Abstract
Batteries have gained importance as power sources for electric vehicles. The main problem with the battery technology available today is that the design of the battery system has not been optimized for different applications. No comprehensive battery models are available today that can accurately predict the performance of the battery system. This thesis presents a modeling technique for batteries employing neural networks. The advantage of using neural networks is that the effect of any variable of the performance of the battery need not be known apriori. The neural network develops the model by corelating experimental data. A software model was developed and tested for lead acid batteries using this technique. The results obtained from the model when compared to experimental data showed that the technique was successful in modeling the performance of a lead acid battery module.
Arikara, Muralidharan Pushpakam (1993). Advanced battery modeling using neural networks. Master's thesis, Texas A&M University. Available electronically from
https : / /hdl .handle .net /1969 .1 /ETD -TAMU -1993 -THESIS -A699.