Adaptive Intelligent Systems applied to two-wheeled robot and the effect of different terrains on performance

摘要

This work discuss two different intelligent controllers: Online Neuro Fuzzy Controller (ONFC) and Proportional-Integral-Derivative Neural Network (PID-NN). They were applied to maintain the equilibrium and to control the position of a two-wheeled robot prototype. Experiments were carried out to investigate the equilibrium control and movement of the two-wheeled robot first on flat terrain, then in other situations, where terrain may not be flat, horizontal surface. The effectiveness of each controller was verified by experimental results, and the performance was compared with conventional PID control scheme applied for the prototype.