Optimisation of series electric hybrid wheel loader energy management strategies using dynamic programming

Abstract

Hybrid non-road mobile machines offer a solution to the low efficiency of diesel-powered machines, but their energy management strategies depend heavily on machine-specific duty cycles. This paper uses experimental data from a 5.7-ton diesel wheel loader performing an industry-standard Y-cycle to optimise and compare four control strategies for fuel savings. Dynamic programming is first applied to determine the optimal power split between the diesel generator and battery. The resulting optimal control sequence is then used to tune thermostat control, power follower control, equivalent consumption minimisation (ECMS), and adaptive equivalent consumption minimisation (A-ECMS) strategies. Simulations in MATLAB Simulink, using both measured Y-cycle data and an artificial half-load cycle, evaluate the performance of each strategy. Results show that A-ECMS achieves fuel consumption within 0.37% of the dynamic programming optimum, followed by thermostat and power follower control. Additionally, different loading conditions influence the relative effectiveness of the management strategies.