改良式灰狼演算法應用於三動力複合動力車之最佳化能源管理控制器開發
No Thumbnail Available
Date
2022
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
本研究旨於開發一新型能量管理策略，以灰狼演算法 (Grey Wolf Optimizer, GWO)與粒子群演算法 (Particle Swarm Optimization, PSO)相結合所開發 之 改良式灰狼演算法 (Improved Grey wolf Optimizer, IGWO)，並以此為核心應 用於 三動力複合動力車 之最佳化能量管理， 再 透過硬體嵌入式系統 (Hardware-In-the-Loop, HIL)進行即時 (Real-Time)運算，驗證本研究開發之能量管理系統可於真實環境下運行。改良式灰狼演算法結合灰狼演算法中三項灰狼行為，分別為： ：(1) 取得獵物位置 、 (2)包圍獵物、(3)攻擊獵物，及粒子群演算法 中 粒子 移動更新 方式，結合兩種演算法之優點，並能彌補相互間缺點， 以 得到更加 精確之最佳解。本研究除開發IGWO外，也 將 開發 其他 三種控制策略 與 IGWO進行能源效益比較 ，分別為 ：(1)規則控制庫 (Rule-base, RB)、 (2)最小等效油耗法 (Equivalent Consumption Minimization Strategy, ECMS)以及原始灰狼演算法 (Grey Wolf Optimizer, GWO)進行比較 。 基本規則庫依設計者的專家工程經驗與元件物理特性所撰寫模式切換之策略，共五種策略 (純電動模式、 純引擎模式、 複合動力模式、 ISG回充模式及煞車回充模式最小等效油耗法透過巢狀 迴圈進行全域格點搜索 (Global Grid並透過輸入各項行車條件，利用窮舉法運算，找出全部可行解，再透過內插方式計算 出最小油耗時動力分配 比 ；灰狼演算法是透 過 模仿 灰狼捕獵 行為 透過包圍環繞方式， 進而找到最佳獵物位置， 亦 同時為最佳解。各控制策略
(RB、 ECMS、 GWO及 IGWO)運行一次 WLTP行車型態之等效燃油消耗分別為 [1479.0g, 1167.0g, 977.2g, 968.2g]；運行一次NEDC行車型態等效燃油消耗分別為 [561.7g, 490.3g, 393.0g, 391.9g]，在一次WLTP中能耗改善以Rule-base為基準分別為 [21.095% , 33.520% , 34.536%]，在一次NEDC中能耗改善以Rule-base為基準分別為 [12.711% , 30.033 , 30.229%]，由此可知，將最佳化方法導入三動力複合動力車進行動力分配，能有效降低整車能源消耗。本研究透過 兩 台快速雛型控制器，建立一即時模擬平台。驗證由 IGWO為核心開發之能量管理系統於真實環境下應用之可行性，在兩種行車型態下，於電腦模擬與HIL環境運行之等效 燃 油 消 耗 相似度 高達 90%以上由此可知，從上述結果可客觀判斷，此能量管理系統未來應用於實車有非常高之可行性 。
The purpose of this study is to develop a new energy management strategy, an Improved Grey Wolf Optimal (IGWO), developed by combining Grey Wolf Optimizer (GWO) and Particle Swarm Optimization (PSO), This algorithm is applied to the optimized energy management system of the three-power Hybrid Vehicles. The Hardware-In-the-Loop system (HIL) is used to perform real-time calculations to verify that the energy management system developed in this study can operate under real-world conditions. IGWO is combines three grey wolf behaviors in the GWO: (1) Tracking, chasing, and approaching the prey, (2) Pursuing, encircling, and harassing the prey until it stops moving, (3) Attack towards the prey, and the particle update method in the PSO to combine the advantages of both algorithms to compensate for each other's deficiencies and obtain a more accurate optimal solution. The other three energy management strategies were developed that to compare with IGWO, it is Rule Base (RB), Equivalent Consumption Minimization Strategy (ECMS) and Grey Wolf Optimizer (GWO), respectively. The RB is designed by expert experience and components physic characteristic, includes five modes: pure motor, pure engine, hybrid, charge and regenerative braking; The ECMS via for loop calculate Global grid search of various driving conditions by exhaustive method to find all feasible solutions, through interpolation acquire minimum fuel consumption; The GWO mimic the behavior of gray wolves when hunting, and then find the best prey position, it is best solution also.The equivalent fuel consumption of Rule base, ECMS, GWO and IGWO were 1479.0g, 1167.0g, 977.2g and 968.2g when each control strategy is run once in the WLTP driving cycle. The equivalent fuel consumption of each control strategy were 561.7g, 490.3g, 393.0g and 391.9g, when each control strategy is run once in the NEDC driving cycle. The improvement percentage of energy consumption were 21.095%, 33.520% and 34.536%, respective, for ECMS, GWO and IGWO compared to Rule base control strategy under a WLTP driving cycle. The improvement percentage of energy consumptionwere 12.711%, 30.033% and 30.229%, respective under a NEDC driving cycle. It can be seen that introducing the optimization method into the hybrid vehicle for power distribution can effectively reduce the energy consumption of the vehicle. In this study, a real-time simulation platform was established by using two rapid prototype controllers. To verify the feasibility of the energy management system developed by IGWO in the real environment, in the two driving modes, the equivalent fuel consumption similarity between computer simulation and HIL environment operation is as high as 90%. It can be judged objectively that this energy management system has a very high feasibility to be applied to real vehicles in the future.
The purpose of this study is to develop a new energy management strategy, an Improved Grey Wolf Optimal (IGWO), developed by combining Grey Wolf Optimizer (GWO) and Particle Swarm Optimization (PSO), This algorithm is applied to the optimized energy management system of the three-power Hybrid Vehicles. The Hardware-In-the-Loop system (HIL) is used to perform real-time calculations to verify that the energy management system developed in this study can operate under real-world conditions. IGWO is combines three grey wolf behaviors in the GWO: (1) Tracking, chasing, and approaching the prey, (2) Pursuing, encircling, and harassing the prey until it stops moving, (3) Attack towards the prey, and the particle update method in the PSO to combine the advantages of both algorithms to compensate for each other's deficiencies and obtain a more accurate optimal solution. The other three energy management strategies were developed that to compare with IGWO, it is Rule Base (RB), Equivalent Consumption Minimization Strategy (ECMS) and Grey Wolf Optimizer (GWO), respectively. The RB is designed by expert experience and components physic characteristic, includes five modes: pure motor, pure engine, hybrid, charge and regenerative braking; The ECMS via for loop calculate Global grid search of various driving conditions by exhaustive method to find all feasible solutions, through interpolation acquire minimum fuel consumption; The GWO mimic the behavior of gray wolves when hunting, and then find the best prey position, it is best solution also.The equivalent fuel consumption of Rule base, ECMS, GWO and IGWO were 1479.0g, 1167.0g, 977.2g and 968.2g when each control strategy is run once in the WLTP driving cycle. The equivalent fuel consumption of each control strategy were 561.7g, 490.3g, 393.0g and 391.9g, when each control strategy is run once in the NEDC driving cycle. The improvement percentage of energy consumption were 21.095%, 33.520% and 34.536%, respective, for ECMS, GWO and IGWO compared to Rule base control strategy under a WLTP driving cycle. The improvement percentage of energy consumptionwere 12.711%, 30.033% and 30.229%, respective under a NEDC driving cycle. It can be seen that introducing the optimization method into the hybrid vehicle for power distribution can effectively reduce the energy consumption of the vehicle. In this study, a real-time simulation platform was established by using two rapid prototype controllers. To verify the feasibility of the energy management system developed by IGWO in the real environment, in the two driving modes, the equivalent fuel consumption similarity between computer simulation and HIL environment operation is as high as 90%. It can be judged objectively that this energy management system has a very high feasibility to be applied to real vehicles in the future.
Description
Keywords
複合動力車系統, 能量管理策略, 灰狼演算法, 改良式灰狼演算法, 最小等效油耗策略, Hybrid Power Vehicle System, Energy Management Strategy, Grey Wolf Optimal, Improved Grey Wolf Optimal, Equivalent Consumption Minimization Strategy