Noise suppression zeroing neural network for online solving the time-varying inverse kinematics problem of four-wheel mobile manipulators with external disturbances

Abstract

A novel noise suppression zeroing neural network (NSZNN) is presented for the trajectory tracking problem on a four Mecanum wheeled mobile manipulator (FMWMM) by solving its time-varying inverse kinematics (TVIK) problem. The holistic kinematic model of the FMWMM is developed, which can receive synergistic control of the mobile manipulator. Different from the situation without external interference addressed in our previous work, this paper considers a variety of common time-varying interferences by studying the basic principles of various noises, and proves the NSZNN model’s of the validity and superiority, which solves the TVIK problem of the FMWMM with external disturbances through theoretical analyses. Compared with the existing gradient neural network (GNN) and the traditional zeroing neural network (ZNN), the most representative hybrid noise is selected to conduct a large number of experiments to substantiate the high efficiency and robustness of the NSZNN model. Finally, the NSZNN model is verified on the FMWMM via a robot operating system (ROS) by a successful execution of the trajectory tracking task.