Research On Driving Condition Adaptation Based Energy Management Strategy For Plug-in Hybrid Electric City Buses

As the hot spot of current research and application,plug-in hybrid electric vehicles(PHEVs)combine the merits of both conventional hybrid and pure electric vehicles,with remarkable energy saving and emissions reduction effect.Energy management strategy is one of the key technologies for PHEVs,which has a significant effect on its fuel economy.Existing widely used on-board energy management strategies are based on fixed rules,which lack adaptability to the complex and time-varying real-world driving conditions and cause the PHEV cannot give full play to their potential fuel savings.In order to solve the shortage of existing on-board strategies and improve their adaptability to the real-world driving condition and fuel economy,this paper studies the driving condition adaptation based PHEV energy management strategy.The research content is as follows.1)The key problems of driving condition based PHEV energy management optimization were studied.With a new structure four-mode plug-in hybrid city bus as the research object,a vehicle simulation model has been developed for the energy management algorithm research.Energy management optimization methods for the four-mode PHEV were developed based on dynamic programming(DP),Pontryagin’s minimum principle(PMP),and the equivalent consumption minimization strategy(ECMS).Research confirmed that the optimal state of charge(SOC)depleting trajectory is the key of charge-depleting energy management optimization,and the optimal equivalence factor of ECMS is the key of charge-sustaining energy management optimization of PHEV under various driving conditions.2)A driving condition based SOC control trajectory planning model was proposed for charge-depleting energy management optimization of PHEVs.The optimal battery power consumption law was studied using DP method based on a variety of typical driving conditions.The driving condition features that have the most influence on the optimal SOC trajectory were pointed out,by utilizing the methods of correlation coefficient analysis.Then,the SOC trajectory planning model was proposed based on average velocity,velocity deviation and net drive energy demand.Accordingly,the relationship to map the optimal SOC control trajectory and the driving condition was established,which laid a foundation for the driving-condition-oriented optimization of PHEV charge depletion control.Tests based on comprehensive cycles that have significant driving condition variations showed that the average difference between the optimal SOC trajectory and that planned by the proposed method is less than 3.57%.It’s significantly better than the mileage-oriented SOC linearly planning method.And it’s suitable for vehicle application while other global optimization methods cannot meet the real time requirements of on-board control.3)As for the charge-sustaining energy management optimization problem of PHEV,an optimal equivalence factor approximation method was proposed for equivalent consumption minimization strategy based on two novel driving condition parameters,named as vehicle indicated energy demand and energy-based vehicle torque-speed statistical distribution.Compared with the offline calculated optimal equivalence factors,the maximum approximation error of the proposed method is within ±1.6%,which solves the problem of driving condition based accurate estimation of the optimal equivalence factor.4)A driving condition forecasting method for different driving scenarios was established.Realized route recognition and driving-condition-based trip deviation based on the information from GPS and PHEV operation status。As for commuter driving,an instance based driving condition self-learning and prediction method was established.Tests showed that the proposed method can provide an effective prediction of driving condition parameters that used for energy management optimization,such as average velocity,velocity standard deviation,net drive energy demand,indicated energy demand etc..The 25% to 75% distribution range of prediction error is within ±5% and the maximum error is less than ±15%.Compared with the existing research,the proposed method in this paper can autonomously learn and predict the condition without additional external equipment.What’s more,the method can be used for control optimization of other types of vehicles,with huge potential application value.As for random driving,a driving condition forecasting method was presented with information from GPS/GIS and intelligent transportation system(ITS).And a net driving energy demand forecast model based on the support vector regression algorithm was established.Tests showed that the maximum prediction error is less than 6.76%,while the prediction time consumption in 50 MHz vehicle controller is no more than 0.34 s.It solves the difficulty of accurate prediction of driving conditions in real-time.5)A driving condition adaptation based PHEV energy management optimization strategy was developed,which includes two sub-strategies.One is the optimized charge depleting(OP-BCD)strategy,which uses the forecasted driving condition features for SOC control trajectory planning.Adaptive ECMS,combined with an equivalence factor feedforward and feedback controller,was used to realize a driving condition adaptation based real-time control by tracking planned SOC trajectory.The second one is the optimized charge sustaining(OP-CS)strategy.It uses forecasted driving condition features to approximate the optimal equivalence factor and realizes real-time adaptive control for charge sustaining.The proposed energy management optimization method fulfills the requirements of vehicle real-time control and the hardware performance limitation of the vehicle controller.Compared with existing on-board energy management strategies,the proposed method can effectively improve the PHEV fuel economy under complex real-world driving conditions.Compared with the optimized on-board rule-based energy management strategy,hardware-in-the-loop tests showed that the proposed strategy can improve the fuel economy up to 7.98%,vehicle tests showed the fuel economy can be improved up to 6.22%.