Energy-Efficient Predictive Cruise Control in Heavy-Duty Electric Trucks with Optimization of the Penalty Factor as a Function of Payload

Heavy-duty Class 8 battery electric trucks not only offer a significant reduction in greenhouse gas (GHG) emissions compared to conventional diesel trucks but can also provide significant savings in fuel costs. To further enhance the energy and freight efficiency, Predictive Cruise Control (PCC) algorithms can be developed that generate optimal acceleration profiles for the vehicle by minimizing a cost function which combines both energy consumption and deviation from the desired velocity. A critical component of the cost function is the penalty factor, which governs the trade-off between energy use and travel time, which are two conflicting objectives in freight logistics. Selecting an appropriate penalty factor is essential, as freight deliveries are time-sensitive but minimizing energy consumption remains a priority. Moreover, variations in payload significantly affect vehicle dynamics and energy usage, making it critical to adapt the penalty factor to different payload conditions and maintain consistent performance. This study presents a method for optimally selecting the penalty factor for various payload scenarios. A validated powertrain simulator which is calibrated using data from an actual electric truck, was used to conduct 100 simulations across a spectrum of payloads, from no load to fully loaded. The resulting discrete search space of energy and time was used to perform a brute-force (exhaustive) search to determine the optimal penalty factor for each scenario. The proposed algorithm incorporates adjustable weightings of the penalty factor for energy and time preferences. This allows flexibility for the driver or fleet operator to prioritize either objective. The results demonstrate that using a fixed penalty factor is suboptimal for heavy-duty electric trucks. In contrast, the optimal selection of the penalty factor significantly improves consistency across different payloads. A reduction of the variation in travel time to within approximately 3% across all loading conditions was observed. This work shows the importance of adaptive penalty tuning in PCC for real-world deployment in freight applications, ensuring both energy efficiency and timely deliveries under varying payload demands.