Retired Lithium-Ion Power Battery Screening And Fault Diagnosis

The development of new energy electric vehicles is one of the key ways to improve energy problems and solve the environmental crisis,which has become an international consensus and national strategy.With the rise of a new round of scientific and technological revolution and industrial transformation,the new-energy automobile industry is entering a stage of accelerated development,which will contribute to the economic growth of various countries,reduce greenhouse gas emissions and improve the global ecological environment.Lithium-ion batteries（LIBs）are the heart of new energy vehicles and have great potential for development.Compared with other batteries,lithium-ion batteries have the advantages of high voltage,light weight,high energy and density,long service life,small self-discharge coefficient,no memory effect and wide range of temperature adaptation.With the increase of the installed quantity of lithium-ion batteries,the corresponding decommissioning quantity is increasing,and safety accidents caused by battery faults occur frequently.Therefore,reclassification and fault diagnosis of retired lithium-ion batteries are important problems to be solved.In order to solve the above problems,this paper discusses in detail and gives solutions.The main research and innovation are as follows:In order to solve the problems of low efficiency and poor consistency in the screening of retired lithium-ion batteries,a screening method based on naive bayes（NB）was proposed.Firstly,12 lithium iron phosphate（LiFePO₄）battery modules were subjected to life aging cycle experiment until their capacity was reduced to 80%,and they were disassembled into 225 cells.Secondly,the characteristic charging scheme of retired lithium-ion batteries was designed,and the incremental capacity（IC）curve was drawn according to the voltage and capacity data,and the curve noise was filtered by kalman filter（KF）algorithm.Finally,the peak IC curve coordinates were extracted as the capacity characteristics of retired batteries,and a naive bayes classifier（NBC）model was constructed to achieve accurate classification of retired batteries.The experimental results show that the screening accuracy can reach 96.96%.Then,the consistency difference of retired batteries before and after separation was compared and analyzed.The standard deviation was used as the basis for quantification.The standard deviation of charging and discharging conditions of classified batteries were all less than 4%,indicating that the proposed naive bayes screening method could achieve consistent and accurate screening of retired batteries.On this basis,this paper uses X-ray diffraction（XRD）,scanning electron microscope（SEM）,energy dispersive spectrometer（EDS）and other microscopic angle experiments,combining microscopic angle analysis with macro data to further verify the accuracy of classification.In view of the difficulty in diagnosing minor faults of lithium-ion batteries,a two-dimensional sample entropy（Samp En）fault detection method based on phase plane was proposed.Firstly,the phase plane is formed by taking the voltage value of the series battery pack as the horizontal coordinate and the voltage first-order difference ordinate,where the point is defined as the cell state point.Then,the sample entropy of the battery state points is calculated through the sliding window to diagnose the fault of the battery.Finally,fault simulation experiments with different fault degrees and temperatures were designed to verify the accuracy and robustness.The experimental results show that the proposed fault diagnosis method combines the phase plane and sample entropy to fill the deficiency of sample entropy.By analyzing the complexity of the battery state points,the fault diagnosis is realized.At the same time,by simulating the fault of different temperature and degree,the method has better accuracy and robustness.