Browse Publications Technical Papers 2018-36-0331

Modeling and estimation of the state of charge of lithium-ion battery based on artificial neural network 2018-36-0331

The Fórmula SAE Elétrico is a student competition of monoblock vehicles with propulsion from electric engines, according to the regulations of the organization; the vehicle must be completely electric, not having any connection with a electric network to recharge it. The main thing being researched by the teams is the optimization and the management of the energy storage systems (battery banks) in the electric vehicles. However, there are some parameters which negatively influences the search for such optimization, like: the approximations associated to the battery mathematic models and the determination of the SoC (State of Charge), loss of useful life due to the quantity of cycles and the wear and tear, caused by multiple factors like ambient temperature, pressure, humidity and others. With the goal to build a cellular battery model, it was developed a Neural Network Artificial (ANN), with a learning capacity and adaptation to the charge cycles, discharge and rest to the Lithium-Ion battery. According to the results, it was modeled a system to estimate the SoC. To find the values of every parameter used in the training of the ANN, a set of experiments were made in a measuring bench composed of a four quadrant source, a thermal bath and a data acquisition system. The experiments were made manipulating the current and temperature with charge cycles, discharge and rest of the cells, in which were collected voltage, temperature and current data injected in the cell. With that said, this paper has the objective of obtaining a model and a estimation from the SoC based in Neural Network Artificial (ANN).


Subscribers can view annotate, and download all of SAE's content. Learn More »


Members save up to 18% off list price.
Login to see discount.
Special Offer: Download multiple Technical Papers each year? TechSelect is a cost-effective subscription option to select and download 12-100 full-text Technical Papers per year. Find more information here.