State of Charge (SOC) of a storage battery gives the capacity remaining in the battery to meet the load demands. SOC of a Lead Acid battery is predicted based on the temperature compensated value of electrolytes' specific gravity (Sp. gr.). Since measuring specific gravity is not possible in an automobile under dynamic conditions, Open Circuit Voltage (OCV) is used as the parameter to predict the SOC. But OCV can indicate SOC accurately only after a sufficient period of rest of a battery in any condition. Also it varies between batteries due to several factors like temperature, ageing, electrolyte volume, internal construction, etc. Predicting the SOC of battery theoretically depends on number of equations developed to accommodate these variables.
Hence for a real time estimation of SOC of battery, a Heuristic algorithm is suggested. The initial State of Battery is estimated by temperature compensated OCV Vs SOC characteristics and followed by dynamic prediction of SOC using Coulomb or Energy Measurement. The drawback of cumulative error in energy measurement is also overcome in this algorithm. For improving the accuracy of dynamic prediction, the heuristic algorithm also talks about compensating the predicted SOC value with respect to factors such as rate of discharge, rate of charge, self discharge and temperature. Finally, the algorithm also talks about “A Self learning feature”, to predict OCV based the initial SOC of Lead Acid battery more accurately irrespective of the vehicle quiescent current differences.