected, it automatically switches to self-learning mode, compensating for nonlinear effects by updating the control model in real time. In response to external interference, by comparing interference intensity and locating the interference source, it can adopt targeted strategies such as maintaining the current state or triggering optimization, thereby enhancing the system's anti-interference capabilities in complex environments and significantly reducing the impact of sudden operating condition changes on control effectiveness.

[0019] Traditional PID parameter setting relies on manual experience and is difficult to adapt to changing operating conditions. This invention, through a pre-set nonlinear characteristic database and adaptive algorithm, automatically optimizes parameters and autonomously restores the initial state when the deviation is less than a threshold, without the need for manual intervention. The model iteration capability in self-learning mode enables the system to continuously optimize control strategies under repetitive operating conditions, gradually improving control effectiveness. This is particularly suitable for small-batch, high-variety precision manufacturing scenarios.

[0020] When the optimization conditions are met, the present invention will automatically shut down the redundant control module to reduce unnecessary computing power and energy consumption; at the same time, through precise interference compensation and parameter adjustment, it avoids energy waste caused by excessive control, thereby improving the system operation efficiency and meeting the development needs of industrial energy conservation and green manufacturing.

[0021] In summary, the present invention realizes high-precision, high-dynamic, and high-adaptability control of macro-micro composite drive systems in nonlinear and strong interference environments through the closed-loop control logic of "perception-analysis-learning-compensation", providing reliable technical support for high-end manufacturing fields such as semiconductor packaging, optical processing, and precision measurement. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 This is a workflow diagram of the nonlinear dynamic characteristics detection module of the present invention; Figure 2 This is a workflow diagram of the first deviation value calculation module of the present invention; Figure 3 This is a workflow diagram of the nonlinear feature analysis module of the present invention; Figure 4 This is a working diagram of the external interference compensation module of the present invention; Figure 5 This is a workflow diagram of the control parameter optimization module of the present invention. DETAILED DESCRIPTION

[0023] The following will clearly and completely describe the technical solution of the present invention in conjunction with the accompanying drawings. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0024] The present invention provides a method for controlling a precision macro-micro hybrid drive system based on a self-learning nonlinear PID algorithm. The specific implementation is described in detail with reference to the accompanying drawings. The system includes a macro-micro hybrid drive system, a real-time operation data acquisition module, a nonlinear dynamic characteristics detection module, a first deviation value calculation module, a nonlinear characteristic analysis module, a self-learning mode switching module, an external interference compensation module, a control parameter optimization module, and a redundant control module. These modules cooperate through signal transmission and data processing to achieve precise control of the macro-micro hybrid drive system.

[0025] During actual operation, the real-time operation data acquisition module is located at the core of the system and is used to collect real-time operation data of the macro-micro hybrid drive system. These data include but are not limited to the system's input signals, output responses, operation state variables, etc. The real-time operation data acquisition module is directly connected to the nonlinear dynamic characteristics detection module, which is responsible for performing preliminary analysis on the collected data to determine whether the system has nonlinear dynamic characteristics. Figure 1 As shown in the figure, when a variable to be analyzed is detected in real-time operating data, the nonlinear dynamic characteristics detection module extracts key indicators and compares them with corresponding indicators in a pre-set nonlinear characteristic database. A calculation formula is used to evaluate the similarity between the key indicators and the corresponding indicators in the database. If the similarity is greater than or equal to a preset similarity threshold, the system is deemed to have nonlinear dynamic characteristics. This process relies on high-quality data provided by the real-time operating data acquisition module, and the nonlinear dynamic characteristics detection module performs a preliminary judgment through logical operations.

[0026] When the nonlinear dynamic characteristic detection module confirms that the system has nonlinear dynamic characteristics and that the characteristics affect the operating state of the system, the first deviation value calculation module starts working. Figure 2As shown, the first deviation value calculation module determines the direction of the system's input signal based on real-time operating data and obtains the angle between the input signal direction and the target position direction, starting from the system's current position. Simultaneously, multiple real-time operating data sets are continuously collected and, based on these data, a determination is made as to whether the deviation between the system and the target position is increasing. If the angle is less than or equal to a preset angle and the deviation between the system and the target position is increasing, it is determined that the nonlinear dynamic characteristics are affecting the system's operating state. At this point, the first deviation value calculation module obtains a first deviation value based on the real-time operating data and transmits it to the nonlinear characteristic analysis module.

[0027] The task of the nonlinear characteristic analysis module is to further analyze the nonlinear characteristics of the system, including input-output characteristics and dynamic response characteristics. Figure 3 As shown, the nonlinear characteristic analysis module first obtains the system's input-output characteristics and dynamic response characteristics based on real-time operating data. For the input-output characteristics, the system's gain variation range is determined and compared with the standard gain range in a pre-set nonlinear characteristic database. If the difference is greater than or equal to a first pre-set difference value, the input-output characteristics are deemed to meet the input-output conditions. For the dynamic response characteristics, the nonlinear characteristic analysis module obtains the system's response time, overshoot, oscillation frequency, and steady-state error. Based on these parameters, the module determines whether the system's response speed is lower than a pre-set response speed, stability is lower than a pre-set stability threshold, vibration amplitude is higher than a pre-set amplitude, and accuracy is lower than a pre-set accuracy threshold. If all conditions are met, the dynamic response characteristics are deemed to meet the dynamic response conditions. When both the input-output characteristics and the dynamic response characteristics meet the corresponding conditions, the nonlinear characteristic analysis module sends a signal to the self-learning mode switching module.

[0028] After receiving the signal from the nonlinear characteristic analysis module, the self-learning mode switching module adjusts the control parameters of the system and switches to the self-learning mode. Specifically, the proportional coefficient is increased by 10%, the integral coefficient is reduced by 5%, and the differential coefficient is increased by 8%. The adjusted parameters are then transmitted to the control parameter optimization module through the internal communication interface. At the same time, the external interference compensation module starts working, such as Figure 4As shown, in the self-learning mode, the external interference compensation module compares the real-time interference intensity with the interference intensity at the last moment before entering the self-learning mode based on the external interference characteristics of the system. The real-time interference intensity is calculated by the average value of the interference intensity in the first time period after the system switches to the self-learning mode, while the interference intensity at the last moment before entering the self-learning mode is calculated by the average value of the interference intensity in the last first time period before the system switches to the self-learning mode. If the real-time interference intensity is greater than the interference intensity at the last moment before entering the self-learning mode, it is determined whether the position of the interference source is in the input signal direction of the system. If the position of the interference source is in the input signal direction of the system, the control system maintains the current operating state and operating parameters; otherwise, the system meets the optimization conditions.

[0029] When the system meets the optimization conditions, the control parameter optimization module starts working. Figure 5 As shown, the control parameter optimization module determines whether the system's operating state meets the optimization criteria based on the updated control model. If so, the system's control parameters are adjusted to their optimal values, and the redundant control modules are disabled. Under normal operation, the redundant control modules provide additional control capabilities, but under optimization conditions, their functionality is deemed redundant and therefore disabled to conserve resources. Furthermore, if the deviation between the system and the target state is less than a preset deviation threshold, the control system's operating state is restored to its initial state and parameters.

[0030] Throughout the control process, the various modules work closely together through signal transmission and data exchange. The real-time data acquisition module provides basic data support for subsequent modules. The nonlinear dynamic characteristics detection module and the first deviation value calculation module are responsible for preliminary analysis and deviation calculation. The nonlinear characteristic analysis module further refines the determination of nonlinear characteristics. The self-learning mode switching module and the external interference compensation module jointly address nonlinear dynamic characteristics and external interference in complex environments. The control parameter optimization module and the redundant control module adjust the system to the optimal operating state under optimized conditions. The modules have clear functional divisions and close collaborative relationships, ensuring that the macro-micro hybrid drive system can achieve rapid response and high-precision control in complex and changing working environments.

[0031] The above examples describe in detail the specific implementation of a control method for a precision macro-micro hybrid drive system based on a self-learning nonlinear PID algorithm, covering the entire process from data acquisition to control parameter optimization. Through these steps, the system can effectively handle complex nonlinear dynamic characteristics, improve response speed and control accuracy, and adapt to diverse application scenarios.

[0032] In order to better enable relevant personnel in this technical field to fully understand and implement the present invention, the specific implementation principle of the present invention is supplemented below with reference to a specific application scenario.

[0033] In practical applications, macro-micro hybrid drive systems are deployed in high-precision optical component processing equipment to achieve precise positioning and trajectory control of complex surfaces. This scenario requires the system to be able to respond quickly in a dynamically changing working environment while maintaining extremely high positioning accuracy and stability. To this end, the real-time operation data acquisition module first obtains real-time operation data such as the system's input signal, output response, and operating state variables through the sensor array. After filtering and preprocessing, this data is transmitted to the nonlinear dynamic characteristics detection module for analysis. For example, when nonlinear dynamic characteristics due to load changes occur during system operation, the nonlinear dynamic characteristics detection module extracts key indicators and compares them with corresponding indicators in a preset nonlinear feature database. If the similarity reaches or exceeds the preset similarity threshold, it is determined that the system has nonlinear dynamic characteristics.

[0034] Subsequently, the first deviation calculation module determines the direction of the system's input signal based on real-time operating data and calculates the angle between the input signal direction and the target position direction. For example, at a certain moment, the system's current position is the origin, the target position's direction vector is (1, 0), and the input signal's direction vector is (0.707, 0.707). In this case, the calculated angle is 45°, which is less than the preset angle of 60°. Simultaneously, multiple consecutive acquisitions of real-time operating data indicate a gradual increase in the deviation between the system and the target position. Therefore, the first deviation calculation module determines that nonlinear dynamic characteristics are affecting the system's operating state and further calculates the first deviation value based on the real-time operating data.

[0035] Next, the nonlinear characteristic analysis module conducts a detailed analysis of the system's input-output characteristics and dynamic response characteristics. Taking the input-output characteristics as an example, assume that the system's gain variation range over a certain period of time is [0.8, 1.2], while the standard gain range in the preset nonlinear characteristic database is [0.9, 1.1]. Since the difference between the two is greater than a first preset difference value of 0.1, the input-output characteristics are considered to meet the input-output conditions. Regarding the dynamic response characteristics, the nonlinear characteristic analysis module obtains a response time of 0.5 seconds, an overshoot of 15%, an oscillation frequency of 20 Hz, and a steady-state error of 0.02 mm. These parameters are, respectively, lower than the preset response speed of 0.6 seconds, lower than the preset stability threshold of 10%, higher than the preset amplitude of 25 Hz, and lower than the preset accuracy threshold of 0.03 mm. Therefore, the dynamic response characteristics also meet the dynamic response conditions. Based on this, the nonlinear characteristic analysis module sends a signal to the self-learning mode switching module.

[0036] After receiving the signal, the self-learning mode switching module adjusts the control parameters of the system, increasing the proportional coefficient by 10%, reducing the integral coefficient by 5%, and increasing the differential coefficient by 8%. At the same time, the external interference compensation module starts working. For example, in the first time period after the system switches to the self-learning mode, the average value of the external interference intensity is 0.12N, while the average value of the interference intensity in the last first time period before entering the self-learning mode is 0.10N. Since the real-time interference intensity is greater than the interference intensity at the last moment before entering the self-learning mode, the external interference compensation module further determines whether the location of the interference source is in the direction of the input signal of the system. If the interference source is in the direction of the input signal, the control system maintains the current operating state and operating parameters; otherwise, the system meets the optimization conditions.

[0037] When the system meets the optimization criteria, the control parameter optimization module evaluates the system's operating status based on the updated control model. For example, suppose the system's target state is the position coordinates (10, 10) and the current state is (9.98, 9.99). The deviation between the two is 0.02 mm, which is less than the preset deviation threshold of 0.05 mm. At this point, the control parameter optimization module adjusts the system's control parameters to the optimal values ​​and disables redundant control modules to conserve resources. Furthermore, if the deviation between the system and the target state continues to decrease, the control system's operating state returns to its initial state and parameters.

[0038] Throughout the entire process, each module closely collaborates through signal transmission and data exchange. For example, the real-time data acquisition module provides high-quality data support for the nonlinear dynamic characteristics detection module. The nonlinear dynamic characteristics detection module and the first deviation value calculation module jointly complete preliminary analysis and deviation calculation. The nonlinear characteristic analysis module further refines the judgment of nonlinear characteristics. The self-learning mode switching module and the external interference compensation module work together to cope with complex nonlinear dynamic characteristics and external interference. The control parameter optimization module and the redundant control module ensure that the system achieves optimal operation under optimized conditions. The modules have clear functional divisions and close collaborative relationships, ensuring that the macro-micro hybrid drive system can achieve rapid response and high-precision control in complex and changing working environments.

[0039] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A precision macro-micro composite drive system control method based on a self-learning nonlinear PID algorithm, characterized in that: The control method includes: Collecting real-time operating data of the macro-micro hybrid drive system, and determining whether the system has nonlinear dynamic characteristics based on the real-time operating data; When it is detected that the system has nonlinear dynamic characteristics and the characteristics affect the operating state of the system, a first deviation value between the current operating state of the system and the target state is obtained; When the first deviation value is greater than or equal to a preset deviation threshold, extracting nonlinear characteristics of the system based on the real-time operation data, and determining whether the nonlinear characteristics meet a preset nonlinear condition, the nonlinear characteristics including input and output characteristics and dynamic response characteristics of the system; If the nonlinear characteristics meet the preset nonlinear conditions, the control parameters of the system are adjusted and the system is switched to a self-learning mode; In the self-learning mode, the control model of the system is updated based on the real-time operation data and external interference of the system is compensated; After the external disturbance compensation is completed, judging whether the system meets the optimization conditions based on the control model; If the system meets the optimization conditions, the system control parameters are adjusted to the optimal values ​​and the system's redundant control modules are shut down.

2. The control method of a precision macro-micro composite drive system based on a self-learning nonlinear PID algorithm according to claim 1 is characterized in that: Collecting real-time operating data of the macro-micro hybrid drive system and determining whether the system has nonlinear dynamic characteristics based on the real-time operating data, including: When a variable to be analyzed is detected in the real-time operation data, key indicators in the variable to be analyzed are extracted, and the key indicators are compared with corresponding indicators in a preset nonlinear feature database; Evaluate the similarity between the key indicator and the corresponding indicator in the preset nonlinear feature database through a calculation formula; When the similarity is greater than or equal to a preset similarity threshold, it is determined that the system has nonlinear dynamic characteristics.

3. The control method of a precision macro-micro composite drive system based on a self-learning nonlinear PID algorithm according to claim 1 is characterized in that: When it is detected that the system has a nonlinear dynamic characteristic and the characteristic affects the operating state of the system, obtaining a first deviation value between the current operating state of the system and the target state includes: Determine the input signal direction of the system based on the real-time operation data, and obtain the angle between the input signal direction and the target position direction with the current position of the system as the starting point; continuously collecting a plurality of the real-time operation data, and determining whether a deviation between the system and the target position increases based on the plurality of the real-time operation data; If the angle is less than or equal to the preset angle and the deviation between the system and the target position increases, it is determined that the nonlinear dynamic characteristics have an impact on the operating state of the system; When the nonlinear dynamic characteristics have an impact on the system operation state, a first deviation value is obtained based on the real-time operation data.

4. The control method of a precision macro-micro composite drive system based on a self-learning nonlinear PID algorithm according to claim 1 is characterized in that: When the first deviation value is greater than or equal to a preset deviation threshold, a nonlinear characteristic of the system is extracted based on the real-time operation data, and it is determined whether the nonlinear characteristic meets a preset nonlinear condition, wherein the nonlinear characteristic includes input and output characteristics and dynamic response characteristics of the system, including: Acquiring input and output characteristics and dynamic response characteristics of the system based on the real-time operation data; Analyzing whether the input-output characteristics meet input-output conditions, and analyzing whether the dynamic response characteristics meet dynamic response conditions; When the input-output characteristic satisfies an input-output condition and the dynamic response characteristic satisfies a dynamic response condition, the nonlinear feature satisfies a preset nonlinear condition.

5. The control method of a precision macro-micro composite drive system based on a self-learning nonlinear PID algorithm according to claim 4 is characterized in that: Analyzing whether the input and output characteristics meet the input and output conditions includes: Determining a gain variation range of the system based on the input-output characteristics, and comparing the gain variation range with a standard gain range in a preset nonlinear characteristic database; When the difference between the gain variation range and the standard gain range is greater than or equal to a first preset difference value, it is determined that the input-output characteristic meets the input-output condition.

6. The control method of a precision macro-micro composite drive system based on a self-learning nonlinear PID algorithm according to claim 4 is characterized in that: Analyzing whether the dynamic response characteristics meet the dynamic response conditions includes: Acquire the response time, overshoot, oscillation frequency and steady-state error of the system based on the dynamic response characteristics; determining whether a response speed of the system is lower than a preset response speed based on the response time, determining whether a stability of the system is lower than a preset stability threshold based on the overshoot, determining whether a vibration amplitude of the system is higher than a preset amplitude based on the oscillation frequency, and determining whether an accuracy of the system is lower than a preset accuracy threshold based on the steady-state error; When the response speed of the system is lower than the preset response speed, the stability of the system is lower than the preset stability threshold, the vibration amplitude of the system is higher than the preset amplitude, and the accuracy of the system is lower than the preset accuracy threshold, the dynamic response characteristic meets the dynamic response condition.

7. The control method of a precision macro-micro composite drive system based on a self-learning nonlinear PID algorithm according to claim 1, characterized in that: If the nonlinear characteristics meet the preset nonlinear conditions, the control parameters of the system are adjusted and the system is switched to a self-learning mode, including: When the nonlinear characteristics meet the preset nonlinear conditions, the proportional coefficient of the control system increases by 10%, the integral coefficient decreases by 5%, the differential coefficient increases by 8%, and the system is switched to the self-learning mode.

8. The control method of a precision macro-micro composite drive system based on a self-learning nonlinear PID algorithm according to claim 1, characterized in that: The method further comprises: When the deviation between the system and the target state is less than the preset deviation threshold, the operating state of the control system is restored to the initial operating state and initial operating parameters.

Images