Pil implementation of adaptive gain sliding mode observer and ann for soc estimation

Abstract

Nowadays, Electric Vehicles have gained a lot of interest among academic researches and the industrial actors, however, for a vast adoption of these tools, tasks such as their autonomy prolongation as well as ensuring their battery security are of great importance. Tasks that are accomplished via a survey of their battery state of charge (SOC) and state of health (SOH). In this paper, we present two advanced algorithms; Artificial Neural Network and Adaptive Gain Sliding Mode Observer (AGSMO), based on a combined battery equivalent circuit model (CBECM) to estimate the SOC of a lithium-ion battery for electric vehicles. To verify the effectiveness of each algorithm, PIL (Processor In the Loop) tests were implemented using an STM32F429ZI discovery board. The experimental results prove that both algorithms have good performance in battery SOC estimation, with a slight edge of AGSMO over the ANN due to the limitation of training data.