ArticlePDF Available

XLarge-scale wind power stochastic optimization scheduling method considering flexible load peaking

Authors:

Abstract

Large-scale access to the uncertainty of wind power, has brought new challenges to the power system scheduling. This paper sees flexible load as a peaking resource and presents a large-scale wind power stochastic optimization considering the peak flexible load regulation This method takes into account the probability characteristics of wind power and the peak regulation characteristics of flexible load, constructs the mathematical expression revealing the risk of power system operation, measures the system's safety and economy comprehensively, and establishes a large scale wind power stochastic scheduling model considering flexible peak load regulation. Imperialistic competition algorithm is used to find the solution to the model and 38 machine example simulation verifies the correctness and validity of the model. The results show that the method in this paper can coordinate the operation cost and risk of power system effectively, and the method can also improve the system's operation efficiency and enhance the assimilative capacity of wind power.
2013 11 Vol.28 No. 11
28 卷第 11 TRANSACTIONS OF CHINA ELECTROTECHNICAL SOCIETY Nov. 2013
考虑柔性负荷调峰的
大规模风电随机优化调度方法
刘涤尘
(武汉大学电气工程学院 武汉 430072
摘要 不确定性风电大规模接入,给电力系统调度带来了新挑战。本文将柔性负荷作为一种
调峰资源,提出了考虑柔性负荷调峰的大规模风电随机优化调度方法。该方法考虑了风电的概率
特性,计及柔性负荷的调峰特点,构建了揭示电力系统运行风险的数学表达方式,综合度量了系
统的安全性和经济性,建立了考虑柔性负荷调峰的大规模风电随机调度模型;利用帝国竞争算法
对模型进行求解,38 机算例仿真验证了所建立模型的正确性和有效性。结果表明:考虑柔性负
荷调峰的大规模风电随机调度方法可以有效协调电力系统的运行成本与风险,提高了系统的运行
效益,增强了系统对风电的消纳能力。
关键词:风力发电 柔性负荷 风险约束 不确定性优化调度 帝国竞争算法
中图分类号:TM315
Large-Scale Wind Power Stochastic Optimization Scheduling Method
Considering Flexible Load Peaking
Yang Nan Wang Bo Liu Dichen Zhao Jie Wang He
Whuhan University Whuhan 430072 China
Abstract Large-scale access to the uncertainty of wind power, has brought new challenges to the
power system scheduling. This paper sees flexible load as a peaking resource and presents a large-scale
wind power stochastic optimization considering the peak flexible load regulation This method takes
into account the probability characteristics of wind power and the peak regulation characteristics of
flexible load, constructs the mathematical expression revealing the risk of power system operation,
measures the system's safety and economy comprehensively, and establishes a large scale wind power
stochastic scheduling model considering flexible peak load regulation. Imperialistic competition
algorithm is used to find the solution to the model and 38 machine example simulation verifies the
correctness and validity of the model. The results show that the method in this paper can coordinate the
operation cost and risk of power system effectively, and the method can also improve the system's
operation efficiency and enhance the assimilative capacity of wind power.
KeywordsWind power, flexible load, risk constraints, uncertainty optimized scheduling,
imperialistic competition algorithm
1 引言
近年来我国风电的战略地位将由目前的补充能
源逐步上升为一种替代能源乃至主导能源,可以预
见,我国即将形成一个风电大规模接入的互联大电
[1,2]。然而,出力具有随机性的风电大规模利用,
在给电网的安全稳定运行带来新挑战的同时,也导
致系统各个环节的运行成本上升。因此,需将风电
大规模开发、输送和消纳纳入电力发展统一规划,
研究大规模风电调度的新方法[3]
国家自然科学基金(51207113,国家电网公司大电网重大专项资
助项目课题(SGCC-MPLG029-2012)和武汉大学博士自主科研基
金(274703)资助项目。
收稿日期 2013-01-31 改稿日期 2013-05-14
DOI:10.19595/j.cnki.1000-6753.tces.2013.11.032
232 2013 11
现有的大规模风电调度,是将风电与火电“打
捆”运行[4-6]这种调度方法没有考虑风电出力的概
率特性,因此难以反映系统实际运行的风险水平,
无法统筹协调风电出力波动的严重性和可能性,因
此,基于不确定性理论的风电调度方法更符合实际。
文献[7]在风速预测的基础上,应用随机规划理论建
立了含风电场电力系统动态经济调度模型。文献[8]
以燃煤机组的发电成本最小和污染气体排放量最小
为目标构建了考虑环境因素的风电多目标随机调度
模型。但上述文献均未将电力系统需求侧调峰资源
纳入风电随机调度模型之中。
柔性负荷(Flexible LoadFL)包括可中断负
荷和激励负荷,因其具有良好的调峰性能使之日益
受到学者关注[9]。但现有研究多倾向利用可中断负
荷优化备用市场[10]解决输电线路阻塞问题[11],对
于利用柔性负荷平抑风电出力波动性并将之融入系
统随机调度模型的研究较少。
本文建立了柔性负荷的成本费用函数,并将柔
性负荷作为一种可调度的调峰资源,纳入到电力系
统的调度体系之中,针对风电出力的不确定性,构
建了揭示系统运行风险的数学表达方式,综合度量
了系统的运行风险水平,最后建立了考虑柔性负荷
调峰的大规模风电随机调度模型,并通过帝国竞争
算法对模型进行求解。基于 38 机算例仿真,验证了
模型及算法的正确性和有效性。与现有调度方法相
比,该模型既充分体现了柔性负荷的潜在调峰效益,
又能够使调度决策具备兼顾电力系统安全性、经济
性和可靠性的全局视角。
2 考虑风电概率特性的风险备用分析
2.1 风电场的出力特性及其概率分布
研究表明,描述短期风速概率特性多用正态分
[12,13]
2
2
1()
() exp
2π2
v
vv
vv
qv


1
式中,v为预测平均风速;v为风速; v
为预测风
速误差的标准差。
在忽略风电场尾流和电气损耗前提下风电场的
输出功率与风电场的风速有关,两者的数学关系为[14]
in out
3
3
WR in
WWRinR
33 33
Rin Rin
WR R out
00
()
vv vv
Pv
Pv v P v v v
vv vv
Pvvv


≤,
≤≤
<>
<
2
式中,v为风机轮毂高度处的风速;vin 为切入风速;
vout 为切出风速;vR为额定风速;PW为风机的输出
功率;PWR 为额定输出功率。
根据风速概率分布和风电功率输出特性可知,
风电的输出功率是一个混合型随机变量,分别求解
其离散部分、连续部分的概率密度函数 W
()qP
2.1.1 离散部分
1)当 in
0vv<out
vv>
in
out
out
in
0
(0) ( )d ( )d 1
v
vv
vvv
vv
vv
qqvvqvv


  
 
 
 

3
式中, ()v
为风速 v的概率分布函数。
2)当 R out
vvv<
out
R
out
WR
() ()d
vR
v
vvv
vv vv
qP q v v






4
2.1.2 连续部分
1)当 in R
vvv≤≤ 时,式(2连续可微,根
据随机变量函数分布的求解理论[15]
WWW
() (()) ()
v
qP q vP v P
 5
式中,v(PW) 为式(2)的反函数。
2)当 PW0PWPWR 时,根据式2)可
q(PW)=0 6
综上所述,风电场输出功率的概率密度函数为
q(PW)=
WWWR
out
in
W
out R
WWR
2
1/3
W
2/3
W
1/3 2
WWR
00
10
1()exp
32π2
0
vv
vv
vv
PPP
vv
vv P
vv vv PP
Pb v
a
Pb
a
PP





 
 
 
 






















<>
<<
7
式中, WR
33
Rin
P
avv
3
in
WR
33
Rin
v
bP
vv
2.2 旋转备用的风险指标
为保证安全稳定运行,电力系统需保持一定容
量的旋转备用,包括正旋转备用和负旋转备用两种。
为研究系统备用的风险水平,计算合理的系统备用
容量,本文提出旋转备用的风险指标,主要分为正
28 卷第 11 考虑柔性负荷调峰的大规模风电随机优化调度方法 233
旋转备用风险指数和负旋转备用风险指数两种。
正旋转备用风险是由实际可利用的风电出力少
于风电平均出力导致的。其风险指数为上述情况导
致正旋转备用不足的概率,定义为
dWWWGpWWLp
{| }QqPP P R P P R<< 8
式中,Qd为系统正旋转备用风险指数; W
P为风电
的平均出力;RGp 为系统可以提供的正旋转备用;
RLp 为系统常规机组的正旋转备用需求。
由式(7)和式(8)可得
W
WLpGp
WW Gp Lp
d
WW Gp Lp
()d
()d
P
PR R
qP P R R
Q
qP P R R



<
9
负旋转备用风险是由风电的实际出力大于平均
出力导致的。其风险指数为上述情况导致负旋转备
用不足的概率,定义为
uWWWGnWWLn
{| }QqPP PR PPR>< 10
式中,Qu为系统负旋转备用风险指数;RGn 为系统
可以提供的负旋转备用;RLn 为系统常规机组的负旋
转备用需求。
由式(7)和式(10)可得
W
Gn W Ln
W W Gn Ln
u
W W Gn Ln
()d
()d
P
RPR
qP P R R
Q
qP P R R



<
11
3 考虑柔性负荷调峰的大规模风电随机
调度模型
3.1 柔性负荷的成本费用函数
传统的需求侧控制只在负荷高峰时段进行,即
可中断负荷控制。本文提出了柔性负荷控制策略,
在原有可中断负荷控制基础上增加激励负荷控制策
略,用户与电力公司签订负荷控制协议,在系统峰
时或谷时的固定时段内减少或增加他们的用电需求
并获取相应补偿,使柔性负荷同时具有“削峰”、
“填谷”的双重调峰效益。
可中断负荷的赔付函数通常用线性模型描述[16,17]
然而线性赔付模型的费率是一个与中断负荷容量无
关的常数,不利于提高用户增加中断负荷容量的积
极性,且无法体现用户的停电意愿。本文提出柔性
负荷的非线性成本费用模型可有效弥补上述缺陷。
可中断负荷的补偿成本函数为
2
III Ij1I 2I 2II
1
() ( )
N
t t t t jt jt jt j
j
F, UP P P


UP 12
式中,N为系统可中断负荷用户数量; II1
[,,
tt
U
U
II
,]
j
tNt
UU为可中断负荷用户的状态向量,其中
I0
jt
U
表示不中断 j用户的负荷; 1
Ijt
U表示中断
j用户的负荷; II1 I I
[,,, ]
ttjtNt
PPP
P为中断负荷的
容量向量;
1
2为赔偿系数; Ij
为可中断负荷用
户的停电意愿因子[18]
激励负荷的激励成本函数为
2
HHH H1H 2H 2HH
1
() ( )
D
ttt kt kt kt ktk
k
F, UPPP


UP
13
式中,系统激励负荷用户数量为 DHH1
[,,
tt
U
U
HH
,]
kt Dt
UU为激励负荷用户的状态向量,其中
H0
kt
U
表示不增加 k用户的负荷, H1
kt
U
表示增
k用户的负荷; HI1 I H
[,, , ]
ttktDt
PPP
P为增加负
荷的容量向量;
1
2为激励系数; Hk
为激励负
荷用户的增加负荷意愿因子。
3.2 目标函数
系统发电侧成本函数为
GGG G G G G1 G
1
()[()(1)()]
M
t tt ititit it it itit
i
F, UYPUUSP

UP
14
式中,M为系统中的发电机组数; GG1 G
[,,,
ttit
PPP
G]
M
t
P为机组的出力向量; G
2
GG
() it
it it i i it i
YP P P
 

为第 i号发电机组在 t时段的运行成本,其中
i
i
i为机组运行成本参数; /
G01
() (1e)
i
it it i i
SP S S
 
为第 i号发电机组在 t时段的启停成本, 0i
S1i
S
i为启停成本参数,
为发电机组的停机时间;
GG1 G G
[,,, ]
tt itMt
UUU
U 表示第 i号机组在t时段
的运行状态, G0
it
U
表示发电机组处于停机状态,
G1
it
U
表示发电机组处于开机状态。
基于前文分析,考虑柔性负荷调峰的优化调度
模型的运行成本应当包括:可中断负荷的补偿成本、
激励负荷的激励成本、系统发电成本,即
III GGG H HH
1
min [ ( , ) ( , ) ( , )]
T
ttt t t t t t t
t
FF F F

UP U P U P
15
式中,T为系统调度周期。
3.3 约束条件
3.3.1 系统功率平衡约束
GG W L II H H
111
MND
it it t t jt jt kt kt
ijk
UP P P UP U P

 

16
234 2013 11
式中,PLt为系统在t时段的发电负荷;PWt为风电
机组在t时段的输出功率。
3.3.2 常规发电机组约束
1)发电机组出力上下限约束
Gmin G Gmaxiiti
PPP≤≤ 17
式中, Gmini
PGmaxi
P分别为发电机组出力下限和上
限。
2)最大启停次数约束
24
GG1
1
it it i
t
UU M
18
式中,Mi为机组在调度周期内最大允许启停次数。
3)机组爬坡速率约束
dGG1 r
60 60
iitit i
rPP r
≤≤ 19
式中,rdirri分别为机组每分钟有功输出的最大下
降速度和最大上升速度。
3.3.3 柔性负荷约束
1)柔性负荷的限值约束
Imin I Imaxjjtj
PPP≤≤ 20
式中,PIjmin 为用户 jt时段可中断负荷的下限值,
PIjmax 为用户 jt时段可中断负荷的上限值。
2)激励负荷的限值约束
Hmin H Hmaxkktk
PPP≤≤ 21
式中,PHkmin 为用户 kt时段增加负荷的下限值,
PHkmax 为用户 kt时段增加负荷的上限值。
3.3.4 旋转备用风险约束
1)正旋转备用的风险约束:
为保证系统可靠运行,必须将正旋转备用风险
指数约束在允许的范围内,不能超过给定的风险槛
值。
d
Q
22
式中,
为旋转备用风险槛值,系统调度部门可利
用年总费用最小法获取可靠性标准后换算而得,通
常取 010%之间。
系统在实际运行过程中,要求系统提供的旋转
备用需满足常规机组的备用需求,即 RGpRLp。则
式(22)可变为
WLpGp
dGp W W
() ()d
PR R
QR qP P


23
式中,系统可提供的正旋转备用为
Gp G G max I I max L
11
MN
it i jt j
ij
RUP UPP



2)负旋转备用的风险约束:
系统同样需要将负旋转备用风险指数约束在允
许的范围之内。
WGnLn
uGn W W
() ()d
PR R
QR qP P


24
式中,系统可提供的负旋转备用为
Gn L G G min H H min
11
MD
it i kt k
ik
RP UP UP

 

3.4 模型分析
为进一步分析模型的本质特征,将模型抽象表
达为紧凑型公式如下:
IGHIGH
IGHIGH
IGHIGH
IGHIGH
min ( , , , )
st. ( , , , ) 0
(, , , )0
( , , , )
tttttt
tttttt
tttttt
tttttt
F
h
g
q
U U U P ,P ,P
UU U P,P,P
U U U P ,P ,P
U U U P ,P ,P
25
式中, IGHIGH
(, , , )
tttttt
hU U U P ,P ,P 为决策变量的等式
约束; IGHIGH
(, , , )
tttttt
gU U U P ,P ,P 为决策变量的不等
式约束; IGHIGH
(, , , )
tttttt
qU U U P ,P ,P 为决策变量的风
险约束。
对模型进一步分析可知:
1)若不计柔性负荷调度,且 0
,则模型
变为传统的大规模风电确定性调度模型。
2)若计及柔性负荷调度,且 0
,则模型
变为考虑柔性负荷调峰的大规模风电确定性调度模
型。
3)若计及柔性负荷调度,且 0
,则模型
为考虑柔性负荷调峰的大规模风电随机调度模型。
4 算例仿真
4.1 算法
本文采用帝国竞争算法Imperialistic Competition
AlgorithmICA[19]对模型进行求解。ICA 算法是
一种借鉴了人类政治社会殖民阶段帝国之间相互竞
争并占领其殖民地过程的一种全局性优化的进化算
法。该算法的种群个体称为国家,根据权力的大小
分为两类:殖民地和帝国,算法主要包括:帝国形
成、吸收殖民地、帝国竞争、帝国消亡等四个主要
环节。
1)帝国形成。由 h维决策变量组成国家
country=[x1,x2,,xh],其函数值为 fcountry,定义第 n
个国家的标准化权力为
28 卷第 11 考虑柔性负荷调峰的大规模风电随机优化调度方法 235
1
max{ }
(max{})
ni
m
nm
ii
m
i
ff
P
f
f
26
根据标准化权力的大小,将权利最大的 kC个国
家作为帝国,其余国家作为殖民地随机分配给各帝
国。
2吸收殖民地。帝国周围的殖民地向帝国靠
近,设殖民地移动的距离 l服从均匀分布 l
(0, )
D
Ul
,其中,
1
D
l为帝国与其殖民地之
间的距离;殖民地移动方向和其与帝国的连线呈偏
移夹角
其服从分布
(,)U
其中,
为偏
移夹角调整参数。在吸收殖民地过程中,若出现帝
国的标准化权力值小于殖民地的情况,则将该帝国
的位置与其殖民地所在的位置进行交换。
3)帝国竞争。定义帝国总权力值为
C
Acol col
()max{}
kk i
ki
k
Cf f f f

  27
式中,fk为帝国 k的目标函数值;
为权重参数;
colk
f
为帝国 k占有殖民地的目标函数平均值。
从总权力最弱的帝国中挑选出若干(一般为一
个)弱小的殖民地,按一定概率分配给其他 C1k
帝国,第 j个帝国的占有概率为
C
A
1
A
1
j
jk
i
i
C
P
C
28
4帝国消亡。当权力较小的帝国经过帝国竞
争后,其拥有的所有殖民地全部都被权力更强的帝
国占有,则定义该帝国已经灭亡,并消除其位置。
当帝国竞争结束后,仅存在一个帝国,且其余所有
殖民地都由该帝国占有,则算法停止,输出最优解。
否则,返回(2
4.2 算例
本文以具有并网风电场的 38 机电力系统[20,21]
为例,对模型进行仿真验证。ICA 算法参数设置为:
国家数量为 400,初20 个,
=1.75
=0.2
π/2
常规机组正旋转备用需求为系统最大负荷
8%,负旋转备用需求为系统最小负荷的 2%,旋
转备用风险指标为 0.01。可中断负荷的赔偿系数
1
=0.02$/[(MW)2·h]2
=4$/(MW·h),激励负荷的
激励系数 1
=0.03$/[(MW)2·h]2
=7$/(MW·h),风
场共有 600 台额定功率为 1MW 的风力发电机,假
设风力发电机组不提供旋转备用且不考虑其强迫停
运的可能性,风机的相关参数为 vin=3m/svout=
25m/svR=15m/s,风速预测误差的标准差为 0.5
单台风机实测出力数据如图 1所示。
1 风力发电机功率输出曲线
Fig.1 Wind turbine power output curve
柔性负荷用户参数见表 1
1 柔性负荷用户参数
Tab.1 Flexible load users’ parameters
用户 高峰时段 低谷时段
I 可中断负荷上限/MW
H 激励负荷上限/MW
1 0.05 90 0.08 72
2 0.12 132 0.24 162
3 0.25 150 0.34 180
根据 3.4 节分析,确立三种调度方式如下:
方式 1不计柔性负荷调度且令
=0此时风
电的正负旋转备用需求为风电出力的最大波动幅
值,本文取其装机容量的 30%
方式 2:计入柔性负荷调度,且令
=0
方式 3:计入柔性负荷调度,且令
=0.01
4.3 仿真结果分析
4.3.1 发电费用
忽略柔性负荷用户内部的负荷分配,计算三种
运行方式的总成本见表 2
2 系统运行总成本
Tab.2 Power generation costs
(单位:k$
发电成本 可中断负荷
调度费
激励负荷
调度费 总成本
方式 1 191 934 0 191 934
方式 2 190 876 11 20 190 907
方式 3 189 757 11 20 189 788
由表 2可知,柔性负荷参与调峰之后,系统总
成本较其未参与调峰时降低了 0.54%,其中,可中
断负荷节约系统发电成本 742k$激励负荷节约系
统发电成本 316k$;而考虑风电出力随机性,采用
236 2013 11
计入柔性负荷调峰的随机调度模型,则使系统总成
本降低了 1.12%
仿真结果表明,柔性负荷具有削峰填谷的调峰
效益,可以降低机组因参与调峰产生的运行成本。
考虑风电出力随机性的调度模型协调了电力系统运
行的安全性和经济性,降低电力系统的运行总成本。
为研究柔性负荷参数对计算结果的影响,另设
计两种调度方式如下:
方式 4将算例中柔性负荷的赔偿/激励系数
1
2
1
2增加为原来的 10 倍,用户意愿因子
I
H不变,考虑柔性负荷调度,且令
=0
方式 5:将算例中的用户意愿因子
I
H增加
为原来的 5倍,柔性负荷的赔偿/激励系数不变,
虑柔性负荷调度,且令
=0
计算两种方式下的系统总成本见表 3
3 系统运行总成本
Tab.3 Power generation costs
(单位:k$
发电成本 可中断负荷
调度费
激励负荷
调度费 总成本
方式 4 191 461 66 122 191 649
方式 5 190 655 9 7 190 671
由表 3可知,方式 4的总成本比方式 2增加了
0.39%,而方式 5的总成本比方式 2降低了 0.12%
结果表明,在现有的柔性负荷由电力系统统筹调用
的模式下,赔偿/激励系数和用户意愿因子会影响电
网调用柔性负荷的成本和积极性,并间接影响柔性
负荷的调峰效益,其中赔偿/激励系数与柔性负荷的
调峰效益负相关,用户意愿因子与其正相关。需要
指出的是,上述结论是在电力系统直接调用柔性负
荷的非对称模式下得出,在体现柔性负荷用户参与
意愿的电力市场模式下,上述参数对其调峰效益的
影响尚待研究。
4.3.2 风电穿透功率
基于调峰平衡约束,计算三种运行方式的风电
穿透功率[22],结果见表 4
4 风电穿透功率
Tab.4 The penetrating power of the wind power
方式 1 方式 2 方式 3
风电穿透功率/MW 1 727 1 780 1 966
由表 4可知,发挥柔性负荷的调峰效益,可以
缓解风电的负调峰效应,从而降低了调峰平衡约束
对系统消纳风电能力的限制。在调度模型中充分考
虑风电注入电网的随机性因素,可以合理优化电力
系统的备用需求,有效提高系统对风电的消纳能力。
4.3.3 风险旋转备用
计算随机调度模型中风速预测误差标准差和系
统风险阈值取不同数值时系统的正旋转备用需求,
计算结果如图 2所示。
2 正旋转备用需求曲线
Fig.2 Positive spinning reserve demand curve
由图 2可知,系统的正旋转备用需求与风险阈
负相关,与风速预测误差标准差 σv正相关,其
原因是,系统风险阈值越大,系统安全风险越高,
故所需的旋转备用也越少;而风速预测标准差增大,
表明风电出力的波动性增加,因此需要系统提供更
多的备用容量以平抑风电的出力波动。
由于系统正旋转备用需求的增加势必提高系统
的运行成本,因此系统的运行成本亦与风险阈值 λ
负相关,与风速预测误差标准差 σv正相关。
由于风电出力概率分布具有对称性,故负旋转
备用需求情况与正旋转备用类似,本文不再赘述。
4.3.4 算法性能分析
为对比 ICA 算法性能,本文分别采用禁忌粒子
群算法[23] Taboo Particle Swarm Optimization
TPSO标准遗传算法[24]
Genetic AlgorithmGA
对调度方式 3进行求解,其收敛曲线如图 3所示,
由图可知,ICA 算法在收敛速度上较 TPSOGA
有明显优势,与 GA 算法相比,ICA 算法具有较强
的全局搜索能力,更容易跳出局部最优解。
3 算法收敛曲线
Fig.3 Convergence characteristics of the algorithms
28 卷第 11 考虑柔性负荷调峰的大规模风电随机优化调度方法 237
5 结论
本文将柔性负荷作为一种可调度的调峰资源,
基于风险约束理论,建立考虑柔性负荷调峰的大规
模风电随机优化调度模型,并通过帝国竞争算法对
模型进行求解,研究结果表明:
1)本文所建模型,体现了柔性负荷作为调峰
资源的经济价值,充分考虑了大规模风电给电力系
统注入的不确定性因素,可辅助调度人员从安全性、
经济性和可靠性的全局视角进行统筹协调,与现有
调度模型相比,取得了较好的经济效益。
2)模型充分体现了柔性负荷的削峰填谷效益,
有助于系统优化备用需求,从而有效提高系统对风
电的消纳能力。
3)电力系统的运行总成本除与常规机组、柔
性负荷用户费用参数等传统因素相关外,还与系统
风险阈值、风速预测误差标准差密切相关。
本文所述的调度方法可以为电力系统调度人员
合理地安排发电计划提供参考。
参考文献
[1] 陈宁, 于继来. 兼顾系统调频需求的分布式风电分
散自治调控策略[J]. 电工技术学报, 2008, 23(11):
123-130.
Chen Ning, Yu Jilai. Strategy for decentralized
autonomous regulation of distributed wind power
considering system frequency regulation demand[J].
Transactions of China Electrotechnical Society, 2008,
23(11): 123-130.
[2] 吴义纯, 丁明, 李生虎. 风电场对发输电系统可靠
性影响的评估[J]. 电工技术学报, 2004, 19(11): 72-76.
Wu Yichun, Ding Ming, Li Shenghu. Reliability
Assessment of wind farms in generation and
transmission systems[J]. Transactions of China
Electrotechnical Society, 2004, 19(11): 72-76.
[3] 张学清, 梁军, 张利, . 计及风光电源的一种地
区电网电动汽车充电调度方法[J]. 电工技术学报,
2013, 28(2): 28-35.
Zhang Xueqing, Liang Jun, Zhang Li, et al. Approach
for plug-in electric vehicles charging scheduling
considering wind and photovoltaic power in Chinese
regional power grids[J]. Transactions of China
Electrotechnical Society, 2013, 28(2): 28-35.
[4] Ruey Hsun Liang, Jian Hao Liao. A fuzzy-
optimization approach for generation scheduling with
wind and solar energy systems[J]. IEEE Transactions
on Power Systems, 2007, 22(4): 1665-1674.
[5] Walters D C, Sheble G B. Genetic algorithm solution
of economic dispatch with valve point loading[J].
IEEE Transactions on Power Systems, 1993, 8(3):
1325-1332.
[6] 王卿然, 谢国辉, 张粒子. 含风电系统的发用电一
体化调度模型[J]. 电力系统自动化, 2011, 35(5):
15-18, 30.
Wang Qingran, Xie Guohui, Zhang Lizi. An
integrated generation-consumption dispatch model
with wind power[J]. Automation of Electric Power
Systems, 2011, 35(5): 15-18, 30.
[7] 吴俊, 李国杰, 孙元章. 基于随机规划的并网风电
场最大注入功率计算[J]. 电网技术, 2007, 31(14):
15-19.
Wu Jun, Li Guojie, Sun Yuanzhang. Maximum
injection power calculation of wind farms connected
to power systems based on stochastic programming[J].
Power System Technology, 2007, 31(14): 15-19.
[8] 邱威, 张建华, 刘念. 含大型风电场的环境经济调
度模型与解法[J]. 中国电机工程学报, 2011, 31(19):
8-16.
Qiu Wei, Zhang Jianhua, Liu Nian. Model and solution
for environmental/economic dispatch considering
large-scale wind power penetration[J]. Proceedings of
the CSEE, 2011, 31(19): 8-16.
[9] Bhattacharya K, Bollen M H J, Daalder J E. Real time
optimal interruptible tariff mechanism incorporating
utility-customer interactions[J]. IEEE Transactions on
Power Systems, 2000, 15(2): 700-706.
[10] 葛炬, 张粒子. 可中断负荷参与的备用市场帕累托
优化模型[J]. 电力系统自动化, 2006, 30(9): 34-37.
Ge Ju, Zhang Lizi. Pareto optimality model of reserve
market including interruptible load as participants[J].
Automation of Electric Power Systems, 2006, 30(9):
34-37.
[11] 周春明, 江辉, 何禹清, 可中断负荷参与阻塞管
理的多目标模糊优化[J]. 电网技术, 2008, 32(9):
27-32.
Zhou Chunming, Jiang Hui, He Yuqing, et al. A
multi-objective fuzzy optimization of congestion
management with participation of interruptible loads[J].
Power System Technology, 2008, 32(9): 27-32.
238 2013 11
[12] Moghimi Hadji M, Vahidi B. A solution to the unit
commitment problem using imperialistic competition
algorithm[J]. IEEE Transactions on Power Systems,
2012, 27(1): 117-124.
[13] Matevosyan J, Soder L. Minimization of imbalance
cost trading wind power on the short-term power
market[J]. IEEE Transactions on Power Systems,
2006, 21(3): 1396-1404.
[14] 雷亚洲, 王伟胜, 印永华, . 基于机会约束规划
的风电穿透功率极限计算[J]. 中国电机工程学报,
2002, 22(5): 32-36.
Lei Yazhou, Wang Weisheng, Yin Yonghua, et al.
Wind power penetration limit calculation basedon
chance constrained programming[J]. Proceedings of
the CSEE, 2002, 22(5): 32-36.
[15] 齐民友. 概率论与数理统计[M]. 1 . 北京:高等教
育出版社, 2002.
[16] 刘新东, 陈焕远, 姚程. 计及大容量燃煤机组深度
调峰和可中断负荷的风电场优化调度模型[J]. 电力
自动化设备, 2012, 32(2): 95-99.
Liu Xindong, Chen Huanyuan, Yao Cheng. Economic
dispatch considering deep peak-regulation and
interruptible loads for power system incorporated
with wind farms[J]. Electric Power Automation
Equipment, 2012, 32(2): 95-99.
[17] 于娜. 电力需求响应参与系统运行调控问题的研究
[D]. 哈尔滨:哈尔滨工业大学, 2009.
[18] Fahrioglu M, Alvarado FUsing utility information to
calibrate customer demand management behavior
models[J]IEEE Transactions on Power Systems,
2001, 16(2)317-322
[19] Moghimi Hadji M, Vahidi B. A solution to the unit
commitment problem using imperialistic competition
algorithm[J]. IEEE Transactions on Power Systems,
2012, 27(1): 117-124.
[20] Yang Hong Tzer, Yang Pai ChuanA parallel genetic
algorithm approach to solving the unit commitment
problem: implementation on the transputer
networks[J]IEEE Transactions on Power Systems,
1997, 12(2)661-668.
[21] 娄素华, 余欣梅, 熊信艮, . 电力系统机组启停
优化问题的改进DPSO 算法[J]. 中国电机工程学报,
2005, 25(8): 30-35.
Lou Suhua, Yu Xinmei, Xiong Xinyin, et al. Unit
commitment using improved discrete particle swarm
optimization algorithm[J]. Proceedings of the CSEE,
2005, 25(8): 30-35.
[22] 杨宏, 刘建新, 苑津莎. 风电系统中常规机组负调
峰能力研究[J]. 中国电机工程学报, 2010, 30(16):
26-31.
Yang Hong, Liu Jianxin, Yuan Jinsha. Research of
peak load regulation of conventional generators in
wind power grid[J]. Proceedings of the CSEE, 2010,
30(16): 26-31.
[23] 陈道君, 龚庆武, 张茂林, . 考虑能源环境效益
的含风电场多目标优化调度[J]. 中国电机工程学报,
2011, 31(13): 10-17.
Chen Daojun, Gong Qingwu, Zhang Maolin, et al.
Multi-objective optimal dispatch in wind power
integrated system incorporating energy-environmental
efficiency[J]. Proceedings of the Chinese Society for
Electrical Engineering, 2011, 31(13): 10-17.
[24] Kazarlis S A, Bakirtzis A G, Petridis V. A gcnctic
algorithm solution to the unit connnitmcnt problcm[J].
IEEE Transactions on Power Systems, 1996, 11(1):
83-92.
作者简介
男,1987 年生,博士,研究方向为电力系统分析,电力
系统规划与控制。
(通讯作者)男,1978 年生,博士,副教授,主要从事电
力系统自动监测,电力系统运行与规划等方面的研究。
... The introduction and specification of the stochastic factors in the SCUC model are the focal points for the power system with the integration of the large-scale wind power. At present, the common methods in the uncertainty modelling are scenario method [7][8][9], chance-constrained method [10,11], robust optimisation [12][13][14], and so on. The DCPF constraints of the traditional SCUC model are mainly adopted as the security constraints. ...
... For the purpose of adding the proper uncertainty into the SCUC model, three methods have already been proved effectively. They are the chance-constrained method [7][8][9], the scenario method [10,11], and the robust optimisation [12][13][14], respectively. Meanwhile, the chance-constrained method is adopted in this study because of the computation convenience. ...
Article
Full-text available
With the more precise day-ahead scheduling strategy for large-scale wind power integration, a model of stochastic security-constrained unit commitment (SCUC) under AC constraints is built and its corresponding solution is given here. A unit commitment model under AC power flow is developed based on network security constraints considering the uncertainty of wind power. In order to analyse this model in time, an ordinal optimisation is presented. The proposed method is verified by the numerical test on a modified IEEE-118 test system. The results show that the risk of voltage over-limit in large-scale wind power integration can be reduced effectively. Meanwhile, the validity of day-ahead generation schedule is enhanced. Furthermore, the computational efficiency of the proposed algorithm is improved significantly compared with that of the traditional one.
... The derivation processes of relations (13-16) and the introduction of the wind power uncertainty in the model are described in Appendix. The effectiveness of using the chance-constrained method to describe the uncertainty of wind power generation in a power dispatching problem has been used for Yang et al. (2013). ...
Article
Full-text available
It is of great significance for the development of intermittent renewable energy and the marketization of electricity to study an efficient and accurate algorithm for solving the unit commitment problem with security constraints. Based on the existing constrained ordinal optimization (COO) algorithm, an improved ordinal optimization (OO) algorithm is proposed, which can be applied to solve SCUC problems. In order to solve the rough model and the accurate model, the discrete variable identification strategy and the effective safety constraint reduction strategy are adopted, respectively. Compared with the traditional object-oriented algorithm, the improved object-oriented algorithm not only gives full play to the computational efficiency of the traditional object-oriented algorithm but also further improves the compactness and reduces the computational redundancy. IEEE 118 simulation results verify the correctness and effectiveness of the algorithm.
... The concept of IDR was officially published in the literature [2] for the first time in 2015. Literature [3][4][5][6][7][8][9][10][11][12] conducted a model analysis of the IDR response randomness model, the resource coordination operation mechanism and its benefit evaluation, and the demand side economic model in the integrated energy system. However, the current research doesn't consider the constraints of the user's actual maximum response, and there is no specific model calculation for the user terminal equipment. ...
Article
Full-text available
The current Integrated Demand Response (IDR) strategies cannot accurately evaluate the user’s maximum response which ignore the user terminal equipment response model. At the same time, the optimization target of the current IDR strategies is mostly electric energy, and other energy response requirements such as natural gas are not considered. This paper establishes an IDR strategy model based on energy models for different equipment terminals of residential users. And calculate the theoretical maximum response potential value of different users according to the response strategy formulated by different energy response requirements with the GA-SQP hybrid algorithm. The example analyses the user response potential in typical scenarios and time periods. The results verify the effectiveness and feasibility of the proposed method.
... . (8) Where N is the unit of controllable DG (FC and MT); P is the unit of uncontrollable DG (wind power and Photovoltaic system) ...
Article
The uncertainty of wind farm power and load both have a certain impact on the economic dispatch of the power system. First, this study deals with the fuzzy processing of power prediction error and load forecasting error of wind farms based on fuzzy random theory. On this basis, a multi‐objective fuzzy stochastic dispatch model of a power system with flexible load is established under the carbon trading mechanism. And, the nonlinear cost of flexible load response and the cost of carbon emission compensation is introduced to the multi‐objective function of the model. In addition, the fuzzy chance constraint of the spinning reserve is added to the constraint conditions. By introducing the variable, the fuzzy random dispatch model is transformed into a clear equivalent model. Finally, the discrete bacterial colony chemotaxis algorithm is used to process the model, and the optimal solution to the multi‐objective function is obtained by a compromise strategy based on a small degree of satisfaction. In the simulation, a classic IEEE10 system and a wind farm are taken as examples. The results show that compared with the other three traditional dispatch models, the total power generation cost is reduced by US$13 670, US$5610 and US$86 010, respectively using the model proposed in this article.
Article
Large-scale wind power integrated into power system brings about a great challenge to traditional power flow analysis and economic dispatch decision. This book mainly focuses on these topics to address uncertainties. In this chapter, we will give the brief introduction and the whole flowchart of this book.
Article
Wind power output fluctuation affects its efficiency and power system security. Using energy storage device can reduce the influence of the wind power output fluctuation. The characteristics and efficiency of different energy storage technologies are researched in order to find the available devices for stabilizing wind power output fluctuation. This paper collects and studies various technical parameters of existing energy storage devices, and makes technical and economic comparison of the methods connected to the grid, functions and scale effect. By use of a lot of data analysis and the findings of other related topics, this paper finds out that power grid is the only energy storage system can gather a huge ability of regulate, which is currently the most effective and economical energy storage system. Ultimately it proves that China should develop wind power in the three-north region on a large scale to reduce pollution in energy industry and enhance energy security, while the construction of large power grid is the way of the development of wind power in China. ©, 2015, Power System Protection and Control Press. All right reserved.
Article
This paper introduces a reliability assessment model of wind power integration in power systems based on the Monte-Carlo simulation approach, which considers not only the randomicity of the wind, weather effect on wind power generation forced outage rate, but also the transmission network failure rate and capability constraints. This model is applied to simulate power system including wind power generation. Power system is dispatched in term of the principle that wind power is employed sufficiently, according to meeting system security constraints. The reliability index is calculated and will provide the valuable reference for wind farm operation and planning.
Article
This paper introduces a new method for calculating wind power penetration limits in power system utilizing chance constrained programming. Wind power penetration limits are regarded as the maximum installed capacity of wind farm constrained by network and equipments limits, i.e., transmission lines capability, system spinning reserve, the output limits of conventional generators etc. To deal with the stochastic characteristics of wind farm outputs and loads, the constraints are expressed by means of probability, and stochastic simulation and genetic algorithm is used to solve the problem. The results on IEEE30 system demonstrate the advantages of the proposed approach.
Article
The uncertainty of wind power has widespread impact on connecting power grid. Consequently it becomes the cause of limiting the wind power to connect into gird. The capability of negative peak load regulation of conventional generators is one of the most important effective factors. The research on model and algorithm of calculating the limit of capability of negative peak load regulation becomes an important topic. The paper studies the mechanism of the negative peak load regulation of conventional generators based on active power balance equation, a new model is proposed to calculate the limit of capability of negative peak load regulation, and a practical two-tier algorithm is given. The simulation results prove the correctness of the model and algorithm.
Article
More and more wind generators distribute in electric power system. It is impossible to regulate them by dispatch center of transmission network like conventional electrical sources. It needs to be regulated decentralizedly and autonomously. However, autonomous regulation of wind generators separated fully from transmission network will lead system to be worse in some modes. So it needs to study new strategies for regulation of wind turbine and make the regulation to be limitedly autonomous, which will cooperate with system operation and help power system recover. This paper proposed a strategy for decentralized autonomous regulation of distributed wind power considering system frequency regulation demand. According to system frequency and operation information of wind generators, regulation regions of wind generators are divided into interconnection region, normal region, abnormal region, urgent region and clearing region. Further, three new regulation modes, which are abnormal step-regulation, urgent step-regulation and fault step-regulation, are proposed. Example system results show that wind generators can regulate their outputs in order to support the system according to the system frequency regulation demand.
Article
This paper presents a discrete binary particle swarm optimization (DPSO) method improved for solving unit commitment problem. The proposed method alters the updating rule of the particles position, introduces the No-hope/Re-hope criterion and heuristic pseudo-mutation mechanism to iterative process. As overcome the demerit of focusing on the local optimum and ensure state variables valid. The feasibility of the proposed method is demonstrated for two different systems, and it is compared with other methods. The results show that it is advantageous in terms of the solution quality and computation efficiency.
Article
According to the extensive development of plug-in electric vehicles(PEVs) and renewable resources, such as wind power and photovoltaic power, an approach for PEVs charging scheduling in regional power grids considering the uncertain outputs of wind and photovoltaic power is proposed. Firstly, in order to reduce the difference between the peak and the valley for equivalent load and purchasing power cost, a multi-objective non-linear mixed integer optimization model for PEVs charging scheduling is established. Secondly, the fuzzy theory is introduced to this paper to fuzzy the output of wind power and photovoltaic power. Therefore, the multi-objective fuzzy optimization model is reformulated as a single objective non-linear optimization problem. Finally, the data of example on regional power grid is analyzed to prove to the validity of model and the feasibility of solving for problems with improved particle swarm algorithm. An effective way is provided for the optimal dispatch of PEVs.
Article
An optimal dispatch model is put forward for the power system incorporated with wind farms, which considers, from the view point of economic operation, the deep peak-regulation of high-capacity coal-fired unit and the interruptible loads. The deep peak-regulation fee is defined in three aspects: coal consumption, unit power compensation and life consumption. The interruptible load fee is defined in two aspects: load reservation and real load loss. The optimization object is the minimum summation of deep peak-regulation fee, interruptible load fee and net loss, with which the dispatch model of power system with wind farms is optimized by the interior point method. With New England test system as an example, its loads and fees are analyzed for 4 typical cases and the results of optimal dispatch for daily generation schedule show that, the cost is reduced up to 4.9%, proving the rationality and effectiveness of the proposed algorithm.
Article
With the large-scale wind power connected into the power systems, the influence of wind farms penetration should be considered in power system dispatch. A clean development of the power industry was required because of the increasingly serious global climate change and the strategy of sustainable development of human society. Based on the optimal dispatch in traditional power system, the concept of "energy-environmental efficiency" was introduced to modify the optimal dispatch model in wind power integrated system, and the multi-objective optimal dispatch model was proposed on the basis of comprehensively considering the minimum of the resource consumption, the best energy-environmental efficiency and the highest system stability. A hybrid particle swarm and tabu search optimization algorithm with fuzzy technology was presented to solve the optimization model. Experimental results show that the proposed optimization dispatch model is reasonable and the algorithm is feasible.
Article
Integrating the power generated by large-scale wind farm into power grid introduces an extra factor of uncertainties for generation dispatch. Based on multiobjective chance constrained programming, an environmental/economic dispatch model was formulated for the system incorporating wind farm. Minimizing both the fuel cost and emission of atmospheric pollutants of thermal generators were considered as objective functions. The up spinning reserve and the down spinning reserve constraints were included in the model. According to the cumulative distribution function of wind power, the stochastic optimization problem was transformed to a deterministic one. A two-stage approach incorporating multiobjective optimization and decision making was developed to solve the problem. In this approach, the Pareto optimal set could be pruned in order to help the decision maker to select a final solution. The proposed generation dispatch method was tested on a system consisting of 40 thermal generators and one large-scale wind farm, the results obtained demonstrate the effectiveness of the proposed model and algorithm.
Article
Reserve plays an important role in maintaining power system reliability and security. It has practical significance and applicable value taking interruptible load for one kind of reserves to ensure power system reliability and optimize the resource in the electricity market environment. At present, interruptible load has participated in some reserve markets together with generator, but the existing optimization method only takes the total costs of reserves into account and doesn't consider the difference of the reserves coming from the generator and the interruptible load. The paper analyzes the localization of the current optimization method for the reserve market with interruptible load and generator as participants. Pareto optimality method for the reserve market is brought forward, which can obtain the maximal utility and realize the optimization of two kinds of reserves from interruptible load and generator.