Model-free speed management for a heterogeneous platoon of connected ground vehicles

Abstract

Motivated by military applications, this work considers connected platoons of ground vehicles of potentially different sizes and presents a model-free approach for optimizing the speed of the platoon to adjust the tradeoff between fuel economy and mobility as measured by travel speed. The motivation to seek a model-free solution is twofold: (1) vehicle models that are typically assumed to be available in model-based methods are not available on-board for military vehicles; (2) a model-free solution can offer robustness to modeling errors. Therefore, in this paper, the problem of optimizing the tradeoff between fuel economy and mobility of a mixed platoon is formulated as an optimization problem and solved using the model-free Nelder-Mead approach. To explore the performance characteristics of this approach, a case study is performed with two different size vehicles that are representative of military trucks, both in simulation and in a novel networked engine-in-the-loop setup. The results show that the proposed approach can achieve the desired balance between fuel economy and mobility in a model-free manner despite the nonlinearity caused by gear shift deadzones, albeit at the expense of relatively slow convergence time. In addition, a design guideline for the parameters in the Nelder-Mead approach is also discussed.