Quantum neural network for State of Charge estimation

Abstract

State of Charge (SoC) estimation is one of the most important parts of Battery Management System (BMS). Inaccurate estimation of SoC may cause overcharge or overdischarge which could lead permanent damage to battery cells. Neural Network (NN) models can yield quite accurate SoC estimation. However, the computation effort is also quite huge and it takes long time training. To improve the performance of NN, a new battery SoC estimation method based on Quantum Neural Network (QNN) is proposed. Results show that QNN is more computation efficient and yields more accurate results, when compared to the conventional NN and other methods such as Coulometric Counting (CC) and Open Circuit Voltage (OCV) prediction methods.