Study On Sorting Method Of Retired Lithium-ion Battery Based On Data Mining

With the booming development of new energy vehicle industry,the number of retired lithium-ion batteries is increasing day by day.How to recycle and reuse large-scale retired batteries has become an urgent problem to be solved..Reasonable sorting and reorganization of retired lithium-ion batteries is one of the key technologies for recycling and reuse,which has extremely important practical significance.With the advent of the era of big data,data has become the breakthrough point of technological change.Through data mining,we can find the knowledge hidden in the data,which can guide us to make better decisions.In this paper,battery data from the EV laboratory is used to sort retired batteries,which is guided by means of big data.Taking retired lithium-ion power batteries as the research object,this paper proposes a sorting method for retired batteries based on K-means clustering algorithm and analytic hierarchy process(AHP).By analyzing the influence of the inconsistency of the performance indexes of the retired lithium-ion battery on the reconstituted battery pack,the sorting parameters of the battery were determined.Different clustering algorithms were used to sort the data of 500 groups of retired lithium-ion batteries,and a clustering algorithm suitable for the sorting of retired batteries was obtained by comparing the sorting results.When K-means algorithm was used for sorting,the optimal cluster number was obtained through the comprehensive evaluation of the sum of squares of errors(SSE)and the Calinski-Harabasz variance ratio,meanwhile the performance of each group of batteries was analyzed.In order to make the consistency of the reconstituted battery better meet the needs of the actual scene,the sorting method was improved by parameter weighting,and the weight coefficients of each battery sorting index were determined by AHP according to different use scenarios.It was verified by actual cases of energy storage,low-speed electric vehicle and standby power supply.In this paper,the effectiveness of various parameters of retired batteries participating in clustering was analyzed.By comparing the clustering results to reflect the contribution of each parameter,a new idea was provided for analyzing the importance of various parameters of batteries in evaluating battery performance.Based on the above study,this paper starts from the use requirements of retired batteries,proposes a sorting method suitable for large-scale retired lithium-ion batteries and gives an improvement scheme based on data mining methods.