Estimation Of SOH And SOC For VRLA Batteries Uesd In Electricity And Communication Systems And Battery Fault Diagnosis

Lead-acid batteries as a storage device are used in all aspects of people's production and life.Valve-regulated lead-acid battery is a kind of lead-acid battery,which is widely used in electricity and communication system as energy storage device because of its advantages of maintenance,and it is also the object of this study.The information about state of health and state of charge is the key evaluate performance parameter of VRLA battery.The English name of health status is state of health,referred to SOH;state of charge is also known as the remaining capacity,referred to SOC.We can find the poor VRLA battery in a battery pack by monitoring the real-time SOH and SOC information of the battery pack,and we need to replace and update the poor one in time,which is beneficial to the maintenance and management of the battery pack.The study on the estimation of real-time SOH and SOC and as well as the derivative failure analysis of the VRLA battery,has become the research hotspot of battery management and application.First of all,this paper briefly introduces the related concepts of VRLA battery and its related application in power and communication system,and summarizes the research status about the estimation of SOH and SOC of VRLA battery.Practice shows that Thevenin model can well characterize the static and dynamic characteristics of VRLA battery.Generally,the parameters of the model are used as the input of SOH and SOC estimation.In the SOH estimation,the SOH = f(R1)curve equation is first fitted and the SOH is roughly estimated,then the SOH is estimated by the least squares method and the error is further reduced.Finally,the RBF neural network algorithm is used to estimate the SOH,and the estimation accuracy is very high.In the SOC estimation,the estimation is divided into two processes of charging and discharging.Considering the changing parameters of the Thevenin model,the E-SOC curve is fitted in stages,and the SOC is estimated by the Kalman filter algorithm.The last part of this paper is VRLA battery failure analysis,combined with the link of the battery structure and model,a series of destruction experiments of battery every parts are carred out,and the internal fault of VRLA battery is analyzed by measuring the variation of the parameter value of Thevenin model,which provides some auxiliary functions to evaluate the SOH of VRLA battery.In this paper,the NPP VRLA battery produced by Guangzhou Napu Company is used as the experimental object.Battery charging and discharging experiments are carried out under the control temperature 25? in the power and communication system.At the same time,some SOH estimation related preparation experiments are carried out.Through the experimental data acquisition for Matlab software simulation to achieve SOH and SOC estimates,it is found that the maximum error of SOH estimation by the least squares method is within 10%,while the maximum error of SOH estimation by RBF neural network optimization is within 5%,the average one is within 1%.The error of SOC estimation by Kalman filter can be reduced below 3%.If this SOH and SOC estimation algorithm of VRLA battery is used in the battery online management system,it can play a certain and effective battery maintenance role.