Dual-Mode Adaptive Control With Entropy-Triggered Reactivation for Heavy-Haul Trains

Abstract

This paper addresses the tracking control problem for automatic train operation of heavy-haul trains (HHTs) under uncertain operating environments. To mitigate the limitations of conventional simplified modeling and the high computational burden from continuous adaptation, this work presents an integrated control framework with three synergistic advancements. First, a refined longitudinal dynamical model is constructed that differentiates four subunit types, i.e., leading locomotive, inner locomotive, rear locomotive, and inner wagons, with explicit bounded coupler-deformation constraints. By capturing practical compressive–tensile characteristics and locomotive–wagon dynamic interactions, the model delivers more realistic HHT dynamic descriptions and lays a reliable basis for subsequent controller synthesis. Second, a dual-mode adaptive controller is developed to jointly improve adaptability and computational efficiency. Under satisfied triggering conditions, adaptive parameters can be frozen and reused; adaptation is reactivated once performance degradation is perceived, enabling robust and economical operation for varied uncertain scenarios. Third, a sliding-window entropy-triggered reactivation function (ETRF) is proposed to supervise the transitions between stationary and nonstationary adaptive modes. The ETRF comprehensively evaluates speed-tracking errors, mileage errors, and parameter convergence states and embeds a minimum dwell-time constraint to suppress chattering and ensure stable mode transitions. Simulation studies validate the overall effectiveness and practical merits of the proposed integrated scheme for HHT control.