可快速训练的燃料电池混合动力拖拉机能量管理策略  

Rapidly Trainable Energy Management Strategy for Fuel Cell Hybrid Tractor

作　　者：郭博宇 赵静慧 闫梅 冯旭 李梦林 GUO Boyu;ZHAO Jinghui;YAN Mei;FENG Xu;LI Menglin(Hebei Key Laboratory of Special Carrier Equipment,Yanshan University,Qinhuangdao 066000,China;College of Vehicle and Energy,Yanshan University,Qinhuangdao 066000,China;State Key Laboratory of Intelligent Agricultural Power Equipment,Luoyang 471000,China)

机构地区：[1]河北省特种运载装备重点实验室,河北秦皇岛066000 [2]燕山大学车辆与能源学院,河北秦皇岛066000 [3]智能农业动力装备全国重点实验室,河南洛阳471000

出　　处：《拖拉机与农用运输车》2024年第5期52-57,共6页Tractor & Farm Transporter

基　　金：国家自然科学基金资助项目(52202484);河北省自然科学基金资助项目(F2024203111);河北省高等学校科学技术研究项目-青年拔尖人才项目(BJK2024156);智能农业动力装备全国重点实验室开放课题(SKLIAPE 2023018)。

摘　　要：专注于燃料电池混合动力拖拉机(FCHET)的节能控制问题,提出了一种快速训练并实时应用的能量管理策略。该策略通过学习过往相似工况下的功率分配控制序列的最优解,从而得出一次矩阵计算即可完成功率分配的计算模型。为验证方法的优化性能和计算效率,基于Markov理论构建了一条标准循环工况用于仿真验证。仿真结果显示,所提方法的氢气消耗量仅比基于动态规划(DP)的策略高1.61%,同时相较于传统的模型预测控制(MPC)策略节省了1.54%的等效氢气消耗量,并且运算时间远低于MPC策略,节省了约85.80%的时间。This paper is focused on the energy-saving control issues of Fuel Cell Hybrid Electric Tractors(FCHET),presenting a rapidly trainable and real-time applicable energy management strategy.The strategy derives a computational model capable of achieving rate allocation in a single matrix computation by learning from the optimal power distribution control sequences of past similar operating conditions.To validate the optimization performance and computational efficiency of the proposed method,a standard cycle condition is constructed based on Markov theory for simulation verification.Simulation results indicate that the hydrogen consumption of the proposed method is only 1.61%higher than strategies based on Dynamic Programming(DP),while saving 1.54%equivalent hydrogen consumption compared to traditional Model Predictive Control(MPC)strategies,with significantly lower computational time than MPC strategies.

关 键 词：能量管理策略 多目标优化 动态规划 极限学习机 

分 类 号：U469.72[机械工程—车辆工程]