A novel self-tuning proportional–integral–derivative controller based on a radial basis function network for trajectory tracking of service robots

Abstract

To solve the trajectory tracking of service robots in autonomous navigation, a novel self-tuning proportional–integral–derivative controller identified by a radial basis function neural network (radial basis function proportional–integral–derivative controller) is presented. The error regarding the lateral distance and directional deviation angle of the service robot is taken as the control deviation in the radial basis function proportional–integral–derivative controller. During the trajectory tracking, the proportional–integral–derivative parameters of the proposed controller can be adaptively adjusted online by using a radial basis function identification network. To keep the tracking effect of the service robot from being influenced by the initial values (i.e. the initial proportional–integral–derivative parameters and their learning rates) of the radial basis function proportional–integral–derivative controller, a chaos small-world algorithm is introduced to optimize them. The simulation results of the trajectory tracking show that the proposed controller can realize online adjustment of proportional–integral–derivative parameters according to actual conditions of service robots and is characterized by strong noise and disturbance suppression capability. The optimization of the radial basis function network controller based on chaos small-world algorithm can further improve the trajectory tracking precision. Additionally, experiments in the indoor environment further support the validity of the proposed radial basis function proportional–integral–derivative controller for trajectory tracking of service robots.