Rigid Flexible Coupling Dynamic Modeling And Compensation Control For Heavy-duty Robot

Compared with conventional light-duty industrial robots,heavy-duty robots have higher requirements for the performance of control systems in order to achieve the typical capabilities such as high-speed,high-precision,and high-load.The output torque of the traditional PID controller can not meet the requirements of rapid and real-time change of the joint desired driving torque during high-speed and heavy-load motion.At the same time,as the load-to-weight ratio of heavy-duty robot continuously increases,the characteristics of its flexible components become more apparent,resulting in elastic vibration of the system in the event of start,stop,and sudden speed changes,affecting its trajectory tracking accuracy,and may even lead to system instability.Relying on Jiangsu Province Transformation Project of Scientific and Technological Achievements "Development and Industrialization of Core Technology of High Speed and Heavy-Duty Industrial Robots(BA2015004)",this dissertation aims to research the rigid-flexible coupling dynamic modeling and compensation control of 150 KG heavy-duty robots.The main contents are as follows:In order to obtain an accurate dynamics model for six-degree-of-freedom heavy-duty robot,this dissertation first used the iterative Newton-Euler method to theoretically deduce its dynamic model.After that,a periodic excitation trajectory based on an improved finite term Fourier series is designed to accurately identify unknown dynamic parameters in theoretical model.In addition,a load inertia parameter identification method is designed.The influence of load effect is further considered on the basis of the dynamic model of the robot body to establish a complete dynamic model for heavy-duty robot under load conditions.Aiming at the problem of the low control accuracy and poor dynamic performance of the heavy-duty robot using classical PID closed loop control algorithm.this dissertation studied and implemented the torque feedforward control method based on dynamic model.Through the identified dynamic model,the expected output torque of each joint of the robot is calculated and superimposed with the control output of the current loop to achieve real-time compensation of the torque.The trajectory tracking accuracy and the dynamic performance of the robot are effectively improved.In addition,a torque compensation control method based on load inertia matching is proposed to effectively improve the control accuracy of heavy-duty robots under load conditions.Aiming at the problem of joint elastic vibration of heavy-duty robot due to joint flexibility,a joint elastic vibration suppression compound control method combining state observer and singular perturbation theory is proposed.The rigid-flexible coupling dynamic model of heavy-duty robot is established firstly under consideration of joint flexibility.After that,a high-gain state observer was designed to realize real-time observation of state variables such as the connecting rod angular speed based on the motor side angle,angular velocity and output torque without the aid of an external sensor.Then,the singular perturbation method is used to transform the complex high-order rigid-flexible coupling dynamic model into a reduced-order subsystem.The speed difference feedback control law and PID control law are respectively designed for the obtained fast and slow subsystems to effectively suppress the joint elastic vibration of the robot and further improves its trajectory tracking accuracy.The stability of the system is proved combining with Lyapulov stability theory.Based on the research above,this dissertation focuses on 150 KG heavy-duty robots developed independently in cooperation with companies.The parameter identification of robot dynamics model,load inertial parameter identification,torque feedforward control,load compensation control and joint elastic vibration suppression were studied experimentally and good experimental results were obtained.The feasibility and effectiveness of the method indicate that the relevant research in this paper for heavy-duty robots has certain engineering application value.