Learning Posicast Control of a Pneumatic Servo

To apply a pneumatic actuator more effectively in various automation systems, a pneumatic servo technology should be developed. Recently, from such a viewpoint, the effects of some advanced control strategies as an optimal control or an adaptive control have been confirmed. However, they seem to need some time until put into practical use because of their relatively complicated control algorithms. From the practical application of view, it should be as simple as possible. In this paper, we take notice of the simplicity of a posicast control and propose an advanced posicast control scheme with learning funcion to execute the self tuning. First, we describe the typical features of proposed control scheme. Next, we apply it to the positioning and force control of a pneumatic servo system comprising a pneumatic cylinder and electro-pneumatic proportional control valves. The following is experimentally confirmed. 1) The addition of learning function to the original posicast control scheme allows a finite time settling step response even in a certain degree of high order system such as a pneumatic system. Moreover, it is considered to allow the system to adapt to the change in system parameters, the control input saturation and so on. 2) The proposed control scheme can attain the higer speed and more accurate positioning and force control compared with an usual PI control scheme, so it is well available to a pneumatic servo control. Generally, a posicast control may be effective for a low damping system such as a pneumatic servo, because it basically utilizes a oscillation property of the controlled system.