A study on neural network control of explosion-proof 2-link pneumatic manipulator

This research develops with the control on dynamics of the explosion-proof 2-link pneumatic manipulator. In the explosion-proof pneumatic control system, it is generally difficult to prevent the hunting, due to both the time-delay of pneumatic signals and nonlinear friction. Instead of the conventional PD control, a new control scheme with neural network (NN) is proposed to realize the precise position control. Through simulation and experiment, it is verified that the hunting may be suppressed and stabilization of dynamics can be realized by using the NN-PD control.