The Research Of Battery SOC Estimation For Plug-In Hybrid Electric Vehicles

As we all know, energy shortage and environmental degradation problems are becoming increasingly serious. Vehicle exhaust emission has become a major source of environmental pollution and greenhouse gas emissions. The development of new vehicles is facing unprecedented challenges. Under this background, hybrid electric vehicle (HEV), which combines the advantages of conventional fuel vehicle and electric vehicle, becomes the main flow of automobile industry for its high fuel economy and low emissions in the 21st century. In recent years, plug-in (Plug-In) hybrid electric vehicle (PHEV) is emerged on that basis. It can use the power grid to charging the vehicle battery. Compared with the traditional hybrid vehicles, PHEV has many advantages:low-cost, minimal emissions and it can raise the utilization rate of power. Plug—In Hybrid Electric Vehicle produced great social and economic benefits.PHEV's main energy source is the battery, compared to the traditional HEV, the battery of PHEV has greater capacity. The residual capacity of battery is characterized by battery state of charge (SOC) which is an important parameter. Accurately estimating the battery The SOC value is not only an important basis to develop energy management strategy (EMS), but also the precision of estimate affect the life and cost of battery. SOC estimation is difficult to test accurately due to the nonlinearity of batteries. It has become a bottleneck restricting the development of electric vehicles. Therefore, the study of the PHEV battery SOC estimation is essential for the promotion of industrialization PHEV. In this paper, SOC estimation of PHEV is researched, and the main contents are described as follows:Firstly, this paper gives the development backgrounds and classification of HEV and the improve situation and characteristics of PHEV. Then the foreground of PHEV, the key technologies of PHEV and the main methods, present problem, improvement of SOC estimation are analyzed emphatically. Obtaining accurate battery SOC is one of the important prerequisite to achieve optimal PHEV energy management and optimal control of vehicle. In allusion to the deficiencies of traditional SOC estimation, and conside the factors of battery SOC, the present SOC estimation methods are expatiated. RBF neural network and support vector regression are used to gauge battery SOC separately, and estimating performance of these two types of algorithm. Support vector regression uses its two basic algorithms:ε-SVR algorithm and v-SVR algorithm. The simulation results demonstrate that the both algorithms can be very close to the actual value and the average estimation error is less than 5% which satisfy the practical requirements, but the average estimate of v-SVR algorithm is the best.Finally, the control system of SOC estimation monitoring system was designed by TMS320F2812. Based on the application of SOC estimation, the hardware and software of the monitoring system were designed separately. The hardware part includes detection and display circuit, which improved the monitoring system. The detailed design of the control software and the main flow chart were given in the software part.