| In recent years,the problems of environmental degradation and energy shortage are becoming more and more serious.Countries around the world are seeking for the energy transformation of the automobile industry.New energy vehicles have achieved rapid development under the promotion of relevant policies in China.The state of the power battery is related to the reasonable operation of the power machinery,the driving range of the vehicle and even the safety of the passengers.Since the state change of power battery is strongly coupled by environmental temperature,charge-discharge rate and cycle number,the battery capacity declines in different paths.The decline of ability increases the difficulty of estimating the battery state of charge.The estimation of SOH and SOC of power battery is hotspot and difficult point in battery management system.The main research contents of this paper are as follows:（1）The composition and working principle of Li Fe PO4 battery were studied,and the attenuation mechanism of Li Fe PO4 battery was summarized from three aspects:the reduction of effective lithium ions,the loss of active materials of positive and negative electrodes and the decomposition of electrolyte.（2）The common performance parameters of lithium-ion batteries were analyzed.SOC and SOH that define different way superior to domestication.The battery test platform select the CC-CV charge-discharge strategy,designed the circulation of the Li Fe PO4 monomer battery charging and discharging test and cycles,discharge rate,temperature effect on the battery voltage’s performance resistance test.Combined with the electrochemical mechanism of Li Fe PO4 battery,the effects of cycle number,discharge rate and temperature on the battery performance parameters were analyzed.（3）The working principle of BP neural network were explained,signal forward transfer from two aspects of subprocesses and error back propagation subprocesses learning process of BP neural network was deduced.The charge and discharge circuit test data obtained will represent the status of the battery discharge voltage and resistance as input samples of estimated battery SOH.The design has set up a BP neural network model for the estimation of battery SOH.Adding capacity factors to the battery SOC estimation model,selecting the battery in the whole life cycle SOH step length for 5 discharge voltage input sample data as the battery SOC estimation model.The BP neural network model to estimate the battery SOC,analysis summed up the BP neural network model for evaluating the effect of the battery SOC and SOH and can be optimized.（4）BP neural network is analyzed,the necessity of integration with the global optimization algorithm,and the rationality of the immune algorithm.The BP neural network fusion,appeared for immune algorithm optimization effect and choose the same as the BP neural network model of network structure and the data sample.IA-BP neural network models for estimating battery SOH and SOC were built respectively.By comparing the IA-BP neural network model with the BP neural network model,under the same test samples,the SOH and SOC estimation showed a better prediction effect in terms of prediction accuracy and iteration efficiency after being optimized by immune algorithm. |
