Neural Network Based Modelling and Steering Control of an Intelligent Vehicle Under Dynamic Sensitive Conditions. Issue 1 (April 2020)

• Record Type: Journal Article
• Authors: Zhang, K; Wang, S J; Ji, L; Wang, C
• Is Part Of: IOP conference series. Volume 825:Issue 1(2020)
• Publication Date: 2020-04
• Subjects: Materials science -- Periodicals; 620.1105
• DOI: 10.1088/1757-899X/825/1/012043
• Languages: English
• ISSNs: 1757-8981

Abstract

Abstract: A neural network based modelling and steering control method of an intelligent vehicle under dynamic sensitive conditions is proposed in this paper. Four radius function neural networks are used to compensate the modelling error of a traditional dynamic vehicle steering model. The designed control law is consisted of four terms: a neural network term, a proportion term, an integral term and a robust term. Stability is analysed through Lyapunov function and the training law is generated at the same time. Test results show that the designed steering controller works quite well and is of great value in path tracking of an intelligent vehicle when the vehicle dynamics getting significant.