LAUSR

In an electrohydraulic system, the load is an external factor that cannot obtain similar results by repetitive experiments compared with other internal factors. It can work a great change in an extremely short time determined by the characteristics of the loaded object, which is more difficult to be identified timely and compensated precisely. This paper first built the mathematic model of the electrohydraulic system containing load characteristics and illustrated most kinds of basic loads that possibly occur in it. Then, a practice nonlinear variable load composed of these basic loads is presented, and the corresponding effect to the electrohydraulic system is analyzed. Based on the summary of load, a nonlinear variable load compensation controller (NVLCC) with neural networks is proposed to identify the mathematic model of the electrohydraulic system. The nonlinear variable load is considered as a part of the model to be globally identified, and an adaptive sliding mode control method is combined to quickly converge the errors of system state variables. The compensation effectiveness is demonstrated by both rigid step response and dynamic sine response in the simulations, and the actual improvements are verified with the experiments using three different typical linear loaded objects and a representative nonlinear loaded object.