Research In Pneumatic Position Servo System Based On Cmac Neural Network

Pneumatic servo technique having already successfully applied in industry fields as a auto-control mean.But the position servo system has some drawbacks,such as low output stiffness,low natural frequency, low-response, the nonlinear of discharge when air pass the valve port, air cylinder existence friction and time-variation etc. hard to adapt many special work conditions.Considerated in the nonlinear friction, time delay, flow nonlinear of the system, this dissertation established the nonlinear mathematics model of the servo valve control air cylinder in the pneumatic position servo system. On the theory research the character of dynamic and static in different gain, different frequency, different damping ratio about the system, get the influence degree of the character about the system.Aim at the single nerve unit self adapting PID control can make up the weak quality of the low-response and instability about the system, this dissertation put forward a compound control algorithm of the single nerve unit self adapting PID add nonlinear differential signal tracker and the cerebella model neural network controller(CMAC), use the single nerve unit PID replace the ordinary PID control, the nerve unit on-line adjust PID control parameter, make use of the self study and self adapting ability of the neural network, completion real-time control of the system.The CMAC realize feedforward control, passing train to acquire inverse model of the complications nonlinear controlled object, the single nerve unit PID controller realize feedback control, assurance the stability of system and check turbulence.On the simulation research adopted the MATLAB/Simulink and AMESim software, and compared in the system nature of which controlled by PID. The result certificate which, the pneumatic position servo system which adopted CMAC control algorithm has good character of dynamic and static. The bandwidth is 5 Hertzian, the tracking accuracy is ninety-five percent. CMAC control strategy has simple composition, easily realize, can adapt change of environment, has stronge robust, greatly improve the dynamic and static performance of the system.Based on theory research and simulation analysis, an experimental research carry on the pneumatic position servo system about point to point on±50 millimeter,±100 millimeter and±300 millimeter. The pneumatic position servo system can realize the air cylinder piston precision positioning on the whole stroke arbitrarily position, the mean absolute positioning accuracy is 0.0844 millimeter,the maximal error is 0.2 millimeter, has high engineering application value.