Health Assessment And Remaining Useful Life Prediction Of Li-ion Battery

The functionality and reliability of Li-ion battery as major energy storagedevice has received more and more attentions from a wide spectrum ofstakeholders including federal/state policymakers， business leaders， technicalresearchers， environmental groups and general public. Failures of Li-ion battery not onlyresult in serious inconvenience and enormous costs， but also increase the risk of inducingcatastrophic consequences. In order to prevent severe failure from happening andoptimize the Li-ion battery maintenance schedules， breakthroughs in prognosticsand health monitoring of Li-ion battery must be achieved.The paper presents as：(1) The analysis of Li-ion battery datasets from NASA Ames Research Center.the analysisinvolves the impact of temperature against Li-ion battery health， the impact of EIS tests againstLi-ion battery health， the impact of depth of discharge against Li-ion battery health and theinvestigation of randomness in battery health degradation.(2) The research of Gaussian mixture model. Firstly， Principal Component Analysis (PCA)is utilized for feature extraction from the raw data collected under normal conditon，Then， theGaussian Mixture Model (GMM) is built based on the feature extraction.Lastly the testing datacomes in and constructs a new GMM. The distance or similarity of this GMM and the onegenerated in the training process will indicate the current health status of battery， as representedby Confidence Value (CV). The result of the experiment is satisfactory.(3)The research of ARIMA-PF fusion prognostic framework.It is composed of ARIMAAlgorithm and PF Algorithm. Firstly， monitor the lithium ion battery online， then run thecorresponding algorithm based on short-term forecasts or long-term forecasts Requirements， weget the forecast maps which transverse and longitudinal coordinates stand for the cycle andcapacity respectively. The result of the experiment indicates the proposed prognostic frameworkcan predict lithium ion battery RUL accurately and fast.