The Design And Implementation Of Lithium Battery Formation Quality Evaluation And Intelligent Screening System

Since the invention of the battery, lithium-ion battery is quickly and widely used, for example, in daily life, electric vehicles, aerospace, military and other fields. On one hand, at present, the use of lithium-ion battery more exists in the form of combination. So we must maintain consistency of it. On the other hand, even if in the same batch of batteries, because of the different chemical reactions inside the battery, the influence of external factors and the influence of uncertain factors in the production process etc. are likely to lead to the difference in batteries after the production. Therefore, it is necessary to screen lithium-ion battery in line with national standards and classify the same performance to a level. Aiming at these problems, researching in relevant literature and analysing the performance of battery is important in designing and implementing the system.The content of this dissertation is mainly on the following aspects in details:(1)From the lithium battery formation process, for the SEI film on the electrode surface after lithium battery, the paper have the research of the effect of SEI film on the performance of lithium-ion battery and lithium-ion battery characteristic parameters,establish the relationship model between the two at first. Then, extract the characteristic parameters in the allowed error range. Focuses on the analysis and processing of data, calculate capacity using improvement current integration method based on Peukert equation and select resistance.(2)For the screening for lithium-ion battery, by using the characteristic parameters of lithium-ion batteries designed support vector machine classifier suitable for lithium-ion battery, including process sample data, build classifier models and optimize the classification model parameters. Using samples of training train the classifier and analysis and comparison the result come from commonly used kernels based on the accuracy rate of test data. Then using the highest test classification accuracy rate to establish a classification model. After it, using a grid search method, genetic algorithms, improved genetic algorithm to optimize classification model parameters. At the same time,MATLAB simulation tool is used to compare and analyse algorithms. Finally, completing the design and implementation of classifier using the parameters get from improved genetic algorithm.(3) Implement the SVM classifier in the lithium battery formation PC software. While achieving classify the qualified lithium-ion batteries into different levels with voltage, resistance, capacitance three performance parameters according to the different requirements of users.The system has been thoroughly tested, including the whole test, the module test and the used to verify the accuracy of the system. The test results show that the system is stable, can select lithium-ion battery correctly and meet the requirements, and the user interface of software is simple and friendly, meet the requirements of original design.