IoT And GAN-LSTM Network Oriented Health Status Assessment Platform For Lithium Iron Phosphate Batteries

Lithium iron phosphate batteries have broad application prospects and research value in new energy heavy trucks,tractor trailers and other high-power vehicle power supply applications,hybrid energy supply and traditional electronics industry.Accurate battery health state(State of Health,SOH)estimation in various application scenarios is of great importance.Generally,when the remaining capacity of the power battery reaches 70-80%,it can not continue to be used in the engineering environment,and it will choose to be retired for disposal.At this time the battery can still be downgraded for energy stor-age,power backup and other scenarios to achieve maximum utilization of residual energy.Therefore,efficient,fast and accurate assessment of the SOH of lithium iron phosphate batteries is of great importance.With the help of Io T and data-driven ideas,this paper an-alyzes the charging data of lithium iron phosphate batteries in detail,establishes the SOH evaluation model of lithium iron phosphate batteries through deep learning methods,and carries out practical applications.In terms of model selection,this paper first uses GAN(Generative Adversarial Network)for battery charging feature data generation,then uses LSTM(Long Short Term Memory Recurrent Neural Network)to learn the process of bat-tery SOH decay,and finally evaluates the battery state by collecting Io T data.The main work of this paper includes the following points:First,the research background and significance of SOH of lithium iron phosphate bat-teries are introduced,and the estimation methods of SOH at home and abroad are discussed in detail.On this basis,the advantages and defects of the current estimation methods of Li Fe PO4 battery model are analyzed,and the research content of this paper is presented.Next,this paper presents an in-depth analysis of the charging characteristics of lithium iron phosphate batteries.Firstly,the working principle of lithium iron phosphate battery is explained,and relevant parameters and concepts are introduced.Then,the Io T platform is built to collect data uploading and data processing from the battery equipment side with the lithium iron phosphate battery in the actual working condition as the research target.Then,the charging process data based on the whole life cycle and the aging process are studied in detail.Then,data processing and feature extraction were carried out.Fea-ture learning and generation were also carried out through GAN networks.Subsequently,the time-series data were trained by LSTM network,and finally the evaluation results of Li Fe PO4 battery were output.Finally,in order to make the model applicable to practical work,this paper introduces the calculation method and process of model invocation,so that the research results of the model do not just stay in the research stage,but apply them to practical work.