Pneumatic Robot Position Servo Control System And Control Strategy

Since pneumatic technology has a serials of advantages of high drive velocity, simple contracture, anti-pollution, easy to maintenance, low cost, it has been widely used in the industrial fields of medical instrument etc. Pneumatic technical has been the indispensable means to realize automation. As the frontier research area, pneumatic technology has been paid more attention. However, because of the inherent strong linear feature and the influence of friction, it' s very difficult to achieve ideal result of perfect precision and stability, restraining its further application in the industry. Therefore, focusing on the pneumatic positioning system, the study should improve, through the suitable method of control, the cushioning characteristic and speed up the velocity, realize the arbitrary positioning as well. It certainly has been the important and practicable area of research.The pneumatic servo positioning system, which is composed of rodless cylinder and the servo valve, is studied on mathematical modeling and control in this paper. Firstly, pneumatic servo positioning system is analyzed systemically, and mathematical model is build. The pneumatic system is a nonlinear time-variant system because of the compressibility of air, the air mass flow rate through the valve, the friction force between the piston and the cylinder, etc.Secondly, the components of the pneumatic servo system and the principles of software and hardware are introduced. Lastly, according to the characteristics of the experimental system, research on control strategy is a very important part, in which feasibility is validated by simulation and experiments .BP neural network can easily approximate arbitrary nonlinear mappings under certain circumstance and has learning and adaptive ability, and it is suitable to identify and control nonlinear systems. Therefore, based on conventional PID control and neural control, neural network PID control method is adopted as the strategy to keep a good dynamic process and a better steady-state behavior. The simulated resultsshow neural network using for control is feasible and valid. The experiment system enables the piston to position at any point within the stroke, which can get the precision within 0. 1mm. The precision can satisfy positioning needs of manipulator completely, possessing the high engineering values.