Automatic tuning of predictive control in a hydrostatic drive train system in nominal operation

Abstract

This paper presents an automatic model based predictive control tuning and operation of a hydrostat drive train system under nominal conditions. An industrial-friendly version of the predictive control strategy is applied whereas the model is simplified as an approximation of first order plus dead time and the control design parameters are automatically tuned. Implementation on a SISO drive train is experimentally achieved with successful performance in closed loop tests. From the identified characteristics of the real life drive train, a simulated MIMO system is extracted and automatic tuning is again performed. A limitation of the present work is that constraints are not yet included, only safety limits.