Optimization of control parameters for fluid power systems by genetic algorithms

Abstract

This thesis presents a genetic algorithm-based method for optimizing control parameters for fluid power systems. Genetic algorithms are general-purpose optimization methods based on natural evolution and genetics. A genetic algorithm seeks control parameters that maximize a measure that evaluates system performance. Gains of a state feedback controller for an electropneumatic position control system and gains of a PID-PD cascade controller for an electrohydraulic speed control system were optimized by a genetic algorithm within reasonable number of experiments. It is concluded that genetic algorithms are efficient for optimizing control parameters for fluid power system.