Design And Realization Of Portable Intelligent Testing System For Lithium Ion Power Battery

As one of the major power sources in the 21 st century,lithium-ion power batteries are widely used in military and civilian fields such as electric vehicles and aerospace.In order to ensure the reliability and safety of the lithium battery during use,it is necessary to accurately detect and estimate its key parameters.However,the remaining capacity(SOC,State Of Charge)cannot be directly obtained through measurement such as current,voltage,etc.It needs to be estimated by combining relevant parameters and mathematical methods.Therefore,based on the research of SOC estimation,aiming at the shortcomings of the existing detection systems such as bulky and expensive,it is of great practical significance to design a portable intelligent detection system to detect the performance parameters of lithium batteries.This article mainly completes the work as follows:Aiming at the problem that the non-linearizable parameters in the non-linear model of the lithium battery Thevenin are difficult to identify,based on the limited measurement experimental data,a gradient descent algorithm is adopted to construct the objective function and successively approach the set threshold to obtain the target identification value.Compared with the results of the least squares algorithm,the former method is simple and accurate.Through Simcape,the identified Thevenin model is simulated and verified,and improved on the basis of it,and the new model is more applicable through error analysis.As an important performance parameter of lithium battery,SOC is used to estimate by extended Kalman filter(EKF),but in the process of linearization,Taylor's expansion of high-order items tended to result in decreased precision and divergence.The unscented Kalman filter(UKF)solves the problem of mean and covariance of the system through unscented transformation,and has a significant improvement over the EKF accuracy.Through iterative methods,the accuracy of the SOC is further improved by using the result of the unscented Kalman filter algorithm as the initial value in the algorithm prediction process.Through the constant current discharge of lithium battery,pulsed discharge,QC/T897-2011 experiment to verify and analyze the relevant algorithms,a joint algorithm combining IUKF and Ampere-hour integration method is obtained which can be applied directly Engineering practice.A portable lithium battery intelligent detection system was designed and implemented.The system consists of hardware and software,and supports the collection of lithium battery performance parameters,including voltage,current,temperature,and SOC estimation of the lithium battery's remaining charge.The acquisition accuracy meets the QC/T897-2011 standard.The system has a complete human-computer interaction interface,which can view the waveform diagrams and stored data of the required data.At the same time,it includes serial ports and USB external interfaces,which improves the system's practicality.