| The environmental pollution and energy crisis has been pushed to the forefront years ago, although the government has launched a series of policies, regulations and the improvement measures to limit the development of industry enterprises, but the effect need long time to work. And people are not only to restrain their own behavior but also to explore new energy, do not maintaining the status and also innovating. And electric cars become a direction to solve the problem of energy and environment, Battery Management System(BMS) is the core part of the electric car and Battery SOC(State of Charge) estimation in the BMS plays an important role.So how to develop a set of battery management system with perfect function is extremely urgent. The BMS need to achieve real-time information of battery, control charge and discharge normally and reasonably, makes the battery performance stability and prolong the life of cycle, makes electric vehicles safe, stable and efficient operation requirements. The first task is to calculate SOC of the battery and it is the most direct response of battery pack about the state and the residual capacity of battery pack, and find out the most direct physical quantities of the heterogeneity of battery pack. Through SOC values can be concluded that how much mile to drive and choose the appropriate balance plan to extend battery life. Today online estimation of SOC is a technical difficulty, there is interference lead to error when battery information collected, and a low estimation accuracy. According to the above problem, this paper mainly research work as follows: 1. The battery work for the establishment of the equivalent circuit modelFirst conduct a pulse power battery performance test, achieve the relationship between SOC and the open circuit voltage, and the relationship between SOC and ohm internal resistance On the basis of the original Thevenin battery model increases a RC network. The two RC network mapping the electrode polarization and concentration polarization phenomenon. According to the voltage recover curve through Matlab to fit the double exponential curve, and verify the reliability of the second order RC network, also calculate the battery under different states of polarization resistance value. 2. Extended Kalman Filter to dynamic estimate SOCThe Kalman Filter state equation spreads as the first order Taylor expansion, achieve the Extended Kalman Filter can estimate nonlinear system. State space equations are established according to the Li-battery model. Joined the battery Health status(the State-of-the Health, SOH) in the process of operation as a factor, the system is more close to the real condition. Using the Extended Kalman Filtering algorithm to estimate the SOC. Estimated value fluctuated near the true value, constant current and constant resistance discharge conditions, estimate relative error up to 0.015%, can be an accurate reflection of the SOC of the battery status in real time. 3. Virtual platform for estimating SOCAcquisition circuit combined with LabVIEW programming achieved the virtual instrument platform for estimating SOC. Including PCI- 6251 data acquisition card with a high precision, the data collected in the form of differential acquisition through Extended Kalman Filter estimation method to estimate the SOC values. Combined with virtual instrument to achieve real-time data acquisition and estimate can be obtained directly, people can observe data intuitively and clearly by the fineness. |
PDF Full Text Request