Speed, low production cost and clarity are of main importance.in developing mechanical equipment which deal with humans such as rehabilitation systems. Therefore, pneumatic systems are suitable systems in this field. However, controlling these systems can be a challenging issue, due to the nonlinearities that arose in these systems: For instance, friction, fluid compression and delays of valves performances. various methods are available to design a controller for pneumatic systems. Varseyeld

more »

... d Bone [1] designed a discrete PID controller using a model which had been obtained by identification methods. They used their proposed controller to control the position of a servo-pneumatic operator, then, they examined the controller performance for different inputs. the obtained results of their work lacked accuracy because of two reasons: the inaccuracy of their mathematical model and using a PID controller. In [1] solenoid on/off switches have been used. One of the methods of transferring the control signals to command for such valves is to use a PWD algorithm which Ahn and Yokota used in their work [2]. They controlled the position of a pneumatic operator using a smart switching controller via on/off solenoid switches. In this method and ones used later on, inexactness in modeling and ignoring some parts of the model causes inaccuracy in trajectory following. Moreover, the proposed controllers did not have a satisfactory precision as well. Since exact modelling of pneumatic system are complicated and approximations are always within such systems, it makes sense to use robust control methods. For all the aforementioned reasons, in this paper, a sliding mode controller is used to follow a trajectory. After applying an optimization on the controller, its performance is compared with a PID controller. Using a pulse width modulation algorithm, differential equations of a double sided single action operator and solenoid valves, the pneumatic mathematical equations are simulated in MATLAB/Simulink computer software. Afterwards, a PID controller and a sliding mode controller are designed for the trajectory following task. In order to get the best performance of the designed controllers they are then optimized by genetic algorithm. Finally, the simulated results are compared and advantages/disadvantages of each controller are discussed.