Grinding and polishing control method and control system based on sliding mode control

Abstract

The invention discloses a grinding and polishing control method and system based on sliding mode control, and relates to the field of robot industrial control. The method comprises the steps that a force signal of a force sensor at the tail end of a mechanical arm is collected and decomposed into a normal force component perpendicular to the surface of a workpiece; determining a deviation value between the normal force component and preset target pressure, and constructing a sliding mode surface expression of sliding mode control; determining a force control component according to the force control gain coefficient; position deviation and speed deviation are determined according to the position signals of the joint encoders and a preset polishing track; determining a position control component according to the position deviation, the speed deviation, the position gain coefficient and the speed gain coefficient; constructing a selection matrix, and performing orthogonal decomposition and superposition on the force control component and the position control component to obtain a comprehensive control instruction; and the driving signals are converted into driving signals of the joint motors, and the driving signals are transmitted to the mechanical arm for grinding and polishing. By implementing the method, force control of robot grinding and polishing can be optimized, and response delay is reduced.

Description

Technical Field

[0001] The present application relates to the field of robot industrial control, and in particular to a grinding and polishing control method and control system based on sliding mode control. Background Art

[0002] With the continuous improvement of industrial automation level, robots are increasingly used in the field of grinding and polishing. Robotic arm grinding and polishing technology can be used for surface treatment of hardware parts, engine covers, leather products, etc., which is of great significance to improving product quality and production efficiency. During the grinding and polishing process, the end of the robot arm needs to maintain a stable contact pressure with the workpiece surface and move accurately along the trajectory of the workpiece surface, which places high demands on the force and position control of the robot arm.

[0003] In related technologies, the mainstream control method for grinding and polishing with a robot arm adopts impedance control. This method establishes an impedance model between force and position and uses the three-loop control architecture of the robot arm to achieve a closed position loop. In specific implementation, by adjusting impedance parameters such as the damping coefficient and stiffness coefficient, the end of the robot arm generates appropriate pressure when it contacts the workpiece surface. At the same time, the motion trajectory is adjusted through the position loop feedback of the robot arm to achieve smooth motion control during grinding and polishing.

[0004] However, as an indirect force control method, impedance control has an obvious hysteresis characteristic in its force control response. When the end of the robot arm moves along the surface of the curved workpiece, the force response delay causes a large deviation between the actual pressure and the set pressure. This pressure fluctuation will affect the processing quality and surface uniformity of grinding and polishing. Summary of the invention

[0005] The present application provides a grinding and polishing control method and control system based on sliding mode control, which are used to optimize the force control of robot grinding and polishing and reduce response delay.

[0006] In the first aspect, the present application provides a grinding and polishing control method based on sliding mode control, which is applied to a control system, and the method includes: collecting the force signal of the force sensor at the end of the robotic arm, and decomposing the force signal into a normal force component perpendicular to the surface of the workpiece; determining the deviation value between the normal force component and the preset target pressure, and constructing a sliding surface expression of sliding mode control based on the deviation value; determining the force control component according to the sliding surface expression and the force control gain coefficient; determining the position deviation and speed deviation according to the collected position signals of each joint encoder and the preset grinding trajectory; determining the position control component according to the position deviation, speed deviation, position gain coefficient and speed gain coefficient; constructing a selection matrix based on the normal direction and tangential direction of the workpiece surface, and orthogonally decomposing and superimposing the force control component and the position control component based on the selection matrix to obtain a comprehensive control instruction; converting the comprehensive control instruction into a drive signal for each joint motor, and transmitting the drive signal to the robotic arm to perform grinding and polishing based on sliding mode control.

[0007] In the above embodiment, the control system implements force / position hybrid control based on sliding mode control, collects force signals and decomposes the normal force components, performs force control in combination with the sliding surface expression, and realizes precise trajectory tracking through position and velocity feedback; when grinding the surface of a flat workpiece, compared with impedance control, although the initial downward pressure overthrow force is slightly larger, the pressure stability is better, and the fluctuation range is only within ±1N; when grinding the surface of a curved workpiece, the pressure tracking accuracy is significantly better than that of impedance control, and the fluctuation range is controlled within ±2N, which effectively overcomes the problem of response lag of impedance control.

[0008] In combination with some embodiments of the first aspect, in some embodiments, the step of determining the deviation value between the normal force component and the preset target pressure, and constructing the sliding surface expression of the sliding mode control based on the deviation value, specifically includes: calculating the deviation value between the normal force component and the preset target pressure and the deviation integral term within a preset time window; weightedly combining the deviation value, the deviation integral term and the sliding mode coefficient to obtain the sliding surface expression.

[0009] In the above embodiment, the control system calculates the force deviation and the integral term and performs weighted combination to construct the sliding surface expression, thereby achieving rapid response adjustment of the grinding pressure, making full use of the characteristics of the sliding mode control, improving the control accuracy while ensuring the stability of the system, and making the pressure fluctuation during surface grinding significantly smaller than that of the impedance control method, effectively improving the grinding quality.

[0010] In combination with some embodiments of the first aspect, in some embodiments, before the step of weighted combination of the deviation value, the deviation integral term and the sliding mode coefficient to obtain the sliding surface expression, the method also includes: constructing a grinding dynamic model including the surface curvature of the workpiece and the posture of the grinding tool; determining the Lyapunov function based on the grinding dynamic model; solving the derivative expression of the Lyapunov function, determining the range of the adaptive law parameters, and determining the sliding mode coefficient based on the range of the adaptive law parameters.

[0011] In the above embodiment, the control system determines the optimal sliding mode coefficient based on the grinding dynamic model and Lyapunov function, realizes the adaptive optimization of system parameters, ensures the convergence and robustness of the control system through theoretical analysis, shows good stability in practical applications, and can maintain good pressure tracking effect even when grinding curved workpieces.

[0012] In combination with some embodiments of the first aspect, in some embodiments, after converting the comprehensive control instructions into drive signals of each joint motor and transmitting the drive signals to the robotic arm to perform grinding and polishing based on sliding mode control, the method also includes: collecting surface quality data of the workpiece surface during the grinding process; determining the key grinding area of ​​the workpiece surface based on the surface quality data and a preset grinding judgment threshold; determining the posture angle, local residence time and grinding speed of the grinding tool based on the key grinding area.

[0013] In the above embodiment, the control system monitors the surface quality in real time and dynamically adjusts the grinding parameters, realizing intelligent optimization of the processing process, adaptively adjusting the tool posture, dwell time and grinding speed according to the actual processing effect, ensuring the consistency of processing quality in different areas and improving grinding efficiency.

[0014] In combination with some embodiments of the first aspect, in some embodiments, the steps of determining the posture angle, local residence time and grinding speed of the grinding tool according to the key grinding area specifically include: constructing a dynamic model of the grinding process including the material removal rate according to the key grinding area; calculating the local residence time of each local interval based on the dynamic model; calculating the trajectory curvature of each local interval in combination with the surface features of the workpiece, and determining the posture angle and speed constraints of the grinding tool according to the trajectory curvature; solving the speed planning scheme that meets the speed constraints to obtain the grinding speed.

[0015] In the above embodiment, the control system establishes a material removal rate model, optimizes grinding parameters in combination with the workpiece surface features, achieves precise speed and posture planning, fully considers the workpiece geometric characteristics, and ensures the uniformity and stability of processing quality.

[0016] In combination with some embodiments of the first aspect, in some embodiments, before the step of collecting the force signal of the force sensor at the end of the robotic arm and decomposing the force signal into a normal force component perpendicular to the surface of the workpiece, the method also includes: controlling the robotic arm to contact the surface of the workpiece at a preset speed for a preset movement, and collecting the force-controlled displacement characteristic curve during the contact process; determining the stiffness coefficient of the workpiece surface based on the force-controlled displacement characteristic curve; determining the material damping coefficient of the workpiece surface based on the contact transient response characteristics, and generating a material property model based on the material damping coefficient and the stiffness coefficient; and determining the force control gain coefficient and the preset target pressure based on the material property model.

[0017] In the above embodiment, the control system pre-identifies the surface characteristics of the workpiece and establishes a material model, thereby achieving optimal configuration of control parameters, and determining appropriate control parameters through force-controlled displacement characteristic analysis, thereby improving the system's adaptability to workpieces of different materials.

[0018] In combination with some embodiments of the first aspect, in some embodiments, the step of determining the force control gain coefficient and the preset target pressure based on the material property model specifically includes: constructing a grinding database including material properties and grinding parameters based on the material property model; collecting real-time operation data including force data and position data during the grinding process, and determining the grinding quality index based on the real-time operation data; when the grinding quality index is lower than the preset qualified index, extracting the optimal grinding parameters for similar workpiece materials from the grinding database; and determining the force control gain coefficient and the preset target pressure based on the optimal grinding parameters.

[0019] In the above embodiment, the control system constructs a polishing database and optimizes control parameters in real time, thereby realizing dynamic optimization of the machining process and ensuring timely optimization of the control strategy based on historical data and real-time feedback.

[0020] In a second aspect, an embodiment of the present application provides a control system, comprising: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the control system to execute the method described in the first aspect and any possible implementation method of the first aspect.

[0021] In a third aspect, an embodiment of the present application provides a computer program product comprising instructions, which, when the computer program product is run on a control system, enables the control system to execute the method described in the first aspect and any possible implementation of the first aspect.

[0022] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, comprising instructions, which, when executed on a control system, enable the control system to execute the method described in the first aspect and any possible implementation of the first aspect.

[0023] It is understandable that the control system provided in the second aspect, the computer program product provided in the third aspect, and the computer storage medium provided in the fourth aspect are all used to execute the method provided in the embodiment of the present application. Therefore, the beneficial effects that can be achieved can refer to the beneficial effects in the corresponding method, which will not be repeated here.

[0024] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: 1. Due to the adoption of a force / position hybrid control strategy based on sliding mode control, combined with technical means such as force sensor signal decomposition, sliding surface construction and orthogonal decomposition, it is possible to achieve effective coordination between force control and position control, effectively solving the problems of impedance control response lag and large pressure fluctuations in the prior art, thereby achieving high-precision and high-stability grinding and polishing process control.

[0025] 2. Due to the adoption of an adaptive control solution based on real-time monitoring of surface quality, by collecting surface quality data, identifying key grinding areas, and optimizing parameters such as tool posture, dwell time and grinding speed accordingly, the grinding strategy can be dynamically adjusted according to the actual processing effect, effectively solving the problems of poor grinding uniformity and low efficiency in the existing technology, and thus realizing differentiated grinding control.

[0026] 3. Due to the adoption of a parameter optimization scheme based on material property identification, by conducting force-controlled displacement characteristic tests in advance, determining the stiffness coefficient and damping coefficient, and establishing a material property model, the force control gain and target pressure can be accurately set, effectively solving the problem of strong experience dependence and poor adaptability of parameter setting in the existing technology, and thus realizing adaptive control of different workpiece materials. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 (a) is a control block diagram based on impedance control during the grinding and polishing process of the robot arm in an embodiment of the present application; Figure 1 (b) is a control block diagram of the traditional force / position hybrid control during the grinding and polishing process of the robot arm in the embodiment of the present application; Figure 1 (c) is a control block diagram of the improved force / position hybrid control based on sliding mode control in the grinding and polishing process of the robot arm in the embodiment of the present application; Figure 2 (a) is a schematic diagram of a scene in which a robotic arm grinds and polishes a flat workpiece surface in an embodiment of the present application; Figure 2 (b) is a schematic diagram of a scene in which a robotic arm grinds and polishes a curved workpiece in an embodiment of the present application; Figure 3 (a) is a schematic diagram of the control curves of position and force in various directions of grinding and polishing of a robot arm based on impedance-controlled force / position hybrid control on a flat workpiece surface in an embodiment of the present application; Figure 3 (b) is a schematic diagram of the control curves of position and force in various directions of grinding and polishing of a robot arm based on force / position hybrid control on a flat workpiece surface in an embodiment of the present application; Figure 3 (c) is a schematic diagram of control curves of position and force in various directions of grinding and polishing of a robot arm based on force / position hybrid control of sliding mode control on a flat workpiece surface in an embodiment of the present application; Figure 4 (a) is a schematic diagram of the control curves of the position and force in various directions of the robot arm grinding and polishing based on impedance control on the surface of the curved workpiece in an embodiment of the present application; Figure 4 (b) is a schematic diagram of the control curves of position and force in various directions of grinding and polishing of a robotic arm based on force / position hybrid control on a curved workpiece surface in an embodiment of the present application; Figure 4 (c) is a schematic diagram of control curves of position and force in various directions of grinding and polishing of a robotic arm using force / position hybrid control based on sliding mode control on a curved workpiece surface in an embodiment of the present application; Figure 5 It is a flow chart of a grinding and polishing control method based on sliding mode control in an embodiment of the present application; Figure 6 is another flow chart of the grinding and polishing control method based on sliding mode control in an embodiment of the present application; Figure 7 It is a schematic diagram of the structure of a physical device of the control system in an embodiment of the present application. DETAILED DESCRIPTION

[0028] The terms used in the following embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to be used as limitations to the present application. As used in the specification of the present application, the singular expressions "one", "a kind of", "above", "the" and "this" are intended to also include plural expressions, unless there is a clear indication to the contrary in the context. It should also be understood that the term "and / or" used in the present application refers to any or all possible combinations comprising one or more of the listed items.

[0029] In the following, the terms "first" and "second" are used for descriptive purposes only and are not to be understood as suggesting or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features, and in the description of the embodiments of the present application, unless otherwise specified, "plurality" means two or more.

[0030] To facilitate understanding, multiple control methods for the grinding and polishing process of the robotic arm in the scenarios corresponding to the embodiments of the present application are introduced below.

[0031] See also Figure 1 (a), Figure 1 (a) is a control block diagram based on impedance control during the grinding and polishing process of the robot arm in an embodiment of the present application; the control block diagram shows the implementation structure of the traditional impedance control method. Figure 1 The input end in (a) includes the desired position xd and the desired force Fd, which are controlled by calculating the deviation from the actual position x and the actual force F. IK in the block diagram represents the inverse kinematics module, which is used to convert the position command in Cartesian space into joint space. The transfer function in the middle (z^2-1.9801z+0.9801)^-1 represents the impedance characteristic of the design, which is used to establish the dynamic relationship between position and force. Two PID controllers are used for closed-loop control of the position loop and the force loop respectively, and their outputs are superimposed and act on the robot. The system dynamics equation M(q)q̈+C(q,q̇)q̇+G(q)=τ describes the kinematic characteristics of the robot, where M(q) is the inertia matrix, C(q,q̇) is the Coriolis force and centrifugal force term, and G(q) is the gravity term. This control scheme realizes indirect coupling control of force and position by adjusting the impedance parameters, but there is a problem of lag in force control response.

[0032] See also Figure 1 (b), Figure 1 (b) is a control block diagram of traditional force / position hybrid control during the grinding and polishing process of the robot arm in an embodiment of the present application; the control block diagram shows the implementation scheme of direct force / position hybrid control. Figure 1 In (b), J represents the Jacobian matrix, and FK (q) represents forward kinematics, which is used for position feedback calculation. The position control channel contains proportional gain Kpp (5.0) and differential gain Kpd (0.1) to realize PD control. The force control channel contains force feedback gain Kfp (0.001) and integral link Kfi / s (0.2), where ±200 represents the saturation limit of force control. The outputs of the two control channels are superimposed and act on the robot arm. The system allocates the tasks of force control and position control through the S matrix, realizing decoupled control in different directions. Compared with impedance control, this scheme has faster force response characteristics, but large force overshoot may occur during dynamic processes.

[0033] See also Figure 1 (c), Figure 1 (c) A control block diagram of the improved force / position hybrid control based on sliding mode control during the grinding and polishing process of the robot arm in an embodiment of the present application; the control block diagram shows the improved force / position hybrid control scheme based on sliding mode control. Figure 1 A dynamic compensation module CDNN (q, q̇) q̇ + GSNN (q) based on RBF neural network (Radial Basis Function Neural Network) is added to the top of the block diagram of (c) to estimate and compensate for the nonlinear dynamic characteristics of the system. The controller introduces the sliding mode control term sgn (x) and the switching gain λ to enhance the robustness of the system. The position control loop maintains the PD control structure (Kpp + Kpi / s) similar to the traditional scheme, but adds an adaptive adjustment mechanism. In addition to the basic feedback control, the force control loop also introduces a nonlinear control term based on sliding mode. The orthogonal decomposition of force / position control is achieved by selecting the matrix S, while considering the dynamic characteristics M (q) of the system. This scheme improves the system's ability to suppress external disturbances through sliding mode control, improves dynamic performance through neural network compensation, and comprehensively improves the force control accuracy and stability during the grinding process. The various gain parameters of the controller (such as 5.0, 0.1, 0.001, 0.2, etc.) are optimized to ensure good dynamic performance while ensuring system stability.

[0034] Next, we will introduce the polishing scenarios and experimental data corresponding to the embodiments of this application. Figure 2 ,in Figure 2 (a) is a schematic diagram of a scene in which a robotic arm grinds and polishes a flat workpiece surface in an embodiment of the present application; Figure 2 (b) is a schematic diagram of a scene in which a robotic arm grinds and polishes a curved workpiece in an embodiment of the present application. Figure 2 (a) and Figure 2 (b) shows the control software simulation diagrams of the robot arm facing the straight workpiece surface and the curved workpiece surface during the grinding and polishing process. Figure 3 and Figure 4 The control curves of the robot arm applying three control algorithms on the surface of a flat workpiece and a curved workpiece are shown in the figure, which can intuitively display the relevant control fluctuations and responses.

[0035] See also Figure 3 ,in Figure 3 (a) is a schematic diagram of the control curves of position and force in various directions of grinding and polishing of a robot arm based on impedance-controlled force / position hybrid control on a flat workpiece surface in an embodiment of the present application; Figure 3(b) is a schematic diagram of the control curves of position and force in various directions of the robot arm grinding and polishing based on force / position hybrid control on the flat workpiece surface in an embodiment of the present application; Figure 3 (c) is a schematic diagram of the control curves of position and force in various directions during grinding and polishing of a robotic arm using force / position hybrid control based on sliding mode control on a flat workpiece surface in an embodiment of the present application.

[0036] correspond Figure 3 In the simulation, the compressive stress is set to 20N. The simulation experiment shows that the effects of the three algorithms applied to grinding on the flat workpiece surface seem to be similar. The force fluctuations during the grinding process are also small. The subtle differences are reflected in the pressure of the robot arm grinding head. The impedance control is relatively stable during the initial downward pressure, and the overshoot force is only about 18N. The overshoot force of the classic force / position hybrid control reaches 230N, and the overshoot force of the force / position hybrid control based on SMC (sliding mode control) also reaches 180N. However, since the impedance control is an indirect force control, its response speed is slow and the tracking accuracy is naturally unsatisfactory. On the flat workpiece surface, its pressure is not exactly 20N, but gradually tends to 20N from a slightly lower 3N. In contrast, the upper and lower deviations of the classic force / position hybrid control are only within ±0.05N, and the force / position hybrid control based on SMC is only within ±1N.

[0037] See also Figure 4 ,in Figure 4 (a) is a schematic diagram of the control curves of the position and force in various directions of the robot arm grinding and polishing based on impedance control on the surface of the curved workpiece in an embodiment of the present application; Figure 4 (b) is a schematic diagram of the control curves of position and force in various directions of the robot arm grinding and polishing on the surface of the curved workpiece based on the force / position hybrid control in the embodiment of the present application; Figure 4 (c) is a schematic diagram of the control curves of position and force in various directions during grinding and polishing of a robotic arm using force / position hybrid control based on sliding mode control on a curved workpiece surface in an embodiment of the present application.

[0038] For grinding and polishing work on curved workpieces, the comparison between the effect of force / position hybrid control algorithm and impedance control algorithm immediately shows obvious superiority.

[0039] correspond Figure 4 As shown in (a), when the curved workpiece slides, the pressure of the robot arm is difficult to constantly track the set force. During the downward movement, due to tracking delay, its force is less than the set force by 7N~10N. When it moves up from the bottom of the valley, its force is greater than the set force by 11N.

[0040] In comparison, reference Figure 4 (b) and Figure 4 (c) Force / position hybrid control is a direct force control with extremely high real-time response speed, so the tracking accuracy is also guaranteed. The tracking force fluctuation of the classic force / position hybrid control is within ±0.5N, while the tracking force fluctuation of the force / position hybrid control based on SMC is only within ±2N, which is slightly larger than the classic force / position hybrid control algorithm. This is caused by the chattering characteristics of the sliding mode control itself.

[0041] In summary, during the grinding of a flat workpiece surface, the impedance-controlled downward pressure overshoot force is the smallest, followed by the SMC-based force / position hybrid control, while the classic force / position hybrid control has the largest downward pressure overshoot force. In terms of pressure smoothness, the impedance control is slightly inferior to the traditional force / position hybrid control and the SMC-based force / position hybrid control. During the grinding of a curved workpiece surface, the impedance-controlled downward pressure overshoot force is the smallest, followed by the SMC-based force / position hybrid control, while the classic force / position hybrid control has the largest downward pressure overshoot force. In terms of pressure smoothness, the impedance control is far inferior to the traditional force / position hybrid control and the SMC-based force / position hybrid control.

[0042] Therefore, the grinding and polishing control method based on sliding mode control in the present application, when grinding the surface of a flat workpiece, compared with impedance control, although the initial downward pressure overthrust force is slightly larger, the pressure stability is better, and the fluctuation range is only within ±1N; when grinding the surface of a curved workpiece, the pressure tracking accuracy is significantly better than that of impedance control, and the fluctuation range is controlled within ±2N, effectively overcoming the problem of impedance control response lag.

[0043] For ease of understanding, the following describes the process of the method provided by this implementation in combination with the above scenarios and related experimental data. Figure 5 , which is a flow chart of a grinding and polishing control method based on sliding mode control in an embodiment of the present application.

[0044] S501, collecting force signals from the force sensor at the end of the robot arm, and decomposing the force signals into normal force components perpendicular to the surface of the workpiece.

[0045] Among them, the force sensor at the end of the robot arm refers to a multi-dimensional force / torque sensor installed between the end effector of the robot arm and the tool, which is used to measure the contact force between the tool and the workpiece in real time; the force signal represents the three-dimensional force and three-dimensional torque data including size and direction collected by the sensor; the normal force component refers to the force component perpendicular to the surface of the workpiece, which is used to indicate the actual grinding pressure.

[0046] When the robot arm is performing grinding and polishing operations, the control system needs to monitor and control the grinding pressure in real time. Specifically, the control system first collects force signal data at a preset sampling frequency (such as 1kHz) through the end force sensor, and then decomposes the collected three-dimensional force signal into normal and tangential components based on the workpiece surface normal vector, where the normal component represents the actual grinding pressure for subsequent force control.

[0047] In some embodiments, the collection and decomposition of force signals can be achieved in a variety of ways: Optionally, the force signals are collected and decomposed by the following steps: configuring the sampling parameters of the force sensor; collecting the original force signal data; performing signal filtering processing; obtaining the surface normal vector according to the workpiece CAD model; decomposing the force signal projection to obtain the normal force component. Optionally, the force signals are collected and decomposed by the following steps: real-time detection of the force sensor state; collecting and calibrating the original data; determining the workpiece surface normal through contact detection; calculating the projection of the force signal on the normal direction. It is understandable that other signal collection and processing methods can also be used to achieve the collection and decomposition of force signals, which are not limited here.

[0048] S502: Determine a deviation value between the normal force component and a preset target pressure, and construct a sliding surface expression of sliding mode control based on the deviation value.

[0049] Among them, the preset target pressure represents the expected grinding pressure value, which is used to ensure the processing quality; the deviation value refers to the difference between the actual pressure and the target pressure; the sliding surface expression refers to the mathematical expression that describes the dynamic characteristics of the system state deviation, which is used to construct the sliding mode control law.

[0050] The control system needs to build a controller based on force deviation to achieve pressure tracking control. Specifically, the control system first calculates the deviation between the actual normal force and the target pressure, and at the same time calculates the integral term of the deviation within the preset time window, and then weightedly combines the deviation value, the integral term and the preset sliding mode coefficient to construct a sliding mode surface expression that reflects the dynamic characteristics of the system.

[0051] In some embodiments, the construction of the sliding surface expression can be achieved in a variety of ways: Optionally, it can be constructed by the following steps: calculating the force deviation and the rate of change of the deviation; designing the integral sliding surface form; determining the sliding mode coefficient; combining to obtain the sliding surface expression. Optionally, it can be constructed by the following steps: determining the sliding surface structure based on Lyapunov stability analysis; introducing an integral term to suppress steady-state errors; optimizing the sliding mode coefficient to suppress chattering. It is understandable that other control theory methods can also be used to construct the sliding surface expression, which is not limited here.

[0052] S503: Determine the force control component according to the sliding surface expression and the force control gain coefficient.

[0053] The force control component represents the control output for adjusting the grinding pressure; the force control gain coefficient refers to the parameter used to adjust the response characteristics of the controller.

[0054] The control system needs to generate a force control law based on the sliding mode control theory. Specifically, the control system uses the constructed sliding surface expression, combined with the preset force control gain coefficient, to design an equivalent control law and a switching control law that satisfy Lyapunov stability, and combines the two to obtain the force control component.

[0055] In some embodiments, the calculation of the force control component can be implemented in a variety of ways: Optionally, the calculation is performed by the following steps: designing an equivalent control term; designing a switching control term; combining to obtain a continuous control law; introducing a boundary layer to eliminate chattering. Optionally, the calculation is performed by the following steps: adjusting the gain coefficient based on an adaptive law; designing an exponential approach law; constructing a continuous control law. It is understandable that other sliding mode control methods can also be used to determine the force control component, which is not limited here.

[0056] S504, determining the position deviation and the speed deviation according to the collected position signals of the encoders of each joint and the preset grinding trajectory.

[0057] Among them, the joint encoder signal represents the actual angular position of each joint of the robot arm; the preset grinding trajectory refers to the expected tool motion trajectory; the position deviation represents the difference between the actual position and the expected position; and the speed deviation represents the difference between the actual speed and the expected speed.

[0058] The control system needs to achieve precise trajectory tracking control. Specifically, the control system first converts the joint angles into the Cartesian space position of the end effector through forward kinematics, then compares it with the preset trajectory to obtain the position deviation, and at the same time, derives the position signal and compares it with the expected speed to obtain the speed deviation.

[0059] In some embodiments, the calculation of position and velocity deviation can be achieved in a variety of ways: Optionally, the calculation is performed through the following steps: collecting encoder data; performing forward kinematic transformation; calculating position deviation; numerically differentiating to obtain actual velocity; calculating velocity deviation. Optionally, the calculation is performed through the following steps: estimating velocity based on an observer; calculating error by coordinate transformation; filtering to eliminate noise effects. It is understood that other kinematic analysis methods can also be used to calculate motion deviations, which are not limited here.

[0060] S505. Determine the position control component according to the position deviation, the speed deviation, the position gain coefficient and the speed gain coefficient.

[0061] Among them, the position control component represents the control output for trajectory tracking; the position gain coefficient refers to the parameter for adjusting the position response characteristics; the speed gain coefficient refers to the parameter for adjusting the speed response characteristics; the position deviation and speed deviation reflect the degree of deviation between the actual motion state and the expected trajectory.

[0062] The control system needs to achieve accurate position tracking control. Specifically, the control system adopts a PD control structure, multiplies the position deviation by the position gain coefficient to obtain a proportional term, multiplies the speed deviation by the speed gain coefficient to obtain a differential term, and superimposes the two terms to obtain a position control component to achieve precise control of the end position of the robot arm.

[0063] In some embodiments, the calculation of the position control component can be implemented in a variety of ways: Optionally, the calculation is performed through the following steps: adaptively adjusting the gain coefficient; calculating the proportional control term; calculating the differential control term; combining to obtain the control amount; and performing saturation limiting. Optionally, the calculation is performed through the following steps: using fuzzy rules to adjust the gain; introducing nonlinear control terms; and designing feedforward compensation control. It is understood that other control algorithms can also be used to calculate the position control component, which is not limited here.

[0064] S506, constructing a selection matrix based on the normal direction and the tangential direction of the workpiece surface, and performing orthogonal decomposition and superposition on the force control component and the position control component based on the selection matrix to obtain a comprehensive control instruction.

[0065] Among them, the selection matrix represents the matrix that distributes control weights in the normal and tangential directions; the normal direction refers to the direction perpendicular to the workpiece surface; the tangential direction refers to the direction parallel to the workpiece surface; the comprehensive control instruction refers to the final control output combining force control and position control.

[0066] The control system needs to coordinate force control and position control. Specifically, the control system first constructs a selection matrix based on the surface characteristics of the workpiece, mainly using force control in the normal direction and position control in the tangential direction, then orthogonally decomposes the force control component and the position control component through the selection matrix, and finally superimposes the decomposed control quantities to obtain a comprehensive control instruction.

[0067] In some embodiments, the coordination of control components can be achieved in a variety of ways: Optionally, it can be achieved by the following steps: constructing an orthogonal projection matrix; decomposing force control components; decomposing position control components; and obtaining control instructions by weighted superposition. Optionally, it can be achieved by the following steps: adaptively adjusting weight coefficients; constructing a hybrid control law; and optimizing the control allocation strategy. It is understandable that other control allocation methods can also be used to achieve force-position hybrid control, which is not limited here.

[0068] S507, converting the comprehensive control instruction into a driving signal of each joint motor, and transmitting the driving signal to the robot arm to perform grinding and polishing based on sliding mode control.

[0069] Among them, the drive signal refers to the voltage or current signal that controls the movement of each joint motor; the joint motor represents the actuator that drives the movement of each joint of the robotic arm; and the comprehensive control instruction represents the desired motion and force in Cartesian space.

[0070] The control system needs to convert the control instructions into actual drive signals. Specifically, the control system first converts the comprehensive control instructions in the Cartesian space into motion instructions in the joint space through inverse kinematics, then converts the motion instructions into corresponding drive signals according to the characteristics of each joint motor, and finally transmits the drive signals to the robot controller through the communication interface to perform motion control.

[0071] In some embodiments, the execution of control instructions can be implemented in a variety of ways: Optionally, it can be implemented by the following steps: calculating inverse kinematics solutions; generating joint trajectories; calculating drive signals; executing motion control; monitoring execution status. Optionally, it can be implemented by the following steps: online planning of optimal trajectories; dynamic compensation control; real-time adjustment of control parameters. It is understandable that other motion control methods can also be used to implement the execution of control instructions, which are not limited here.

[0072] The following is a more detailed description of the process of the method provided by this implementation. Figure 6 , is another flow chart of the grinding and polishing control method based on sliding mode control in an embodiment of the present application.

[0073] S601, controlling the robot arm to contact the workpiece surface at a preset speed and perform a preset movement, and collecting a force-controlled displacement characteristic curve during the contact process.

[0074] Among them, the preset speed refers to the calibrated speed when the robot arm contacts the surface of the workpiece, which is used to ensure the stability of the contact process; the preset movement refers to the standard test movement of the robot arm along the surface of the workpiece; the force-controlled displacement characteristic curve refers to the curve that records the relationship between force and displacement during the contact process, which is used to characterize the material properties of the workpiece; the contact process refers to the entire process from the tool never contacting to fully contacting the workpiece surface.

[0075] The control system needs to identify the material characteristics of the workpiece before starting the grinding operation. Specifically, the control system first controls the robot arm to approach the workpiece surface at a low speed of 0.1mm / s, and after detecting contact, it performs a preset detection movement along the workpiece surface. During this process, the force sensor data and position encoder data are synchronously collected at a sampling frequency of 1kHz to generate a force-displacement curve describing the material characteristics.

[0076] In some embodiments, the characteristic curve can be collected in a variety of ways: Optionally, the characteristic curve can be collected by the following steps: setting a contact detection threshold; controlling low-speed approach motion; detecting the initial contact point; executing a preset detection trajectory; collecting force displacement data; filtering and processing data. Optionally, the characteristic curve can be collected by the following steps: multi-point grid detection motion; online sensor calibration; adaptive adjustment of sampling parameters; and construction of characteristic curves. It is understandable that other detection methods can also be used to collect characteristic curves, which are not limited here.

[0077] S602. Determine the stiffness coefficient of the workpiece surface according to the force-controlled displacement characteristic curve.

[0078] Among them, the stiffness coefficient refers to the parameter that characterizes the ability of the workpiece surface to resist deformation; the force-controlled displacement characteristic curve represents the deformation characteristics of the material under load; and the workpiece surface refers to the surface material of the workpiece to be processed.

[0079] The control system needs to determine the material stiffness characteristics based on the test data. Specifically, the control system performs piecewise linear fitting on the collected force-displacement curve, calculates the slope of the curve to obtain the local stiffness value, and then takes the average value within the working range as the equivalent stiffness coefficient of the workpiece surface.

[0080] In some embodiments, the stiffness coefficient can be determined in a variety of ways: Optionally, the stiffness coefficient can be determined by the following steps: data preprocessing and denoising; piecewise linear fitting; calculating local stiffness; and determining equivalent stiffness by statistical analysis. Optionally, the stiffness coefficient can be determined by the following steps: establishing a nonlinear stiffness model; fitting parameters by the least squares method; and verifying the model accuracy. It is understood that other parameter identification methods can also be used to determine the stiffness coefficient, which is not limited here.

[0081] S603, determining the material damping coefficient of the workpiece surface based on the contact transient response characteristics, and generating a material characteristic model according to the material damping coefficient and stiffness coefficient.

[0082] Among them, the contact transient response characteristics refer to the dynamic response process of the material under the action of impact load; the material damping coefficient represents the ability of the material to dissipate mechanical energy; the material property model refers to the mathematical model that describes the mechanical properties of the material, which is used to predict the dynamic response characteristics of the material during the processing process; the stiffness coefficient represents the ability of the material to resist deformation.

[0083] The control system needs to establish a complete material dynamics model. Specifically, the control system analyzes the oscillation attenuation characteristics during the contact process, uses the logarithmic attenuation method to calculate the damping ratio, and combines the known stiffness coefficient to establish a second-order dynamic model including mass, stiffness and damping characteristics as a material characteristic model.

[0084] It should be noted that the material property model is based on the force-displacement characteristic curve data and contact transient response data at a preset speed during the training process. The stiffness coefficient is obtained through piecewise linear fitting of the force-displacement curve, and the damping coefficient is extracted from the oscillation response using the logarithmic decay method. The second-order dynamic model is constructed in combination with the mass parameters. The training standard is that the fitting error between the model prediction value and the measured force-displacement curve is minimized. The material property model contains the second-order dynamic equation of mass-stiffness-damping, which describes the dynamic response characteristics of the workpiece material under load. By inputting contact force and displacement, the material property model can output the dynamic response prediction of the material, which is used to determine the force control gain coefficient and the target pressure value.

[0085] In some embodiments, the material property model can be constructed in a variety of ways: Optionally, the material property model can be constructed by the following steps: collecting impact response data; calculating the oscillation period; determining the attenuation coefficient; identifying system parameters; and verifying the model accuracy. Optionally, the material property model can be constructed by the following steps: identifying parameters using the frequency domain analysis method; establishing a nonlinear dynamic model; optimizing model parameters; and evaluating model performance. It is understood that other modeling methods can also be used to construct the material property model, which is not limited here.

[0086] S604: Determine a force control gain coefficient and a preset target pressure based on the material property model.

[0087] Among them, the force control gain coefficient refers to the parameter used to adjust the response characteristics of the force controller; the preset target pressure represents the expected grinding pressure value; and the material property model includes the dynamic response characteristics of the workpiece surface.

[0088] The control system needs to optimize the control parameters according to the material characteristics. Specifically, the control system performs closed-loop stability analysis based on the material characteristic model, calculates the force control gain range that ensures system stability through the pole configuration method, and determines the target pressure value that can ensure processing quality without damaging the workpiece based on the hardness and strength characteristics of the material.

[0089] In some embodiments, the control parameters can be determined in a variety of ways: Optionally, the control parameters can be determined by the following steps: stability analysis; calculation of critical gain; optimization of control parameters; simulation verification of system performance; determination of safe pressure range. Optionally, the control parameters can be determined by the following steps: establishment of a fuzzy rule base; adaptive gain adjustment; online optimization of target pressure; evaluation of control effect. It is understandable that other parameter optimization methods can also be used to determine the control parameters, which are not limited here.

[0090] In some embodiments, the control system constructs a grinding database including material properties and grinding parameters based on the material property model; collects real-time operation data including force data and position data during the grinding process, and determines the grinding quality index based on the real-time operation data; when the grinding quality index is lower than the preset qualified index, extracts the optimal grinding parameters for similar workpiece materials from the grinding database; based on the optimal grinding parameters, determines the force control gain coefficient and the preset target pressure.

[0091] Among them, the material property model refers to the mathematical model that describes the mechanical properties of the workpiece material; the grinding database refers to the data set that stores the material properties and the corresponding grinding parameters; the real-time operation data refers to the force and position information collected during the grinding process; the grinding quality index is used to indicate the processing quality level; the preset qualified index refers to the standard threshold for quality assessment; the optimal grinding parameters refer to the parameter combination with the best performance in the historical data.

[0092] The control system needs to continuously optimize the control parameters during the grinding process. Specifically, the control system first establishes a database containing different material characteristics and corresponding grinding parameters, collects processing data in real time and calculates the quality index. When the quality does not meet the standard, it retrieves the optimal parameters of similar cases from the database based on the current workpiece material characteristics, dynamically adjusts the force control gain and target pressure, and realizes parameter adaptive optimization.

[0093] In some embodiments, the polishing parameters can be optimized in a variety of ways: Optionally, the optimization is performed by the following steps: constructing a hierarchical database structure; real-time feature extraction and matching; parameter similarity calculation; optimal parameter screening; online parameter adjustment. Optionally, the optimization is performed by the following steps: establishing a deep learning model; quality assessment prediction; parameter space search; iterative optimization update. It is understandable that other intelligent algorithms can also be used to optimize the polishing parameters, which are not limited here.

[0094] S605, collecting the force signal of the force sensor at the end of the robot arm, and decomposing the force signal into a normal force component perpendicular to the surface of the workpiece.

[0095] Referring to step S501 , the control system collects the force signal and determines the normal force component.

[0096] S606: Calculate the deviation value between the normal force component and the preset target pressure and the deviation integral term within the preset time window.

[0097] Among them, the normal force component refers to the actual pressure value perpendicular to the workpiece surface; the preset target pressure represents the expected grinding pressure; the deviation value refers to the instantaneous difference between the actual pressure and the target pressure; the preset time window refers to the time range of the integral calculation, which is used to eliminate steady-state errors; the deviation integral term represents the cumulative effect of the pressure error within the time window.

[0098] The control system needs to calculate the pressure control error to build a sliding mode controller. Specifically, the control system calculates the difference between the actual normal force and the target pressure in real time, and uses a sliding time window (such as 0.1 seconds) to integrate the deviation, and improves the tracking accuracy and anti-interference ability of the system through the integral term.

[0099] In some embodiments, the deviation calculation can be implemented in a variety of ways: Optionally, the calculation is performed through the following steps: real-time acquisition of force signals; filtering to eliminate high-frequency noise; calculating pressure deviation; setting an integral window; calculating integral value; dynamically updating the integral upper limit. Optionally, the calculation is performed through the following steps: adaptively adjusting the window size; integral calculation with clipping; considering compensation for the effect of time lag; optimizing the integral strategy. It is understandable that other error calculation methods can also be used to determine the pressure deviation, which is not limited here.

[0100] S607: Perform weighted combination of the deviation value, the deviation integral term and the sliding mode coefficient to obtain a sliding surface expression.

[0101] Among them, the sliding mode coefficient refers to the weight parameter used to adjust the dynamic characteristics of the system; the sliding surface expression represents the deviation characteristics between the system state and the expected trajectory; and the weighted combination refers to the linear combination of each item according to different weights.

[0102] The control system needs to construct a suitable sliding surface to achieve robust control. Specifically, the control system linearly combines the pressure deviation, the deviation integral term and the preset sliding coefficient to construct a sliding surface expression that can reflect the dynamic characteristics of the system and satisfy the Lyapunov stability, providing a basis for the subsequent control law design.

[0103] In some embodiments, the construction of the sliding surface can be achieved in a variety of ways: Optionally, the construction is carried out through the following steps: determining the sliding surface structure; designing the integral term weight; optimizing the sliding coefficient; verifying the stability condition; adjusting the dynamic characteristics. Optionally, the construction is carried out through the following steps: introducing nonlinear terms; adaptively adjusting the weights; constructing a composite sliding surface; analyzing the convergence performance. It is understandable that other control theory methods can also be used to construct the sliding surface expression, which is not limited here.

[0104] In some embodiments, before step S607, the control system first constructs a grinding dynamic model including the surface curvature of the workpiece and the posture of the grinding tool; determines the Lyapunov function based on the grinding dynamic model; solves the derivative expression of the Lyapunov function, determines the adaptive law parameter range, and determines the sliding mode coefficient according to the adaptive law parameter range.

[0105] Among them, the grinding dynamic model refers to the mathematical model that describes the interaction process between the tool and the workpiece; the workpiece surface curvature represents the curvature information of the surface geometric features; the grinding tool posture refers to the spatial position relationship of the tool relative to the workpiece surface; the Lyapunov function is a scalar function used to represent the stability of the system; the adaptive law parameters refer to the parameters used to adjust the adaptability of the controller; the sliding mode coefficient represents the key design parameters of the sliding mode controller.

[0106] The control system needs to ensure the stability of the system before constructing the sliding surface. Specifically, the control system first establishes a dynamic model that takes into account the surface curvature and tool posture, and constructs a Lyapunov function based on the model that can reflect the energy change of the system. The stability condition is obtained by solving its derivative expression, and then the parameter range of the adaptive law is determined, and finally the sliding coefficient that ensures the stability of the system is obtained.

[0107] It should be noted that the grinding dynamic model is based on the workpiece surface curvature data and the grinding tool posture parameters during training, and the dynamic equations considering geometric features are established in combination with mechanical theory, and the model parameters are optimized through Lyapunov stability analysis. The training standard is the parameter constraint condition to ensure the stability of the system. The grinding dynamic model describes the nonlinear dynamic equations of the interaction process between the tool and the workpiece surface, including surface geometric features and tool kinematic constraints. By inputting the tool state and surface features, the grinding dynamic model can output the system dynamic response prediction, which is used for the design and parameter optimization of the sliding mode controller.

[0108] In some embodiments, the control parameters can be determined in a variety of ways: Optionally, the control parameters can be determined by the following steps: establishing a nonlinear dynamics equation; selecting a Lyapunov candidate function; stability analysis; solving parameter constraints; optimizing the sliding mode coefficient. Optionally, the control parameters can be determined by the following steps: simplifying model construction; piecewise linearization processing; robustness analysis; adaptive law design. It is understandable that other control theory methods can also be used to determine the parameters, which are not limited here.

[0109] S608. Determine the force control component according to the sliding surface expression and the force control gain coefficient.

[0110] Referring to step S503 , the control system determines the force control component.

[0111] S609, determining the position deviation and speed deviation according to the collected position signals of the joint encoders and the preset grinding trajectory.

[0112] Referring to step S504 , the control system determines the position deviation and the speed deviation.

[0113] S610, determining a position control component according to the position deviation, the speed deviation, the position gain coefficient and the speed gain coefficient.

[0114] Referring to step S505 , the control system determines the position control component.

[0115] S611. Construct a selection matrix based on the normal direction and the tangential direction of the workpiece surface, and perform orthogonal decomposition and superposition of the force control component and the position control component based on the selection matrix to obtain a comprehensive control instruction.

[0116] Referring to step S506, the control system generates a comprehensive control instruction.

[0117] S612, converting the comprehensive control instructions into driving signals of the motors of each joint, and transmitting the driving signals to the robot arm to perform grinding and polishing based on sliding mode control.

[0118] Referring to step S507, the control system will issue a comprehensive control instruction.

[0119] S613. Collect surface quality data of the workpiece surface during the grinding process.

[0120] Among them, surface quality data refers to the measurement data that characterizes the surface processing status of the workpiece, including parameters such as roughness and smoothness; the grinding process refers to the entire cycle of the robotic arm performing the grinding operation; the workpiece surface refers to the surface area of ​​the workpiece to be processed, which is used to evaluate the processing effect.

[0121] The control system needs to monitor the grinding effect in real time to achieve quality control. Specifically, during the grinding process, the control system collects real-time status data of the workpiece surface, including surface morphology, reflective properties, contact force characteristics, etc., through the configured visual sensors, force sensors and other multimodal sensing devices, and performs data preprocessing and feature extraction.

[0122] In some embodiments, the quality data can be collected in a variety of ways: Optionally, the data can be collected by the following steps: configuring a multimodal sensor; synchronously collecting image data; collecting force tactile data; extracting surface features; constructing a quality feature vector; and data fusion processing. Optionally, the data can be collected by the following steps: online optical measurement; dynamic force characteristic analysis; vibration signal processing; and establishing a feature database. It is understandable that other sensing methods can also be used to collect surface quality data, which is not limited here.

[0123] S614: Determine a key grinding area on the workpiece surface according to the surface quality data and a preset grinding judgment threshold.

[0124] Among them, the preset grinding judgment threshold refers to the standard parameter used to evaluate the surface quality; the key grinding area represents the surface area that needs to be processed intensively; and the surface quality data includes various characteristic parameters of the workpiece surface.

[0125] The control system needs to identify areas that are not processed enough to optimize the grinding strategy. Specifically, the control system compares and analyzes the collected surface quality data with the preset quality standards, identifies areas where the quality parameters are below the threshold through a clustering algorithm, and determines the scope of the local area that needs to be focused on grinding and the processing requirements based on the workpiece geometric features.

[0126] In some embodiments, the key areas can be determined in a variety of ways: Optionally, the key areas can be determined by the following steps: setting quality evaluation indicators; data normalization processing; feature clustering analysis; threshold judgment classification; regional boundary extraction; generating a processing area map. Optionally, the key areas can be determined by the following steps: establishing a deep learning model; feature recognition classification; quality assessment scoring; determining the priority processing order. It is understandable that other image processing and pattern recognition methods can also be used to determine the key polishing areas, which are not limited here.

[0127] S615. Determine the posture angle, local dwell time and grinding speed of the grinding tool according to the key grinding area.

[0128] Among them, the posture angle refers to the spatial direction of the grinding tool relative to the workpiece surface; the local dwell time indicates the duration of processing in a specific area; the grinding speed refers to the movement speed of the tool relative to the workpiece; the key grinding area contains the location information and quality requirements that need to be focused on.

[0129] The control system needs to optimize local processing parameters to improve processing efficiency. Specifically, the control system calculates the optimal tool posture that meets the processing requirements based on the geometric characteristics and quality requirements of the key grinding area, combined with the material removal rate model, and dynamically allocates processing time according to quality differences, and performs speed planning to ensure the processing effect.

[0130] It should be noted that the material removal rate model is based on the process parameters such as grinding pressure, speed, tool posture and the corresponding material removal data during training, and the mapping relationship between parameters and removal efficiency is established through regression analysis. The training standard is to minimize the error between the predicted removal rate and the actual processing effect. Model itself: a mathematical model that describes the relationship between process parameters and material removal effect, taking into account multiple influencing factors such as pressure, speed, posture, etc. Model use: input the desired material removal amount, and output the required combination of processing parameters to optimize local dwell time and motion planning.

[0131] In some embodiments, the optimization of processing parameters can be achieved in a variety of ways: Optionally, the optimization is achieved through the following steps: establishing a material removal model; analyzing surface curvature characteristics; optimizing tool posture angles; calculating minimum dwell time; velocity trajectory planning; simulation verification scheme. Optionally, the optimization is achieved through the following steps: constructing a processing efficiency model; multi-objective parameter optimization; generating space-time trajectories; and evaluating processing effects. It is understandable that other optimization algorithms can also be used to determine processing parameters, which are not limited here.

[0132] In some embodiments, the control system constructs a dynamic model of the grinding process including the material removal rate according to the key grinding area; calculates the local residence time of each local interval based on the dynamic model; calculates the trajectory curvature of each local interval in combination with the surface features of the workpiece, and determines the posture angle and speed constraints of the grinding tool according to the trajectory curvature; solves the speed planning scheme that meets the speed constraints to obtain the grinding speed.

[0133] Among them, material removal rate refers to the amount of material removed per unit time; the dynamic model refers to the mathematical model that describes the dynamic characteristics of the grinding process; the local interval refers to the segmented processing area on the workpiece surface; the trajectory curvature refers to the curvature characteristics of the tool motion path; the speed constraint refers to the speed limit considering the process requirements; the speed planning scheme is used to guide the tool motion control.

[0134] The control system needs to optimize the motion trajectory based on the key grinding area. Specifically, the control system establishes a dynamic model including the material removal rate, calculates the processing time required for each interval, takes into account the surface curvature characteristics of the workpiece, determines the tool posture and speed constraints that meet the processing requirements, and finally solves the speed planning scheme that meets the constraints through the optimization algorithm.

[0135] In some embodiments, trajectory planning can be achieved in a variety of ways: Optionally, planning can be achieved through the following steps: establishing a material removal model; interval time allocation; posture constraint analysis; speed boundary calculation; trajectory optimization solution; simulation verification. Optionally, planning can be achieved through the following steps: task decomposition and sorting; multi-target trajectory planning; real-time collision detection; speed trajectory smoothing. It is understandable that other path planning methods can also be used to achieve the generation of motion trajectories, which are not limited here.

[0136] In the embodiment of the present application, due to the use of a force / position hybrid control scheme based on sliding mode control, combined with technical means such as workpiece material characteristic identification, adaptive parameter optimization and real-time quality monitoring, it is possible to achieve fast and accurate pressure tracking control while ensuring system stability, effectively solving the problems of response lag, large pressure fluctuations, and parameter adjustment dependence on experience in impedance control in the prior art, thereby realizing an intelligent grinding and polishing system that is adaptable to different workpiece materials and has self-learning optimization capabilities. The experimental results show that this scheme can control the pressure fluctuation during the grinding of curved workpieces within the range of ±2N, which is significantly better than the traditional method. At the same time, through online quality monitoring and parameter optimization, the processing efficiency and quality consistency are further improved.

[0137] The following describes the control system in the embodiment of the present invention from the perspective of hardware processing. Figure 7 , which is a schematic diagram of the structure of a physical device of the control system in an embodiment of the present application.

[0138] It should be noted that Figure 7 The structure of the control system shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0139] like Figure 7 As shown, the control system includes a central processing unit (CPU) 701, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 702 or the program loaded from the storage part 708 to the random access memory (RAM) 703, such as executing the method described in the above embodiment. In the RAM 703, various programs and data required for system operation are also stored. The CPU 701, the ROM 702 and the RAM 703 are connected to each other through the bus 704. The input / output (I / O) interface 705 is also connected to the bus 704.

[0140] The following components are connected to the I / O interface 705: an input section 706 including an audio input device, a button switch, etc.; an output section 707 including a liquid crystal display (LCD) and an audio output device, an indicator light, etc.; a storage section 708 including a hard disk, etc.; and a communication section 709 including a network interface card such as a LAN (Local Area Network) card, a modem, etc. The communication section 709 performs communication processing via a network such as the Internet. A drive 710 is also connected to the I / O interface 705 as needed. A removable medium 711, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 710 as needed so that a computer program read therefrom is installed into the storage section 708 as needed.

[0141] In particular, according to an embodiment of the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present invention includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a computer program for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through the communication part 709, and / or installed from a removable medium 711. When the computer program is executed by the central processing unit (CPU) 701, various functions defined in the present invention are performed.

[0142] It should be noted that specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, apparatus, or device.

[0143] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present invention. Each box in the flowchart or block diagram may represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the box may also occur in an order different from that marked in the accompanying drawings.

[0144] Specifically, the control system of this embodiment includes a processor and a memory. The memory stores a computer program. When the computer program is executed by the processor, the grinding and polishing control method based on sliding mode control provided in the above embodiment is implemented.

[0145] As another aspect, the present invention further provides a computer-readable storage medium, which may be included in the control system described in the above embodiment; or may exist independently without being assembled into the control system. The above storage medium carries one or more computer programs, and when the above one or more computer programs are executed by a processor of the control system, the control system implements the grinding and polishing control method based on sliding mode control provided in the above embodiment.

[0146] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

[0147] As used in the above embodiments, the term "when..." may be interpreted to mean "if..." or "after..." or "in response to determining..." or "in response to detecting...", depending on the context. Similarly, the phrases "upon determining..." or "if (the stated condition or event) is detected" may be interpreted to mean "if determining..." or "in response to determining..." or "upon detecting (the stated condition or event)" or "in response to detecting (the stated condition or event)", depending on the context.

[0148] Those skilled in the art can understand that to implement all or part of the processes in the above-mentioned embodiments, the processes can be completed by computer programs to instruct related hardware, and the programs can be stored in computer-readable storage media. When the programs are executed, they can include the processes of the above-mentioned method embodiments. The aforementioned storage media include: ROM or random access memory RAM, magnetic disk or optical disk and other media that can store program codes.

Claims

1. A grinding and polishing control method based on sliding mode control, characterized in that: Applied to a control system, the method comprises: Collecting a force signal from a force sensor at the end of the robot arm, and decomposing the force signal into a normal force component perpendicular to the surface of the workpiece; Determining a deviation value between the normal force component and a preset target pressure, and constructing a sliding surface expression of sliding mode control based on the deviation value; Determining a force control component according to the sliding surface expression and the force control gain coefficient; Determine the position deviation and speed deviation based on the collected position signals of the encoders of each joint and the preset grinding trajectory; Determining a position control component according to the position deviation, the speed deviation, the position gain coefficient and the speed gain coefficient; Constructing a selection matrix based on the normal direction and the tangential direction of the workpiece surface, and performing orthogonal decomposition and superposition on the force control component and the position control component based on the selection matrix to obtain a comprehensive control instruction; The comprehensive control instruction is converted into a driving signal of each joint motor, and the driving signal is transmitted to the robot arm to perform grinding and polishing based on sliding mode control.

2. The method according to claim 1, characterized in that The step of determining the deviation value between the normal force component and the preset target pressure, and constructing a sliding surface expression of sliding mode control based on the deviation value, specifically includes: Calculating a deviation value between the normal force component and the preset target pressure and a deviation integral term within a preset time window; The deviation value, the deviation integral term and the sliding mode coefficient are weightedly combined to obtain a sliding mode surface expression.

3. The method according to claim 2, characterized in that Before the step of weighted combination of the deviation value, the deviation integral term and the sliding mode coefficient to obtain a sliding mode surface expression, the method further includes: Construct a grinding dynamic model including the workpiece surface curvature and grinding tool posture; Based on the polishing dynamic model, determining a Lyapunov function; The derivative expression of the Lyapunov function is solved to determine the range of adaptive law parameters, and the sliding mode coefficient is determined according to the range of adaptive law parameters.

4. The method according to claim 1, characterized in that: After the step of converting the comprehensive control instruction into a driving signal of each joint motor and transmitting the driving signal to the robot arm to perform grinding and polishing based on sliding mode control, the method further includes: Collect surface quality data of workpiece surface during grinding; Determining a key grinding area on the surface of the workpiece according to the surface quality data and a preset grinding judgment threshold; According to the key grinding area, the posture angle, local dwell time and grinding speed of the grinding tool are determined.

5. The method according to claim 4, characterized in that The step of determining the posture angle, local dwell time and grinding speed of the grinding tool according to the key grinding area specifically includes: According to the key grinding area, a dynamic model of the grinding process including material removal rate is constructed; Based on the kinetic model, calculating the local residence time of each local interval; Calculating the trajectory curvature of each of the local intervals in combination with the workpiece surface features, and determining the posture angle and speed constraint conditions of the grinding tool according to the trajectory curvature; A speed planning scheme satisfying the speed constraint is solved to obtain the polishing speed.

6. The method according to claim 1, characterized in that Before the step of collecting the force signal of the force sensor at the end of the robot arm and decomposing the force signal into a normal force component perpendicular to the surface of the workpiece, the method further includes: Control the robot arm to contact the workpiece surface at a preset speed and perform a preset movement, and collect the force-controlled displacement characteristic curve during the contact process; Determining the stiffness coefficient of the workpiece surface according to the force-controlled displacement characteristic curve; Determining a material damping coefficient of the workpiece surface based on contact transient response characteristics, and generating a material property model according to the material damping coefficient and the stiffness coefficient; Based on the material property model, a force control gain coefficient and a preset target pressure are determined.

7. The method according to claim 6, characterized in that The step of determining the force control gain coefficient and the preset target pressure based on the material property model specifically includes: Based on the material property model, construct a polishing database including material properties and polishing parameters; Collecting real-time operation data including force data and position data during the grinding process, and determining a grinding quality index according to the real-time operation data; When the grinding quality index is lower than a preset qualified index, extracting optimal grinding parameters of similar workpiece materials from the grinding database; Based on the optimal grinding parameters, a force control gain coefficient and a preset target pressure are determined.

8. A control system, characterized in that: The control system includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the control system to execute the method described in any one of claims 1-7.

9. A computer-readable storage medium comprising instructions, characterized in that: When the instructions are executed on a control system, the control system is caused to execute the method according to any one of claims 1 to 7.

10. A computer program product, characterized in that When the computer program product is run on a control system, the control system is caused to execute the method according to any one of claims 1 to 7.

Images