Study Of Energy Consumption Optimal Control Strategy For Axle-split4WD Hybrid Electric Vehicle

Hybrid Electric Vehicle (HEV) energy management control strategy grabs muchattention for a long time as a key technology of HEV. This dissertation focues on energyconsumption minimization using an Axle-split Four Wheel Drive Hybrid Electric Vehicle(A4WDHEV) as a case study. Equivalent (fuel) Consumption Minimization Strategy (ECMS)has been improved with respect to optimization method and driving condition recognition.Firstly, through A4WDHEV structure and vehicle mode shift control analysis, the HEVand its key components model has been built on theoretical and experiment modelling. ADynamic Programming based Control Strategy (DPCS) for A4WDHEV is designed tooptimize vehicle energy consumption. Simulation results indicate the theoretical fuel savelimit while providing database for other control strategy improvement.After a brief introduction of the basic theory of Equivalent fule ConsumptionMinimization Strategy (ECMS), the equivalent factors implication and equivalent fuelconsumption penaty method are discussed systematically. The ECMS is applied toA4WDHEV for energy save optimization with numerical computation acceleration methodsuch as pattern search and parallel computation. The simulation results under multiple drivecycles show that ECMS can achive above20%fule saving compare to engine-onlyconventional vehicle.The influence to ECMS caused by equivalent factors variation and the correction forSOC maintenance is analysed. A new optimal method for equivalent factors calculationunder certain drive cycle is proposed. This method refers to DPCS optimal control outputand uses DIRECT algorithm to solve the optimization problem. A SOC based corcttionfunction optimization method is provided to keep the advantage of fule save while correctingequivalent consumption to maintain SOC. Then, present a new Global optimization basedECMS (GECMS) according to these two methods sutdied above. Simulation results showthat the fuel consumption using GECMS can be decreased by about6.8%under multipledrive cycles compared to ECMS.An Adaptive GECMS (AGECMS) is proposed to deal with the fact that ECMS has poordrive cycle adaptability.Four typical Guangzhou drive cycles are formed through real roadtest data acquisition as the example data for LVQ neural network drive condition identifierdesign. After training, the LVQ identifier can achive about100%identify accuracy. Add theroad condition identifier to GECMS, an Adjustable Global optimization based Equivalent Consumption Minimization Strategy (AGECMS) is put forward. Simulation software isdeveloped for the comparison study of DPCS, ECMS, GECMS and AGECMS forA4WDHEV energy consumption optimization. Mutiple drive cycle simulation resultsindicate that AGECMS is2.76%above to GECMS with respect to fule saving.The overallfuel saving is9.5%compared with original ECMS in a drive-cycle-average sense.Finally, related to an industrial A4WDHEV sedan development project, the GECMSHardware In Loop (HIL) simulation is conducted for evaluation of the control effectiveness.Some core parts of the original HEV HCU software are modified to realize the GECMS.After the construction of HIL platform, a comparison test is carried out for GECMS andoriginal HEV strategy. Under circumstances that the optimization effect is restricted to theHCU software architecture, the test results show that GECMS increase the fuel economy by3.83%compared with the original HEV strategy.