A pragmatic approach to distributed nonlinear model predictive control: Application to a hydrostatic drivetrain

Abstract

SUMMARY The global control of large-scale production machines composed of interacting subsystems is a challenging problem due to the intrinsic presence of high coupling, constraints, nonlinearity, and communication limitations. In this work, a pragmatic approach to distributed nonlinear model predictive control (DNMPC) is presented with guaranteed decrease in cost. Furthermore, in order to tackle time-varying process dynamics, a learning algorithm is developed, thereby improving the performance of the global control. The proposed control framework is experimentally validated on a hydrostatic drivetrain, which exhibits nonlinear dynamics, strongly interacting subsystems. The experimental results indicate that good tracking performance and disturbance rejection can be obtained by the proposed DNMPC.