| Since pneumatic technology has an obvious advantages of high drive velocity and non-pollution, it has been widely used in the industrial fields. Pneumatic technology has been one of the indispensable technologies in the up-to-date automation. However, because of the inherent strong non-linear feature and low frequency, it's very difficult to achieve ideal result of control precision and stability, restraining its further application in the industry. Therefore, deeply research on the pneumatic system, adoption of the suitable control strategy improving the precision of position has been the important research and practicable value.The mathematical model of the pneumatic servo positioning system, which is composed of rodless cylinder and the pneumatic servo valve, is build in terms of the gas peculiarity and the theory of heat and mechanics and the reasonable hypothesis and presumption. the research on the peculiarity of system by means of emulation in the paper. On the basis, the corresponding research on control strategy has been studied for the pneumatic servo positioning system. Adoption to classical control method (BP) is difficult to achieve the ideal control result because of the serious non-linear in the system , therefore, neural network control strategy is introduced , neural network controller based on BP-PID-RBF has been designed and Emulated by means of a serial of advantages of self-learning and adaptive and non-linear reflexion ability. The result proves that the control strategy can obviously improve the dynamic and static capability, realizing the precise position at any point within the stroke. Which has the character of the high precise position and fast echo and good trail result. And the same time, the difficult question of choosing parameter of the PID controller has been completely solved,the defect, given rise to the neglecting friction causing non-precision of the mathematical model, has been basically compensated. |
PDF Full Text Request