Multi-Gradient Optimization for MiFREN-Based Data-Driven Sliding Mode Platooning Control

This paper explores the application of the distributed multi-input fuzzy rule emulating network (MiFREN)-based data-driven sliding mode platooning control (MB-DDSMPC) for vehicle platooning systems (VPSs). Initially, the nonlinear VPSs are transformed into a linear data model by using the compact format dynamic linearization technique. Furthermore, considering the unknown nonlinearities arising from the operation of VPSs, a MiFREN approach and an extended state observer are introduced to estimate the nonlinearities, respectively. Additionally, the multi-gradient descent method is incorporated to optimize the weight updates for the MiFREN algorithm, further improving the accuracy and adaptability of the proposed control scheme. To enhance the robustness of VPSs, a sliding mode surface is constructed for the consensus error, and an MB-DDSMPC scheme is proposed to address the robust tracking problem for VPSs. Meanwhile, in order to achieve the asymptotic stability in the velocity of VPSs, we determine the threshold values for the sliding mode parameters by constructing a Lyapunov function with respect to the consensus error. The effectiveness of the designed MB-DDSMPC scheme is demonstrated through simulation comparisons.