T‐S fuzzy model based generalized predictive control of vehicle yaw stability

Abstract

Purpose -The purpose of this paper is to design a steering control for vehicles to protect the vehicle from spin and to realize improved cornering performance. Design/methodology/approach -The improved cornering performance is realized based on Takagi-Sugeno fuzzy model and generalized predictive control (GPC). A new approach to establish model of the vehicle is presented on the basis of fuzzy neural network. The network which inputs and outputs are composed of five layers of forward structure is utilized to build the structure and parameters of T-S fuzzy model through learning from training data. In this way, the vehicle dynamic system is divided into many linear sub-systems, and the system output is the weighted-sum of these sub-systems' outputs. A CARIMA model can be derived from the presented fuzzy model, and GPC is applied to deal with the control problem of vehicle stability. Findings -Vehicle model can be divided into local linear models, corresponding controller can be developed. Simulation results show that fuzzy model based on GPC can be applied to improve stability of the vehicle effectively. Research limitations/implications -As an exploration of a new approach, the training data are from simulation, and the result of the paper will be applied in actual vehicle trials. Practical implications -The paper presents useful advice for developing a vehicle stability controller. Originality/value -The paper presents a new approach to establish a model of the vehicle on the basis of fuzzy neural network, which is valuable for establishing a new controller for vehicle stability.

Stability Constraints for Robust Model Predictive Control

“…A large diversity of real processes can be modeled using CARIMA models, for instance, unmanned seaplanes [29], stirred tank reactors [30], vehicle yaw [31], gas engines [32], and distillation columns [33,34]. Therefore, the development of control design techniques that are suitable for those models is relevant.…”

“…Proof. The proof of this theorem is stated as the sequence of expressions, from expression (25) to expression (31).…”