ARTICLE IN PRESS

Renewable Energy 32 (2007) 11651174 www.elsevier.com/locate/renene

Predictive analysis on electric-power supply and demand in China

Min Huang, Yong He, Haiyan Cen
College of Biosystems Engineering and Food Science, Zhejiang University, Hangzhou 310029, PR China Received 18 November 2005; accepted 6 April 2006 Available online 2 June 2006

Abstract In order to analyze the electric-power demand and supply in China efciently, this paper presents a GreyMarkov forecasting model to forecast the electric-power demand in China. This method takes into account the general trend series and random uctuations about original time-series data. It has the merits of both simplicity of application and high forecasting precision. This paper was based on historical data of the electric-power requirement from 1985 to 2001 in China, and forecasted and analyzed the electric-power supply and demand in China by the GreyMarkov forecasting model. The forecasting precision of Grey-Markov forecasting model from 2002 to 2004 is 99.42%, 98.05% and 97.56% respectively, and in GM(1,1) Grey forecasting model, it is 98.53%, 94.02% and 88.48%, respectively. It shows that the GreyMarkov forecasting models has higher precision than GM(1,1) Grey forecasting model. The forecast values from 2002 to 2013 were as follows: 16106.7, 18541.3, 20575.7, 23940.5, 24498.0, 26785.1, 27977.2, 29032.2, 31247.5, 33428.8, 35865.4, and 38399.3 TW h. The results provide scientic basis for the planned development of the electric-power supply in China. r 2006 Elsevier Ltd. All rights reserved.
Keywords: GM(1,1); Markov chain; GreyMarkov; Forecast; Electric-power demand

1. Introduction Along with economic growth of nearly 10% per year over the last two decades, electricpower use has been rapidly increasing in China. During the 1980s, because the developing speed of the economy was quicker than that of the electric-power industry, China had a
Corresponding author. Tel.: +86 57186971143; fax: +86 571 86971143.

E-mail address: yhe@zju.edu.cn (Y. He). 0960-1481/$ - see front matter r 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.renene.2006.04.005

ARTICLE IN PRESS
1166 M. Huang et al. / Renewable Energy 32 (2007) 11651174

scarcity of electric-power. To deal with the situation, China rst innovated the investment system. In 1981, the Longkou power plant was constructed with joint venture by the central government and local units in Shandong province. In 1987, State Department advanced a twenty-words policy about power system innovation, which translated to English was government and corporation should be separated, a province is an entity, unite the electric-power network, attemp the system united, raise funds to build power plants. In 1984, a Japanese company constructed a new hydropower station in the Yunnan province, which was called as Lubuge hydropower station. Because it was the rst power plant constructed by foreign capital, it was also called as Lubuge shock wave which inuenced the investment system of electric-utility industry very much. From 1985, State Department set up Huaneng Power International Corporation and other power corporations to raise foreign capital. With the new policies, Chinese electric-utility industry developed very quickly. After the total electric-power equipped capacity exceeded 1 TW in 1987, the total electric-power equipped capacity exceeded 2 TW in 1995 and exceeded 3 TW in 2000. In 2004, the additional electric-power equipped capacity was 0.51 TW, which was at the rst place of the world. The total electric-power equipped capacity was 4.47 TW. Chinese electric-power demand and supply began to keep a balance from 1996 [1]. The electric generation production of 1985 was 4117.6 TW h, and that of 2004 was 20418.08 TW h. It means that there was 4.98 times increase from 1985. The electric generation production had grown at 8.82% per year between 1985 and 2004. The total electric-power equipped capacity and the generation production of China were at the second place of the world, from 1996 untill now. While great development was accessed in Chinese electric-utility industry, there were many problems which inuenced the development of the society badly. Electric-power was a resource that could not be stored. The supply quantity had to equal the consumption. So neither the shortage of supply nor the short of consumption would inuence the electricutility industry seriously. Perhaps, the whole occupation would be decient due to lack work, and maybe insufcient supply would affect the function of different elds in society. In late 1980s, consumption of electricity in our country kept increasing between 9% and 10%. It reached 16% and became the top in history in 1991, while in 1992 and 1993, it kept above 11%. This period was called as over-economic. The lack of electricity appeared in the whole society. The power cut appeared even in Beijing. The government started to adjust the quo-status in 1993, which made the economy in China come into soft-land. The consumption of the whole country was decreasing, and reached the bottom of 2.8%. It led to the free of electric equipments and the whole coal industry fell into decit. The government made a stipulation on prohibiting building new electric factory during three years. However, the increase of consumption in 1999 recovered 6.2% and 9% in 2000. Since 2000, the lack of electricity pervaded from the delta in Changjiang River to Zhujiang River even the whole country. There were 12 provinces to cut power and limit the electric use. The lack of electric equipment was 20.35 GW. In 2003, the power of 22 provinces cut, the lack of electric equipment was 44.85 GW, which doubling compared to last year. The power of 24 provinces cut till 2004, and the largest lack of electricity reached 30 GW. It diffused in the whole country, including the undeveloped areas such as Guangxi, Yunnan, Guizhou, Shanxi, Neimenggu, Gansu, Qinghai, Ningxia et al. The summer of 2004, was considered to have experienced as the most serious lack of power since more than twenty years. The consumption of electricity was 19,458.4 TW h from January to November in this year. It increased 15.13% more than that in last year at the same time approaching to the top value.

ARTICLE IN PRESS
M. Huang et al. / Renewable Energy 32 (2007) 11651174 1167

There are many factors that cause lack of electricity, such as insufcient investment in electricity, the distinction in the speed of new electrical product and the demand increase, the rapid increase of the industry and electrical industry with high energy consumption, the low ability of electrical transport and the accidents, the quick increase in the load of airconditioning, the overwrought supply caused by the controversy in the price of electricity and coal, the shortage of water, etc. Because of many undermining factors, including economical development, the structure of industry, the incoming lever of citizen, climate and the national policy, etc., uctuation in the consumption of electricity and the increase of load appeared obvious. Efcient methods to predict the electrical demand in middle and long term are scarce. We attempt to study a reasonable approach for development of electricity using scientic methods. Thus, according to the data of the total consumption of electricity, this paper proposed a GreyMarkov forecasting model to forecast and analyze the electric-power supply and demand in China. Deng initially presented the Grey system theory in 1982 [2]. The Grey forecasting model adopts the essential part of the Grey system theory. The GM(1,1) Grey forecasting model can be used in circumstances with relatively little data and it can use a rst-order differential equation to characterize an unknown system. So the GM(1,1) Grey forecasting model is suitable for forecasting the competitive environment where decision makers can refer only to a limited historical data [3]. But the forecasting precision for data sequences with large random uctuation is low. The Markov-chain forecasting model can be used to forecast a system with randomly varying time series. It is a dynamic system which forecasts the development of the system according to transition probabilities between states which reect the inuence of all random factors. So the Markov-chain forecasting model is applicable to problems with random variation, which could improve the GM(1,1) model [4]. The disadvantage of the model is that it demands that the process be a stationary one, that is, the trend curve is a horizontal line. If the trend curve is nonhorizontal, the forecast is less accurate because the states of the ordinary Markov probability matrix forecast are stationary. To improve the Markov-chain forecasting model, this paper builds a GM(1,1) Grey forecasting model to describe the historical data of the electric-power demand in China to calculate the uctuating trend curve rst [5]. The rationale of GreyMarkov forecasting model is as follows: rst a GM(1,1) Grey forecasting model is built to calculate the uctuating trend curve of the historical data series, then specify some states around the trend curve, a Markov transition matrix can be built to nd out the transition probability, nally these two models should be combined to forecast accurately by the historical time-series data. This forecasting method can make full use of the information given by historical data, and increase greatly the forecasting precision of random uctuating sequences. 2. Research methodology 2.1. GM(1,1) Grey forecasting model Step 1: Assume the original series to be X(0), X 0 fX 0 1; X 0 2; . . . ; X 0 ng. (1)

ARTICLE IN PRESS
1168 M. Huang et al. / Renewable Energy 32 (2007) 11651174

Step 2: A new sequence X(1) is generated by the accumulated generating operation (AGO). X 1 fX 1 1; X 1 2; . . . ; X 1 ng, where X 1 k
k P i1

(2)

x0 i; k 1; 2; . . . n:

Step 3: Establish a rst-order differential equation. 1 dx =dt az u,

(3)

where z1 k ax1 k 1 ax1 k 1; k 1; 2; . . . n 1: a denotes a horizontal adjustment coefcient, and 0oao1. The selecting criterion of a value is to yield the smallest forecasting-error rate [6]. ^ ^ Step 4: From step 3, we can have x1 k 1 and Y k. u u ^ (4) x1 k 1 x0 1 eak , a a where
! a BT B1 BT Y ; u z1 2 6 1 6 z 3 B6 6 4 z1 n 2 3 1 7 1 7 7; 7 5 1 T ^ Y k x0 2; x0 3; . . . ; x0 n .

^ y

(5) Step 5: Inverse accumulated-generation operation (IAGO). Because the Grey forecasting model is formulated using the data of AGO rather than original data, IAGO can be used to reverse the forecasting value. Namely ^ ^ ^0 x0 k x1 k x1 k 1; k 2; 3; . . . ; n. Step 6: From Step 5, a trend curve equation can be built: (7) ^ k denotes the forecasting data [7]. where g x 1 m=a1 e ; k 0; 1; . . . ; n 1:Y
0 a

(6)

^ ^ Y k x0 k 1geak ,

2.2. Partition of states by Markov-chain forecasting model ^ The values of X 0 k 1 are distributed in the region of the trend curve Y k that may be divided into a convenient number of contiguous intervals. When X 0 k 1 falls in interval i, one of S such intervals, it may be regarded as corresponding to a state Ei in an m order Markov unstable sequence, Ei can be signied as follows: E i E 1i ; E 2i , where i 1,2,yS, S is the amount of states. ^ E 1i Y k Ai . ^ E 2i Y k Bi . (9) (10) (8)

ARTICLE IN PRESS
M. Huang et al. / Renewable Energy 32 (2007) 11651174 1169

^ Y k is a time function, so E1i and E2i will vary with the time series. It means that the state Ei is dynamic. To establish S (the amount of states), E1i and E2i , depends on the study object and the original data series [8]. 2.3. Calculate the transition probability P For Markov-chain series, the transition probability from state Ei to state Ej can be established using an equation as follows: Pij m M ij m Mi i; j 1; 2; . . . ; S, (11)

where Pij m is the transition probability of state Ej transferred from state Ei for m steps (in this paper, 1 step stands for 1 year), m is the number of transition steps each time, M ij m is the number of original data of state Ej transferred from state Ei for m steps, Mi is the number of original data points in state Ei. These Pij m values can be presented as a transition probability matrix R(m). 2 3 p11 m p12 m p1j m 6 p m p m p m 7 6 21 7 22 2j 7 i; j 1; 2; . . . ; S. (12) Rm 6 6 7 4 5 pi1 m pi2 m pij m The state transition probability Pij m reects the statistical law of each state transition in a system, which is the foundation of Markov probability matrix forecast. The future development of the system can be forecasted by studying the state transition probability matrix R(m). Generally, it is necessary to observe the one-step transition matrix R(1). Suppose the object to be forecasted is in state E Q 1pQpS, row Q in matrix R(1) should be considered. If max PQj 1 PQL 1j 1; 2; . . . ; S; 1pQpS, then what will most probably happen in the system at the next moment is the transition from state EQ to state EL. It is difcult to determine the future transition of the state, if two or more transition probabilities in the row Q of matrix R(1) are the same. Therefore the transition probability matrix of two-step transition matrix R(2) or multi-step transition matrix R(m), where mX3, should be considered [9]. 2.4. Calculate the forecasting data After the determination of the future-state transition of a system, i.e. the determination of Grey elements E1i, E2i, the changing interval of the forecast value is between E1i and E2i. ^ The most probable forecast value, Y k 1, is considered to be the middle value of the determined state interval, that is ^ ^ Y k 1 1=2E 1i E 2i Y k 1=2Ai Bi . (13)

3. The GreyMarkov forecasting model for the electric-power demand in China There are many factors which could inuence the electric-power demand, such as the economy, the industry framework, peoples income level, the weather, the governments

ARTICLE IN PRESS
1170 M. Huang et al. / Renewable Energy 32 (2007) 11651174 Table 1 The Electric-power consumption of China from 1985 to 2004 (unit: TW h) Number Year Amount Number Year Amount Number Year Amount Number Year Amount 1 1985 4117.6 6 1990 6230.4 11 1995 10023.4 16 2000 13471.4 2 1986 4507.0 7 1991 6804.0 12 1996 10764.3 17 2001 14633.5 3 1987 4985.2 8 1992 7589.2 13 1997 11284.4 18 2002 16200 4 1988 5466.8 9 1993 8426.5 14 1998 11598.4 19 2003 18910 5 1989 5865.3 10 1994 9260.4 15 1999 12305.2 20 2004 21735

policy and so on. Some factors are clear, and others are not clear. So the time series of electric-power demand of a country shows large random uctuations. As Table 1 shows, the historical data series of the electric-power consumption of China from 1985 to 2002 is rising, but uctuating randomly. So this paper forecasts and analyzes the electric-power demand by GreyMarkov forecasting model [10]. 3.1. Build the GM(1,1) Grey forecasting model Based on the historical data of the electric-power consumption in China from 1985 to 2001, a trend curve equation is built by GM(1,1) Grey forecasting model. It is as follows: ^ ^ 0 Y k X k 1 4331:308566e0:078455k k 1; 2; 3; . . . ; n; (14)

where k is the series number of the year, and k 1 means that it is 1985. The forecast value of 2002 calculated by the equation is 16438.15 TW h. 3.2. Partition of states by Markov-chain forecasting model According to the actual electric-power consumption data, four states, i.e., four ^ 0 contiguous intervals are established about the curve of X k 1. According to(8)(10), the four states intervals can be got as follows: E1 : E2 : E3 : E4 : ^ E 11 Y k 0:130Y ; ^ k 0:045Y ; E 12 Y ^ k; E 13 Y ^ E 14 Y k 0:045Y ; ^ E 21 Y k 0:045Y ; ^ k; E 22 Y ^ E 23 Y k 0:045Y ; ^ E 24 Y k 0:130Y ;

where Y denotes the average value of the historical electric-power consumption from 1985 to 2001. The historical data series, the regressed curve and the states intervals are showed in Fig. 1.

ARTICLE IN PRESS
M. Huang et al. / Renewable Energy 32 (2007) 11651174 1171

16000

Y Electric power requirement

14000

12000

10000

8000

6000

4000 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002

X Year

Fig. 1. The forecasting curve of the electric-power requirement in China.

3.3. Calculate the transition probability P After observing the states intervals and the historical data series, the number of the historical data in every interval can be got. They are as follows: M 1 3 M 2 6 M 3 3 M 4 5, where Mi denotes the number of the historical data in the interval i, and i 1,2,3,4. The numbers of one-step transiting to E1, E2, E3, E4 from E1 are as follows: M 11 1 2 M 12 1 0 M 13 1 0 M 14 1 1. Next M ij 1, where i 2,3,4 and j 1,2,3,4 can be calculated in the same way. Then calculate the one-step transition probability to every states interval and present them in the transition matrix R(1) as follows: 22 0 0 13 60 65 R1 6 1 6 43
1 5 3 3 4 5 1 3 0 5 3 1 5 0 3 1 5 3 07 57 7 1 7. 35 3 5

The electric-power consumption of 2001 year is in the E4 state interval, so observe the fourth row of the transition matrix. Because max P4j P44 j 1; 2; 3; 4 as the matrix R(1) shows, the electric-power demand of 2002 year is most probably in the E4 state interval, namely between 15,319.28 and 16,894.12 TW h.

ARTICLE IN PRESS
1172 M. Huang et al. / Renewable Energy 32 (2007) 11651174 Table 2 Forecast values from 2003 to 2013. Year Forecast Value (TW h) GreyMarkov 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 18541.3 20575.7 23940.5 24498.0 26785.1 27977.2 29032.2 31247.5 33428.8 35865.4 38399.3 GM(1,1) 17779.7 19230.8 20800.4 22498.0 24334.1 26320.2 28468.3 30791.7 33304.7 36022.9 38962.9

3.4. Calculate the forecast value Calculate the most probable forecast value of 2002 years electric-power demand by Eq. (13). ^ Y 18 1=2E 14 E 24 1=215; 319:28 16; 894:12 16; 106:7. As Table 1 shows, the actual electric-power consumption in China in 2002 is 16,200 TW h. So the forecasting precision is 99.42%. 3.5. Calculate the forecast value from 2003 to 2013 Calculate the forecast value from 2003 to 2013 by GreyMarkov forecasting model following the above steps. Table 2 shows the forecasting values which are calculated by Grey-Markov forecasting model and GM(1,1) Grey-forecasting model. 3.6. Comparison of forecast precision between the Grey Markov forecasting model and the GM(1,1) Grey forecasting model As the above, we can get the forecast values from 2002 to 2004 calculated by GreyMarkov forecasting model and GM(1,1) Grey-forecasting model. From Table 1, we can get the reality amount of the electric-power demand in China from 2002 to 2004. We compared the forecast values between the two models and the results are presented in Table 3. Table 3 shows that the GreyMarkov forecasting model is better for forecasting the electric-power demand in China. The forecast values of GreyMarkov forecasting models are more precise than GM(1,1) Grey forecasting model, for data sequence with large random uctuation. 4. Conclusions The GreyMarkov forecasting model can fully utilize the information of the historical time-series data with large random uctuation. As the case study shows that the accuracy

ARTICLE IN PRESS
M. Huang et al. / Renewable Energy 32 (2007) 11651174 Table 3 Comparison of forecast results with two different methods. Year Reality (TW h) GM(1,1) Value (TW h) Precision 2002 2003 2004 16,200 18,910 21,735 16,438.15 17,779.7 19,230.8 98.53% 94.02% 88.48% GreyMarkov Value (TW h) Precision 16,106.7 18,541.3 21,204.7 99.42% 98.05% 97.56% 1173

of GreyMarkov forecasting model in forecast value from 2002 to 2004 is higher than those of GM(1,1) Grey forecasting model in this paper. Based on the above analysis, the GreyMarkov forecasting model appeals to be intrinsically better because it has merits of both simplicity of application and high forecasting precision. This model has integrated the advantages of GM(1,1) Grey forecasting model and Markov-chain forecasting model. The forecasting result of this forecasting method greatly depends on state-intervals partitioning. There is no standard rule to divide the states intervals. Generally speaking, the number of states should be decided according to the data and the demands of the problem. If the historical data are not abundant, the number of states should be fewer, so that the transition between the states can be clearly revealed. If the data are abundant, the number of states should be increased and the forecasting precision would be increased correspondingly. Because the electric-power demand in China is inuenced by many factors such as the economy, industry framework, peoples income level, weather and governments policy, there is a certain development trend for the historical time-series data, and the data uctuate randomly. So it is suitable to forecast the electric-power demand by the GreyMarkov forecasting model. The GreyMarkov forecasting model could be applied to forecast other time series problems with large random uctuation. Acknowledgements This study was supported by the Teaching and Research Award Program for Outstanding Young Teachers in Higher Education Institutions of MOE, PRC, Natural Science Foundation of China (Project No. 30270773), Specialized Research Fund for the Doctoral Program of Higher Education (Project no. 20040335034) and Natural Science Foundation of Zhejiang (Project no. RC02067). References
[1] Zhao XZ. Electricity industry revolution and development in China. Chin Power Enterprise Manage 2000;12:45. [2] Deng JL. Control problems of Grey System. Wuhang: Huazhong University of Science and Technology Press; 1990. [3] Chao HW. Predicting tourism demand using fuzzy time series and hybrid grey theory. Tourism Manage 2004;25:36774. [4] He Y, Bao YD. GreyMarkov forecasting model and its application. Syst Eng (Theory Practice) 1992;9(4):5963.

ARTICLE IN PRESS
1174 M. Huang et al. / Renewable Energy 32 (2007) 11651174 [5] Zhang SJ, He Y. A GreyMarkov forecasting model for forecasting the total power requirement of agricultural machinery in Shangxi Province. J Shanxi Agric Univ (Nat Sci Edi) 2001;21(3):299302. [6] Wen JC, Huang KH, Wen KL. The study of a in GM(1,1) model. J Chin Inst Eng 2000;23(5):5839. [7] Bao YD, Wu YP, He Y. A new forecasting model based on the combination of GM(1,1) model and linear regression. Syst Eng (Theory Practice) 2004;24(3):958. [8] Zhang SJ, He Y. Forecast of the grain yield based on the trend-state mathematical model. J Zhejiang Univ (Agric Life Sci) 2001;27(6):6736. [9] He Y. A new forecasting model for agricultural commodities. J Agric Eng Res 1995;60:22735. [10] National Bureau of Statistic of China. China statistical year book. Beijing: China Statistics Press; 2004.