Affiliations: Department of Electronics and Communication Engineering, MIT School of Engineering and Sciences, MIT Art, Design and Technology University, Pune, Maharashtra, India
Correspondence:
[*]
Corresponding author: Kunal Subhash Khandelwal, Department of Electronics and Communication Engineering, MIT School of Engineering and Sciences, MIT ADT University, Pune, Maharashtra 412201, India. E-mail: kunalkhandelwal13@gmail.com.
Abstract: Lithium-ion batteries are widely utilized in a variety of transportation sectors, including highways, airplanes, and defensive military applications because of their positive features, which include a low self-discharge rate, raised operating voltage, prolonged cycle life, and high energy density. Nevertheless, during operation, the battery undergoes adverse reactions that may eventually cause material aging and capacity deterioration. For this reason, the prediction of remaining useful life (RUL) for LiBs is important and necessary for ensuring reliable system operation. Moreover, accurate RUL prediction can effectively offer maintenance strategies to certify the system’s dependability and safety. The impedance of the battery plays a vital role in the degradation process and hence its measurement accounts for the changes in the battery’s internal parameters as aging occurs. The objective of this paper is to create a forecasting model for the lifespan of batteries through the application of data analysis. Moreover, it aims to use a Regression Neural Network (RNN) based mathematical model to assess the degradation of the battery across diverse operational scenarios. The co-efficient of the RNN model is found by solving the RNN equations. In this research, MATLAB is employed for data analysis, utilizing open-source battery data sourced from the NASA dataset. The conclusive prediction outcomes indicate the effectiveness of the proposed methodology in accurately forecasting the battery RUL.
Keywords: Battery, remaining useful life prediction, regressionneural network, transportation, NASA dataset
