Study On Energy Efficiency Optimization For Electric Vehicles

With the increasingly serious energy and environmental issue, battery electricvehicles (BEVs) are considered to be one of the effective solutions in the transport fielddue to the characteristics of zero emission and zero fuel consumption. The low mileageissue has become a key factor in limiting their rapid development. Therefore, energyefficiency optimization becomes an urgent research topic for electric vehicles.Regarding convex optimization and multi-objective optimization as research tools, thisdissertation focuses on five issues from perspectives of charging and discharging,including optimal valley-filling charging strategy for vehicle-grid system,multi-objective charging optimization in device level, optimal torque distributionstrategy for electric drive system, optimal sizing of powertrain and regenerative brakingstrategy optimization.To address on the confliction issue of charging control between EVs and the grid, aday-ahead pricing scheme is designed by solving a minimum-cost optimization problem.EV owners respond to the pricing scheme by autonomously optimizing their individualcharge patterns. This device-level response induces a valley-filling effect in the grid atthe system level. The proposed strategy offers three advantages: coordination (by thevalley-filling effect), practicality (no requirement for a bidirectionalcommunication/control network between the grid and EV owners), and autonomy (usercontrol of EV charge patterns). The proposed strategy achieves the valley-fillingcharging effect at28%less generation cost than the uncoordinated charging strategy.To address on the issue of conflicting optimization objectives between chargingcost and battery life, a P2D-based SEI formation model is simplified and improved inorder to describe the battery degradation quantitatively. By solving the correspondingmulti-objectives optimization problem, the Pareto optimal charging patterns areintroduced to describe the trade-off between the two conflicting objectives.To address on the issue of torque distribution optimization, the input-outputcharacteristic of a type of PMSM is proved convex based on the analysis of its powerloss model. In the spatial torque distribution optimization, the optimal “equivalent lossgradient distribution” strategy is derived for EVs with different PMSMs and optimal “equivalent torque distribution” strategy for those with identical PMSMs. In thetime-domain torque distribution optimization, i.e. the P&G (pulse and gliding) drivingoptimization, the optimal “constant-velocity” strategy is theoretically derived bytransforming an optimal control problem to a convex optimization problem. In addition,an improved P&G strategy is proposed for electric drive systems with clutch, which has8.5%of the energy-saving potential.To experimentally verify the research results on P&G driving strategies, a vehiclecontrol system is designed and developed for an electric vehicle with single PMSM.Two key issues which have great impact on the system energy efficiency are studiedduring the development process, including optimal powertrain sizing problem andoptimal regenerative braking strategies problem. The Pareto fronts are introduced tobalance the conflicting objectives of powertrain sizing optimization. Severalregenerative braking strategies with both good braking efficiency and brakingperformance are proposed for front-wheel and rear-wheel electric vehicles. Theconstraints and the scope of application of every regenerative braking strategy are cleardefined.