Application Of Model Free Adaptive Control In Pose Control System Of Pneumatic 3-UPU Parallel Robot

The 3-UPU parallel robot has the advantages of large structural stiffness,high control accuracy and small inertia of end parts.Therefore,it is widely used in many fields,such as intelligent control,automobile manufacturing,parallel machine tool and chemical production.Taking the pneumatic 3-UPU parallel robot as the research goal,this paper makes simulation and Experimental Research on the position and attitude control of the robot.The main work and research results are as follows:Firstly,the hardware selection and platform construction of the pneumatic 3-UPU parallel robot system are carried out,and the kinematics analysis is carried out according to the geometric characteristics of the robot to obtain the forward and inverse solutions of the working space and position of the mechanism.The mechanism model of pneumatic position servo system is established,and the discrete mathematical model of the system is established.Secondly,because the pneumatic position servo system can not obtain an accurate mathematical model.Therefore,the tight format model free adaptive method determined only by the historical input and output data of the system is used to control the pose of the system.simultaneously,aiming at the difficulty of parameter tuning and Optimization in the traditional model free adaptive control,a tight format model free adaptive control parameter optimization method based on genetic algorithm is proposed.The simulation results show that the model free control is better than the traditional PID control,and the model free adaptive control optimized by genetic algorithm has faster response speed and better tracking performance than the model free adaptive control.Finally,the robot pose control system is developed based on Lab Windows/CVI software platform,and the robot pose control is verified by experiments.The experimental results show that compared with the traditional PID control,the rise time and adjustment time of step response are shortened by about 15% and 25% respectively,and the cumulative error of sinusoidal tracking is reduced by about 13.6%.