Research On The Estimating Residual Capacity Of Lead-Acid Battery In HEV

Issues about environment protection and energy sources security are focusedon the automobile in recent years. Many schemes are proposed for this (for PureElectric vehicle and Fuel Battery Vehicle),Then the times of a quasi-green vehicle—Hybrid electric vehicle comes about and becomes the focus and hotspot ofresearch. It is essential that EV competition were the battery competition for EV,economically,fleetly ascertaining capability of batteries become one of researchhotspots. Capability of batteries is in the final analysis the residual capacity andState Of Charge( SOC) problem. Since the actual SOC of lead-acid battery isinfluenced by many factors,such as temperature,cycle of discharge and aging ofbattery,it is difficult to obtain ideal results by traditional methods. The applicationof fuzzy logic and ANN method for the SOC estimation is described. The researchachievements and innovations in this paper are as follows:1. Analyzed the principle of the reaction and special characteristic of theLead-acid batteriesLead-acid battery is always the first choice of Hybrid Electric Vehicle (HEV)due to its advantages such as high Power Level , long life, low price and highperformance. Having analyzed the electronic mechanism and special characteristicof the Lead-acid batteries during charge and discharge detailed,obtained the chargeand discharge figure, the result is the current, open circuit volt, terminal volt,temperature and unbalanced between two batteries are all affected the residualcapacity and state of charge (SOC). Therefore, estimating the residual capacity ofbattery is very difficult as well as practical.2. Researched and compared methods of estimating the lead-acid batteryresidual capacityDue to the complicated performance of battery in the different workingsituation, how to estimating the residual capacity and determining the state ofcharge (SOC) of lead-acid battery is the main reason restraining electric vehiclepractical.Principle of the reaction of the lead-acid battery,the characteristics andexisted problems of different kinds of state of charge measuring methods arecompared,such as useing voltage or resistance of the batteries to get the residualcapacity,ampere-hour method by computing electric current, etc. Then put forwardthe way of fuzzy logic and ANN to estimate the residual capacity of the batteries inthis research. Combining fuzzy logic and ANN has some advantages for thecomplicated unlined model of battery, and the estimating result is more accurate, inthe mean while the computing procedure is more quick.3. Estimated the residual capacityCompared the residual capacity estimating methods of lead-acid battery inHEV including ANN, fuzzy logic method and compensated fuzzy neural networktechnology methods, the fuzzy neural network model of the lead-acid battery isdeveloped. The residual capacity of battery in any situation during charge anddischarge is estimated. The simulink result by Matlab tell us the residual capacityof the battery is obtained conveniently and quickly using this new model, and theresult which meet the need supply a new estimating method.4. Qualitative analyzed and compared the advantage of compensated fuzzyneural network technology methodsFuzzy neural network technology methods is always the central issue in thefield of residual capacity estimating. This paper described compensated fuzzyneural network technology methods and optimized the construction under theformer foundation. The compensated fuzzy neural network model algorithmsuccessfully identified the studied problem. The result identification shows that thealgorithm based on compensated fuzzy neural network is an effective method forcomplex nonlinear system. Through pre-treating the trained data, using thecompetitive learning to make the self-adapted clustering of input data, anddetermining parameters of the membership function and the numbers of fuzzy rules,etc. The algorithm has more accurate precision and faster convergent speed.