Safe and Energy-Efficient Control Allocation for Multi-Unit Heavy Vehicles

Doktorsavhandling, 2026

Electrification, increased variation, and the complexity of multi-unit vehicle combinations have created a need for novel control allocation strategies. Traditional vehicle combinations are propelled solely by the towing vehicle—trucks or tractors—with braking forces distributed in proportion to the axle loads, which is commonly the safest approach. Conversely, modern combinations may be propelled by multiple units, with regenerative braking performed by various electrified units, not necessarily in proportion to the axle loads. An optimal approach for allocating propulsion and braking forces among vehicle units is power loss minimization. This thesis presents a hierarchical, distributed control architecture that effectively manages over-actuation by decoupling upper-level combination control from lower-level unit control. This distributed framework minimizes power losses across multiple units, demonstrating that selectively disengaging electric motors reduces overall energy consumption, while integrating state-of-charge control ensures equal depletion rates among batteries across units in an energy-efficient manner.

From an energy-efficiency perspective, concentrating propulsion or regenerative braking on a single vehicle unit may be optimal. However, such allocations can compromise safety by inducing motion instabilities, such as jackknifing and trailer swing. Consequently, ensuring combination stability while allocating unit forces efficiently is essential. This thesis introduces safe operating envelopes for vehicle combinations in both the force and slip domains. Using single-track, two-track, and high-fidelity models, along with phase-portrait analyses, this work systematically establishes adaptive slip limits and safe force boundaries. These envelopes are tested in simulations and real-world vehicle tests, confirming their effectiveness in preventing yaw instabilities. Furthermore, predictive safety barriers—ranging from advanced Model Predictive Control formulations to simpler proportional-heuristic controllers—are developed and tested in simulations, demonstrating promising performance in proactively avoiding yaw instabilities.

Finally, this study compares the evaluated controllers and details the envisioned functional architecture and signal interfaces, establishing a robust foundation for the safe and efficient operation of future multi-unit electrified heavy vehicles.