Bioinspired Reference Model and Fully Actuated System Approach-Based Neuroadaptive Control for Uncertain Active Suspension Systems With Input Dead Zones

In this article, we propose a neuroadaptive control method for quarter-vehicle active suspension systems, incorporating a bioinspired reference model and a fully actuated system approach, specifically designed to address systematic uncertainties and input dead zone constraints. Specifically, the radial basis function neural networks are employed to approximate the systematic uncertainties and input dead zones. Inspired by the inherent features and advantages of fully actuated systems, this article proposes the first fully actuated system-based continuous adaptive control approach for active suspension systems. Moreover, a biologically inspired, bioinspired reference model is deliberately employed to take advantage of favorable nonlinear stiffness and damping effects. Different from existing fully actuated system approaches, the designed control method can guarantee the state converges to zero rather than a small region. The experimental results confirm that the designed control method achieves significant enhancements in transient performance, with improvement of up to 31% or more, and also results in control energy saving of up to 32% or more when compared with existing control methods.