Nonlinear modeling of an electrohydraulic actuation system via experiments and its characterization by means of neural network

The friction and end-cushioning effects in the actuator have been taken care in consideration in the modeling of the actuation method related to an electrohydraulic system. A much simpler friction model that retains all the features of the existing models has been developed for this purpose. In addition, a systematic experimental characterization procedure has been proposed that has been utilized in a supplementary manner for the development of the elaborate nonlinear system model. An artificial neural-network model has been constructed by training with experimental data keeping in mind the variation of discharge through the proportional valve with pressure and command signal. All the nonlinear subsystem models thus obtained have been incorporated simultaneously in MATLAB/SIMULINK to monitor the actuation dynamics. The variations of the theoretical and investigated (via experiments) displacements of the piston against different command signals have been found to be quite close to each other.