Multi-scenario optimization approach for fuzzy control of a robot-car model

A simple dynamic model of a robot-car has been built in the Matlab/Simulink environment [1], which expanded with a minimal dynamic part [2]. A fuzzy route controller was developed and its performance compared to the PID control [2]. The multi-scenario simulation with five different spatial target points is using in order to represent the all expected scenarios during the controller optimization. The multi-objective fitness evaluation of the driving has also been developed based on kinematic and dynamic characteristics. The optimization of the Fuzzy route controller is performed on the multi-scenario simulation using previously implemented heuristic optimization methods [3]. The multi-scenario optimum is compared with the single-scenario optimums, and evaluated in that way.