Assessing the potential of prediction in energy management for ancillaries in heavy-duty trucks

Abstract

The degree of importance of prediction in the control of ancillary systems for conventional heavy-duty trucks is investigated, with focus on fuel economy. An optimal control law that utilizes prediction is compared with a suboptimal causal control law. The incentive for this investigation is that the suboptimal control law is less complex to develop and implement for the considered system. The results are not general since only a limited amount of ancillary systems have been modeled, only two different drive cycles have been evaluated and simplified mathematical component models for a specific test truck have been used in the optimization problem. Nevertheless, preliminary results from this investigation indicate that simple suboptimal control laws can yield close to the same improvements in fuel efficiency as a predictive controller, when compared to a baseline control law. Tuning complexity is expected to rise when more ancillaries are included in the energy management problem. This can be a valid argument to incorporate prediction in future work.