Neural Network-Interacted Robust Coordinated Control of Full-Vehicle Active Air Suspension With Uncertainties and Time Delays

Abstract

This paper proposes a neural network-interacted robust coordinated control (NNIRCC) scheme to address the problem of full-vehicle active air suspension (AAS) systems subject to uncertainties and different time-varying actuator delays. The NNIRCC scheme consists of neural network-interacted (NNI) approximator, projector-based estimator and robust coordinated control term. The NNI approximator based on the radial basis function is employed to capture the nonlinearities caused by the unmodeled dynamics of adjustable air spring. Meanwhile, an interactive updating algorithm is designed to manipulate the weights of the NNI approximator so as to improve the approximation accuracy. Moreover, projector-based nonlinear estimators are designed to handle the prevalent sensitive parameter variations (such as vehicle body mass and its moments of inertia). Furthermore, delay compensators are developed and integrated into the synthesized coordinated control law to mitigate the impact of different time-varying input delays caused by force actuators. The asymptotic stability of closed-loop system is rigorously proven by employing a Lyapunov-Krasovskii functional, guaranteeing the boundedness of both tracking and estimation errors within a finite time. Additionally, co-simulation results are provided and analyzed, illustrating the feasibility and efficiency of the proposed NNIRCC scheme.

Keywords

• Full-vehicle active suspension system
• adaptive robust control
• iterative learning approximator
• uncertainties
• delay-free coordinated laws