Implementation of neural network approach in MIMO electrohydraulic servosystem control

The objective of this study is to implement a practical neural controller for a MIMO nonlinear hydraulic control system. A model reference approach is used in which the controller forces the hydraulic system to follow predetermined reference outputs. Experiments were conducted to illustrate the feasibility and benefits of the neural network approach in comparison with the traditional PID control strategy. This study shows that a neural controller can control a highly coupled and nonlinear hydraulic system within a limited frequency range where system dynamics do not dominate the response. A conventional PID controller was not able to provide adequate control even in this range.