Font Size: a A A

Research On State Estimation Method Of Electric Vehicle Power Battery

Posted on:2021-05-08Degree:MasterType:Thesis
Country:ChinaCandidate:K LiuFull Text:PDF
GTID:2392330629451468Subject:Electrical engineering
Abstract/Summary:PDF Full Text Request
Battery status estimation is an important part of the battery management system of electric vehicles,and it is also the basis of the battery management system.Only with an accurate estimation of the current state of the battery can the battery be managed reasonably,thereby improving the efficiency of the battery,saving energy,increasing the cruising range,increasing the battery life,and saving costs.In this paper,the lithium ion battery with nickel,cobalt and manganese as the cathode material is the research object,the battery state-of-charge,state-of-energy,and state-of-power joint estimation technology are studied.The research focus is mainly on the following aspects.First of all,from the perspective of the external characteristics of the battery,this paper designs corresponding experiments to study the relationship between the actual capacity of the battery and the temperature and discharge rate,and analyzes the influencing factors and acquisition methods of the OCV-SOC curve of the battery.Based on these analyses,this paper compares common battery equivalent circuit models,considering accuracy and complexity,and selects an improved second-order Thevenin equivalent circuit model as the state estimation model.In order to improve the accuracy of parameter identification,this paper studies the characteristics of battery polarization effects.It is found that to improve the accuracy of the improved second-order Thevenin model,the time constants of the two RC links in the model need to be orders of magnitude different.Based on this,a fusion constraint factor recursive least squares method is used to identify the parameters of the improved second-order Thevenin model.The experiment proves that the improved second-order Thevenin model established in this paper has higher adaptability under various working conditions and high model accuracy.Secondly,this article briefly introduces the principles of the five basic SOC estimation algorithms that are widely used at present,and designs three types of experiments to analyze and compare the performance of the five algorithms for vehicle SOC estimation.The first type of experiment is used to verify and compare the adaptability of the five algorithms to different working conditions;the second type of experiment is used to compare the initial value correction and noise filtering ability of the five algorithms;the third type of experiment is used to verify the algorithm's ability to resist parameter disturbance.Three types of experiments simulate the main conditions that affect the accuracy of SOC estimation when the five algorithms are onboard.This provides a certain reference for the selection of electric vehicle SOC estimation algorithms.Based on the research on the performance of five algorithms and improving the SOC estimation accuracy as the goal,this paper proposes a sliding mode observer algorithm combined with Kalman filtering.The joint algorithm can synthesize the advantages of Kalman filter and sliding mode observer at the same time,while filtering noise,it also has strong robustness to modeling errors.This paper designs the corresponding simulated working conditions for experiment.The experimental results prove that the proposed algorithm has higher SOC estimation accuracy than the extended Kalman filter and sliding mode observer in a complex vehicle-mounted environment.Finally,based on the SOC estimated by the joint algorithm,this paper uses the Open Circuit Voltage(OCV)as the medium to establish a SOC,SOE,and SOP joint estimation strategy.Experimental verification was carried out under the different working conditions,and the experiment proved that the joint estimation strategy has high accuracy.There are 44 figures,13 tables and 97 references in this paper.
Keywords/Search Tags:Battery modeling, battery state joint estimation, Kalman filter, sliding mode observer
PDF Full Text Request
Related items