Adaptive Moving-Target Tracking Control of a Vision-Based Mobile Robot via a Dynamic Petri Recurrent Fuzzy Neural Network

In this study, an adaptive moving-target tracking control (AMTC) scheme via a dynamic Petri recurrent fuzzy neural network (DPRFNN) is constructed for a vision-based mobile robot with a tilt camera. In this study, the dynamic model of a vision-based mobile robot system, including a nonholonomic mobile robot and a tilt camera based on the concepts of mechanical geometry and motion dynamics, is developed first. Then, a continuously adaptive mean shift algorithm is adopted for the moving-object detection, and a model-based conventional sliding-mode control (CSMC) strategy is introduced. In order to relax the control design dependent on detailed system information and alleviate chattering phenomena caused by the inappropriate selection of uncertainty bounds, it further designs a model-free AMTC scheme with a DPRFNN to imitate the CSMC strategy. In the DPRFNN, the concept of a Petri net and the recurrent frame of internal feedback loops are incorporated into a traditional fuzzy neural network to alleviate the computation burden of parameter learning and to enhance the dynamic mapping of network ability. This five-layer DPRFNN is utilized for the major role in the proposed AMTC scheme. The corresponding adaptation laws of network parameters are established in the sense of projection algorithm and Lyapunov stability theorem to ensure the network convergence, as well as robust control performance without detailed system information and the compensation of auxiliary controllers. In addition, the effectiveness of the proposed AMTC scheme is verified by numerical simulations under different target tracking, and its superiority is indicated in comparison with the CSMC system. Furthermore, experimental results are also provided to verify the validity of the proposed AMTC scheme in practical applications.