Multilayer neural network-based battery module SOH diagnosis

Today, lithium batteries are used in a variety of applications such as cell phones, electric vehicles, unmanned submarines, and energy storage systems (ESS) as the primary power source. Therefore, for stable use, it is important for the device and the system to quickly detect any fault occurring in the battery and accurately diagnose it. Battery faults can be diagnosed from the battery's state of health (SOH) which reflects its operating condition. In this paper, a system is proposed to diagnose battery cell faults by means of a multilayer neural network (MNN) state classifier. In this method, the MNN state classifier utilizes the discharge voltage data obtained by operating the lithium battery cell at high temperature. We concluded from experimental results that the proposed battery SOH monitoring method diagnoses the state of the battery very well.