Study On Adaptive Control Method Of The Robotic Manipulator Servo System For Radial Tire Making Process

With the development of highway construction and automobile industry,tires,especially radial tires are widely used in daily life.The rapid rise of intelligent control technology and robot technology,many kinds of robotic manipulators are widely used in the radial tire production.The dead zone,friction,unmodeled dynamics,parameter uncertainty and external disturbance in servo system seriously affect robotic manipulator positioning accuracy,running speed and running track accuracy,which leads to generate various biases in tire production process,eventually affect the tire product quality,make the tire size,density and rigid distribution uniformity reduced,and cause safety risks to the use of tires.In addition,the robotic manipulator control performance problems can also cause lost work and rework,which brings adverse effects on the tire production cycle,production capacity and production rhythm.Therefore,improve the control performance of robotic manipulator servo system is very important to ensure the quality of tires and the development of the tire industry.In this dissertation,theoretical research work on control methods such as funnel adaptive control,robust adaptive asymptotic tracking control,adaptive control based on parameter estimation error and multi-sensor adaptive weight information fusion are carried out,in view of the dead zone,friction,unmodeled dynamics,parameter uncertainty and external disturbance of the selective compliance assembly robot arm(SCARA)servo system used in radial tire production line.To improve the control performance of the robotic manipulator servo system and provide theoretical guidance for practical production.The main work is given as follows:(1)Study on funnel adaptive control methodIn view of the problems in SCARA servo system used in radial tire production mixing process,which affect system control performance such as dead zone and unmodeled dynamics,an funnel adaptive control method based on echo state neural network(ESN)is proposed,to improve the accuracy of the mixing raw material grasping and release.The control method restrains the adverse effect of the dead zone on the control system by establishing the inverse transformation model of the robotic manipulator;ESN is used to approximate the unmodeled dynamics of the system;the system order limit is relaxed by improving the traditional funnel function.On the above basis,the funnel adaptive controller is designed,and the corresponding parameter adjustment rules are given.Lyapunov function is used to prove the stability of the closed-loop system.Compared with traditional neural network,ESN used in this control method has better robustness and accuracy,has shorter run time,and does not need to adjust the weight between the input layer and the implicit layer.The control method relaxes the strict limit of 1 or 2 system orders by adjusting the time-varying control gain in the funnel function,and this adjustment also can avoid the singularity problem in the prescribed performance function(PPF)when error conversion.The simulation results verify that the control method achieves the effect of bringing the robotic manipulator joints speed and trajectories quickly converge to the reference signal and remain stable,and the tracking error is always limited to the bounds of a given funnel function.(2)Study on robust adaptive asymptotic tracking control methodIn view of the problems in SCARA servo system used in radial tire composition production process,which affect system control performance such as friction and external disturbance,an adaptive asymptotic tracking control method based on the robust integral of the sign of error(RISE)is proposed,to improve the positioning accuracy of the robotic manipulator when tire components are assembled.The control method approximates the unmodeled dynamics of the system by adjusting weights online with the radial basis function(RBF)neural network;the improved PPF is used to converse tracking error,the new tracking error remains within the specified boundary after the conversion.The adaptive asymptotic tracking controller based on RISE feedback is designed finally.Lyapunov function is used to prove the stability of the closed-loop system.Compared with traditional PPF,the improved PPF used in this control method avoids the singularity problems;the RISE feedback can restrain the adverse effects of neural network approximation error and disturbance on system performance,so that the system tracking error is globally bounded and asymptotically converges.The simulation results verify that the control method achieves the effect of bringing the robotic manipulator joints speed and trajectories quickly converge to the reference signal and remain stable,and the tracking error can quickly converge to zero or near-zero minimal regions.(3)Study on adaptive control method based on the parameter estimation errorIn view of the problems in SCARA servo system used in radial tire production forming process and sulphide process,which affect system control performance such as unmodeled dynamics and parameter uncertainty,the adaptive parameter estimation algorithm and adaptive controller design method based on parameter estimation error are proposed,to improve the movement trajectory accuracy of the robotic manipulator.The mathematical model of the robotic manipulator is transformed,the neural network is used to approximate the unmodeled dynamics in servo system;the auxiliary variables are designed and filtered to extract and characterize the parameter estimation error,then put them into the adaptive law as forgetting factors,to update he parameter adaptive law and design the adaptive controller to ensure that the system control error and the neural network weight estimation error converge at the same time;the sliding mode is introduced in the parameter adaptive law,and the limited time adaptive controller is designed to ensure that the control error and parameter estimation error can converge quickly.Lyapunov function is used to prove the stability of the closed-loop system.Compared with the traditional parameter estimation method,the adaptive parameter estimation method based on parameter estimation error does not need to calculate system state differential,persistent excitation(PE)conditions can be monitored online,parameter estimation value can more quickly and accurately converge to the true value.The simulation results verify that the proposed method can converge the parameter estimation faster and achieve better system control effect.(4)Study on the improvement method of the multi-sensor adaptive weight information fusion algorithmIn view of the problems that the visual sensor and force sensor measurement data of SCARA robotic manipulator used in all processes of radial tire production,which is affected by noise,vibration,and other disturbance factors,the improved method of the multi-sensor adaptive weight information fusion method based on discrete Kalman filter is proposed,to improve sensor information fusion accuracy,smoothness and algorithm stability.The mathematical model of the robotic manipulator is linearized to simplify the system description and improve the dynamic identification speed of parameters;the discrete Kalman filtering is used to optimize the acquisition path of fusion data sources,so as to reduce the adverse impact of random interference such as noise and vibration on the system;the multi-sensor adaptive weight information fusion algorithm is improved and the sensor measurement variance is calculated to reduce the negative effects of sensor variance jitter on the information fusion effect;the value of the contact force with the outside environment at the joint of the robotic manipulator is obtained finally by combining the motor signals(joint angle,joint velocity,and joint current).This algorithm replaces the traditional algorithm of using the acquired sensor measurements directly as the fusion data source with the sensor original measurement data has been filtered by discrete Kalman.The simulation results verify that the proposed algorithm can improve sensor information fusion speed,smoothness and accuracy.