Neural Cognitive Computing Mechanisms

Abstract

In this chapter, a mobile robotic system is designed under the vision–brain hypothesis, taking the wheeled mobile robotic (WMR) system as an example. Based on the hypothesis and results of Chap. 3, robots can selectively detect and tracking objects and the robot path-planning problem have been solved. Therefore, an adaptive neural network (NN)-based tracking control algorithm is enough to design the full state constrained WMR system. To deal with the brain-inspired tracking task requirements of the WMR system, it is necessary to take the full state constraints problem into account and based on the assumptions and lemmas given in this chapter, the uniform ultimate boundedness for all signals in the WMR system can be guaranteed to ensure the tracking error converges to zero. Numerical experiments are presented to illustrate the good performance of our control algorithm. Moreover, a partial reinforcement learning neural network (PRLNN)-based tracking algorithm is proposed to enhance the WMR system performance. As the major neural cognitive computing mechanisms the enhanced WMR system, PRLNN adaptive control solve the WMR tracking problem with the time-varying advance angle. The critic NN and action NN adaptive laws for decoupled controllers are designed using the standard gradient-based adaptation method. The Lyapunov stability analysis theorem is employed to test whether the uniform ultimate boundedness of all signals in the system can be guaranteed, and in addition, a numerical simulation is also presented to verify the effectiveness of the proposed control algorithm.