A control parameter adjustment method and system for a three-degree-of-freedom parallel robot

Abstract

The present invention relates to the field of parallel robot control technology, and in particular to a control parameter adjustment method and system for a three-degree-of-freedom parallel robot. The method comprises the following steps: performing terminal operation space analysis on the parallel robot, and optimizing the working posture space to obtain optimized operation space data; extracting high-speed motion residual vibration characteristics of the parallel robot by optimizing the operation space data, and performing intelligent control parameter adjustment strategy processing to obtain an intelligent vibration control parameter adjustment strategy; performing real-time vibration monitoring of the terminal effector of the parallel robot to generate real-time terminal vibration characteristic data; performing intelligent feedforward compensation control and adjustment processing on the real-time terminal vibration characteristic data by an intelligent vibration control parameter adjustment strategy to realize intelligent parameter adjustment control operation of the three-degree-of-freedom parallel robot. The present invention effectively suppresses the residual vibration of the parallel robot after high-speed motion through intelligent vibration control parameter adjustment, thereby improving work efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of parallel robot control, and in particular to a control parameter adjustment method and system for a three-degree-of-freedom parallel robot. Background Art

[0002] Parallel robots have been widely used in the field of industrial automation due to their advantages such as high rigidity, high load-bearing capacity, high speed and high precision, especially in situations where high-speed and high-precision operations are required, such as fast grasping, placement, assembly, etc. As an important type of parallel robot, the three-degree-of-freedom parallel robot has unique advantages in the field of precision operation and micro-operation due to its compact structure and flexible movement. In many industrial application scenarios, the manipulator needs to complete rapid movement from one position to another in a very short time and accurately perform the predetermined operation at the target position. During high-speed movement, due to the combined influence of factors such as inertial force, sudden change of driving torque, clearance of the transmission system, and elastic deformation of the structure itself, the clamping head often produces obvious vibration. This vibration does not disappear immediately after the manipulator reaches the target position, but will continue to decay for a period of time, that is, residual vibration. The existence of residual vibration causes the clamping head to produce unnecessary jitter near the target position, which seriously reduces the positioning accuracy, may lead to grasping failure or placement error, affect product quality and production efficiency, and may even damage the workpiece or the manipulator itself. However, the traditional control parameter adjustment method for a three-degree-of-freedom parallel robot is mainly linear control, such as PID control. Although basic positioning control can be achieved, it has limited effect on suppressing residual vibration. This is because the parameters of PID control are usually fixed and cannot adapt to dynamic changes during high-speed motion. When the robot performs rapid movements, the dynamic characteristics of the system will change, and fixed PID parameters are difficult to effectively suppress residual vibration. Summary of the invention

[0003] Based on this, the present invention provides a control parameter adjustment method and system for a three-degree-of-freedom parallel robot to solve at least one of the above technical problems.

[0004] To achieve the above object, a control parameter adjustment method for a three-degree-of-freedom parallel robot comprises the following steps:

[0005] Step S1: Establish a manipulator dynamics model for the parallel robot to generate a manipulator dynamics model; perform terminal operation space analysis based on the manipulator dynamics model to generate terminal workspace range data; perform mechanism singular configuration identification on the manipulator dynamics model through the terminal workspace range data to obtain mechanism singular configuration point data;

[0006] Step S2: optimizing the working posture space of the end working space range data through the singular configuration point data of the mechanism to obtain the optimized working space data; deploying the end monitoring sensor network on the parallel robot to obtain the sensor monitoring network; performing modal vibration test on the parallel robot based on the optimized working space data using the sensor monitoring network to generate modal vibration characteristic parameters; performing high-speed motion residual vibration feature extraction on the parallel robot based on the optimized working space data using the sensor monitoring network to generate motion task residual vibration characteristic data;

[0007] Step S3: performing adaptive rule adjustment strategy processing on the residual vibration characteristic data of the motion task to generate a vibration suppression control rule strategy; performing control parameter intelligent adjustment strategy processing on the vibration suppression control rule strategy through modal vibration characteristic parameters to obtain an intelligent vibration control parameter adjustment strategy;

[0008] Step S4: Use the sensor monitoring network to perform real-time vibration monitoring of the end effector of the parallel robot to generate real-time end vibration characteristic data; use the intelligent vibration control parameter adjustment strategy to perform intelligent feedforward compensation control adjustment processing on the real-time end vibration characteristic data to generate feedforward compensation control adjustment parameters; use the feedforward compensation control adjustment parameters to smoothly switch the control parameters of the parallel robot to realize the intelligent parameter adjustment control operation of the three-degree-of-freedom parallel robot.

[0009] The present invention can accurately identify the singular configuration points of the mechanism by establishing an accurate manipulator dynamics model and performing end workspace analysis, and optimize the work posture space accordingly. This avoids the robot from working near the singular configuration, thereby effectively preventing the mechanical structure damage and control failure of the robot caused by the sharp increase in torque in this area, and improving the safety and reliability of the robot. Traditional control methods usually do not consider the influence of singular configurations, which easily leads to uncontrollable vibration of the robot near the singular configuration, and even damages the robot. However, this method fundamentally solves this problem by pre-identifying and avoiding singular configurations. The optimized workspace also makes the robot's movement more stable and reduces unnecessary energy consumption. The modal vibration test and high-speed motion residual vibration feature extraction based on the sensor monitoring network can accurately identify the natural frequency and vibration type of the robot. The modal vibration feature parameters obtained by actual measurement can more accurately describe the vibration characteristics of the robot, thereby achieving more effective vibration suppression. At the same time, the high-speed motion residual vibration feature extraction can capture the unique vibration mode of the robot during high-speed motion, provide input for the adaptive rule adjustment strategy, and further improve the pertinence and effectiveness of vibration suppression. The adaptive rule is adjusted according to the residual vibration characteristic data of the motion task, and can adapt to different motion trajectories and load conditions in real time, ensuring that the vibration suppression strategy always maintains the optimal state. The intelligent vibration control parameter adjustment strategy uses the modal vibration characteristic parameters to finely adjust the control parameters, further improving the accuracy of vibration suppression. Compared with the traditional fixed parameter PID control, this dual adjustment mechanism can better cope with dynamic changes in high-speed motion and achieve faster and more accurate residual vibration suppression. By real-time monitoring of the vibration state of the end effector, the occurrence of residual vibration can be captured in time, and the corresponding control signal can be generated through feedforward compensation control adjustment to accurately compensate for the residual vibration. The smooth switching of control parameters ensures the stability and continuity of the control process and avoids secondary vibration caused by parameter mutation. This feedforward compensation control strategy can predict and compensate for residual vibration before it occurs, thereby minimizing the impact of residual vibration on positioning accuracy. Therefore, a control parameter adjustment method for a three-degree-of-freedom parallel robot of the present invention reduces the impact of residual vibration on positioning accuracy by real-time monitoring of the vibration state of the end effector and performing feedforward compensation control adjustment, thereby improving the stability and reliability of the robot during high-speed operation.

[0010] Preferably, step S1 comprises the following steps:

[0011] Step S11: measuring the structural parameters of the parallel robot to obtain the structural parameters of the parallel robot;

[0012] Step S12: identifying the joint friction force according to the structural parameters of the parallel robot and generating joint friction force data;

[0013] Step S13: establishing a manipulator dynamics model based on the parallel robot structural parameters and joint friction force data to generate a manipulator dynamics model;

[0014] Step S14: Performing terminal working space analysis according to the manipulator dynamics model to generate terminal working space range data;

[0015] Step S15: using the end workspace range data to identify the singular configuration of the mechanism of the manipulator dynamics model, and obtaining the singular configuration point data of the mechanism.

[0016] The present invention starts with the structural parameter measurement of the parallel robot and obtains accurate robot geometric dimension information. By incorporating the joint friction force into the dynamic model, the robot's motion trajectory and force conditions can be more accurately predicted, thereby improving the performance of the control system. Combining accurate structural parameters and joint friction force data, a high-precision manipulator dynamics model is established. The model can more accurately reflect the actual motion characteristics of the robot. Based on the accurate dynamic model, the workspace analysis can more accurately determine the robot's reachable working range and prevent the robot from entering an uncontrollable area. At the same time, accurate singular configuration recognition can effectively prevent the robot from working near a singular configuration, avoid problems such as a sharp increase in torque and control failure, and improve the safety and reliability of the robot.

[0017] Preferably, step S2 comprises the following steps:

[0018] Step S21: optimizing the working posture space of the end working space range data by using the singular configuration point data of the mechanism to obtain optimized working space data;

[0019] Step S22: marking the key positions of the driving joints of the parallel robot based on the manipulator dynamics model, and deploying the end effector monitoring sensor network to obtain a sensor monitoring network;

[0020] Step S23: performing a modal vibration test on the parallel robot using a sensor monitoring network according to the optimized working space data to generate modal vibration characteristic parameters;

[0021] Step S24: Plan the high-speed motion task trajectory for the parallel robot, and use the sensor monitoring network to extract residual vibration features to generate motion task residual vibration feature data.

[0022] The present invention optimizes the terminal workspace through the singular configuration point data of the mechanism, avoids the robot working in the singular configuration area, thereby fundamentally reducing the vibration risk caused by the singular configuration and improving the safety of the robot. Based on the dynamic model, the key positions of the driving joints are marked, and the sensor monitoring network is deployed in combination with the acceleration sensor and the laser displacement sensor, so as to realize the comprehensive and accurate monitoring of the vibration state of the robot. The acceleration sensor can capture high-frequency vibration information, while the laser displacement sensor can measure the low-frequency displacement change. The combination of the two can provide more complete vibration data. Based on the modal vibration test after the optimization of the workspace, the natural frequency and vibration mode of the robot can be more accurately identified, and the interference of the singular configuration on the test results can be avoided. The test is carried out on the optimized workspace, eliminating the influence of the singular configuration, and improving the accuracy and reliability of the test results. By planning the high-speed motion task trajectory and extracting the residual vibration features, vibration data closer to the actual working conditions can be obtained.

[0023] Preferably, step S23 includes the following steps:

[0024] Step S231: Selecting working test points according to the optimized working space data to generate spatial test point position data;

[0025] Step S232: Predicting the modal vibration shape of the parallel robot through the spatial test point position data to generate main modal vibration shape data;

[0026] Step S233: Based on the main modal vibration shape data and the spatial test point position data, a sensor monitoring network is used to perform a preliminary modal test on the parallel robot to generate preliminary modal vibration characteristic parameters;

[0027] Step S234: using a preset multi-degree-of-freedom vibrator model to perform subspace control parameter identification on the preliminary modal vibration characteristic parameters, and perform vibration modal parameter association to generate modal vibration characteristic parameters.

[0028] The present invention selects test points based on the optimized workspace, avoids testing near singular configurations, and thus ensures the validity and reliability of the test results. Test points are selected in a targeted manner according to the optimized workspace, which reduces unnecessary tests and improves test efficiency. The estimated modal vibration type information can help determine the optimal position of the sensor, as well as the optimal frequency and amplitude of the excitation signal, thereby improving test efficiency and accuracy. The combination of preliminary modal testing and subspace control parameter identification effectively improves the accuracy of modal parameter identification. Preliminary modal testing can quickly obtain preliminary modal parameters, while subspace control parameter identification can perform more detailed identification on this basis. Using a preset multi-degree-of-freedom oscillator model to identify subspace control parameters can effectively reduce computational complexity and improve identification efficiency. Compared with complex finite element models, the preset oscillator model is more concise and efficient, and can significantly reduce computational cost and time while ensuring a certain degree of accuracy, making modal parameter identification more practical.

[0029] Preferably, step S233 includes the following steps:

[0030] Based on the spatial test point position data, the parallel robot is subjected to terminal pulse excitation test, and the vibration response signal is collected using the sensor monitoring network to obtain the pulse excitation response data;

[0031] Perform a preliminary estimation of the attitude natural frequency based on the pulse excitation response data and generate multi-attitude natural frequency estimation data;

[0032] The multi-frequency sinusoidal excitation signal is processed according to the multi-posture natural frequency estimation data, and the signal is superimposed using a preset pseudo-random excitation signal to obtain the multi-frequency sinusoidal excitation signal data;

[0033] Based on the multi-frequency sinusoidal excitation signal data, the controller drives the joint motor of the parallel robot to perform sinusoidal frequency sweep motion, and uses the sensor monitoring network to perform sinusoidal frequency sweep vibration response to obtain sinusoidal frequency sweep excitation response data;

[0034] The frequency response fitting of the sine sweep excitation response data is performed through the main modal vibration data, and the natural frequency damping ratio is calculated based on the multi-posture natural frequency estimation data to obtain the natural frequency damping ratio data;

[0035] The multi-posture natural frequency estimation data, natural frequency damping ratio data and main modal vibration shape data are integrated into preliminary vibration modal parameters to obtain preliminary modal vibration characteristic parameters.

[0036] The present invention adopts pulse excitation test and sensor monitoring network, which can quickly obtain the vibration response data of the robot under different postures, and make an initial estimation of the posture natural frequency, which can quickly provide preliminary frequency information and improve the test efficiency. Based on the multi-posture natural frequency estimation data, a multi-frequency sinusoidal excitation signal is generated, and a pseudo-random excitation signal is superimposed, which can effectively excite the multi-order mode of the robot and improve the signal-to-noise ratio. Compared with single-frequency sinusoidal excitation, multi-frequency excitation can excite multiple modes at the same time, reducing the test time; and the superposition of pseudo-random signals can further enhance the richness of the excitation signal and improve the accuracy of modal recognition. The controller is used to drive the joint motor to perform sinusoidal sweep frequency motion, and the vibration response data is collected in combination with the sensor monitoring network, so that more refined frequency response characteristics can be obtained. Sine sweep frequency excitation can cover a wider frequency range and scan with a smaller frequency step size, so as to obtain more detailed frequency response information. The sinusoidal sweep frequency excitation response data is frequency-fitted by the main modal vibration shape data, and the natural frequency damping ratio is calculated in combination with the multi-posture natural frequency estimation data, so that the robot's key modal parameters such as natural frequency, damping ratio and vibration shape can be accurately identified.

[0037] Preferably, step S24 includes the following steps:

[0038] Step S241: Analyze the actual work tasks of the parallel robot, and process the typical high-speed motion mode to generate high-speed motion mode data;

[0039] Step S242: planning a high-speed motion task trajectory for the high-speed motion mode data by optimizing the working space data, and setting motion parameters, thereby obtaining high-speed motion task sequence data;

[0040] Step S243: Based on the high-speed motion task sequence data, the parallel robot is controlled to sequentially perform the high-speed motion tasks, and a sensor monitoring network is used to synchronously collect multi-channel terminal vibrations to generate residual vibration raw signal data;

[0041] Step S244: performing time-frequency domain transformation processing on the residual vibration original signal data to generate residual vibration time-frequency spectrum data;

[0042] Step S245: extracting the vibration signal decay time according to the residual vibration time-frequency spectrum data to generate vibration decay time data;

[0043] Step S246: extracting the dominant frequency of the residual vibration time-frequency spectrum data to generate residual vibration dominant frequency data;

[0044] Step S247: Calculating the vibration energy attenuation rate of the residual vibration time-frequency spectrum data to generate vibration energy attenuation rate data;

[0045] Step S248: Perform residual vibration feature clustering processing based on the vibration attenuation time data, the residual vibration dominant frequency data, and the vibration energy attenuation rate data, and perform motion task mapping processing based on the high-speed motion task sequence data to generate motion task residual vibration feature data.

[0046] The present invention analyzes and extracts typical high-speed motion patterns based on actual work tasks, ensuring the practicality and pertinence of subsequent trajectory planning and vibration testing. Compared with traditional random trajectory testing, the test method based on actual work tasks can more accurately reflect the vibration characteristics of the robot during actual work, thereby more targeted vibration suppression. High-speed motion trajectory planning is carried out on the basis of optimizing the workspace, avoiding the robot from moving near singular configurations, thereby ensuring the validity and safety of the test results. By using a sensor monitoring network to synchronously collect multi-channel terminal vibrations, more comprehensive vibration information can be obtained, and the errors caused by single-point measurement can be avoided. Multi-channel data acquisition can more completely reflect the vibration state of the robot terminal. By performing time-frequency domain transformation processing on the collected vibration data, key characteristic parameters such as the decay time, dominant frequency and energy decay rate of the vibration signal can be extracted. Time-frequency domain analysis can more clearly show the frequency components and time evolution laws of the vibration signal, thereby more accurately extracting the characteristic information of the residual vibration. The extracted characteristic parameters are clustered and mapped to the corresponding motion tasks, and the correlation relationship between the motion tasks and the residual vibration characteristics can be established.

[0047] Preferably, step S3 comprises the following steps:

[0048] Step S31: dividing the residual vibration characteristic data of the motion task into vibration characteristic intervals to generate characteristic interval division data;

[0049] Step S32: fuzzy variables are defined for the residual vibration characteristic data of the motion task through the characteristic quantity interval division data, and input fuzzy variable data and output fuzzy variable data are obtained respectively; wherein the input fuzzy variable data includes vibration attenuation time data, dominant frequency data and energy attenuation rate data; and the output fuzzy variable data includes velocity feedforward gain data and acceleration feedback gain data;

[0050] Step S33: Performing input-output fuzzy control rule mapping on the vibration attenuation time data, the dominant frequency data, the energy attenuation rate data, the velocity feedforward gain data and the acceleration feedback gain data through a preset vibration mode fuzzy control rule library to generate a vibration suppression control rule strategy;

[0051] Step S34: Processing the vibration control performance requirement index of the parallel robot to generate vibration control performance requirement index data;

[0052] Step S35: performing control parameter intelligent adjustment strategy processing on the vibration suppression control rule strategy and the vibration control performance requirement index data through the modal vibration characteristic parameters to obtain an intelligent vibration control parameter adjustment strategy.

[0053] The present invention divides the residual vibration characteristic data of the motion task into characteristic quantity intervals, defines fuzzy variables, and converts the complex vibration characteristic data into fuzzy language variables, so that the control rules are easier to understand and express, and more in line with the actual situation. The introduction of fuzzy variables can better handle the uncertainty and ambiguity of vibration characteristic data. The input and output fuzzy control rule mapping is performed using a preset vibration mode fuzzy control rule library, and a nonlinear mapping relationship between vibration characteristic quantities and control parameters is established. The application of the fuzzy control rule library avoids complex mathematical modeling, simplifies the design of the controller, and can effectively handle nonlinear systems. The vibration control performance requirement index is introduced, and the modal vibration characteristic parameters are integrated into the intelligent adjustment strategy of the control parameters, so as to realize the refined adjustment of the control parameters. The introduction of the control performance requirement index makes the control target clearer, and the addition of the modal vibration characteristic parameters can optimize the control parameters according to the vibration characteristics of the robot itself, thereby improving the pertinence and effectiveness of vibration suppression.

[0054] Preferably, step S35 includes the following steps:

[0055] Step S351: extracting residual vibration control parameters based on the vibration control performance requirement index data to obtain multi-dimensional residual vibration control space parameters;

[0056] Step S352: Calculating the sensitivity of vibration control parameters to the multi-dimensional residual vibration control space parameters using the modal vibration characteristic parameters to generate residual vibration control parameter weight data;

[0057] Step S353: performing vibration control performance target processing according to the residual vibration control parameter weight data and the vibration control performance requirement index data, and constructing a vibration control parameter adjustment model using a particle swarm algorithm and a vibration suppression control rule strategy to generate a vibration control parameter adjustment model;

[0058] Step S354: initializing the fuzzy control parameter particle swarm of the vibration control parameter adjustment model to generate initialization control parameter particle swarm data;

[0059] Step S355: Calculate the first-order contribution effect index of the initialized control parameter particle swarm data, and evaluate the particle fitness to generate a control parameter fitness value;

[0060] Step S356: Based on the control parameter fitness value, the vibration control parameter adjustment model is used to update the particle position, and the control parameter is iteratively optimized to obtain an intelligent vibration control parameter adjustment strategy.

[0061] The present invention extracts multidimensional residual vibration control space parameters based on vibration control performance requirement indicators, and provides a search space for subsequent control parameter optimization. The construction of multidimensional parameter space can more comprehensively consider various influencing factors, so as to find a better control parameter combination. The residual vibration control parameter sensitivity is calculated using modal vibration characteristic parameters, and the control parameter weight data is generated, which provides guidance for the optimization process of the particle swarm algorithm. The calculation of the control parameter sensitivity can identify the key parameters that have a greater impact on the vibration suppression effect, thereby improving the optimization efficiency and avoiding blind search. Combining the residual vibration control parameter weight data, the vibration control performance requirement indicator data and the vibration suppression control rule strategy, a vibration control parameter adjustment model is constructed, which provides a framework for the application of the particle swarm algorithm. The model combines fuzzy control rules with the particle swarm algorithm to realize rule-based intelligent optimization. The intelligent vibration control parameter adjustment strategy is obtained by iterative optimization of the control parameters, which can adaptively adjust the control parameters according to different vibration conditions, thereby achieving more accurate vibration suppression.

[0062] Preferably, step S4 comprises the following steps:

[0063] Step S41: using the terminal device to obtain the manipulator feedforward control parameters; wherein the manipulator feedforward control parameters include feedforward monitoring time domain data, feedforward control time domain data and control compensation data;

[0064] Step S42: When the parallel robot performs real-time high-speed operation, the sensor monitoring network is used to perform real-time vibration monitoring of the end effector based on the feedforward monitoring time domain data, the feedforward control time domain data, and the control compensation data to obtain real-time vibration monitoring state data;

[0065] Step S43: extracting terminal vibration features from the real-time vibration monitoring state data to generate real-time terminal vibration feature data;

[0066] Step S44: performing intelligent feedforward compensation control adjustment processing on the real-time terminal vibration characteristic data through an intelligent vibration control parameter adjustment strategy to generate feedforward compensation control adjustment parameters;

[0067] Step S45: setting a control parameter buffer for the parallel robot, and performing smooth switching of control parameters according to the feedforward compensation control adjustment parameters, so as to realize intelligent parameter adjustment control operation of the three-degree-of-freedom parallel robot.

[0068] The present invention uses a terminal device to obtain the feedforward control parameters of the manipulator, including feedforward monitoring time domain data, feedforward control time domain data and control compensation data. Based on the feedforward control parameters and the sensor monitoring network, the end effector of the parallel robot is subjected to real-time vibration monitoring during real-time high-speed operation. The real-time vibration monitoring can timely capture the vibration state of the robot end. Intelligent feedforward control can predict and compensate for the upcoming vibration based on the real-time vibration characteristic data, thereby effectively reducing the residual vibration. By setting a control parameter buffer and smoothly switching the control parameters according to the feedforward compensation control adjustment parameters, the sudden change of the control parameters is avoided and the stability of the control system is ensured. Through the synergistic effect of real-time vibration monitoring, intelligent feedforward compensation control and smooth switching of control parameters, the residual vibration of the parallel robot is effectively suppressed, the positioning accuracy and stability of the robot during high-speed operation are significantly improved, and the reliability and safety of the robot's operation are ensured.

[0069] The present invention also provides a three-degree-of-freedom parallel robot control parameter adjustment system, which executes the three-degree-of-freedom parallel robot control parameter adjustment method as described above, and the three-degree-of-freedom parallel robot control parameter adjustment system comprises:

[0070] The robot dynamics modeling module is used to establish the manipulator dynamics model of the parallel robot and generate the manipulator dynamics model; perform terminal operation space analysis based on the manipulator dynamics model to generate terminal workspace range data; perform mechanism singular configuration identification on the manipulator dynamics model through the terminal workspace range data to obtain mechanism singular configuration point data;

[0071] The residual vibration extraction module is used to optimize the working posture space of the end workspace range data through the singular configuration point data of the mechanism to obtain the optimized working space data; deploy the end monitoring sensor network on the parallel robot to obtain the sensor monitoring network; use the sensor monitoring network to perform modal vibration test on the parallel robot to generate modal vibration characteristic parameters; use the sensor monitoring network to extract the high-speed motion residual vibration characteristics of the parallel robot to generate the residual vibration characteristic data of the motion task;

[0072] The intelligent vibration control module is used to process the residual vibration characteristic data of the motion task with an adaptive rule adjustment strategy to generate a vibration suppression control rule strategy; the vibration suppression control rule strategy is processed with an intelligent control parameter adjustment strategy through modal vibration characteristic parameters to obtain an intelligent vibration control parameter adjustment strategy;

[0073] The real-time vibration control execution module is used to use the sensor monitoring network to perform real-time vibration monitoring of the end effector of the parallel robot and generate real-time end vibration characteristic data; to perform intelligent feedforward compensation control adjustment processing on the real-time end vibration characteristic data through the intelligent vibration control parameter adjustment strategy and generate feedforward compensation control adjustment parameters; to smoothly switch the control parameters of the parallel robot through the feedforward compensation control adjustment parameters to realize the intelligent parameter adjustment control operation of the three-degree-of-freedom parallel robot. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] Figure 1 A schematic flow chart of the steps of the control parameter adjustment method of the three-degree-of-freedom parallel robot of the present invention;

[0075] Figure 2 for Figure 1 Detailed implementation steps of step S1 in FIG.

[0076] Figure 3 for Figure 1 Detailed implementation steps of step S3 in FIG.

[0077] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0078] The technical method of the present invention is described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0079] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0080] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0081] To achieve this, please refer to Figures 1 to 3 The present invention provides a control parameter adjustment method for a three-degree-of-freedom parallel robot, comprising the following steps:

[0082] Step S1: Establish a manipulator dynamics model for the parallel robot to generate a manipulator dynamics model; perform terminal operation space analysis based on the manipulator dynamics model to generate terminal workspace range data; perform mechanism singular configuration identification on the manipulator dynamics model through the terminal workspace range data to obtain mechanism singular configuration point data;

[0083] Step S2: optimizing the working posture space of the end working space range data through the singular configuration point data of the mechanism to obtain the optimized working space data; deploying the end monitoring sensor network on the parallel robot to obtain the sensor monitoring network; performing modal vibration test on the parallel robot based on the optimized working space data using the sensor monitoring network to generate modal vibration characteristic parameters; performing high-speed motion residual vibration feature extraction on the parallel robot based on the optimized working space data using the sensor monitoring network to generate motion task residual vibration characteristic data;

[0084] Step S3: performing adaptive rule adjustment strategy processing on the residual vibration characteristic data of the motion task to generate a vibration suppression control rule strategy; performing control parameter intelligent adjustment strategy processing on the vibration suppression control rule strategy through modal vibration characteristic parameters to obtain an intelligent vibration control parameter adjustment strategy;

[0085] Step S4: Use the sensor monitoring network to perform real-time vibration monitoring of the end effector of the parallel robot to generate real-time end vibration characteristic data; use the intelligent vibration control parameter adjustment strategy to perform intelligent feedforward compensation control adjustment processing on the real-time end vibration characteristic data to generate feedforward compensation control adjustment parameters; use the feedforward compensation control adjustment parameters to smoothly switch the control parameters of the parallel robot to realize the intelligent parameter adjustment control operation of the three-degree-of-freedom parallel robot.

[0086] In the embodiment of the present invention, reference Figure 1FIG. 1 is a schematic diagram of a process flow of a control parameter adjustment method for a three-degree-of-freedom parallel robot according to the present invention. In this embodiment, the control parameter adjustment method for a three-degree-of-freedom parallel robot includes the following steps:

[0087] Step S1: Establish a manipulator dynamics model for the parallel robot to generate a manipulator dynamics model; perform terminal operation space analysis based on the manipulator dynamics model to generate terminal workspace range data; perform mechanism singular configuration identification on the manipulator dynamics model through the terminal workspace range data to obtain mechanism singular configuration point data;

[0088] In an embodiment of the present invention, a dynamic modeling method, such as the Lagrangian method or the Newton-Euler method, is selected to perform dynamic modeling on a three-degree-of-freedom parallel robot. During the modeling process, it is necessary to determine the connecting rod parameters of the robot, including the connecting rod length, mass, center of mass position, and joint type. For example, each connecting rod of the robot can be regarded as a rigid body, and the kinematic constraint equations between the connecting rods can be established. By solving these equations, the relationship between the displacement, velocity, and acceleration of each joint of the robot can be obtained. Then, according to the dynamic equation of the robot, the relationship between the driving force / torque of each joint of the robot and the end position of the robot can be derived, and finally a manipulator dynamic model is generated. Based on the established dynamic model, the end workspace of the robot is analyzed. By changing the input variables (such as joint angles or displacements) of each joint of the robot, all spatial positions that the end effector of the robot can reach are calculated, and the range of the workspace is plotted. For example, the robot workspace can be discretized, and each discrete point can be subjected to reachability analysis, and finally the end workspace range data is generated. Using the end workspace range data, the singular configuration of the robot's mechanism is identified. A singular configuration of a mechanism refers to a configuration in which the determinant of the Jacobian matrix of the robot is zero or close to zero. In these configurations, some degrees of freedom of the robot will be lost, resulting in a decrease in the robot's motion accuracy or uncontrollable. The data of the singular configuration points of the mechanism can be obtained by calculating the determinant of the Jacobian matrix of the robot and searching for points where the determinant is zero or close to zero within the workspace. For example, a threshold can be set, and when the absolute value of the determinant of the Jacobian matrix is ​​less than the threshold, the configuration is considered to be a singular configuration.

[0089] Step S2: optimizing the working posture space of the end working space range data through the singular configuration point data of the mechanism to obtain the optimized working space data; deploying the end monitoring sensor network on the parallel robot to obtain the sensor monitoring network; performing modal vibration test on the parallel robot based on the optimized working space data using the sensor monitoring network to generate modal vibration characteristic parameters; performing high-speed motion residual vibration feature extraction on the parallel robot based on the optimized working space data using the sensor monitoring network to generate motion task residual vibration characteristic data;

[0090] In an embodiment of the present invention, the end workspace range data is optimized according to the singular configuration point data of the mechanism obtained in step S1. The optimized workspace data can be obtained by eliminating the area containing the singular configuration point or adjusting the workspace boundary to prevent the robot from entering the singular configuration. For example, the area around the singular configuration point can be defined as a restricted area, and these areas can be avoided during path planning. A sensor monitoring network is deployed on the parallel robot. For example, an acceleration sensor can be installed on each connecting rod of the robot to measure the vibration signal of the robot. The number and position of the sensors need to be selected according to the structure and application scenario of the robot. The parallel robot is subjected to modal vibration testing using the deployed sensor monitoring network. For example, the vibration of the robot can be stimulated by means of knocking excitation or white noise excitation, and sensor data can be collected. The collected data is subjected to spectrum analysis to obtain modal vibration characteristic parameters of the robot, such as natural frequency, damping ratio and modal vibration shape. The residual vibration characteristics of the parallel robot under high-speed motion are extracted using the sensor monitoring network. For example, the robot can be controlled to perform a specific high-speed motion task and the sensor data can be recorded. By performing time domain and frequency domain analysis on the collected data, we can obtain residual vibration characteristic data of the motion task, such as residual vibration amplitude, decay time, and frequency component. For example, we can analyze the vibration data after the robot completes point-to-point motion and extract the residual vibration characteristics.

[0091] Step S3: performing adaptive rule adjustment strategy processing on the residual vibration characteristic data of the motion task to generate a vibration suppression control rule strategy; performing control parameter intelligent adjustment strategy processing on the vibration suppression control rule strategy through modal vibration characteristic parameters to obtain an intelligent vibration control parameter adjustment strategy;

[0092] In an embodiment of the present invention, the residual vibration characteristic data of the motion task obtained in step S2 is analyzed to extract key characteristic parameters. These characteristic parameters may include residual amplitude, main frequency, decay time constant, etc. For example, multiple repeated experiments can be performed on point-to-point motion at different speeds and loads, and the residual vibration data after each motion can be recorded, and the above characteristic parameters can be extracted therefrom. Then, based on these characteristic parameters, an adaptive rule adjustment strategy is established. The goal of this strategy is to dynamically adjust the vibration suppression control rule according to different motion tasks. Fuzzy logic control is adopted. The extracted characteristic parameters can be used as the input of the fuzzy logic system, and the output is the vibration suppression control rule. For example, the input can be the residual amplitude and main frequency, and the output can be the damping coefficient of the controller. Through the pre-set fuzzy rules, the parameters of the controller can be adjusted in real time according to the changes in the input characteristic parameters, thereby realizing adaptive control. The modal vibration characteristic parameters obtained in step S2 are used to intelligently adjust the control parameters of the generated vibration suppression control rule strategy. Modal vibration characteristic parameters, such as natural frequency, damping ratio and vibration mode, reflect the natural vibration characteristics of the robot. These parameters can be integrated into the adjustment strategy of the control parameters to improve the control effect. For example, a notch filter can be designed based on the first-order natural frequency of the robot to suppress vibrations near this frequency. Alternatively, the control gain of each joint can be adjusted based on the modal vibration shape so that the control force / torque can suppress vibration more effectively, such as by using a genetic algorithm to search for the optimal control parameters. Finally, an intelligent vibration control parameter adjustment strategy is obtained, which can adaptively adjust the control parameters according to the robot's modal characteristics and motion tasks.

[0093] Step S4: Use the sensor monitoring network to perform real-time vibration monitoring of the end effector of the parallel robot to generate real-time end vibration characteristic data; use the intelligent vibration control parameter adjustment strategy to perform intelligent feedforward compensation control adjustment processing on the real-time end vibration characteristic data to generate feedforward compensation control adjustment parameters; use the feedforward compensation control adjustment parameters to smoothly switch the control parameters of the parallel robot to realize the intelligent parameter adjustment control operation of the three-degree-of-freedom parallel robot.

[0094] In an embodiment of the present invention, a sensor monitoring network is used to perform real-time vibration monitoring on the end effector of the parallel robot. The sensor network can be composed of multiple accelerometers or other types of vibration sensors, which are distributed on the end effector of the robot and are used to collect vibration signals in real time. The collected signal is converted into real-time end vibration characteristic data that can be used for control after data processing, such as filtering and amplification. For example, a Kalman filter can be used to filter the sensor data, remove noise interference, and estimate the real-time vibration state of the end effector, including vibration displacement, velocity and acceleration. The intelligent vibration control parameter adjustment strategy obtained by step S3 performs intelligent feedforward compensation control adjustment processing on the real-time end vibration characteristic data. According to the real-time vibration characteristic data, the intelligent vibration control parameter adjustment strategy can adjust the parameters of the controller in real time, such as the gain of the PID controller, or the compensation coefficient of the feedforward controller. For example, the strength of the feedforward compensation can be adjusted according to the size of the real-time vibration amplitude. The larger the vibration amplitude, the greater the compensation strength. This intelligent feedforward compensation control can effectively predict and offset the upcoming vibration, thereby minimizing the residual vibration. The control parameters of the parallel robot are smoothly switched by the feedforward compensation control adjustment parameters. In order to avoid sudden changes in control parameters that lead to system instability, the control parameters need to be switched smoothly. A feasible solution is to use a low-pass filter to filter the control parameters, or to use a gradient function to smoothly transition the control parameters. For example, a time constant can be set, and the rate of change of the control parameters cannot exceed the reciprocal of the time constant. By smoothly switching the control parameters, the smoothness of the robot's motion and the stability of the control system can be guaranteed, and ultimately the intelligent parameter adjustment control operation of the three-degree-of-freedom parallel robot can be realized.

[0095] Step S1 includes the following steps:

[0096] Step S11: measuring the structural parameters of the parallel robot to obtain the structural parameters of the parallel robot;

[0097] Step S12: identifying the joint friction force according to the structural parameters of the parallel robot and generating joint friction force data;

[0098] Step S13: establishing a manipulator dynamics model based on the parallel robot structural parameters and joint friction force data to generate a manipulator dynamics model;

[0099] Step S14: Performing terminal working space analysis according to the manipulator dynamics model to generate terminal working space range data;

[0100] Step S15: using the end workspace range data to identify the singular configuration of the mechanism of the manipulator dynamics model, and obtaining the singular configuration point data of the mechanism.

[0101] As an example of the present invention, refer to Figure 2 As shown, Figure 1 Detailed implementation steps of step S1 in the flowchart, in this example, step S1 includes:

[0102] Step S11: measuring the structural parameters of the parallel robot to obtain the structural parameters of the parallel robot;

[0103] In an embodiment of the present invention, to measure the structural parameters of a three-degree-of-freedom parallel robot, it is necessary to measure the length, mass, center of mass position of each connecting rod, and the type and position of the joints. High-precision measuring tools, such as laser trackers, three-dimensional coordinate measuring machines, etc., can be used to measure these parameters. For example, a laser tracker can be used to measure the length of the connecting rod, and the center of mass position of the connecting rod can be determined in combination with a CAD model. For the mass of the connecting rod, a high-precision electronic scale can be used for measurement. At the same time, it is necessary to record the type of joint, such as a rotating joint or a moving joint, as well as the installation position and direction of the joint. In order to improve the measurement accuracy, multiple measurements can be performed and the average value can be taken.

[0104] Step S12: identifying the joint friction force according to the structural parameters of the parallel robot and generating joint friction force data;

[0105] In an embodiment of the present invention, a series of different torques / forces are applied to each joint, and the movement speed of the joint is measured. By analyzing the relationship between torque / force and speed, the friction model parameters of the joint, such as Coulomb friction coefficient, viscous friction coefficient, etc., can be identified. A model-based method can also be used to identify friction. For example, by using known robot structural parameters and dynamic equations, combined with experimental data, the friction model parameters are estimated through a parameter identification algorithm, such as the least squares method. Finally, the identified joint friction data, such as the friction model parameters.

[0106] Step S13: establishing a manipulator dynamics model based on the parallel robot structural parameters and joint friction force data to generate a manipulator dynamics model;

[0107] In an embodiment of the present invention, a suitable dynamic modeling method is selected, such as the Lagrangian method or the Newton-Euler method. The Lagrangian method is based on the principle of conservation of energy, and the dynamic equation is derived by calculating the kinetic energy and potential energy of the system. The Newton-Euler method is based on Newton's second law and the Euler equation, and the dynamic equation is derived by analyzing the force conditions of each connecting rod. In the modeling process, the influence of joint friction needs to be considered. For example, friction can be added to the dynamic equation as a generalized force. Finally, a dynamic model describing the relationship between the robot joint driving force / torque and the joint displacement, velocity, and acceleration can be obtained. The model can be represented by a mathematical equation or simulation software.

[0108] Step S14: Performing terminal working space analysis according to the manipulator dynamics model to generate terminal working space range data;

[0109] In an embodiment of the present invention, the end working space refers to the set of all spatial positions that the end effector of the robot can reach. The end working space range of the robot can be determined by numerical calculation or simulation. For example, the angle or displacement of each joint of the robot can be discretized within the allowable range, and then the end effector position corresponding to each discrete point can be calculated. By connecting all the reachable position points, the end working space range of the robot can be obtained. In addition, indicators such as the accessibility and flexibility of the workspace can also be analyzed. For example, the Jacobian matrix of each position in the workspace can be calculated, and its singular value can be analyzed to evaluate the flexibility of the robot at that position.

[0110] Step S15: using the end workspace range data to identify the singular configuration of the mechanism of the manipulator dynamics model, and obtaining the singular configuration point data of the mechanism.

[0111] In an embodiment of the present invention, the end workspace range data generated in step S14 is used to identify the singular configuration of the mechanism of the manipulator dynamics model. The singular configuration of the mechanism refers to a configuration in which the determinant of the Jacobian matrix of the robot is zero or close to zero. In a singular configuration, some degrees of freedom of the robot will be lost, resulting in a decrease in motion accuracy or inability to control. The singular configuration of the mechanism can be identified by calculating the determinant of the Jacobian matrix of the robot and searching for points in the workspace where the determinant is zero or close to zero. For example, a threshold value can be set, and when the absolute value of the determinant of the Jacobian matrix is ​​less than the threshold value, the configuration is considered to be a singular configuration. The identified singular configuration points can be plotted in the workspace to intuitively understand the distribution of the singular configurations of the robot.

[0112] Step S2 includes the following steps:

[0113] Step S21: optimizing the working posture space of the end working space range data by using the singular configuration point data of the mechanism to obtain optimized working space data;

[0114] Step S22: marking the key positions of the driving joints of the parallel robot based on the manipulator dynamics model, and deploying the end effector monitoring sensor network to obtain a sensor monitoring network;

[0115] Step S23: performing a modal vibration test on the parallel robot using a sensor monitoring network according to the optimized working space data to generate modal vibration characteristic parameters;

[0116] Step S24: Plan the high-speed motion task trajectory for the parallel robot, and use the sensor monitoring network to extract residual vibration features to generate motion task residual vibration feature data.

[0117] In the embodiment of the present invention, singular configuration points may cause the robot control performance to deteriorate, so these points need to be eliminated or avoided in the workspace. The area around the singular configuration point can be defined as a restricted area, and these areas can be avoided during path planning. The singular configuration point can also be excluded from the workspace by adjusting the workspace boundary. For example, if a singular configuration point is located at the edge of the workspace, the workspace range can be narrowed so that it does not contain the singular configuration point, and after optimization, the optimized workspace data is obtained. According to the dynamic model, the force conditions of each joint are analyzed, and the joints with greater force or joints that are prone to vibration are marked as key positions. Then, a sensor monitoring network is deployed at these key positions and the end effector. The sensor monitoring network includes an acceleration sensor and a laser displacement sensor. The acceleration sensor is used to measure the vibration acceleration signal of each component of the robot, and the laser displacement sensor is used to measure the displacement information of the end effector. The number and position of the sensors need to be selected according to the structure and application scenario of the robot. For example, an acceleration sensor can be installed on each drive joint, and multiple acceleration sensors and laser displacement sensors can be installed on the end effector. The robot is stimulated in different positions in the optimized working space by using hammering method or exciter excitation method, and sensor data is collected. For example, a force hammer can be used to hit the robot base or connecting rod to stimulate the vibration of the robot, and the response signal of the acceleration sensor is recorded. Then, the collected data is subjected to spectral analysis, for example, the modal vibration characteristic parameters are extracted using the fast Fourier transform (FFT) method. The parallel robot is subjected to high-speed motion task trajectory planning. For example, a point-to-point straight line trajectory or a complex curved trajectory can be planned. Then, the robot is controlled to move at high speed according to the planned trajectory, and the residual vibration feature extraction is performed using the sensor monitoring network deployed in step S22. Residual vibration refers to the vibration that still exists in the end effector after the robot completes high-speed motion. The data of the acceleration sensor and the laser displacement sensor can be recorded, and the characteristics such as the amplitude, frequency and decay time of the vibration can be analyzed.

[0118] Step S23 includes the following steps:

[0119] Step S231: Selecting working test points according to the optimized working space data to generate spatial test point position data;

[0120] Step S232: Predicting the modal vibration shape of the parallel robot through the spatial test point position data to generate main modal vibration shape data;

[0121] Step S233: Based on the main modal vibration shape data and the spatial test point position data, a sensor monitoring network is used to perform a preliminary modal test on the parallel robot to generate preliminary modal vibration characteristic parameters;

[0122] Step S234: using a preset multi-degree-of-freedom vibrator model to perform subspace control parameter identification on the preliminary modal vibration characteristic parameters, and perform vibration modal parameter association to generate modal vibration characteristic parameters.

[0123] In an embodiment of the present invention, several representative test points are selected in the optimized working space. The test points can be selected in a uniformly distributed manner, for example, the workspace is divided into several small areas, and a test point is selected in each area. It is also possible to select test points at some specific locations according to the actual application scenario, for example, the location where the robot often works or the location where vibration is more sensitive, to generate spatial test point position data. Finite element analysis software can be used to simulate the modal vibration shape of the robot at each test point position. For example, the CAD model of the robot can be imported into the finite element analysis software, and parameters such as material properties, boundary conditions and excitation methods can be set. Then, modal analysis is performed to obtain the main modal vibration shape data of the robot at each test point position, including vibration shape frequency and vibration shape vector. The sensor monitoring network deployed in step S22 is used to perform preliminary modal testing on the parallel robot. According to the estimated modal vibration shape, a suitable excitation method and sensor layout scheme are selected. For example, according to the vibration shape vector, the position of the excitation point and the sensor can be selected so that the excitation can effectively excite the target mode and the sensor can accurately measure the vibration response. At each test point, excitation is applied to the robot and sensor data is collected. The preset multi-degree-of-freedom oscillator model is a simplified dynamic model used to approximately describe the vibration characteristics of the parallel robot. The model can be selected according to the number of degrees of freedom of the robot. For example, for a three-degree-of-freedom parallel robot, a three-degree-of-freedom oscillator model can be selected. Model parameters, such as mass, stiffness, and damping coefficient, can be obtained through theoretical calculations, finite element analysis, or experimental measurements. The subspace identification method can identify more accurate model parameters based on preliminary modal test data and preset models, and associate them with preliminary modal vibration characteristic parameters to ultimately generate more accurate modal vibration characteristic parameters. Among them, the preset multi-degree-of-freedom oscillator model is used to approximately describe the dynamic characteristics of the robot. For a three-degree-of-freedom parallel robot, it can be assumed to be a three-degree-of-freedom oscillator, with each degree of freedom corresponding to an independent vibration mode. Each mode can be described by a second-order differential equation: m x''+c x'+kx=F; where m represents equivalent mass, c represents equivalent damping, k represents equivalent stiffness, x represents displacement, and F represents external force.

[0124] Step S233 includes the following steps:

[0125] Based on the spatial test point position data, the parallel robot is subjected to terminal pulse excitation test, and the vibration response signal is collected using the sensor monitoring network to obtain the pulse excitation response data;

[0126] Perform a preliminary estimation of the attitude natural frequency based on the pulse excitation response data and generate multi-attitude natural frequency estimation data;

[0127] The multi-frequency sinusoidal excitation signal is processed according to the multi-posture natural frequency estimation data, and the signal is superimposed using a preset pseudo-random excitation signal to obtain the multi-frequency sinusoidal excitation signal data;

[0128] Based on the multi-frequency sinusoidal excitation signal data, the controller drives the joint motor of the parallel robot to perform sinusoidal frequency sweep motion, and uses the sensor monitoring network to perform sinusoidal frequency sweep vibration response to obtain sinusoidal frequency sweep excitation response data;

[0129] The frequency response fitting of the sine sweep excitation response data is performed through the main modal vibration data, and the natural frequency damping ratio is calculated based on the multi-posture natural frequency estimation data to obtain the natural frequency damping ratio data;

[0130] The multi-posture natural frequency estimation data, natural frequency damping ratio data and main modal vibration shape data are integrated into preliminary vibration modal parameters to obtain preliminary modal vibration characteristic parameters.

[0131] In an embodiment of the present invention, based on the spatial test point position data, the parallel robot is moved to each test point position. At each test point position, a pulse excitation test is performed on the robot end effector. A force hammer can be used to strike the end effector, or other devices capable of generating pulse excitation can be used. At the same time, a sensor monitoring network is used to collect vibration response signals, including data of an acceleration sensor and a laser displacement sensor. According to the pulse excitation response data obtained in step S2331, an initial estimation of the posture natural frequency is performed. The vibration response signal at each test point position is subjected to spectrum analysis, for example, a spectrum diagram of the vibration signal can be obtained using a fast Fourier transform (FFT) method. In the spectrum diagram, the frequency corresponding to the peak is the natural frequency of the robot in the posture. The estimated natural frequency is used as the frequency of the sinusoidal signal to generate multiple sinusoidal signals of different frequencies. At the same time, a pseudo-random excitation signal is preset. The pseudo-random excitation signal can be white noise or other types of random signals, the purpose of which is to excite the robot's vibration response in a wider frequency range. The multi-frequency sinusoidal signal and the pseudo-random excitation signal are superimposed to obtain multi-frequency sinusoidal excitation signal data. The parallel robot joint motor is driven by a controller to perform a sinusoidal sweep motion. The multi-frequency sinusoidal excitation signal is used as the input signal of the controller to control the robot to move according to the frequency and amplitude of the excitation signal. At the same time, the sensor monitoring network is used to collect the sinusoidal swept frequency vibration response data, including the data of the acceleration sensor and the laser displacement sensor. The sinusoidal swept frequency excitation response data obtained in step S2334 is subjected to frequency response fitting. Modal analysis software, such as ME'scope, can be used to fit the frequency response data to obtain more accurate natural frequency and vibration shape information. Then, according to the multi-posture natural frequency estimation data generated in step S2332 and the frequency response function obtained by fitting, the damping ratio corresponding to each natural frequency is calculated. The half-power bandwidth method or other damping ratio identification methods can be used for calculation to obtain the natural frequency damping ratio data. The multi-posture natural frequency estimation data generated in step S2332, the natural frequency damping ratio data calculated in step S2335, and the main modal vibration shape data generated in step S232 are integrated to obtain preliminary modal vibration characteristic parameters.

[0132] Step S24 includes the following steps:

[0133] Step S241: Analyze the actual work tasks of the parallel robot, and process the typical high-speed motion mode to generate high-speed motion mode data;

[0134] Step S242: planning a high-speed motion task trajectory for the high-speed motion mode data by optimizing the working space data, and setting motion parameters, thereby obtaining high-speed motion task sequence data;

[0135] Step S243: Based on the high-speed motion task sequence data, the parallel robot is controlled to sequentially perform the high-speed motion tasks, and a sensor monitoring network is used to synchronously collect multi-channel terminal vibrations to generate residual vibration raw signal data;

[0136] Step S244: performing time-frequency domain transformation processing on the residual vibration original signal data to generate residual vibration time-frequency spectrum data;

[0137] Step S245: extracting the vibration signal decay time according to the residual vibration time-frequency spectrum data to generate vibration decay time data;

[0138] Step S246: extracting the dominant frequency of the residual vibration time-frequency spectrum data to generate residual vibration dominant frequency data;

[0139] Step S247: Calculating the vibration energy attenuation rate of the residual vibration time-frequency spectrum data to generate vibration energy attenuation rate data;

[0140] Step S248: Perform residual vibration feature clustering processing based on the vibration attenuation time data, the residual vibration dominant frequency data, and the vibration energy attenuation rate data, and perform motion task mapping processing based on the high-speed motion task sequence data to generate motion task residual vibration feature data.

[0141] In an embodiment of the present invention, the actual work tasks of the parallel robot are analyzed, such as pick and place, welding, spraying, etc. Typical high-speed motion modes are extracted from the actual work tasks, such as fast movement, fast rotation, etc. These high-speed motion modes are simplified and abstracted, for example, they can be represented by basic motion units such as point-to-point linear motion and circular motion. Finally, high-speed motion mode data is generated, which contains description information of various typical high-speed motions, such as motion trajectory, speed, acceleration, etc. The high-speed motion mode data generated in step S241 is planned for high-speed motion task trajectories based on the optimized working space data obtained in step S21. The planned trajectory needs to avoid singular configuration points in the workspace and ensure that the robot does not collide during the motion process. At the same time, it is necessary to set the motion parameters, such as setting the motion speed, acceleration, starting point and end point, etc. Finally, high-speed motion task sequence data is obtained, which contains a series of high-speed motion tasks executed in sequence, and each task contains detailed trajectory and motion parameter information. Based on the high-speed motion task sequence data obtained in step S242, the parallel robot is controlled to perform high-speed motion tasks in sequence. When the robot performs each high-speed motion task, the sensor monitoring network deployed in step S22 is used to perform multi-channel terminal vibration synchronous acquisition. For example, the data of multiple acceleration sensors and laser displacement sensors on the end effector are collected at the same time, and the collected data is processed in time synchronization to generate residual vibration original signal data. The residual vibration original signal data collected in step S243 is transformed in the time-frequency domain. The time domain signal can be converted into the time-frequency domain signal using methods such as short-time Fourier transform (STFT) or wavelet transform. The time-frequency domain transformation can simultaneously reflect the frequency and time information of the signal, which is helpful for analyzing the time-varying characteristics of the residual vibration. Finally, the residual vibration time-frequency spectrum data is generated. The attenuation of the vibration energy over time in the time-frequency spectrum is analyzed to extract the vibration attenuation time. For example, an energy threshold can be defined, and when the vibration energy decays below the threshold, it is considered that the vibration has stopped. The time taken to record the vibration energy decaying from the initial value to the threshold is the vibration attenuation time. The distribution of the vibration energy in the frequency dimension in the time-frequency spectrum is analyzed to extract the dominant frequency of the vibration. For example, the frequency point with the largest energy in the time-frequency spectrum can be found, and its corresponding frequency is the dominant frequency of the vibration. The vibration dominant frequency corresponding to each high-speed motion task is sorted into residual vibration dominant frequency data. The decay rate of vibration energy over time can be calculated. For example, an exponential decay model can be used to fit the decay curve of vibration energy, and the attenuation coefficient can be calculated. The vibration energy decay rate corresponding to each high-speed motion task is sorted into vibration energy decay rate data. Residual vibration data with similar characteristics are grouped together using methods such as K-means clustering. Then, according to the high-speed motion task sequence data obtained in step S242, the clustered residual vibration feature data is mapped with the corresponding motion task.Finally, the residual vibration feature data of the motion task is generated, which contains the residual vibration features corresponding to each high-speed motion task, such as vibration decay time, dominant frequency, energy decay rate, and cluster category.

[0142] Step S3 includes the following steps:

[0143] Step S31: dividing the residual vibration characteristic data of the motion task into vibration characteristic intervals to generate characteristic interval division data;

[0144] Step S32: fuzzy variables are defined for the residual vibration characteristic data of the motion task through the characteristic quantity interval division data, and input fuzzy variable data and output fuzzy variable data are obtained respectively; wherein the input fuzzy variable data includes vibration attenuation time data, dominant frequency data and energy attenuation rate data; and the output fuzzy variable data includes velocity feedforward gain data and acceleration feedback gain data;

[0145] Step S33: Performing input-output fuzzy control rule mapping on the vibration attenuation time data, the dominant frequency data, the energy attenuation rate data, the velocity feedforward gain data and the acceleration feedback gain data through a preset vibration mode fuzzy control rule library to generate a vibration suppression control rule strategy;

[0146] Step S34: Processing the vibration control performance requirement index of the parallel robot to generate vibration control performance requirement index data;

[0147] Step S35: performing control parameter intelligent adjustment strategy processing on the vibration suppression control rule strategy and the vibration control performance requirement index data through the modal vibration characteristic parameters to obtain an intelligent vibration control parameter adjustment strategy.

[0148] As an example of the present invention, refer to Figure 3 As shown, Figure 1 Detailed implementation steps of step S3 in the flowchart, in this example, step S3 includes:

[0149] Step S31: dividing the residual vibration characteristic data of the motion task into vibration characteristic intervals to generate characteristic interval division data;

[0150] In an embodiment of the present invention, the residual vibration characteristic data of the motion task generated in step S248 is divided into vibration characteristic intervals. The characteristic quantities such as vibration attenuation time, dominant frequency and energy attenuation rate are divided into several intervals according to their value ranges and distribution. For example, the vibration attenuation time can be divided into three intervals of "short", "medium" and "long", the dominant frequency can be divided into three intervals of "low", "medium" and "high", and the energy attenuation rate can be divided into three intervals of "fast", "medium" and "slow". The division of intervals can be adjusted according to experience or cluster analysis results to generate characteristic interval division data.

[0151] Step S32: fuzzy variables are defined for the residual vibration characteristic data of the motion task through the characteristic quantity interval division data, and input fuzzy variable data and output fuzzy variable data are obtained respectively; wherein the input fuzzy variable data includes vibration attenuation time data, dominant frequency data and energy attenuation rate data; and the output fuzzy variable data includes velocity feedforward gain data and acceleration feedback gain data;

[0152] In an embodiment of the present invention, each characteristic quantity and its corresponding interval division are converted into fuzzy variables. For example, the vibration attenuation time is converted into the fuzzy variable "attenuation time", and its membership function can be defined as a triangular function or a trapezoidal function, corresponding to the three intervals of "short", "medium" and "long", respectively. Similarly, the dominant frequency and the energy decay rate are also converted into fuzzy variables. Finally, the input fuzzy variable data (including vibration attenuation time data, dominant frequency data and energy decay rate data) and the output fuzzy variable data (including velocity feedforward gain data and acceleration feedback gain data) are obtained respectively. The interval division of the output fuzzy variable can be determined based on control experience or pre-conducted simulation experiments.

[0153] Step S33: Performing input-output fuzzy control rule mapping on the vibration attenuation time data, the dominant frequency data, the energy attenuation rate data, the velocity feedforward gain data and the acceleration feedback gain data through a preset vibration mode fuzzy control rule library to generate a vibration suppression control rule strategy;

[0154] In an embodiment of the present invention, input-output fuzzy control rule mapping is performed on the input-output fuzzy variable data generated in step S32. The vibration mode fuzzy control rule base is a predefined set of fuzzy rules used to describe the control parameter adjustment strategies corresponding to different combinations of vibration characteristic quantities. For example, a rule may be: if the "decay time" is "long" and the "dominant frequency" is "high" and the "energy decay rate" is "slow", then the "speed feedforward gain" is "large" and the "acceleration feedback gain" is "small". These rules can be learned from a large amount of experimental data based on expert experience or through machine learning algorithms. Finally, a vibration suppression control rule strategy is generated, which describes how to adjust the control parameters according to the vibration characteristic quantities in the form of fuzzy rules.

[0155] Step S34: Processing the vibration control performance requirement index of the parallel robot to generate vibration control performance requirement index data;

[0156] In an embodiment of the present invention, the required vibration control performance indicators, such as residual vibration amplitude, stabilization time, overshoot, etc., are determined according to the specific application scenario. These indicators are quantified and target values ​​or constraints are set. For example, it can be required that the residual vibration amplitude is less than a certain threshold and the stabilization time is less than a certain time limit. Finally, vibration control performance demand indicator data is generated, which includes target values ​​or constraints for each performance indicator.

[0157] Step S35: performing control parameter intelligent adjustment strategy processing on the vibration suppression control rule strategy and the vibration control performance requirement index data through the modal vibration characteristic parameters to obtain an intelligent vibration control parameter adjustment strategy.

[0158] In an embodiment of the present invention, an optimization algorithm, such as a genetic algorithm or a particle swarm algorithm, can be used to search for optimal control parameters so that the robot meets the predetermined vibration control performance requirement indicators. During the optimization process, the modal vibration characteristic parameters can be used to guide the adjustment of the control parameters. For example, the parameters of the controller can be adjusted according to the natural frequency and damping ratio of the robot. Finally, an intelligent vibration control parameter adjustment strategy is obtained, which can adaptively adjust the control parameters according to the modal characteristics and vibration characteristic quantities of the robot to meet the predetermined vibration control performance requirements.

[0159] Step S35 includes the following steps:

[0160] Step S351: extracting residual vibration control parameters based on the vibration control performance requirement index data to obtain multi-dimensional residual vibration control space parameters;

[0161] Step S352: Calculating the sensitivity of vibration control parameters to the multi-dimensional residual vibration control space parameters using the modal vibration characteristic parameters to generate residual vibration control parameter weight data;

[0162] Step S353: performing vibration control performance target processing according to the residual vibration control parameter weight data and the vibration control performance requirement index data, and constructing a vibration control parameter adjustment model using a particle swarm algorithm and a vibration suppression control rule strategy to generate a vibration control parameter adjustment model;

[0163] Step S354: initializing the fuzzy control parameter particle swarm of the vibration control parameter adjustment model to generate initialization control parameter particle swarm data;

[0164] Step S355: Calculate the first-order contribution effect index of the initialized control parameter particle swarm data, and evaluate the particle fitness to generate a control parameter fitness value;

[0165] Step S356: Based on the control parameter fitness value, the vibration control parameter adjustment model is used to update the particle position, and the control parameter is iteratively optimized to obtain an intelligent vibration control parameter adjustment strategy.

[0166] In an embodiment of the present invention, based on the vibration control performance demand index data generated in step S34, such as residual vibration amplitude, stabilization time, overshoot, etc., the residual vibration control parameters related thereto are extracted. These control parameters can be parameters in the controller, such as the gain coefficient of the PID controller, or other parameters that affect the residual vibration, such as the speed and acceleration of the motion trajectory. The extracted control parameters are combined into multi-dimensional residual vibration control space parameters, for example, the proportional gain, integral gain, differential gain and trajectory speed can be combined into a four-dimensional control space parameter. The sensitivity is calculated by numerical calculation or simulation method. For example, the value of a certain control parameter can be changed to observe the change of the vibration control performance index, so as to calculate the sensitivity of the control parameter. The calculated sensitivity value is converted into a weight value to generate residual vibration control parameter weight data. The larger the weight value, the greater the influence of the control parameter on the vibration control performance. The vibration control performance demand index can be converted into an optimization target, for example, minimizing the residual vibration amplitude, minimizing the stabilization time, etc. Then, a vibration control parameter adjustment model is constructed using the particle swarm algorithm and the vibration suppression control rule strategy generated in step S33. The model takes multidimensional residual vibration control space parameters as input, takes vibration control performance index as output, and uses particle swarm algorithm to search for the optimal control parameter combination so that the vibration control performance meets the preset requirements. According to the value range of the multidimensional residual vibration control space parameters, a set of initial control parameter combinations is randomly generated as the initial position of the particle swarm. Each particle represents a control parameter combination. For example, if the control space parameters are four-dimensional, each particle contains four values, corresponding to the proportional gain, integral gain, differential gain and trajectory speed respectively. The generated initial particle swarm data is saved as the initialization control parameter particle swarm data. The first-order contribution effect index is calculated for the initialization control parameter particle swarm data generated in step S354. The first-order contribution effect index can be used to evaluate the contribution of each control parameter to the vibration control performance. Then, the fitness of each particle is evaluated according to the calculated first-order contribution effect index and the vibration control performance demand index data generated in step S34. The fitness function can be defined as the weighted deviation between the performance index and the demand value, where the weight value is provided by the residual vibration control parameter weight data generated in step S352. Finally, a control parameter fitness value is generated to indicate the quality of each particle. The particle position is updated using the vibration control parameter adjustment model constructed in step S353. The particle swarm algorithm updates the position and velocity of each particle according to the fitness value to move it to a more optimal area. The particle position update and fitness evaluation are repeated until the preset termination condition is met, such as reaching the maximum number of iterations or finding a control parameter combination that meets the performance requirements.Finally, an intelligent vibration control parameter adjustment strategy is obtained, which contains the optimal control parameter combination and can adaptively adjust the control parameters according to the modal characteristics of the robot and the vibration control performance requirements.

[0167] Step S4 includes the following steps:

[0168] Step S41: using the terminal device to obtain the manipulator feedforward control parameters; wherein the manipulator feedforward control parameters include feedforward monitoring time domain data, feedforward control time domain data and control compensation data;

[0169] Step S42: When the parallel robot performs real-time high-speed operation, the sensor monitoring network is used to perform real-time vibration monitoring of the end effector based on the feedforward monitoring time domain data, the feedforward control time domain data, and the control compensation data to obtain real-time vibration monitoring state data;

[0170] Step S43: extracting terminal vibration features from the real-time vibration monitoring state data to generate real-time terminal vibration feature data;

[0171] Step S44: performing intelligent feedforward compensation control adjustment processing on the real-time terminal vibration characteristic data through an intelligent vibration control parameter adjustment strategy to generate feedforward compensation control adjustment parameters;

[0172] Step S45: setting a control parameter buffer for the parallel robot, and performing smooth switching of control parameters according to the feedforward compensation control adjustment parameters, so as to realize intelligent parameter adjustment control operation of the three-degree-of-freedom parallel robot.

[0173] In an embodiment of the present invention, a terminal device, such as a host computer or a controller, is used to obtain the feedforward control parameters of the manipulator. The feedforward control time domain data is derived from the high-speed motion task trajectory planning in S242, which describes the ideal motion trajectory and speed of the end effector. The control compensation data is calculated based on the robot dynamics model and the feedforward control data, and is used to compensate for the deficiencies of the feedforward control, such as deviations caused by model errors or unmodeled dynamics. The real-time vibration signal of the end effector is collected using a sensor monitoring network (accelerometers and laser displacement sensors deployed in S22). In order to improve the effectiveness of monitoring, appropriate sampling frequencies and data filtering parameters can be set based on the feedforward monitoring time domain data. For example, if the feedforward monitoring data shows that the residual vibration is mainly concentrated in a certain frequency band, a filter can be set for the frequency band to reduce noise interference. The collected sensor data is fused with the feedforward information. For example, the measured vibration signal can be compared with the feedforward monitoring time domain data to determine whether the vibration exceeds the expected range. The measured end effector posture can also be compared with the feedforward control time domain data, the posture error is calculated, and the correction is combined with the control compensation data. Convert the preprocessed time-domain vibration signal into a time-frequency domain signal. Similar to S244, short-time Fourier transform (STFT) or wavelet transform can be used. STFT can decompose the signal into spectra of different time periods to observe the change of vibration frequency over time. Wavelet transform can provide finer time-frequency resolution, especially suitable for analyzing non-stationary signals. Analyze the attenuation of vibration energy over time in the time-frequency spectrum and extract the vibration attenuation time. An energy threshold can be set. When the vibration energy decays below the threshold, the vibration attenuation is considered to be complete. Since real-time calculation is required, a simplified attenuation model can be used for estimation, such as an exponential decay model. Analyze the distribution of vibration energy in the frequency dimension in the time-frequency spectrum and extract the dominant frequency. The frequency corresponding to the frequency component with the largest energy can be selected as the dominant frequency, or the weighted average of the frequency components can be calculated. In order to improve real-time performance, only the frequency range near the residual vibration main frequency identified in S24 can be focused on. The extracted real-time vibration attenuation time, dominant frequency, energy decay rate and other characteristic parameters are combined into real-time terminal vibration characteristic data. The intelligent vibration control parameter adjustment strategy calculates the corresponding feedforward compensation control adjustment parameters based on the input characteristic data. For example, if the strategy is a fuzzy control rule base, it will be inferred according to the fuzzy rules to obtain the output parameters; the calculated feedforward compensation control adjustment parameters will be output. These parameters can be velocity feedforward gain, acceleration feedforward gain, damping coefficient, etc., which are used to adjust the robot's control instructions. For example, the intelligent vibration control parameter adjustment strategy rule is as follows: "If the vibration amplitude is large and the main frequency is close to the system's natural frequency, increase the velocity feedforward gain and damping coefficient."When the robot performs high-speed motion, if the real-time vibration characteristic data extracted by S43 shows that the vibration amplitude is large and the main frequency is close to the natural frequency of the system, S44 will increase the speed feedforward gain and damping coefficient according to the above rules, and output these adjusted parameters as feedforward compensation control adjustment parameters. A control parameter buffer is set in the robot controller. The buffer is used to store the control parameters currently in use and the control parameters to be updated output by S44. The size of the buffer can be set according to the dimension and update frequency of the control parameters. When S44 outputs new feedforward compensation control adjustment parameters, these parameters are not directly applied to the controller, but are written to the control parameter buffer. Then, a smooth switching algorithm is used to gradually apply the new parameters in the buffer to the controller. For example, suppose the speed feedforward gain of the robot needs to be adjusted. The current speed feedforward gain is 0.5, and the new gain value output by S44 is 0.8. The linear interpolation method is used to set the transition time to 0.1 seconds. Then, in the next 0.1 seconds, the speed feedforward gain will increase linearly at a rate of (0.8-0.5) / 0.1=3 until it reaches the new value of 0.8. By smoothly switching the control parameters, the smoothness of the robot's movement and the stability of the control system can be guaranteed, ultimately realizing the intelligent parameter adjustment and control operation of the three-degree-of-freedom parallel robot and improving the accuracy and efficiency of high-speed operations.

[0174] The present invention also provides a three-degree-of-freedom parallel robot control parameter adjustment system, which executes the three-degree-of-freedom parallel robot control parameter adjustment method as described above, and the three-degree-of-freedom parallel robot control parameter adjustment system comprises:

[0175] The robot dynamics modeling module is used to establish the manipulator dynamics model of the parallel robot and generate the manipulator dynamics model; perform terminal operation space analysis based on the manipulator dynamics model to generate terminal workspace range data; perform mechanism singular configuration identification on the manipulator dynamics model through the terminal workspace range data to obtain mechanism singular configuration point data;

[0176] The residual vibration extraction module is used to optimize the working posture space of the end workspace range data through the singular configuration point data of the mechanism to obtain the optimized working space data; deploy the end monitoring sensor network on the parallel robot to obtain the sensor monitoring network; use the sensor monitoring network to perform modal vibration test on the parallel robot to generate modal vibration characteristic parameters; use the sensor monitoring network to extract the high-speed motion residual vibration characteristics of the parallel robot to generate the residual vibration characteristic data of the motion task;

[0177] The intelligent vibration control module is used to process the residual vibration characteristic data of the motion task with an adaptive rule adjustment strategy to generate a vibration suppression control rule strategy; the vibration suppression control rule strategy is processed with an intelligent control parameter adjustment strategy through modal vibration characteristic parameters to obtain an intelligent vibration control parameter adjustment strategy;

[0178] The real-time vibration control execution module is used to use the sensor monitoring network to perform real-time vibration monitoring of the end effector of the parallel robot and generate real-time end vibration characteristic data; to perform intelligent feedforward compensation control adjustment processing on the real-time end vibration characteristic data through the intelligent vibration control parameter adjustment strategy and generate feedforward compensation control adjustment parameters; to smoothly switch the control parameters of the parallel robot through the feedforward compensation control adjustment parameters to realize the intelligent parameter adjustment control operation of the three-degree-of-freedom parallel robot.

[0179] The present application is that, by establishing a manipulator dynamics model for a parallel robot, the present invention can accurately analyze the end operation space and identify the singular configuration points of the mechanism, thereby optimizing the working posture space, avoiding the robot from operating near a potentially dangerous configuration, and improving the safety and reliability of the robot. By deploying a sensor monitoring network, the parallel robot is subjected to modal vibration testing and high-speed motion residual vibration feature extraction, and the natural frequency and vibration mode of the robot, as well as the unique vibration mode during high-speed motion, can be accurately identified. Through an adaptive rule adjustment strategy and an intelligent vibration control parameter adjustment strategy, dynamic adjustment of vibration suppression control parameters is achieved. Compared with traditional fixed parameter PID control, this dual adjustment mechanism can better cope with dynamic changes during high-speed motion and achieve faster and more accurate residual vibration suppression. By real-time monitoring of the vibration state of the end effector and performing feedforward compensation control adjustment, the present invention can predict and compensate for residual vibration before it occurs, minimizing the impact of residual vibration on positioning accuracy. Real-time vibration control execution can achieve smooth switching of control parameters, ensure the stability and continuity of the control process, and avoid secondary vibration caused by parameter mutations. This feedforward compensation control strategy improves the positioning accuracy and stability of the robot during high-speed operation, and ensures the reliability and safety of the robot's operation.

[0180] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0181] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A control parameter adjustment method for a three-degree-of-freedom parallel robot, characterized in that: The following steps are involved: Step S1: Establish a manipulator dynamics model for the parallel robot to generate a manipulator dynamics model; perform terminal operation space analysis based on the manipulator dynamics model to generate terminal workspace range data; perform mechanism singular configuration identification on the manipulator dynamics model through the terminal workspace range data to obtain mechanism singular configuration point data; Step S2: optimizing the working posture space of the end working space range data through the singular configuration point data of the mechanism to obtain the optimized working space data; deploying the end monitoring sensor network on the parallel robot to obtain the sensor monitoring network; performing modal vibration test on the parallel robot based on the optimized working space data using the sensor monitoring network to generate modal vibration characteristic parameters; performing high-speed motion residual vibration feature extraction on the parallel robot based on the optimized working space data using the sensor monitoring network to generate motion task residual vibration characteristic data; Step S3: performing adaptive rule adjustment strategy processing on the residual vibration characteristic data of the motion task to generate a vibration suppression control rule strategy; performing control parameter intelligent adjustment strategy processing on the vibration suppression control rule strategy through modal vibration characteristic parameters to obtain an intelligent vibration control parameter adjustment strategy; Step S4: Use the sensor monitoring network to perform real-time vibration monitoring of the end effector of the parallel robot to generate real-time end vibration characteristic data; use the intelligent vibration control parameter adjustment strategy to perform intelligent feedforward compensation control adjustment processing on the real-time end vibration characteristic data to generate feedforward compensation control adjustment parameters; use the feedforward compensation control adjustment parameters to smoothly switch the control parameters of the parallel robot to realize the intelligent parameter adjustment control operation of the three-degree-of-freedom parallel robot.

2. The control parameter adjustment method of a three-degree-of-freedom parallel robot according to claim 1, characterized in that: Step S1 includes the following steps: Step S11: measuring the structural parameters of the parallel robot to obtain the structural parameters of the parallel robot; Step S12: identifying the joint friction force according to the structural parameters of the parallel robot and generating joint friction force data; Step S13: establishing a manipulator dynamics model based on the parallel robot structural parameters and joint friction force data to generate a manipulator dynamics model; Step S14: Performing terminal working space analysis according to the manipulator dynamics model to generate terminal working space range data; Step S15: using the end workspace range data to identify the singular configuration of the mechanism of the manipulator dynamics model, and obtaining the singular configuration point data of the mechanism.

3. The control parameter adjustment method of a three-degree-of-freedom parallel robot according to claim 1, characterized in that: Step S2 includes the following steps: Step S21: optimizing the working posture space of the end working space range data by using the singular configuration point data of the mechanism to obtain optimized working space data; Step S22: marking the key positions of the driving joints of the parallel robot based on the manipulator dynamics model, and deploying the end effector monitoring sensor network to obtain a sensor monitoring network; Step S23: performing a modal vibration test on the parallel robot using a sensor monitoring network according to the optimized working space data to generate modal vibration characteristic parameters; Step S24: Plan the high-speed motion task trajectory for the parallel robot, and use the sensor monitoring network to extract residual vibration features to generate motion task residual vibration feature data.

4. The control parameter adjustment method of the three-degree-of-freedom parallel robot according to claim 3, characterized in that: Step S23 includes the following steps: Step S231: Selecting working test points according to the optimized working space data to generate spatial test point position data; Step S232: Predicting the modal vibration shape of the parallel robot through the spatial test point position data to generate main modal vibration shape data; Step S233: Based on the main modal vibration shape data and the spatial test point position data, a sensor monitoring network is used to perform a preliminary modal test on the parallel robot to generate preliminary modal vibration characteristic parameters; Step S234: using a preset multi-degree-of-freedom vibrator model to perform subspace control parameter identification on the preliminary modal vibration characteristic parameters, and perform vibration modal parameter association to generate modal vibration characteristic parameters.

5. The control parameter adjustment method of a three-degree-of-freedom parallel robot according to claim 4, characterized in that: Step S233 includes the following steps: Based on the spatial test point position data, the parallel robot is subjected to terminal pulse excitation test, and the vibration response signal is collected using the sensor monitoring network to obtain the pulse excitation response data; Perform a preliminary estimation of the attitude natural frequency based on the pulse excitation response data and generate multi-attitude natural frequency estimation data; The multi-frequency sinusoidal excitation signal is processed according to the multi-posture natural frequency estimation data, and the signal is superimposed using a preset pseudo-random excitation signal to obtain the multi-frequency sinusoidal excitation signal data; Based on the multi-frequency sinusoidal excitation signal data, the controller drives the joint motor of the parallel robot to perform sinusoidal frequency sweep motion, and uses the sensor monitoring network to perform sinusoidal frequency sweep vibration response to obtain sinusoidal frequency sweep excitation response data; The frequency response fitting of the sine sweep excitation response data is performed through the main modal vibration data, and the natural frequency damping ratio is calculated based on the multi-posture natural frequency estimation data to obtain the natural frequency damping ratio data; The multi-posture natural frequency estimation data, natural frequency damping ratio data and main modal vibration shape data are integrated into preliminary vibration modal parameters to obtain preliminary modal vibration characteristic parameters.

6. The control parameter adjustment method of a three-degree-of-freedom parallel robot according to claim 3, characterized in that: Step S24 includes the following steps: Step S241: Analyze the actual work tasks of the parallel robot, and process the typical high-speed motion mode to generate high-speed motion mode data; Step S242: planning a high-speed motion task trajectory for the high-speed motion mode data by optimizing the working space data, and setting motion parameters, thereby obtaining high-speed motion task sequence data; Step S243: Based on the high-speed motion task sequence data, the parallel robot is controlled to sequentially perform the high-speed motion tasks, and a sensor monitoring network is used to synchronously collect multi-channel terminal vibrations to generate residual vibration raw signal data; Step S244: performing time-frequency domain transformation processing on the residual vibration original signal data to generate residual vibration time-frequency spectrum data; Step S245: extracting the vibration signal decay time according to the residual vibration time-frequency spectrum data to generate vibration decay time data; Step S246: extracting the dominant frequency of the residual vibration time-frequency spectrum data to generate residual vibration dominant frequency data; Step S247: Calculating the vibration energy attenuation rate of the residual vibration time-frequency spectrum data to generate vibration energy attenuation rate data; Step S248: Perform residual vibration feature clustering processing based on the vibration attenuation time data, the residual vibration dominant frequency data, and the vibration energy attenuation rate data, and perform motion task mapping processing based on the high-speed motion task sequence data to generate motion task residual vibration feature data.

7. The control parameter adjustment method of a three-degree-of-freedom parallel robot according to claim 1, characterized in that: Step S3 includes the following steps: Step S31: dividing the residual vibration characteristic data of the motion task into vibration characteristic intervals to generate characteristic interval division data; Step S32: fuzzy variables are defined for the residual vibration characteristic data of the motion task through the characteristic quantity interval division data, and input fuzzy variable data and output fuzzy variable data are obtained respectively; wherein the input fuzzy variable data includes vibration attenuation time data, dominant frequency data and energy attenuation rate data; and the output fuzzy variable data includes velocity feedforward gain data and acceleration feedback gain data; Step S33: Performing input-output fuzzy control rule mapping on the vibration attenuation time data, the dominant frequency data, the energy attenuation rate data, the velocity feedforward gain data and the acceleration feedback gain data through a preset vibration mode fuzzy control rule library to generate a vibration suppression control rule strategy; Step S34: Processing the vibration control performance requirement index of the parallel robot to generate vibration control performance requirement index data; Step S35: performing control parameter intelligent adjustment strategy processing on the vibration suppression control rule strategy and the vibration control performance requirement index data through the modal vibration characteristic parameters to obtain an intelligent vibration control parameter adjustment strategy.

8. The control parameter adjustment method of the three-degree-of-freedom parallel robot according to claim 7, characterized in that: Step S35 includes the following steps: Step S351: extracting residual vibration control parameters based on the vibration control performance requirement index data to obtain multi-dimensional residual vibration control space parameters; Step S352: Calculating the sensitivity of vibration control parameters to the multi-dimensional residual vibration control space parameters using the modal vibration characteristic parameters to generate residual vibration control parameter weight data; Step S353: performing vibration control performance target processing according to the residual vibration control parameter weight data and the vibration control performance requirement index data, and constructing a vibration control parameter adjustment model using a particle swarm algorithm and a vibration suppression control rule strategy to generate a vibration control parameter adjustment model; Step S354: initializing the fuzzy control parameter particle swarm of the vibration control parameter adjustment model to generate initialization control parameter particle swarm data; Step S355: Calculate the first-order contribution effect index of the initialized control parameter particle swarm data, and evaluate the particle fitness to generate a control parameter fitness value; Step S356: Based on the control parameter fitness value, the vibration control parameter adjustment model is used to update the particle position, and the control parameter is iteratively optimized to obtain an intelligent vibration control parameter adjustment strategy.

9. The control parameter adjustment method of a three-degree-of-freedom parallel robot according to claim 1, characterized in that: Step S4 includes the following steps: Step S41: using the terminal device to obtain the manipulator feedforward control parameters; wherein the manipulator feedforward control parameters include feedforward monitoring time domain data, feedforward control time domain data and control compensation data; Step S42: When the parallel robot performs real-time high-speed operation, the sensor monitoring network is used to perform real-time vibration monitoring of the end effector based on the feedforward monitoring time domain data, the feedforward control time domain data, and the control compensation data to obtain real-time vibration monitoring state data; Step S43: extracting terminal vibration features from the real-time vibration monitoring state data to generate real-time terminal vibration feature data; Step S44: performing intelligent feedforward compensation control adjustment processing on the real-time terminal vibration characteristic data through an intelligent vibration control parameter adjustment strategy to generate feedforward compensation control adjustment parameters; Step S45: setting a control parameter buffer for the parallel robot, and performing smooth switching of control parameters according to the feedforward compensation control adjustment parameters, so as to realize intelligent parameter adjustment control operation of the three-degree-of-freedom parallel robot.

10. A three-degree-of-freedom parallel robot control parameter adjustment system, characterized in that: Used to execute the three-degree-of-freedom parallel robot control parameter adjustment method as claimed in claim 1, the three-degree-of-freedom parallel robot control parameter adjustment system comprises: The robot dynamics modeling module is used to establish the manipulator dynamics model of the parallel robot and generate the manipulator dynamics model; perform terminal operation space analysis based on the manipulator dynamics model to generate terminal workspace range data; perform mechanism singular configuration identification on the manipulator dynamics model through the terminal workspace range data to obtain mechanism singular configuration point data; The residual vibration extraction module is used to optimize the working posture space of the end workspace range data through the singular configuration point data of the mechanism to obtain the optimized working space data; deploy the end monitoring sensor network on the parallel robot to obtain the sensor monitoring network; use the sensor monitoring network to perform modal vibration test on the parallel robot to generate modal vibration characteristic parameters; use the sensor monitoring network to extract the high-speed motion residual vibration characteristics of the parallel robot to generate the residual vibration characteristic data of the motion task; The intelligent vibration control module is used to process the residual vibration characteristic data of the motion task with an adaptive rule adjustment strategy to generate a vibration suppression control rule strategy; the vibration suppression control rule strategy is processed with an intelligent control parameter adjustment strategy through modal vibration characteristic parameters to obtain an intelligent vibration control parameter adjustment strategy; The real-time vibration control execution module is used to use the sensor monitoring network to perform real-time vibration monitoring of the end effector of the parallel robot and generate real-time end vibration characteristic data; to perform intelligent feedforward compensation control adjustment processing on the real-time end vibration characteristic data through the intelligent vibration control parameter adjustment strategy and generate feedforward compensation control adjustment parameters; to smoothly switch the control parameters of the parallel robot through the feedforward compensation control adjustment parameters to realize the intelligent parameter adjustment control operation of the three-degree-of-freedom parallel robot.

Images