基於粒子群演算法之主動式能量管理系統應用於雙電力電動車輛

Abstract

本研究為一雙電力電動車輛系統主要實驗於最佳化能源管理系統模擬以及實車被動式電力分配之驗證，實驗項目可分為：(1)雙電力系統之鋰電池組設計與建置、(2)雙電力系統之系統連接器開發與建置、(3)雙電力最佳化能量管理開發與實際被動式電流分配法之控制模擬、(4)雙電力系統車輛於動態負載情況進行雙電力源分配驗證。本研究選用之電動車為長2.3公尺、寬1.2公尺的高爾夫球車輛，搭載36V/3kW之直流馬達。針對原本車輛之儲能系統進行重新設計與建置，透過車輛車廂之空間運用，設計一組36V、180Ah之儲能鋰電池組，於設計過程藉由SOLIDWORKS軟體繪製模組化鋰電池，於軟體內進行等比例尺寸設計，且採用21700鋰電池芯，可提高整體儲能系統能量密度。於限制大小條件考量設計要點包括:1.電池容量最大化、2.減輕電池中重量、減少設計零件數量以及3.快速進行電池抽換功能。設計分為三個階段，第一階段為訂定鋰電池組最大體積規格，第二階段接為設計固定體積下之最佳電芯數量、電池平衡板與電池固定支架，並考量車載環境之耐用性與安全性，第三階段為修正細節與設想工法製作方式，所有的設計都與生產製作相關，過於複雜之設計可能導致生產不易或徒增實驗成本費成本。本研究之最佳化能量管理分配使用為(1)粒子群演算法 (Particle Swarm Optimization, PSO)透過模擬鳥類之覓食習性，如鳥群在空中找尋覓食地點，藉由每一隻鳥彼此之間傳達訊息，使鳥群逐漸往目標方向移動，並找尋最佳解。(2)基本規則庫控制(Rule-Based)係以透過工程經驗與系統特性進行if-else控制語法撰寫，可分為五種模式(待機模式、純電模式、巡航模式、高速模式與充電模式與(3)被動分配法為本研究實測雙電力平台被動式分配特性，將參數建立為一資料庫，藉由當下鋰電池殘電量與需求功率進行雙電力能源分配。在執行五次WLTP行車型態下，基本規則庫、PSO與被動分配法的能耗為為[2005kJ、1502kJ、2578kJ]，將基本規格庫與PSO比較被動式分配法之能耗改善率為[22.2%、41.7%]，綜合上述可知，粒子群演算法具最佳之能量改善效率，未來將實施於實車之雙電力混合車輛。而本文另一項實車雙電力系統被動式分配驗證結果，於鋰電池殘電量60%以上皆為鋰電池單獨輸出，而鋰電池殘電量低於60%時，燃料電池將會加入電流分配，當鋰電池殘電量下降至20%，則可觀察到燃料電池具有回充及單獨輸出之情形，表示被動式混合電力系統具有正常電力分配結果，可提升鋰電池續航力與使用壽命。In this study, the dual electrical power vehicle simulation by optimal energy management strategy and verification power distribution by passive control in real vehicle construction, the experience issue is including: (1) Lithium battery set of the dual electrical power system design and set-up, (2) System connecter of the dual power system develop and set-up, and (3) Optimal energy management strategy of the dual power system design and passive control method simulation in real vehicle and (4) Dual power distribution verifications of the dual power vehicle system in variable load condition.A 36V/3kW DC motor was chosen for the golf cart in our research. The original storage system was redesigned and built, it utilized primary vehicle battery frame design, the 36V/180Ah lithium battery set and the SOLIDWORKS was used to sketch the battery module, we use 21700 battery cell to built 1:1 scale module to impove the energy density of the storage system. The constraints were formulated in whole design. They include: (1) maximize battery capacity, (2) lighten battery weight and reduce component of battery, and (3) exchange battery rapidly. It is divided by three stages in design battery: (1) calculate maximum lithium battery set volume firstly and then (2) design optimal number of battery cell, Battery Management System (BMS) and battery bracket in fixed volume. We consider the durability and safety at the same time in second stage. The third one is to verify details about manufacture, all designs are associated with product procedures, complex design may cause production difficulty or increase cost.In our study, three strategies were chosen as the optimal management distribution, (1) Particle Swarm Optimization (PSO) which mimics the characteristic of birds flock foraging. Each bird has individual consciousness and group experience. They share information between each bird, it let the group toward the goal to move, and find optimal result. (2) Rule-Based (RB) control considers engineering experience and system property utilizing programming language-“if-else”. It divided into five mode: stand-by, pure electricity, cruising, high speed and charge mode, respectively. And (3) Passive control was the passive power distribution characteristic by fuel cells and battery the data set was established by the variable current loads and different battery state-of-charges (SOCs).In fives time WLTP, the energy consumption with the RB, PSO and passive control is 2005kJ, 1502kJ and 2578kJ, respectively. Compared with passive control, the energy consumption improvement rates of the RB and PSO are 22.2% and 41.7%, respectively. It shows PSO has best effect that can applied in real vehicle with dual electrical power system in the future. The passive control method shows the fuel cell supply current distribution when battery SOC was lower than 60% and once battery SOC lower than 20%, fuel cell provides extra current to charge the battery and in charge of power demand individually. It shows the dual electrical power system can prolong battery endurance and life cycle .