Study On Control Strategy Of The Extended-Range Electric Vehicle

The development of the automotive industry and the increaseing in car ownership bring ahuge challenge to the world’s energy and environment. Development of new energy vehiclesis the fundamental way to solve the problem. Due to the battery technology in the short termis difficult to solve life, cost and energy density etc, which lead to the electric vehicle shortdriving range, high cost and difficult to industrialization. Range-Extended electric vehiclewith its own advantages becomes promising new energy automotive products in latest stage.What mainly discussed in this paper is the Range-Extended electric vehicle controlstrategy: choose the key components of Range-Extended electric vehicle, analysis of vehicle’swork mode and complete the preliminary design of control strategy based on matlab/simulink. Build simulation model of vehicle by AVL-cruise software and complete the offlinesimulation. Optimize the control parameters in Multi-objective global optimization by usingparticle swarm optimization algorithm. Study on control strategy with condition recognitionbased on LVQ neural network combined with the self-learning function of neural network.The main contents of the thesis as follows:Firstly, introduce the classification and features of HEV. Research the Present situationof Range-Extended electric vehicle and its control strategy. Introduce the main content of thispaper.Study on the extended-range electric vehicle power system parameters matching.According to the dynamic performance, choose the engine, generator, drive motor, battery andtransmission ratio parameter. Build simulation model by using Cruise software. Verify therationality of matching results by dynamic performance simulation.Focus on analysis of the working condition and control process. Engine uses three workpoint control strategy based on vehicle speed. For the braking condition, according to theexisting distribution algorithms of the motor braking force and friction braking force, putforward a new regenerative braking force distribution control strategy based on the intensityof the drive motor. Complete the preliminary design of the vehicle control strategy model byusing matlab/simulink and complete simulation.Study on the interface program of advisor software with optimization program. Useparticle swarm optimization algorithm, get the global optimal control parameters based on thespecific driving cycle and fuel economy, emission as optimization target. The simulationresults show that the control performance is greatly improved.Study on the adaptive control strategy of LVQ neural network. Get five groups control optimization parameters based on the five kinds of implication operating conditions from thelow-urban speed to high-outskirts speed. From five operating speed, acceleration and otherparameters as samples, training LVQ neural network, get the weight matrix. Establishworking condition recognition model then embed it into the model of vehicle control strategy.The simulation results show that the adaptive control strategy can automatically identify thetype of conditions and choose optimal control parameters to improve fuel economy andemissions.