Bioinspired Intelligent Control of Autonomous Robots with State Estimation

Abstract

Autonomous robots have been widely used to perform various tasks in many fields. Developing efficient and robust control methods for autonomous robots operating in diverse environments with hardware limitations is critical. In addition, accurate state estimation for autonomous robots operating under noises is another essential aspect, as this accuracy significantly influences control performance. This thesis aims to develop multiple practical control strategies with state estimation to tackle the challenges in trajectory tracking control of mobile robots, unmanned underwater vehicles, and unmanned aerial vehicles. First, bioinspired intelligent controllers are proposed based on backstepping and sliding mode control techniques, integrated with a neural dynamics to resolve multiple challenging issues, including speed jump, robot velocity constraints, robustness against disturbances, and control smoothness under noises. Then, the adaptive sliding innovation filter is integrated with the proposed robot control algorithms to provide accurate state estimates for robots to operate under noises and offer extra robustness against modeling errors during estimation processes. Furthermore, the formation control of multiple mobile robots is developed, where distributed estimators are designed to estimate the leader robot's state information, then a distributed leader follower formation control strategy is correspondingly proposed. Finally, the stability of the resulting control systems is proved using the Lyapunov stability theory, and extensive comparison results have demonstrated the efficiency and effectiveness of the proposed tracking control and formation control methods.