Computationally Efficient Non-linear Model Predictive Control for Truck Platoons

A truck platoon consists of several trucks moving together at high speeds with relatively small inter-vehicular spacing. In platoon formations, a key trade-off exists between energy efficiency that depends on the inter-vehicular spacing and safety that is related to the stability of the platoon. In an attempt to capture this trade-off, non-linear model predictive control has been proposed, considering detailed dynamics of on-road truck operation. The non-linear optimization problem that manifests has been solved numerically using the CasADi optimal control framework. The controller capabilities to establish stable platooning have been evaluated for typical accelerating and decelerating maneuvers and for a realistic heavy duty highway drive cycle. The computational efficiency of the approach and its efficacy for real time use have also been studied.