The present investigation proposes an on-board device for the prediction and the optimisation of the energy management of HEVs (Hybrid Electric Vehicles) applied to the ITAN500 prototype. The prediction is performed with a local instance of the SUMO (Simulation of Urban Mobility) traffic simulator and a network simulator for the exchange of information between the ITAN500 and the surrounding vehicles and infrastructure. The results of the numerical and experimental validation presented here confirmed the possibility to achieve a good prediction of the future speed profile of the vehicle in the next 60s with the proposed method based on V2V and V2I communications. The optimisation was performed with the use of pre-loaded optimised maps containing the optimal values of the energy management parameters according to the predicted cycle, fuel tank level, battery SOC (State of Charge) and engine temperature. The power request cycle of ITAN500 in the next 60 s is predicted and then associated to the most similar of the 73 RMDCs previously identified with the use of a clustering algorithm. Three kinds of driving cycles were taken into account for the clustering algorithm. The first two are standard driving cycle adopted for the registration on new cars (NEDC (New European Driving Cycle) and UDDS (Urban Dynamometer Driving Schedule)). Cycles ≠A and ≠B were obtained from simulating random routes of ITAN500 in the Ecotekne campus with the traffic simulator SUMO. Finally, cycle ≠R is a real cycle acquired on ITAN500 with a GPS system along a specific route in the same campus. For each driving cycle, the fuel consumption, the battery usage and the equivalent fuel consumption have been evaluated in three hypothesis. The first one assumes that the cycle is entirely known so that a control optimisation can be performed over the whole cycle. The second approach consists in applying the proposed prediction-RMDCs method. The last one consists in optimising the energy management strategy for the NEDC cycle and then applying it without modification to the other cycles. This corresponds to no knowledge of the driving cycle of the vehicle. The analysis of the results confirmed that the third approach is not useful since the fuel consumption and the battery usage in this case is the highest for all the cycles (except of course driving cycle NEDC) while the proposed methodology has shown to be successful. In fact the consumption of fuel and energy stored in the batteries is very close to the control optimisation case (full knowledge of the cycle) particularly for cycle A and B. Satisfactory results are also obtained for cycle R.
A method for the prediction of future driving conditions and for the energy management optimisation of a hybrid electric vehicle
Vorhersage der künftigen Fahrbedingungen sowie die Energiemanagement-Optimierung eines Hybridfahrzeugs
International Journal of Vehicle Design ; 58 , 2-4 ; 111-133
2012
23 Seiten, 13 Bilder, 7 Tabellen, 19 Quellen
Article (Journal)
English
Automotive engineering | 2012
|Energy Management in a Parallel Hybrid Electric Vehicle for Different Driving Conditions
SAE Technical Papers | 2014
|Approaches to optimisation of energy management in a hybrid electric vehicle
Tema Archive | 2007
|Energy Management in a Parallel Hybrid Electric Vehicle for Different Driving Conditions
British Library Conference Proceedings | 2014
|