Study On Energy Management Strategy Of Series-parallel Hybrid Electric Vehicles Based On Driving Condition Recognition

Hybrid electric vehicles have a number of power sources, how to manage the energy coordination between these power sources to a large extent determines the performance of the vehicle. Thus, the energy management strategy of hybrid vehicle is one of the key technologies in the research field of hybrid electric vehicle. Currently, the car used control strategy failed to consider the effect of vehicle driving cycle, in view of this, this paper take the Toyota Prius as the object to do a research on hybrid electric vehicle energy management strategy based on driving conditions identify.(1)According to the structure and related parameters of hybrid electric vehicle, Models of the engines, the motor, the battery, the planetary gear mechanism, and other major components of the vehicle are established on the simulation platform MATLAB software, Then constitutes a series-parallel hybrid electric vehicle simulation model.(2) The energy management strategy of hybrid electric vehicle based on the logic threshold value is established by the threshold parameters value of the demand power and the SOC value of the battery etc. Secondly, in order to evaluate the advantages and disadvantages of the logic threshold energy management strategies, the global dynamic programming optimization algorithm was introduced to simulate the vehicle under the special conditions, and the result is used to evaluate and optimize the logic threshold energy management strategy and based on recognition of mode of energy strategy in the fuel economy.(3) In view of the typical working conditions of the hybrid vehicle, an online identification algorithm is completed. Through cluster analysis, five typical conditions were selected from the advisor condition library as working condition recognition library. The characteristic parameters of different typical operating conditions are calculated and compared with the current working conditions. Identify the type of driving status of the current vehicle. In real time, one of the typical operating conditions of the vehicle is classified as a typical operating mode, which provides the basis for the vehicle energy management strategy based on the condition identification.(4) By learning the neural network algorithm and the usage of Neural Network Toolbox in MATLAB. Firstly calculated the offline optimal energy management policy of these five typical driving cycles by the dynamic programming algorithm, than the offline optimal energy management strategy of five typical driving cycles are extracted by Neural Networks, Thus Five energy management strategies corresponding to the five typical working conditions are established. Finally, the classification of the vehicle energy management strategy is switched to the category of the current running mode according to the identification strategy.Through the above research work, a better method for series-parallel hybrid electric vehicle energy management strategies is designed. It has positive effects on improving the fuel economy of hybrid electric vehicle.