電動車用鋰電池即時狀態估測與故障偵測之研究

Abstract

高功率鋰電池的發展加速了分散式替代能源系統與新世代電動汽車的發展。作為目前最理想的儲電裝置，鋰電池具有高能量密度、高功率密度、高循環壽命等特性。但與過往攜帶型裝置小型鋰電池不同的是，高功率鋰電池組需要承受大電流、操作環境溫度變化、震動等。在此操作環境下，高串並數的高功率鋰電池組若出現有電池異常或老化等現象，除了會降低整體電池容量與降低功率輸出外，更有機會由於內部結構的毀損造成電池進入危險的熱失控狀態(thermal runaway)。有鑑於此，在大量的的定性實驗後，以基於鋰電池模型的適應性演算為基礎，在不需要實驗室等級電化學設備的協助下，對電池內部重要參數如開路電壓與內電阻等數值進行估計。這些參數具有極重要物理意義，如開路電壓可用於換算電池內部殘電量，電阻抗值顯示了內部電極層間健康狀態，可用於推算目前電池健康狀況與偵測電池組間異常短路等。 本論文主要研究重點包含: 1. 對昇陽半導體所提供之40138F1鋰鐵電池進行詳細的定性測試，針對充放電特性進行分析，並在此基礎上設計電池等效電路模型。 2. 使用自製鋰電池管理系統對電池組進行包含電壓、電流、溫度之監控與紀錄，並應用於以鋰電池為動力之輕型電動代步車上，實際監測記錄道路行駛過程中之電池變化與參數估計。 3. 在MATLAB/Simulink環境下建立並修正具有溫度補償的鋰電池等效電路模型。在LabVIEW環境下建立一組鋰電池監測系統，可用於紀錄與顯示電動車操作時之電壓、電流、溫度變化。

High power Lithium-ion battery has become the major energy storage option for renewable-distributed power generation systems and electrical vehicles. However, the required performance and rigorous operating condition of such utilities make its critical to design a proper battery management system (BMS) and monitor the operating situation carefully. The combination of physical based Lithium-ion dual polarization model and adaptive parametric estimation algorithm with reasonable sampling rate can provide accurate insight on the internal parameters such as open circuit voltage (OCV) and internal resistance, etc. Such parameters provide critical indication to battery states like state of charge (SOC) and state of health (SOH). The SOH is the indication of individual battery age condition and possible failure between cells, and is emphasized in this thesis due to the potential developments for online fault detection system and early warning system. And the estimated parameters act as the indication for the location of degraded cell with the possible failure reason, such indication can assist the maintenance engineer greatly. The major research in this thesis: 1. With detailed qualitative tests on PSI 40138F1 battery, equivalent circuit model is designed based on the characteristics from charge-discharge curve. 2. Using self-developed DAQ board to monitor and record the voltage, current and temperature of tested power battery pack. And the system is deployed on the DC motor EV in field tests to record and estimate the internal parameters variation in road test condition. 3. Constructing a Lithium-ion battery equivalent circuit model in MATLAB/Simulink environment and the model is improved with temperature compensation in Arrhenius form. Constructing a Lithium-ion battery monitoring system in LabVIEW environment which can monitor and record voltage, current and temperature in EV operation