Predictive control of high power hydraulic systems for application to a forestry arm

Manipulator arms for agriculture and forestry are generally hydrostatically powered. The main purpose of this predictive control system is to reduce energy exchanges for large displacements of heavy loads with minimization of execution time. The control system must provide high system velocity, robustness and stability of the system. and a minimal energy cost. As a response to these criteria, a bang-bang control strategy is proposed for the large maneuvering of heavy loads. In the braking phase, the oil flow inversion returns potential energy to the supply circuit. Bond graph modeling was used for a better understanding of the physical phenomena. A predictive algorithm was developed. Experimental results of this predictive system for position control are presented