An intelligent predictive control approach to path tracking problem of autonomous mobile robot

Abstract

An intelligent predictive control approach to the path tracking problem of an autonomous mobile robot is presented based on the analysis of the characteristics of the vehicle model and the traditional predictive control algorithm. A neural network model of the vehicle is used to predict future vehicle posture according to the current posture and control variables. The future tracking error between this predictive posture and the planned ideal path created by the local path planning module can be calculated. The optimal control variables in the next control instant are computed by the online optimization algorithm. The above process is updated and solved repeatedly. The intelligent predictive controller is composed of four principal components: a reference path, a predictive model, a set of online optimization algorithms and a feedback tuning model, which are discussed in detail in the paper. The characteristic of this method is analyzed and the result is provided.