State Monotoring Of Fbl Lithium Iron Phosphate Battery

For the past few years,with the enhancing of people’s environmental consciousness,the progress of energy is becoming one of the key national stratagem.Floating charge battery as a key direction of energy progress,and its correlational study has been widely concerned.Boosting LFP battery is an important backup power source in power engineering.It plays an important role in various places that need continuous power supply in actual production and life.Accurate estimation and monitoring of SOC and SOH of LFP batteries can facilitate the observation and analysis of battery performance,effectually strengthen the availability factor,and ensure the entire electrical system to run safely and stably.Therefore,this article taking the method of obtaining and using SOC and SOH as major research centres,the main work of this paper includes:Firstly,in the aspect of SOC estimation,Thevenin model has high exactitude in simulation,and coefficients ordinary to recognize.So this paper used this model as a foundation,recognized the coefficients with data from actual test.Then,based on the model state space equation,the AEKF is used as the arithmetic to obtain SOC.On the basis of the traditional Kalman filter,the extended Kalman filter solves the nonlinear problem of the battery model by Taylor expansion,and the new information sequence is used to update the noise covariance of kalman filter,which further promotes the precision and ability to accommodate of the algorithm.Secondly,the data-driven SOH estimation method has a large amount of data and a complex training process,and the capacity of FBL battery is hard to get,because FBL battery will connect to the supply and duty after beginning working all the time.In this paper,a SOH Access method based on characteristic choosing and PSO-SVR is using.Before the estimation,the data from LFP battery experiment is analyzed.According to every curve in the process of cell aging subtle changes,to extract the cut-off voltage point in time,the charging capacity and capacity increment curve Ⅱ peak amplitude three characteristic vector,and based on grey correlation degree,verify every feature.Then the PSO was used employed to update the coefficient in SVR,a calculating model with preferable coefficient can be gotten at last.The test show that the model has a good result to estimate SOH after training,and reduce the size of specimen comparing with ANN.What’s more,the pattern of the model conform to the FBL battery,which powers on for a long time.The workers can get the SOH with a few data by using the estimating model.So this method can be used in the daily work of FBL battery.Finally,in order to realize the monitoring of LFP battery.This paper used the Netty as a basis to set up a battery monitoring platform,realized the function of data thransmission from terminal in the dc battery cabinet.And the server has the function of dealing and storage the data.So that the workers can view the battery data and state on the net page at any time.