Hybrid technologies are considered one of the key solutions for CO2 reduction in large off-highway machines that operate for long hours in suburban areas with limited access to energy refueling infrastructure. The effectiveness of CO2 reduction through hybridization largely depends on the hybrid powertrain architecture and its operating strategies. To explore the potential of hybrid powertrains for off-highway applications, this thesis uses a wheel loader as a research platform. In the first part of the thesis, the energy flow in the powertrain during a typical wheel loader operating cycle (V-cycle) is analyzed. A parallel-series hybrid powertrain and its operating strategies are also presented. The benefits of hybridization in wheel loaders are evaluated using both test results from a prototype wheel loader and simulation data. The results show that fuel consumption can be reduced by 32.7% with a rule-based energy management strategy compared to the conventional powertrain. This reduction includes 20.4% from the electrification of the drivetrain and 6.5% from improved system operation, the latter of which could potentially be further enhanced with advanced energy management strategies. In the second part of the thesis, a more sophisticated energy management strategy based on Dynamic Programming (DP) and Pontryagin’s Minimum Principle (PMP) is adopted to fully exploit the potential of the parallel-series hybrid powertrain. The two control inputs – the engine start-stop decision and the optimal torque split between the engine and the P2 electric machine, which corresponds to the engine torque setpoint – are determined using DP and PMP, respectively. The engine torque control using the optimal torque split is implemented as an Equivalent Consumption minimization Strategy (ECMS). To enable real-time implementation of the proposed control policy, short-term machine power demand is predicted using a state detection algorithm developed to account for the repetitive nature of a wheel loader’s duty cycles. Simulation results indicate that fuel consumption can be reduced by 4.0% compared to the rule-based strategy in the V-cycle. The proposed control algorithm was implemented and validated on the prototype hybrid wheel loader. Test results confirm that the DP-PMP-based control strategy operates in real-time and enhances system efficiency. Specifically, the average engine brake specific fuel consumption (BSFC) improved by 0.7% in the V-cycle and by 1.3% in a more realistic mission cycle that combines the V-cycle with driving.
Yuki Kakichi (Thu,) studied this question.