Modeling, control, and optimization of autonomous diesel- and electric-powered off-highway vehicles for energy efficiency

Abstract

Off-highway vehicles are heavily relied upon for industrial transportation and havehighly energy-intensive operation. Automation of off-highway vehicles presents an opportunity to reduce energy consumption without negative impact on productivity through optimization. Prior studies on this topic are lacking a systematic formulation of the optimal control problem for automation. This work introduces novel formulations of the optimal control problem for a typical wheel loader drive cycle for both diesel- and electric-powered vehicles. The formulations include development of control-oriented mathematical models, physical constraints, boundary constraints to create the desired cycle, and multi-objective cost functions. Simulations of the optimized trajectories demonstrate feasibility of implementation and allows deeper analysis of proposed reductions in energy. This work finds that optimal automation for a diesel wheel loader can reduce energy consumption by 42.1% while matching or improving that productivity of a human driver. For electric wheel loaders, this work extends the feasibility of electrification to larger sizes than those found in literature while matching diesel per-day productivity with 1.2 battery charges compared to human drivers and 1.8 battery charges compared to an optimal automated diesel vehicle while demonstrating 74% less energy consumption. The diesel-powered vehicle reduces energy by optimizing the engine operation point via coordination of the driving and working systems, while the electric-powered vehicle further reduces energy consumption with a more efficient architecture. Overall, this work demonstrates the potential for automation of off-highway vehicles to reduce energy consumption, provides a systematic methodology for optimizing automation and evaluating its benefits, and gives a pathway to electrification of mid-size wheel loaders using automation.