Lithium Battery State Of Charge Estimation Based On Improved H-infinity Filter

With the increasingly serious crisis of greenhouse effect and traditional energy,the development of pollution-free and renewable new green energy has become a new trend in the development of various countries.Lithium-ion battery is one of the core of green energy,attracting the attention of scholars and engineers in many countries and regions.Because lithium-ion battery is widely used in various fields,a little careless operation will affect the normal use of the battery and reduce the life.In order to prevent short-circuit and overcharge operations of lithium-ion batteries from reducing the battery life and improving the energy utilization and safety of lithium-ion batteries,it is necessary to estimate the state of charge(SOC)of lithium-ion batteries in real time and limit it within the safe range.Therefore,the research on SOC estimation is of great significance.As the internal parameter of lithium-ion battery,SOC cannot be measured directly by the instrument.The precise SOC value can only be obtained by a series of identification and estimation of the existing parameters.The specific research contents are as follows:1.In this paper,Samsung ICR18650 cylindrical ternary lithium-ion battery is taken as the research object,and the battery test system produced by NEWARE Company is adopted to conduct in-depth analysis of battery characteristics such as storage capacity,voltage,current and resistance.The importance of SOC estimation for lithium-ion battery was clarified,and the difficulties of SOC estimation were discussed.2.Several common lithium-ion battery models were analyzed,and the dual-polarization model was selected through analysis and comparison.The static capacity test and OCV-SOC test were carried out.Then,the least square method with forgetting factor in the mainstream identification method was selected to carry out online identification of lithium-ion battery.Simulation results show that this method can quickly and accurately identify the main component parameters of lithium-ion battery model online.Finally,the improvement direction of variable forgetting factor is proposed.3.An improved H∞ filtering method is proposed,and the sliding mode observer is introduced into H∞ filtering.The dual-polarization equivalent circuit model is established,and EKF,H∞ filter and H∞ filter of joint sliding mode observer are compared by discharge experiment and dynamic stress test.The simulation results show that the H∞ filtering algorithm of the joint sliding mode observer has better practicability and the highest accuracy.4.The strong tracking and multi-information theory are introduced into H∞ filtering to overcome the bad points in the data and improve the tracking performance of the algorithm.To verify the validity of the algorithm,discharge experiment and DST experiment are carried out.The H∞ filter and the improved H∞ filter are compared through the experimental simulation results.The results show that the modified algorithm improves the robustness of the algorithm without requiring a lot of computation.