Fuel consumption optimization of heavy-duty vehicles using genetic algorithms
Paper i proceeding, 2017

The performance of a method for reducing the fuel consumption of a heavy duty vehicle (HDV) is described and evaluated both in simulation and using a real HDV. The method, which involves speed profile optimization using a genetic algorithm, was applied to a set of road profiles (covering sections of 10 km), resulting in average fuel savings of 11.5% and 10.2% (relative to standard cruise control), in the simulation and the real HDV, respectively. Here, a compact representation of road profiles in the form of composite Bézier curves has been used, thus reducing the search space for speed profile optimization, compared to an earlier approach. In addition to outperforming MPC-based methods commonly found in the literature by at least 3 percentage points (in similar settings), the results also show that our simulations are sufficiently accurate to be transferred directly to a real HDV. In cases where the allowed range of speed variation was restricted, the proposed method outperformed standard predictive cruise control (PCC) by an average of around 3 percentage points as well, over the same road profiles.


Sina Torabi

Chalmers, Tillämpad mekanik, Fordonsteknik och autonoma system

Mattias Wahde

Chalmers, Tillämpad mekanik, Fordonsteknik och autonoma system

2017 IEEE Congress on Evolutionary Computation, CEC 2017 - Proceedings