Energy management and component dimensioning of hybrid vehicles require powerful, efficient and representative optimisation tools. Classic dynamic programming is generally used to benchmark on-line strategies. This paper shows that such numerical algorithms often give non-representative results in terms of comfort. Since adding a comfort criterion requires another state variable, it increases the computational burden. Comfort is therefore often neglected in optimal algorithms, which leads to overestimating the potential of the hybrid vehicle. This paper proposes two methods for implementing comfort constraints into dynamic programming algorithms applied to hybrid vehicle energy management. These comfort constraints, applied to the gear shifting control signal, yield results that are more representative of the gear shifting policy of an actual passenger car. Dynamic programming results can thus provide a better reference in order to benchmark suboptimal energy management solutions, especially online strategies. Implementing comfort constraints inherently leads to extra fuel consumption in comparison with basic dynamic programming, regardless of the method considered. However, this overconsumption rarely exceeds 2% for the two methods (when considering reasonable criterion decreases). The comparison between the two methods has revealed that for the same comfort improvement, the extra fuel consumption o( the suboptimal method was in the worst case fairly acceptable. As for the computational effort, implementing comfort constraints using the optimal method led to computation times that on average were 20 times longer than those required by basic dynamic programming, whereas the suboptimal method ensures computation times that are equivalent to those required by basic dynamic programming. This method could therefore be extended to implement constraints on the torque split factor, pollutant emissions etc., without extending the computation time.
Implementation of comfort constraints in dynamic programming for hybrid vehicle energy management
Anwendung der Komfortzwänge bei dynamischer Programmierung von Energiemanagement der Hybridfahrzeuge
International Journal of Vehicle Design ; 58 , 2-4 ; 367-386
2012
20 Seiten, 10 Bilder, 17 Quellen
Aufsatz (Zeitschrift)
Englisch
Implementation of comfort constraints in dynamic programming for hybrid vehicle energy management
Online Contents | 2012
|Energy Management Strategy of Hybrid Electric Vehicle using Stochastic Dynamic Programming
SAE Technical Papers | 2015
|Energy Management Strategy of Hybrid Electric Vehicle using Stochastic Dynamic Programming
British Library Conference Proceedings | 2015
|Hybrid electric vehicle energy management control method based on random dynamic programming
Europäisches Patentamt | 2022
|