Research And Implementation Of Traction Battery Management System On Fuel Cell Electric Vehicle

In modern society, increasingly serious environmental and energy issues require a new generation of vehicles to high energy efficiency, low emission direction. Fuel cell electric vehicle (FCEV) has received wide attention because of its advantages of zero-emission, low noise, but still needs further breakthrough. The main factors that restrict the development of FCEV are power battery and its application technology. Battery management system (BMS) as a key component of FCEV, its good work can not only improve the efficiency of the battery, but also guarantee the safety of use and extend battery life. This article dedicated to the research on the design of battery management system and battery state of charge (SOC) estimation.In this paper, the characteristics of Ni-MH battery were described, the main factors of battery SOC were studied, second-order RC models of nickel-hydrogen battery were established, which can well describe the dynamic characteristics of nickel-hydrogen battery, and the model parameters were obtained through calculation and analysis of experimental data.Hardware including the upper control board and the underlying data acquisition board, used distributed design based on CAN bus. Upper control board completed the CAN communication, data acquisition, SOC estimation and fault alarm, etc., TMS320LF2407 with data processing and control functions as the main chip; Underlying data acquisition board used PIC18F248 as the main chip, completing voltage monitoring, temperature measurement, CAN communication. In view of the shortcomings of the battery pack voltage measurement signal with high common-mode voltage, circuit complexity and big test error, built using discrete components, the design of the battery voltage acquisition circuit used integrated multi-cell battery monitoring chip LTC6802, completing accurate measurement of single voltage (0.25%of the maximum total measurement error) and automatic overcharge protection, undervoltage and overvoltage alarm.In this paper, current integral method and the Kalman filter (Kalman) method were combined to estimate the battery SOC value, which solved the problem that the noise value can not be obtained in Kalman filter method. The noise value in Kalman filter method was calculated by current integral method, used for the amendment of predicted SOC value, so to get the exact battery SOC estimation. In the software design, programme of each module was compiled, to achieve the main function of the battery management system.