| As the pneumatic servo technology is safe, clean, quick and reliable. The development of pneumatic servo technology is becoming more and more important in the field of machinery manufacturing and industrial control. And pneumatic manipulator is a typical case. Comparing with other manipulators using different control modes, it has many advantages, for example:the structure is simple, the cost is low, its quality is light, the ratio of power and volume is high and the capability of anti-jamming is good. At present, many researchers study the position control and achieve some results, while few researchers study the speed control. The throttle control is the main method in the speed control, and the control accuracy of throttle control is poorer than Servo control.The thesis is based on some literature about pneumatic manipulator. The pneumatic manipulator with three degree of freedom introduced from Bosch Rexr-oth is taken as the research object to analyze the speed control problem. At first, the servo system and working principle of pneumatic manipulator have been grasped. The improvement program of servo system hardware is designed so that the speed can be controlled continuously.(The original controller can only get point-to-point position control). Then, the nonlinear of system can be described in some respects, such as the nonlinear of flow of proportional servo valve, friction characteristics and creep characteristics. And it proposes the technology of nonlinear compensation by analyzing the influence about the nonlinear of system. Secondly, the mathematical model of speed control of servo system of pneumatic manipulator is bulit. Based on the mathematical model,the system characteristics affected by parameter changing can be analyzed. The dynamic characteristics of the control system can be corrected based on PID.As pneumatic manipulator has characteristics of repetitive motion, it is a system with characteristics of nonlinear and time-varying. While iterative learning control is applied in system of repetitive motion. So it has good control effect for the problem such as the characteristics of nonlinear, time-varying, interference and cross-coupling. The convergence of the system with the iterative learning control is verified by the thesis. The mode can be built in Simulink and the simulation analysis can be achieved by using the language in Matlab. The PID control and the iterative learning control comparative simulation is used to be the analysis method. The simulation result shows that control accuracy of the system is4%with the PID control while it is1.8%with the iterative learning control.Moreover,the iterative learning control has better robustness compared with the PID control from three different cases. So the system with the iterative learning control has good control effect. |