Abstract
Rugged terrain and abnormal signal transmission in mining areas threaten the stability and reliability of minecart suspension systems. To address this challenge, the reliable control of minecart suspension systems is explored in this article. Different from general vehicle suspension systems, time-varying minecart body mass caused by road-induced vibration and ore loading or unloading is captured by a Markov process. To mitigate the adverse effects of packet dropouts, a redundant-channel compensator based on multisensor data is designed. Furthermore, accounting for the unavailability of partial states, an adjustable partition-dependent switching (APDS) mechanism is developed with observer assistance to achieve desired control performance. Notably, the proposed mechanism removes the crucial requirement of online matrix inversion in nonparallel distributed compensation (non-PDC) controllers, while simultaneously improving the control flexibility and gain-scheduling precision through generalized membership function (MF) partitioning. Moreover, based on a relaxation variable approach, time-varying balanced matrices adapted to the working mode properties are incorporated into the stability conditions. Finally, numerical and hardware-in-the-loop (HIL) experiments demonstrate that the proposed mechanism enhances crucial ride comfort while maintaining strong robustness against both road disturbances and channel noise.