首页|基于蒙特卡洛算法的电动汽车充电负荷概率分布场景三维分析

基于蒙特卡洛算法的电动汽车充电负荷概率分布场景三维分析

扫码查看
电动汽车充电负荷受到车主出行的随机性影响,使负荷随机波动性较大,从而造成了与实际情况不相符的概率分布.针对该问题,提出了基于蒙特卡洛算法的电动汽车充电负荷概率分布场景三维分析方法.根据电动汽车负荷时空分布特性,采用蒙特卡洛算法设计电动汽车出行链结构,通过对各个路段行驶里程来判定蓄电池荷电状态,分析充电需求负荷.计算连续时段集负荷联合分布分位点,抽取负荷分布三维场景,计算独立分布场景、联合分布场景 2 个时段负荷样本相似性,采用K中心聚类法聚类抽取的三维场景负荷数据,更好地表征场景波动性.利用蒙特卡洛统计方法拼接相邻两个连续时段集负荷,通过分析行程起始、出行、抵达时刻荷电状态,实现三维场景分析,有效地避免极值影响,并防止在拼接后出现场景不足情况.由实验结果可知,该方法住宅区负荷概率主要集中在 18∶00~24∶00,工作区负荷概率主要集中在 08∶00~11∶00,与实际概率分布区域一致,验证了该方法的有效性.
Three Dimensional Analysis of Electric Vehicle Charging Load Probability Distribution Scene based on Monte Carlo Algorithm
Abs tract:The charging load of electric vehicles is affected by the randomness of vehicle owners'travel,which makes the load fluctuate randomly,thus resulting in a probability distribution inconsistent with the actual situation.To solve this problem,a three-dimensional analysis method based on Monte Carlo algo-rithm for electric vehicle charging load probability distribution scene is proposed.According to the time-space distribution characteristics of electric vehicle load,the Monte Carlo algorithm is used to design the travel chain structure of electric vehicles.The charging state of the battery is determined by the mileage of each road section,and the charging demand load is analyzed.Via calculating the load joint distribution quantile of the continuous time interval set,extracting the load distribution three-dimensional scene,calcu-lating the similarity of load samples of the independent distribution scene and the joint distribution scene,and clustering the extracted three-dimensional scene load data with the K-center clustering method,the scene volatility was better characterized.When splicing,the Monte Carlo statistical method is used to col-lect loads in two consecutive periods.By analyzing the state of charge at the time of travel start,travel and arrival,the three-dimensional scene analysis is realized to effectively avoid the impact of extreme values and prevent scene shortage after splicing.The experimental results show that the residential area load probability of this method is mainly concentrated in 18:00-24:00,and the working area load probability is mainly concentrated in 08:00-11:00,which is consistent with the actual probability distribution area,which verifies the effectiveness of this method.

Monte Carlo algorithmElectric vehicleCharging loadProbability distributionScene 3D

曹晓庆、杨轩、李林、陈迪、陈芳

展开 >

武汉华源电力设计院有限公司 湖北 武汉 430058

蒙特卡洛算法 电动汽车 充电负荷 概率分布 场景三维

2024

小型内燃机与车辆技术
天津大学

小型内燃机与车辆技术

CSTPCD
影响因子:0.261
ISSN:1671-0630
年,卷(期):2024.53(3)
  • 14