Evaluation of cell state techniques for optimal controller design

Abstract

Cell state mapping is a powerful tool for analyzing the global behavior of nonlinear dynamical systems. This paper examines the main cell state techniques for designing optimal fuzzy controllers. The optimal design techniques are evaluated using the benchmark inverted pendulum problem. The performance measures used to evaluate the controller performance include time-optimality, the range of controllable initial states, and response characteristics. In addition, the techniques are evaluated with respect to time-space requirements, and scalability to larger problems. This study provides valuable insight and guidance on the applicability of cell state techniques to optimal fuzzy controller design. >