基于Stacking融合模型的PHEV复合储能系统实时能量分配策略  

Real-time Energy Distribute Strategy of PHEV Hybrid Energy Storage System Based on Stacking Fusion Model

作　　者：吴忠强[1] 马博岩 WU Zhongqiang;MA Boyan(College of Electrical Engineering,Yanshan University,Qinhuangdao,Hebei 066004,China)

机构地区：[1]燕山大学电气工程学院,河北秦皇岛066004

出　　处：《计量学报》2024年第1期73-81,共9页Acta Metrologica Sinica

基　　金：河北省自然科学基金(F2020203014)。

摘　　要：为了解决插电式混合动力汽车单一电池低比功率、无法响应暂态功率需求的问题,设计了电池和超级电容并联式复合储能系统。同时针对采用动态规划法优化负载电流分配时缺乏实时性的问题,利用不同驱动工况下动态规划优化的结果构成训练集进行训练,并综合GRU网络以及XGBoost算法,提出了一种Stacking集成学习框架下多模型融合的能量分配策略。仿真结果表明,与仅使用单一电池的储能系统相比,基于Stacking融合模型的实时能量分配系统在UDDS和US06两种循环工况下,电池峰值电流分别降低了48.7%和50.8%,有效削弱了电池的峰值电流,提升了电池的整体性能。In order to solve the problem of low specific power of the single-power battery that unable to respond to transient power demand in plug-in hybrid electric vehicle,a hybrid energy storage system with parallel connection between battery and supercapacitor was designed.To the problem of lack real time when using dynamic programming(DP)method to optimize load current distribution,a training set composed of DP optimization results under different driving conditions data sets was used for training,and a multi-model fusion energy distribute strategy under Stacking integrated learning framework was proposed by integrating GRU network and XGBoost algorithm.The simulation results show that compared with the energy storage system using only a single battery,the real-time energy distribution system based on Stacking fusion model can reduce the peak current by 48.7% and 50.8% under UDDS and US06 cycles,respectively,effectively reduces the peak current of the battery and improves the overall performance of the battery.

关 键 词：电学计量 复合储能系统 插电式混合动力汽车 动态规划 XGBoost Stacking融合模型 

分 类 号：TB971[一般工业技术—计量学]