Research On High-performance Motion Control Method For Serial Robotic Manipulator

Industrial robotic manipulators can reduce production costs and replace dangerous,dirty and repetitive manual labor,so they are widely used in various industrial processes such as welding,handling,loading and unloading work.With the advancement of technology and the rapid development of modern industry,there is increasing demand for motion control performance of robots,such as tracking accuracy,dynamic response,and robustness.Industrial robotic manipulators are highly complex nonlinear systems,and there are unavoidable uncertainties such as joint coupling,time-varying parameters,load changes,unmodeled dynamics,and unknown external disturbances.These uncertainties can affect the dynamic characteristics of the robotic system and result in reduced tracking accuracy and even system instability.To this end,this paper takes serial robotic manipulator with uncertainty as the research object,and studies the problems of dynamic parameter identification,nonlinear friction modeling,and high-performance robust motion control technology.The main contents are as follows:Designing a model-based control algorithm is an effective way to improve the motion performance of the robot.Therefore,the dynamic model of the experimental robot platform is established.To simplify the parameter identification process,the open-source software OpenSymoro is used to solve the observation matrix and the minimum inertia parameter set.The excitation trajectory is designed,the excitation trajectory parameters are optimized,and the collected data is preprocessed to reduce the influence of measurement noise.The minimum parameter set is identified by the least-squares method,and the validity of the identification parameters is verified by experiments.Joint nonlinear friction is an important factor affecting the control performance at low speed.To improve the control precision at low speed,this paper chooses a suitable continuous nonlinear friction model to compensate joint friction.According to the measured velocity and friction torque data,the differential evolution algorithm is used to identify the optimal friction model parameters.The fitting results verify the effectiveness of the friction model.Considering the complexity of modeling and parameter identification and the existence of uncertain factors in practical applications,the active disturbance rejection control technique is introduced,and its component tracking differentiator,extended state observer and state feedback combination are introduced in detail.To further improve the control performance,the existing dynamic model information is compensated into the extended state observer,and a new robust control scheme is proposed in combination with the conventional computed torque control,which overcomes the shortcomings of the conventional computed torque control depending on model accuracy and greatly enhances the robustness of the system.At the same time,the tracking differentiator is used to effectively estimate the actual speed.The effectiveness and superiority of the controller are verified by simulation and experiment,compared with several classical controllers.To improve the system response speed and tracking accuracy,a finite time controller based on the dynamic model is designed by using the backstepping method for the uncertain robot system.Considering the complex uncertainties existing in practice,the extended state observer with finite-time convergence characteristics is designed to estimate the uncertainties and eliminate it via feedforward compensation,and the variable structure term is used to compensate the observer's estimation error.Compared with the previously proposed control algorithm,finite-time control has better transient performance,higher control precision,and stronger robustness.The Lyapunov stability theory is used to prove the finite-time stability of the controller in the closed-loop system.The effectiveness and feasibility of the algorithm are verified by simulation and experiments.When there are many joints,the model-based controller is difficult to apply due to the high complexity of the system model,the difficulty of identification and the heavy computational burden.Therefore,the joint independent control algorithm which is more convenient to implement is designed without relying on the dynamic model.At the same time,more comprehensive requirements are placed on the control performance.In addition to the steady-state performance,transient performance such as overshoot,undershoot and convergence speed is also taken into account.The control algorithm is designed via backstepping method combined with a prescribed performance function and a friction model-assisted extended state observer.The friction model-assisted extended state observer is used to estimate the absence of dynamic model information and other uncertainties.The transient and steady-state performance are quantitatively designed and speci fied through prescribed performance functions.A variable gain observer is designed for the peak effect problem in the fixed gain state observer caused by the initial observation error.The performance of the algorithm was verified by two-axis linkage and six-axis linkage experiments of the robot platform.The results show that the control algorithm can constrain the tracking error within the prescribed performance range in both joint space and Cartesian workspace,showing good control performance.