Professional Documents
Culture Documents
AbstractAs a fulcrum in the power market, electricity price Crew and other researchers advanced to made use of economic
affects the reliability and security of the power system indirectly. incentives to induce customers electricity-consumption, i.e. to
The time-of-use (TOU) pricing is one of the important ways to increase the electricity-consumption during the valley period
safeguard the stability of the power system. The TOU pricing and reduce it during the peak period. Thereby the difference
must consider the stakeholders of the government, the power between the peak and the valley could be decreased to relieve
company, and the consumer. With the prerequisite of price-fixing the serious problem of the peak capacity [5]. After researching
level by the government, the power company is allowed to adjust on the relationship between the supply elasticity and the
the power price to realize the goal of power load control. Based demand elasticity and analyzing the operation characteristic of
on this, a TOU pricing model price-fixing by the government, the Nordic electricity market, Doorman and Wangensteen
pricing by the power company, and electricity-consumption by
demonstrated the feasibility of adjusting the demand elasticity
the consumer is established with game theory. The model
simulation is carried out with the Matlab Software, the result of
by imposing additional charges during the peak periods [6].
which indicates that the model is practicable for the power load David, Rsnen, et al. thought that the consumer response
control. could be reflected in the adjustment of consumption periods.
They constructed a DSM TOU pricing model on the basis of
Keywords-TOU pricing; electricity market; game model; investigating the consumer response curve [7,8]. In order to
simulation quantify the consumer response to the TOU pricing, Sheen and
his colleagues quoted price elasticity matrix to express
I. INTRODUCTION consumers demand elasticity [9].
Electricity price is one of the most important elements of In China, along with the deepening of the reform of the
the power market. Many failed cases of marketization of power industry, the research on the theory of power price in the
electric power in western countries are due to the impact of demand side has been springing up. Tang Yudong, Wang
electricity price on the reliability and security of the power Mingfei, et al. considered that prerequisite of calculating the
system. How to calculate the reasonable price and how to fairly price in the demand side was that there was a determinate
allocate the transmission costs to the consumer are hot research power load curve. And they set up a TOU pricing model based
topics. on the goal of DSM [10]. Subsequently, supplementary to
Tangs model was achieved by a series of studies. For instance,
The theoretical research on the TOU pricing of electricity Wu Junji, Ding Ning, et al. optimized the division method of
market was firstly carried out in the generation side [1]. In peak-valley period with semi-gradient fuzzy membership
1981, M. Munasinghe proposed that the overall price level be function. They studied the possibility of the point on the curve
determined by the long-term marginal cost. With a simplified the probability at the peak and at the valley [11]. Wu Qiuwei,
peak-load model sharing between the peak and the valley (off- Wang Lei, et al. probed into the electricity pricing of flat-load
peak), a simple model of TOU pricing was established from the periods and put forward a method to determine the price of
point of production cost [2]. From 1983 to 1987, Oyama flat-load periods by using the market clearing price (MCP) to
measured the marginal cost of electricity with the method of calculate the average purchase electricity price, which was a
mathematical programming. Given the time periods of the feasible approach to marketize the TOU pricing. In this way,
natural peak, peak, waist and the base, the solution of the dual the TOU pricing system suited to the electricity market could
linear programming was obtained. This was a further progress be constructed [12]. In Liu Changs research, he combined the
of the TOU pricing model of the power generation side [3]. electricity price of the demand side with the price of the
The study of TOU pricing from the generation side to the generation side to marketize the TOU pricing and got a model.
demand side arose as the spot price was put forward. Luh, Ho, Through optimizing the objective function of the model, the
et al. put forward an idea of consumer response to describe optimal periods division of peak-valley and corresponding
the impact of the power-load on the TOU pricing, and a DSM- electricity pricing were carried out [13, 14]. Liu Guanqi, Zhang
based model was set up. This was a research on the TOU Jian, et al. surveyed the consumer response to the TOU pricing
pricing in the demand side for the first time [4]. Then M. A. and established a TOU pricing model based on the consumer
pricing should consider factors beyond economic laws for the st. p (q j j ) q j C j (q j ) 0
j =1 j =1
current shortage of electricity [15]. Ding Wei, Yuan Jiahai, et
al. proposed to establish the consumer response matrix to From this model, we can get an
reflect the customer response to the TOU rates, and established p j MC j
j = 1 + , where j = j j is the
q p
a model of consumer satisfaction. After considering the equation
consumer satisfaction of consumption mode and electricity pj p j q j
charges, they improved the TOU pricing model based on 1+
consumer response to price and consumer satisfaction [16]. Liu price elasticity in the jth period. Given R = as Ramsey
Yan, Zhao Juan, et al. analyzed the problems of present TOU
rates in China and thought that the present TOU pricing should index, the above equation can be transformed
be improved in the aspects of other issues of the division of MC j
to p j = ; here pj is the price-fixing level of the
TOU pricing periods, the determination of the ratio of the TOU 1 R / j
rates, and the incentive mechanism of risk-benefit [17, 18]. government in the jth period.
In short, the research on the TOU pricing focused on the Under the prerequisite of this paper, the government only
relationship between the generation side and the demand side needs to develop a unified price level and give a certain right to
to set up TOU pricing model or to make up the model through the power company to establish a pricing structure, so that the
investigating customer response to the pricing, price elasticity power company can control the power load. Therefore, to
of the demand side, and customer satisfaction. The weighted-average the power price with the weight of electric
optimization of these models mainly stressed on the rational
division of peak-valley periods, determination of price, quantity, the governments unified price of p is available:
customer response to the pricing. Most of these models
concerned about the impact of the price on the power-load; 1 n 1 n q j MC j
p= qj pj =
little attention was paid to the game between the customer and nQ j =1 n
1 R / j
the power company. And as the supervisor, the governments n q j j =1
j =1
impact of the regulation on the electricity price, especially the
impact of the TOU pricing on the social welfare was less taken C. A model of customer response
into account. The customer may select the optimal quantity of electricity
under given TOU rates to maximize his profit. For customer i:
II. GAME MODEL OF TOU PRICING
n n
A. Hypothesis max i = rij qij pij qij (Model II)
qi 1"qin
(1) customers: m; periods: n. j =1 j =1
r ij
ma j
(4) with symmetric information.
electricity in the j period is q j =
th i =1
.
(m + 1)b j
B. Price-fixing model of the government
Of n periods, different customers demand elasticity for the D. Bidding model of the power company
electric power is not the same. According to Ramsey-Boiteux
pricing model, suppose the inverse function of the power The goal of modeling the bidding is to realize the power-
company for the market or the customer in the jth period is load control in the case of meeting with the restriction of
pj=pj(qj), and the consumer surplus in the jth period is customer participation. Therefore, for any power-load level A,
qi aj(A), bj(A) should be selected to satisfy the following equation:
S j = p j (q j )dq j p j (q j ) q j . Thus, the problem faced by
0
n
Max = S j (Model I) st. rij p j , j=1,2,,n
j =1
restricted by h j max . Accordingly, under power-i =1 probability of a point is 90% in peak periods, its possibility in
A j the valley periods is of very little possibly this point is in the
load level A, the bidding strategies of the power company peak periods; if the probability of another point is 90% in the
m valley periods, its possibility in the peak periods is of very little
r ij (m + 1) Ah j either possibly this point is in the valley periods; if the
is p j = i =1
+ h j q j ; the equilibrium price probability of a certain point is 50% in the peak periods, its
m possibility in the valley periods is likely to be 50% possibly
m
is p j ( A) = i =1
; the electricity-consumption of We selected the smaller semi-trapezoidal membership
m function as shown in Figure 1(a) and bigger membership
1 1 m A function as shown in Fig. 1(b) to evaluate the points on the
customer i is qij ( A) = rij m rij + .
hj i =1 m curve the probabilities in the peak periods and the valley
periods (only consider the oblique line, i.e. the interval [a, b] on
the x-axis).
E. A model of the optimal power supply of the power
company
For the power company, he must firstly determine the U(x)
bidding strategies at the power-load level A, i.e. to fix the value
of aj(A), bj(A), j=1,2,,n. Based on this, the optimal power
supply level can be established. The determination of the
optimal power supply level is a process of maximizing its
1
profit under the restriction of the governments price-fixing: bx
U ( x) =
ba
n n
max = p j A C j A K 0 Model IV
A
j =1 j =1
(a)
1 n
s.t. pj A p
nA j =1
0 a b x
To solve this linear programming, we can get the optimized U(x)
m n n
r ij m C j
m
level A* =
j =1 i =1
n
j =1
( p) , where ( p) is the
2
2 h j 1
j =1
r ij
mn p
i =1 j =1
namely A* = n
.
hj j =1 (b) 0 a b x
periods.
Power load(10,000kw)
120
B. Model simulating
1) Method 90