Motion Control Design Optimization: Problem and Solutions

Given a particular motion control problem, the question investigated in this paper is how to select the most appropriate control law and its parameters. Here, the selections of both the controller and its parameters are formulated as a single optimization problem. A novel cost function is proposed to capture many practical design considerations. The choices of controllers include the conventional proportional-integral-derivative (PID) control and its variations, the parameterized loop-shaping method, as well as the more recent linear active disturbance rejection method. For the purpose of both performance optimization and ensuring the fairness of the controller comparison and selection, the parameters of each controller are optimized in the comparison study using genetic algorithm (GA). The proposed optimization method is demonstrated in a motion control design case study, where the results convincingly show that the controller, based on the active disturbance rejection concept, is clearly superior to others.