CN114944662B - 基于支持向量聚类的电动汽车集群并网鲁棒优化调度方法 - Google Patents

基于支持向量聚类的电动汽车集群并网鲁棒优化调度方法 Download PDF

Info

Publication number
CN114944662B
CN114944662B CN202210874839.2A CN202210874839A CN114944662B CN 114944662 B CN114944662 B CN 114944662B CN 202210874839 A CN202210874839 A CN 202210874839A CN 114944662 B CN114944662 B CN 114944662B
Authority
CN
China
Prior art keywords
electric vehicle
model
electric
robust optimization
electric automobile
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202210874839.2A
Other languages
English (en)
Other versions
CN114944662A (zh
Inventor
邵晨旭
周吉
郝珊珊
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Liyang Research Institute of Southeast University
Original Assignee
Liyang Research Institute of Southeast University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Liyang Research Institute of Southeast University filed Critical Liyang Research Institute of Southeast University
Priority to CN202210874839.2A priority Critical patent/CN114944662B/zh
Publication of CN114944662A publication Critical patent/CN114944662A/zh
Application granted granted Critical
Publication of CN114944662B publication Critical patent/CN114944662B/zh
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • HELECTRICITY
    • H02GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02JCIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00Circuit arrangements for ac mains or ac distribution networks
    • H02J3/28Arrangements for balancing of the load in a network by storage of energy
    • H02J3/32Arrangements for balancing of the load in a network by storage of energy using batteries with converting means
    • H02J3/322Arrangements for balancing of the load in a network by storage of energy using batteries with converting means the battery being on-board an electric or hybrid vehicle, e.g. vehicle to grid arrangements [V2G], power aggregation, use of the battery for network load balancing, coordinated or cooperative battery charging
    • BPERFORMING OPERATIONS; TRANSPORTING
    • B60VEHICLES IN GENERAL
    • B60LPROPULSION OF ELECTRICALLY-PROPELLED VEHICLES; SUPPLYING ELECTRIC POWER FOR AUXILIARY EQUIPMENT OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRODYNAMIC BRAKE SYSTEMS FOR VEHICLES IN GENERAL; MAGNETIC SUSPENSION OR LEVITATION FOR VEHICLES; MONITORING OPERATING VARIABLES OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRIC SAFETY DEVICES FOR ELECTRICALLY-PROPELLED VEHICLES
    • B60L55/00Arrangements for supplying energy stored within a vehicle to a power network, i.e. vehicle-to-grid [V2G] arrangements
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00Pattern recognition
    • G06F18/20Analysing
    • G06F18/23Clustering techniques
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00Administration; Management
    • G06Q10/06Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063Operations research, analysis or management
    • G06Q10/0631Resource planning, allocation, distributing or scheduling for enterprises or organisations
    • G06Q10/06312Adjustment or analysis of established resource schedule, e.g. resource or task levelling, or dynamic rescheduling
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00Administration; Management
    • G06Q10/06Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063Operations research, analysis or management
    • G06Q10/0631Resource planning, allocation, distributing or scheduling for enterprises or organisations
    • G06Q10/06315Needs-based resource requirements planning or analysis
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/06Energy or water supply
    • HELECTRICITY
    • H02GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02JCIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00Circuit arrangements for ac mains or ac distribution networks
    • H02J3/007Arrangements for selectively connecting the load or loads to one or several among a plurality of power lines or power sources
    • H02J3/0075Arrangements for selectively connecting the load or loads to one or several among a plurality of power lines or power sources for providing alternative feeding paths between load and source according to economic or energy efficiency considerations, e.g. economic dispatch
    • HELECTRICITY
    • H02GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02JCIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J2203/00Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
    • H02J2203/20Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
    • HELECTRICITY
    • H02GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02JCIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J2310/00The network for supplying or distributing electric power characterised by its spatial reach or by the load
    • H02J2310/50The network for supplying or distributing electric power characterised by its spatial reach or by the load for selectively controlling the operation of the loads
    • H02J2310/56The network for supplying or distributing electric power characterised by its spatial reach or by the load for selectively controlling the operation of the loads characterised by the condition upon which the selective controlling is based
    • H02J2310/58The condition being electrical
    • H02J2310/60Limiting power consumption in the network or in one section of the network, e.g. load shedding or peak shaving
    • HELECTRICITY
    • H02GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02JCIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J2310/00The network for supplying or distributing electric power characterised by its spatial reach or by the load
    • H02J2310/50The network for supplying or distributing electric power characterised by its spatial reach or by the load for selectively controlling the operation of the loads
    • H02J2310/56The network for supplying or distributing electric power characterised by its spatial reach or by the load for selectively controlling the operation of the loads characterised by the condition upon which the selective controlling is based
    • H02J2310/62The condition being non-electrical, e.g. temperature
    • H02J2310/64The condition being economic, e.g. tariff based load management
    • YGENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02TCLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T90/00Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation
    • Y02T90/10Technologies relating to charging of electric vehicles
    • Y02T90/16Information or communication technologies improving the operation of electric vehicles
    • Y02T90/167Systems integrating technologies related to power network operation and communication or information technologies for supporting the interoperability of electric or hybrid vehicles, i.e. smartgrids as interface for battery charging of electric vehicles [EV] or hybrid vehicles [HEV]

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Economics (AREA)
  • Strategic Management (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • General Physics & Mathematics (AREA)
  • Power Engineering (AREA)
  • Marketing (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Data Mining & Analysis (AREA)
  • Quality & Reliability (AREA)
  • Game Theory and Decision Science (AREA)
  • Development Economics (AREA)
  • Health & Medical Sciences (AREA)
  • Educational Administration (AREA)
  • Operations Research (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Computation (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • General Engineering & Computer Science (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Public Health (AREA)
  • Water Supply & Treatment (AREA)
  • General Health & Medical Sciences (AREA)
  • Primary Health Care (AREA)
  • Transportation (AREA)
  • Mechanical Engineering (AREA)
  • Electric Propulsion And Braking For Vehicles (AREA)

Abstract

本发明公开了基于支持向量聚类的电动汽车集群并网鲁棒优化调度方法,包括如下步骤:步骤1,针对电动汽车历史充电数据,构建电动汽车的充电场景和调度模型;步骤2,采用支持向量聚类方法构建电动汽车鲁棒优化不确定集;步骤3,基于各时间段内电动汽车的入网时间和入网时的剩余电量具有不确定性,构建基于支持向量聚类的电动汽车鲁棒优化模型,并采用拉格朗日乘子法将模型转换为线性规划模型;步骤4,电动汽车调度鲁棒模型的求解,输出最优决策变量。本发明所提方法能更准确地描述电动汽车充电的不确定性参数,模型在保证经济性的同时能迅速响应分时电价,具有较好的实用性、重要学术意义和工程实用价值。

Description

基于支持向量聚类的电动汽车集群并网鲁棒优化调度方法
技术领域
本发明涉及电力数据分析领域,尤其涉及一种基于支持向量聚类的电动汽车集群并网鲁棒优化调度方法。
背景技术
在新能源汽车技术快速发展的推动下,近些年电动汽车的市场占有率正在逐年增大,并且预计到2025年电动汽车的销量占比将达到1/4。但是,从电力系统运行角度出发,分析电动汽车的迅速发展时,需要考虑到电动汽车充电的时空不确定性,在电动汽车规模化集群并网后会改变电网的负荷水平,无序的充电可能会造成“峰上加峰”的情况,从而影响电力系统运行的安全性、稳定性。
电动汽车作为一种兼具电源与负荷特性的灵活性资源,如何对电动汽车充放电进行合理调度来响应系统需求,减少电网负荷波动,降低用户充电成本。同时,电动汽车的调度优化本质上是考虑多种不确定条件下的优化问题,主要表现在入网位置、时间及充放电功率上的不确定性。针对电动汽车充电的不确定性,可以通过随机规划、机会约束规划、鲁棒优化等方法来展开深入研究。其中,基于鲁棒优化的方法可以通过不确定集构建优化模型,不需要获得不确定参数的概率分布,求取的解能够满足所有约束条件。为解决常规鲁棒优化模型计算过于保守的问题,采用基于数据驱动的鲁棒优化算法进行不确定集构建,利用不确定参数的历史数据构建一个更加接近实际的不确定集合。
发明内容
针对现有优化算法的不足,本发明针对电动汽车入网时长和电池电量的不确定性,以电动汽车的充放电功率作为决策变量,以所有用户最小充电成本为目标函数,构建集群电动汽车调度模型。首先,基于电动汽车历史充电数据,采用支持向量聚类方法,以包含所有数据的最小超球体作为不确定集,广义交叉核作为核函数,计算电动汽车入网时间和充电时长的不确定集,建立集群电动汽车鲁棒优化调度模型。其次,基于边界支持向量计算不确定集的显示表达式,考虑最差情况下的不等式约束,采用拉格朗日乘子法将模型转换为线性规划模型。最后,采用拉格朗日乘子法求解线性规划模型,得到电动汽车的调度策略。
为了达到上述目的,本发明是通过以下技术方案来实现的:
基于支持向量聚类的电动汽车集群并网鲁棒优化调度方法,其特征在于:包括如下步骤:
步骤1,针对电动汽车历史充电数据,构建电动汽车的充电场景和调度模型;
步骤2,采用支持向量聚类方法构建电动汽车鲁棒优化不确定集;
步骤3,基于各时间段内电动汽车的入网时间和入网时的剩余电量具有不确定性,构建基于支持向量聚类的电动汽车鲁棒优化模型,并采用拉格朗日乘子法将电动汽车鲁棒优化模型转换为线性规划模型;
步骤4,电动汽车鲁棒优化模型的求解,输出最优决策变量。
步骤1中以电动汽车的充放电功率为决策变量,以所有用户最小充电成本为目标函数,构建的集群电动汽车调度模型如下:
Figure 100002_DEST_PATH_IMAGE001
(1)
式中:F为调度日的总充电成本,
Figure 100002_DEST_PATH_IMAGE002
为第i辆电动汽车在k时间段内的充放电功 率,
Figure 100002_DEST_PATH_IMAGE003
为第i辆电动汽车在k时间段内的充电时长,
Figure 100002_DEST_PATH_IMAGE004
k时间段的电价,N为样本总 数;
充放电功率和电池剩余电量的关系及其运行上下限如下:
Figure 100002_DEST_PATH_IMAGE005
(2)
Figure 100002_DEST_PATH_IMAGE006
(3)
Figure 100002_DEST_PATH_IMAGE007
(4)
式中:
Figure 100002_DEST_PATH_IMAGE008
Figure 100002_DEST_PATH_IMAGE009
分别为第i辆电动汽车在k时间段入网时的电池剩余电量、离网时 的电池剩余电量,
Figure 100002_DEST_PATH_IMAGE010
为第i辆电动汽车的充电效率,
Figure 100002_DEST_PATH_IMAGE011
为电动汽车的总容量,
Figure 100002_DEST_PATH_IMAGE012
Figure 100002_DEST_PATH_IMAGE013
分别为第i辆电动汽车的最小电池剩余电量和最大电池剩余电量,
Figure 100002_DEST_PATH_IMAGE014
Figure 100002_DEST_PATH_IMAGE015
分别为第i辆电动汽车的最大充电功率和放电功率。
步骤2的具体操作为:设包含N个数据样本的集合
Figure 100002_DEST_PATH_IMAGE016
Figure 100002_DEST_PATH_IMAGE017
在上标位置标表示 数据样本的索引,下标位置表示向量维度,用非线性映射函数
Figure 100002_DEST_PATH_IMAGE018
将数据样本映射到高维 度空间f,将寻找最小球体转化为下述优化问题:
Figure DEST_PATH_IMAGE019
(5)
式中:R为最小球体半径,N为样本总数,
Figure 100002_DEST_PATH_IMAGE020
为松弛量,v为正则化系数,
Figure 100002_DEST_PATH_IMAGE021
为球体中 心,
Figure 100002_DEST_PATH_IMAGE022
为边界支持向量第
Figure 118736DEST_PATH_IMAGE017
个样本,
Figure 100002_DEST_PATH_IMAGE023
为边界支持向量第j个样本;
引入拉格朗日乘子得到拉格朗日方程
Figure DEST_PATH_IMAGE025
,其中
Figure 100002_DEST_PATH_IMAGE026
Figure 100002_DEST_PATH_IMAGE027
为拉格朗日乘子向量:
Figure 100002_DEST_PATH_IMAGE028
(6)
根据KKT条件得到:
Figure DEST_PATH_IMAGE029
(7)
其中
Figure 100002_DEST_PATH_IMAGE030
Figure 100002_DEST_PATH_IMAGE031
为拉格朗日乘子;
将式(7)带入式(6)转换为等式约束的对偶形式如:
Figure 100002_DEST_PATH_IMAGE032
(8)
式中,K为加权的广义交叉核函数,满足等式
Figure 100002_DEST_PATH_IMAGE033
,αj是拉格朗日 乘子向量元素,u、v为核函数自变量,
Figure DEST_PATH_IMAGE034
为加权矩阵,
Figure 100002_DEST_PATH_IMAGE035
为样本范围,
Figure 100002_DEST_PATH_IMAGE036
Figure 100002_DEST_PATH_IMAGE037
分别为样本的 上下界,同时,满足等式关系
Figure 100002_DEST_PATH_IMAGE038
因此定义支持向量集合SV、边界支持向量集合BSV为:
Figure 100002_DEST_PATH_IMAGE039
(9)
对于任意的数据样本,都在超球体的内部因此数据驱动的不确定集合表示为:
Figure 100002_DEST_PATH_IMAGE040
(10)。
步骤3具体操作为:
步骤3.1,在步骤1中构建电动汽车调度模型基础上,增加对各时间段内EV的入网 时间和入网时的剩余电量的不确定性的考虑,引入辅助决策量
Figure 100002_DEST_PATH_IMAGE042
,得到如下鲁棒优化模型:
Figure DEST_PATH_IMAGE043
(11)
式中:N为样本总数,
Figure 100002_DEST_PATH_IMAGE044
为第i辆电动汽车在k时间段内的充放电功率,
Figure 431293DEST_PATH_IMAGE003
为 第i辆电动汽车在k时间段内的充电时长,
Figure 656738DEST_PATH_IMAGE004
k时间段的电价,
Figure 839458DEST_PATH_IMAGE010
为第i辆电动汽车的充电 效率,
Figure 963403DEST_PATH_IMAGE011
为电动汽车的总容量,
Figure 93033DEST_PATH_IMAGE008
为第i辆电动汽车在k时间段入网时的电池剩余电量,
Figure 969722DEST_PATH_IMAGE012
Figure 870813DEST_PATH_IMAGE013
分别为第i辆电动汽车的最小电池剩余电量和最大电池剩余电量,
Figure 669005DEST_PATH_IMAGE014
Figure 602326DEST_PATH_IMAGE015
分别为第i辆电动汽车的最大充电功率和放电功率。
步骤3.2,基于拉格朗日乘子法将电动汽车鲁棒优化模型线性化,具体操作为:
确定不确定集的显示表达式:
Figure 100002_DEST_PATH_IMAGE045
(12)
将式(8)的核函数代入上式,即可得到
Figure 100002_DEST_PATH_IMAGE046
设两个不确定参数的向量形式分别为
Figure DEST_PATH_IMAGE047
Figure 100002_DEST_PATH_IMAGE048
,则不确定性参数的一般形式为
Figure 100002_DEST_PATH_IMAGE049
,决策变量
Figure 100002_DEST_PATH_IMAGE050
其中
Figure 100002_DEST_PATH_IMAGE051
Figure 100002_DEST_PATH_IMAGE052
,约束条件不等式右侧为
Figure 100002_DEST_PATH_IMAGE053
,得 到约束的一般形式为
Figure 100002_DEST_PATH_IMAGE054
,在约束条件左侧
Figure 100002_DEST_PATH_IMAGE056
引入辅助变量
Figure 100002_DEST_PATH_IMAGE057
,在优化问题中引入拉格朗日乘子
Figure 100002_DEST_PATH_IMAGE058
Figure 100002_DEST_PATH_IMAGE059
Figure 100002_DEST_PATH_IMAGE060
,根据KKT条件,转换为以下对 偶问题:
Figure 100002_DEST_PATH_IMAGE061
(13)
将式(13)化简为与
Figure 100002_DEST_PATH_IMAGE063
相关的函数,将其表示为
Figure 100002_DEST_PATH_IMAGE064
,得到电动汽车调度模型的线性 可求解形式:
Figure 100002_DEST_PATH_IMAGE065
(14)。
步骤4 具体为:根据步骤3得到的线性化鲁棒优化模型,输入电动汽车的样本数据,进行鲁棒优化模型的求解,得到电动汽车的最优最经济调度方案。
本发明的有益效果是:本发明所提方法能更准确地描述电动汽车充电的不确定性参数,模型在保证经济性的同时能迅速响应分时电价,具有较好的实用性、重要学术意义和工程实用价值。本发明针对电动汽车的入网时长和电池电量的不确定性,对电动汽车的充放电功率进行优化调度,引导电动汽车合理有序充放电,提高系统对新能源汽车的消纳能力,充分利用电动汽车等负荷侧的可调节资源,实现“用电行为可引导”,促进系统的低碳化、安全化、高效化发展。
附图说明
图1是基于支持向量聚类的电动汽车集群并网鲁棒优化调度方法流程图。
图2是本发明电动汽车调度鲁棒模型求解流程图。
图3是电动汽车充电时间和电池电量的样本分布。
图4正则化参数为0.05下的基于支持向量聚类的不确定集。
图5正则化参数为0.1下的基于支持向量聚类的不确定集。
图6正则化参数为0.15下的基于支持向量聚类的不确定集。
图7是不同调度方案下的负荷水平。
具体实施方式
为使本发明实施例的目的、技术方案和优点更加清楚,下面将结合本发明实施例中的附图,对本发明实施例中的技术方案进行清楚、完整的描述,以令本领域技术人员参照说明书文字能够据以实施。
如图1所示,本发明的基于支持向量聚类的电动汽车集群并网鲁棒优化调度方法,具体步骤如下:
步骤1,针对电动汽车历史充电数据,构建电动汽车的充电场景和调度模型;
步骤2,采用支持向量聚类方法构建电动汽车鲁棒优化不确定集;
步骤3,基于各时间段内电动汽车的入网时间和入网时的剩余电量具有不确定性,构建基于支持向量聚类的电动汽车鲁棒优化模型,并采用拉格朗日乘子法将电动汽车鲁棒优化模型转换为线性规划模型;
步骤4,电动汽车调度鲁棒模型的求解,输出最优决策变量。
步骤1中,假设电动汽车均通过聚合商统一调度,且充电汽车的活动范围位置固定。以电动汽车的充放电功率为决策变量,以所有用户最小充电成本为目标函数,构建集群电动汽车调度模型。
Figure 100002_DEST_PATH_IMAGE066
(1)
式中:F为调度日的总充电成本,
Figure 980525DEST_PATH_IMAGE044
为第i辆电动汽车在k时间段内的充放电功 率,
Figure 505047DEST_PATH_IMAGE003
为第i辆电动汽车在k时间段内的充电时长,
Figure 790535DEST_PATH_IMAGE004
k时间段的电价。
通过以上调度模型,实现聚合商对电动汽车的统一管理,在考虑用户实际用车需求的情况下,用电动汽车集群的充放电行为来响应分时电价,尽量在低价时安排电动汽车充电,起到削峰填谷的作用,同时能够有效降低充电成本。
同时,需要满足电动汽车电池电量的约束,充电功率需要在电动汽车的安全充电范围内,充电时电池电量最高不能超过电动汽车的电池容量,最低不能少于用户的正常出行电量,充电功率和电池电量的关系及其运行上下限如下式所示(2-4)。
Figure 100002_DEST_PATH_IMAGE067
(2)
Figure 100002_DEST_PATH_IMAGE068
(3)
Figure DEST_PATH_IMAGE069
(4)
式中:
Figure DEST_PATH_IMAGE070
Figure DEST_PATH_IMAGE071
分别为第i辆EV在k时间段入网时的SOC、离网时的SOC,
Figure DEST_PATH_IMAGE072
为第i 辆电动汽车的充电效率,
Figure 543858DEST_PATH_IMAGE011
为电动汽车的总容量,
Figure DEST_PATH_IMAGE073
Figure DEST_PATH_IMAGE074
分别为第i辆电动汽车的 最小电池剩余电量和最大电池剩余电量,
Figure DEST_PATH_IMAGE075
Figure DEST_PATH_IMAGE076
分别为第i辆EV的最大充电功率 和放电功率。
步骤2具体操作如下:
首先,由于支持向量聚类方法属于无监督学习,在满足凸优化的情况下可以对复 杂的高维度不确定问题进行建模,找到一个包含所有样本数据点的最小体积封闭球体来描 述样本数据集合。设包含N个数据样本的集合
Figure DEST_PATH_IMAGE077
Figure DEST_PATH_IMAGE078
在上标位置标表示数据样本 的索引,下标位置表示向量维度。使用非线性映射函数
Figure 286817DEST_PATH_IMAGE018
将数据样本映射到高维度空间f,这样寻找最小球体可转化为下述优化问题:
Figure DEST_PATH_IMAGE079
(5)
式中:R为最小球体半径,N为样本总数,
Figure 513399DEST_PATH_IMAGE020
为松弛量,v为正则化系数,
Figure DEST_PATH_IMAGE080
为球体中 心,
Figure 302495DEST_PATH_IMAGE022
为边界支持向量第
Figure 374356DEST_PATH_IMAGE017
个样本,
Figure 548986DEST_PATH_IMAGE023
为边界支持向量第j个样本。
在此基础上,引入拉格朗日乘子得到拉格朗日方程如式(6)所示。根据KKT(Karush-Kuhn-Tucker)条件得到式(7);
Figure DEST_PATH_IMAGE081
(6)
Figure DEST_PATH_IMAGE082
(7)
其中
Figure DEST_PATH_IMAGE083
Figure DEST_PATH_IMAGE085
为拉格朗日乘子,
Figure DEST_PATH_IMAGE086
Figure DEST_PATH_IMAGE087
其中为拉格朗日乘子向量;
将式(7)带入式(6)转换为等式约束的对偶形式如式(8)所示:
Figure DEST_PATH_IMAGE088
(8)
式中,K为加权的广义交叉核函数,满足等式
Figure DEST_PATH_IMAGE089
,αj是拉格朗日乘子向量 元素,u、v为核函数自变量,
Figure 713513DEST_PATH_IMAGE034
为加权矩阵,
Figure 973593DEST_PATH_IMAGE035
为样本范围,
Figure 317987DEST_PATH_IMAGE036
Figure 97855DEST_PATH_IMAGE037
分别为样本的上下界; 同时,满足等式关系
Figure 135081DEST_PATH_IMAGE038
分析可知,当
Figure DEST_PATH_IMAGE090
Figure DEST_PATH_IMAGE092
,此时
Figure DEST_PATH_IMAGE093
在超球体外部,为支持向量; 当
Figure DEST_PATH_IMAGE094
时,
Figure DEST_PATH_IMAGE095
Figure DEST_PATH_IMAGE096
在超球体上,为边界支持向量。因此定义支持向量集 合SV、边界支持向量集合BSV为:
Figure DEST_PATH_IMAGE097
(9)
对于任意的数据样本,都在超球体的内部,因此数据驱动的不确定集合可表示为:
Figure DEST_PATH_IMAGE098
(10)。
步骤3具体为:
参见图2,首先,在步骤1中构建电动汽车充调度模型基础上如式(1)-(4),增加对 各时间段内电动汽车的入网时间和入网时的剩余电量的不确定性的考虑,引入辅助决策量
Figure DEST_PATH_IMAGE099
,得到如下鲁棒优化模型:
Figure DEST_PATH_IMAGE100
(11)
其次,针对电动汽车样本数据,如式(8)求解核函数及函数最小值时对应的
Figure DEST_PATH_IMAGE101
。如式(9)构建的支持向量集合
Figure DEST_PATH_IMAGE102
,边界支持向 量集合
Figure DEST_PATH_IMAGE103
。将式(8)带入式(10),得到不确定集的显示表达式:
Figure DEST_PATH_IMAGE104
(12)
其中,
Figure DEST_PATH_IMAGE105
为不确定集边界上的样本点;
Figure DEST_PATH_IMAGE106
BSV为边界支持向量集合;
将式(8)中的核函数代入上式,即可得到
Figure DEST_PATH_IMAGE107
,表示由N个数据样本的集合D构成的 电动汽车优化调度模型的不确定集。
设两个不确定参数的向量形式分别为
Figure DEST_PATH_IMAGE109
Figure DEST_PATH_IMAGE110
,则不确定性参数 的一般形式为
Figure DEST_PATH_IMAGE111
,决策变量
Figure DEST_PATH_IMAGE112
,其中
Figure DEST_PATH_IMAGE113
Figure DEST_PATH_IMAGE114
, 约束条件不等式右侧为
Figure DEST_PATH_IMAGE116
,b为约束上限。得到约束的一般形式为
Figure DEST_PATH_IMAGE117
, 在约束条件左侧
Figure DEST_PATH_IMAGE118
引入辅助变量
Figure DEST_PATH_IMAGE119
,在优化问题中引入拉格朗 日乘子
Figure DEST_PATH_IMAGE120
Figure DEST_PATH_IMAGE121
Figure DEST_PATH_IMAGE122
。根据KKT条件,转换为以下对偶问题:
Figure DEST_PATH_IMAGE123
(13)
可将式(13)化简为与
Figure DEST_PATH_IMAGE124
相关的函数,将其表示为
Figure DEST_PATH_IMAGE125
,得到电动汽车调度模型的线 性可求解形式:
Figure DEST_PATH_IMAGE126
(14)
同时引入拉格朗日乘子
Figure DEST_PATH_IMAGE127
,根据KKT条件,求解以下等式:
Figure DEST_PATH_IMAGE128
(15)。
步骤4具体为:
根据建立的电动汽车调度鲁棒模型,输入电动汽车的样本数据,通过如下步骤进行鲁棒模型求解,得到电动汽车的最优最经济调度方案。实现最小成本下的电动汽车充放电策略,进一步配合系统削峰填谷。
1)根据电动汽车样本数据,计算加权矩阵
Figure DEST_PATH_IMAGE129
、参数并且构建核函数K
2)求解边界支持向量集合
Figure DEST_PATH_IMAGE130
3)确定不确定集半径R 2 ,构建不确定集
Figure DEST_PATH_IMAGE131
4)线性化鲁棒优化模型,通过引入变量和KKT条件,将模型的最小化优化目标函数及约束条件转换为线形可解的形式;
5)引入拉格朗日乘子求解,得到最优决策变量,即电动汽车的调度方案。
以包含区域中有20辆电动汽车(EV),时间段为24个,样本集的维度为960,总样本数为500的具体电动汽车调度场景为例。区域的分时电价如表1所示:
Figure DEST_PATH_IMAGE132
场景内的基础负荷与分时电价成正相关,不同EV充电用户充电最低需求的SOC值、最大电池容量、EV充电效率、分别满足N(0.6,0.1)、N(70,10)、N(0.8,0.1)的正态分布。
为更直观分析结果,取样本的2个维度,本文取第1辆EV在第10个时间段的充电时间和入网SOC,样本分布如图3所示。
采用的基于支持向量聚类的不确定集建立了一个不对称的包络形状,紧凑的覆盖 样本集。该不确定集的形状更为复杂,能适应复杂的数据分布,符合电动汽车实际运行规 律。支持向量聚类方法构建的不确定集可以通过调整正则化系数
Figure DEST_PATH_IMAGE134
改变不确定集大小,不 同
Figure 788784DEST_PATH_IMAGE134
值对应的不确定集如图4-6所示。
从图4-6可以看出,当v增大时,包络范围减小,识别的异常样本越多,调度策略保守性降低。对于不同地区层级电网的调度需求,调整正则化参数就可以直接控制最终策略的保守程度。因此,基于支持向量聚类的电动汽车集群并网鲁棒优化可以有效的分辨并剔除,当支持向量随着v的增大逐渐增多时,不确定集的边界也越来越平滑,说明调度模型的复杂度可以根据自适应数据进行调整,避免了主观判断带来的误差。
同时,为分析不同方法得到的不确定集下构建得到的四种集群电动汽车鲁棒优化调度模型在优化效果上的差异,图7所示为四种调度模型优化得到的调度方案在平衡负荷的效果对比,从图7中可以看出,基于支持向量聚类不确定集下的调度方案减小峰谷差的幅度最大,在到达峰值和估值区间时,曲线变化率最大,表明了模型响应灵敏迅速。
表2所示为四种调度方案在优化经济性能上的效果对比:
Figure DEST_PATH_IMAGE135
从表2中可以看出,基于多面体不确定集的调度方案总体充电成本最高,盒型与椭球多面体下的调度方案成本相当;本文提出的方法总充电成本最低,本文的不确定集准确的描述了数据样本,剔除了异常情况,能根据EV入网情况合理安排充放电,降低了成本。
综合上述分析,本发明建立的基于SVC的集群EV鲁棒调度模型相比传统鲁棒优化模型,能主动适应实际数据,解决经典不确定集过于保守的问题。

Claims (2)

1.基于支持向量聚类的电动汽车集群并网鲁棒优化调度方法,其特征在于:包括如下步骤:
步骤1,针对电动汽车历史充电数据,构建电动汽车的充电场景和调度模型;
步骤2,采用支持向量聚类方法构建电动汽车鲁棒优化不确定集;
步骤3,基于各时间段内电动汽车的入网时间和入网时的剩余电量具有不确定性,构建基于支持向量聚类的电动汽车鲁棒优化模型,并采用拉格朗日乘子法将电动汽车鲁棒优化模型转换为线性规划模型;
步骤4,电动汽车鲁棒优化模型的求解,输出最优决策变量;
步骤1中以电动汽车的充放电功率为决策变量,以所有用户最小充电成本为目标函数,构建的电动汽车集群调度模型如下:
Figure DEST_PATH_IMAGE001
(1)
式中:F为调度日的总充电成本,
Figure DEST_PATH_IMAGE002
为第i辆电动汽车在k时间段内的充放电功率,
Figure DEST_PATH_IMAGE003
为第i辆电动汽车在k时间段内的充电时长,
Figure DEST_PATH_IMAGE004
k时间段的电价,N为样本总数;
充放电功率和电池剩余电量的关系及其运行上下限如下:
Figure DEST_PATH_IMAGE005
(2)
Figure DEST_PATH_IMAGE006
(3)
Figure DEST_PATH_IMAGE007
(4)
式中:
Figure DEST_PATH_IMAGE008
Figure DEST_PATH_IMAGE009
分别为第i辆电动汽车在k时间段入网时的电池剩余电量、离网时的电 池剩余电量,
Figure DEST_PATH_IMAGE010
为第i辆电动汽车的充电效率,
Figure DEST_PATH_IMAGE011
为电动汽车的总容量,
Figure DEST_PATH_IMAGE012
Figure DEST_PATH_IMAGE013
分 别为第i辆电动汽车的最小电池剩余电量和最大电池剩余电量,
Figure DEST_PATH_IMAGE014
Figure DEST_PATH_IMAGE015
分别为第i辆电动汽车的最大充电功率和放电功率;
步骤2的具体操作为:设包含N个数据样本的集合
Figure DEST_PATH_IMAGE016
Figure DEST_PATH_IMAGE017
在上标位置表示数据样 本的索引,下标位置表示向量维度,用非线性映射函数
Figure DEST_PATH_IMAGE018
将数据样本映射到高维度空间f,将寻找最小球体转化为下述优化问题:
Figure DEST_PATH_IMAGE020
(5)
式中:R为最小球体半径,N为样本总数,
Figure DEST_PATH_IMAGE021
为松弛量,v为正则化系数,
Figure DEST_PATH_IMAGE022
为球体中心,
Figure DEST_PATH_IMAGE023
为边界支持向量第
Figure 333401DEST_PATH_IMAGE017
个样本,
Figure DEST_PATH_IMAGE024
为边界支持向量第j个样本;
引入拉格朗日乘子得到拉格朗日方程
Figure DEST_PATH_IMAGE026
,其中
Figure DEST_PATH_IMAGE027
Figure DEST_PATH_IMAGE028
为拉格朗日乘子向量:
Figure DEST_PATH_IMAGE030
(6)
根据KKT条件得到:
Figure DEST_PATH_IMAGE031
(7)
其中
Figure DEST_PATH_IMAGE032
Figure DEST_PATH_IMAGE033
为拉格朗日乘子向量元素;
将式(7)带入式(6)转换为等式约束的对偶形式如:
Figure DEST_PATH_IMAGE035
(8)
式中,K为加权的广义交叉核函数,满足等式
Figure DEST_PATH_IMAGE036
,αj是拉格朗日乘子 向量元素,u、v为核函数自变量,
Figure DEST_PATH_IMAGE037
为加权矩阵,
Figure DEST_PATH_IMAGE038
为样本范围,
Figure DEST_PATH_IMAGE039
Figure DEST_PATH_IMAGE040
分别为样本的上下 界;同时,满足等式关系
Figure DEST_PATH_IMAGE041
;
因此定义支持向量集合SV、边界支持向量集合BSV为:
Figure DEST_PATH_IMAGE042
(9)
对于任意的数据样本,都在超球体的内部,因此数据驱动的不确定集表示为:
Figure DEST_PATH_IMAGE044
(10);
步骤3具体操作为:
步骤3.1,在步骤1中构建电动汽车调度模型基础上,增加对各时间段内电动汽车的入 网时间和入网时的剩余电量的不确定性的考虑,引入辅助决策量
Figure DEST_PATH_IMAGE045
,得到如下鲁棒优化模 型:
Figure DEST_PATH_IMAGE046
(11)
步骤3.2,基于拉格朗日乘子法将电动汽车鲁棒优化模型线性化,具体操作为:
确定不确定集的显示表达式:
Figure DEST_PATH_IMAGE048
(12)
其中,
Figure DEST_PATH_IMAGE049
为不确定集边界上的样本点;
Figure DEST_PATH_IMAGE050
BSV为边界支持向量集合
将式(8)的核函数代入上式,即可得到
Figure DEST_PATH_IMAGE051
设两个不确定参数的向量形式分别为
Figure DEST_PATH_IMAGE052
Figure DEST_PATH_IMAGE053
,则不确定性参数的一般形式为
Figure DEST_PATH_IMAGE054
,决策变量
Figure DEST_PATH_IMAGE055
其中
Figure DEST_PATH_IMAGE056
Figure DEST_PATH_IMAGE057
,约束条件不等式右侧为
Figure DEST_PATH_IMAGE058
,得到约 束的一般形式为
Figure DEST_PATH_IMAGE059
,b为约束上限,在约束条件左侧
Figure DEST_PATH_IMAGE060
引入 辅助变量
Figure DEST_PATH_IMAGE061
,在优化问题中引入拉格朗日乘子
Figure DEST_PATH_IMAGE062
Figure DEST_PATH_IMAGE063
Figure DEST_PATH_IMAGE064
,根据KKT条件,转换 为以下对偶问题:
Figure DEST_PATH_IMAGE065
(13)
将式(13)化简为与
Figure DEST_PATH_IMAGE066
相关的函数,将其表示为
Figure DEST_PATH_IMAGE067
,得到电动汽车调度模型的线性可求 解形式:
Figure DEST_PATH_IMAGE068
(14)。
2.根据权利要求1所述基于支持向量聚类的电动汽车集群并网鲁棒优化调度方法,其特征在于:步骤4 具体为:根据步骤3得到的线性化鲁棒优化模型,输入电动汽车的样本数据,进行鲁棒优化模型的求解,得到电动汽车的最优最经济调度方案。
CN202210874839.2A 2022-07-25 2022-07-25 基于支持向量聚类的电动汽车集群并网鲁棒优化调度方法 Active CN114944662B (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210874839.2A CN114944662B (zh) 2022-07-25 2022-07-25 基于支持向量聚类的电动汽车集群并网鲁棒优化调度方法

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210874839.2A CN114944662B (zh) 2022-07-25 2022-07-25 基于支持向量聚类的电动汽车集群并网鲁棒优化调度方法

Publications (2)

Publication Number Publication Date
CN114944662A CN114944662A (zh) 2022-08-26
CN114944662B true CN114944662B (zh) 2022-10-21

Family

ID=82910643

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210874839.2A Active CN114944662B (zh) 2022-07-25 2022-07-25 基于支持向量聚类的电动汽车集群并网鲁棒优化调度方法

Country Status (1)

Country Link
CN (1) CN114944662B (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117578488B (zh) * 2023-11-15 2024-05-03 山东大学 考虑不确定性的电动汽车实时灵活性评估方法及系统

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112132351A (zh) * 2020-09-28 2020-12-25 华北电力大学 电动汽车充电负荷聚类预测方法及装置
CN112308386B (zh) * 2020-10-19 2024-01-23 武汉理工大学 一种价格与激励需求响应下电动汽车负荷聚集商调度方法

Also Published As

Publication number Publication date
CN114944662A (zh) 2022-08-26

Similar Documents

Publication Publication Date Title
Huang et al. A clustering based grouping method of nearly zero energy buildings for performance improvements
CN113572157B (zh) 一种基于近端策略优化的用户实时自治能量管理优化方法
CN104578120B (zh) 分布式储能系统的优化配置方法
CN112131733B (zh) 计及电动汽车充电负荷影响的分布式电源规划方法
CN108964050A (zh) 基于需求侧响应的微电网双层优化调度方法
CN110994694B (zh) 计及差异化需求响应的微电网源荷储协调优化调度方法
CN108448620A (zh) 基于综合性能指标的高渗透率分布式电源集群划分方法
CN102855412A (zh) 一种风电功率预测方法及其装置
CN113794199B (zh) 一种考虑电力市场波动的风电储能系统最大收益优化方法
CN114156951B (zh) 一种源网荷储系统的控制优化方法和装置
CN114944662B (zh) 基于支持向量聚类的电动汽车集群并网鲁棒优化调度方法
Lujano-Rojas et al. Contract design of direct-load control programs and their optimal management by genetic algorithm
CN117077960A (zh) 一种区域综合能源系统日前调度优化方法
CN111428766A (zh) 一种高维海量量测数据的用电模式分类方法
CN117621898B (zh) 考虑电网电价的智慧停车场充电桩充电控制方法及系统
CN110991764A (zh) 一种综合能源系统日前滚动优化方法
CN116826710A (zh) 基于负荷预测的削峰策略推荐方法、装置及存储介质
Rodemann A many-objective configuration optimization for building energy management
Hussain et al. Energy management of buildings with energy storage and solar photovoltaic: A diversity in experience approach for deep reinforcement learning agents
CN114462854A (zh) 含新能源和电动汽车并网的分层调度方法及其系统
CN117196884A (zh) 一种虚拟电厂优化控制方法及系统
CN115115145B (zh) 分布式光伏智能住宅的需求响应调度方法及系统
CN115441494A (zh) 基于柔性直流互联系统的换流站容量优化配置方法和装置
Dolatabadi et al. SFNAS-DDPG: A Biomass-based Energy Hub Dynamic Scheduling Approach via Connecting Supervised Federated Neural Architecture Search and Deep Deterministic Policy Gradient
CN118232369B (zh) 一种基于多目标求解的用户侧储能控制方法及系统

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant
CB03 Change of inventor or designer information

Inventor after: Shao Chenxu

Inventor after: Zhou Ji

Inventor after: Hao Shanshan

Inventor after: Jin Wenxing

Inventor after: Xie Hongfu

Inventor before: Shao Chenxu

Inventor before: Zhou Ji

Inventor before: Hao Shanshan

CB03 Change of inventor or designer information