A recurrent emotional CMAC neural network controller for vision-based mobile robots

摘要: Abstract Vision-based mobile robots often suffer from the difficulties of high nonlinear dynamics and precise positioning requirements, which leads to development demand more powerful approximation in controlling monitoring robots. This paper proposes a recurrent emotional cerebellar model articulation controller (RECMAC) neural network meeting such demand. In particular, proposed integrates loop an learning mechanism into (CMAC), is implemented as main component module vision-based robot. Briefly, consists sliding surface, RECMAC, compensator controller. The incorporation structure slide ensures retaining previous states robot improve its dynamic mapping ability. convergence system guaranteed by applying Lyapunov stability analysis theory. was validated evaluated both simulation practical moving-target tracking task. experimentation demonstrated that outperforms other popular network-based control systems, thus it superior approximating highly