CN104112163A - Construction method of electric power forecasting business index - Google Patents

Abstract

The invention provides a construction method of an electric power forecasting business index. The index can be used for forecasting electricity consumption. The method comprises the following steps: tracking electricity-consumption year-on-year growth data of a historical period and each economic leading indicator year-on-year growth data are obtained; according to correlation and hysteresis, an appropriate economic leading indicator is screened; weight of each economic leading indicator is determined; a calculation model of the electric power forecasting business index is established and constructed; and the electric power forecasting business index during the forecast period and the historical period is calculated. An advanced electric power forecasting business index is calculated by the utilization of the current known year-on-year growth value of each economic leading indicator, the tracking electricity-consumption year-on-year growth is forecasted and tracking electricity-consumption is further forecasted. It is conductive to power planning.

Description

A kind of building method of power prediction consumer confidence index

Technical field

The present invention relates to electrical network electricity demand forecasting technical field, specifically a kind of building method of power prediction consumer confidence index.

Background technology

At present, the China's economic leading indicators that Mei Yin Merrill Lynch releases mainly obtain according to the monthly weighted mean of speedup on year-on-year basis of seven indexes such as generated energy, crude steel output, cement output, automobile sales volume, the newly-started area in house, volume of rail freigh and medium-term and long-term credit, and its weight of index that undulatory property is higher is just less.Leading general about two months of the more industrial speedup data of these economic leading indicators, can help anticipation Economy pitch, and predict GDP speedup.Merchandising Manager's index is the index number system of handling monthly survey statistical summaries, working out based on enterprise procurement, the a series of indexes that changed by reflection economic cycle form, it is one of leading indicators of the reflection Economic Climate situation of generally acknowledging in the world, macroeconomy is had to the outstanding forewarning function of prediction in advance, can catch delicately economic development peak-to-valley value and " flex point ".To economic forewarning function, can predict the power consumption development trend of speedup on year-on-year basis according to above-mentioned two indexs by structure power prediction consumer confidence index, contribute to carry out power planning.

Summary of the invention

The object of the present invention is to provide a kind of building method of power prediction consumer confidence index, can realize the speedup on year-on-year basis of tracking power consumption is predicted by structure power prediction consumer confidence index, and then realize lower first phase or next issues of tracking power consumption are predicted.

Technical scheme of the present invention is:

A building method for power prediction consumer confidence index, comprises the following steps:

(1) obtain tracking power consumption speedup data and the each economic leading indicators speedup data on year-on-year basis on year-on-year basis of the N adjacent with time span of forecast historical phase;

(2) determine each economic leading indicators weight of speedup on year-on-year basis;

(3) set up the computation model of constructing power prediction consumer confidence index:

$I_t={\mathrm{\Σ}}_{i=1}^n{\mathrm{\ω}}_i*R_{i,t-a_i},$

Wherein, I _trepresent historical phase or the current power prediction consumer confidence index of time span of forecast, n represents the number of economic leading indicators, ω _irepresent i the economic leading indicators weight of speedup on year-on-year basis, a _irepresent i economic leading indicators on year-on-year basis speedup on the tracking power consumption hysteresis issue of speedup impact on year-on-year basis, represent leading current a _ithe i of a phase economic leading indicators numerical value of speedup on year-on-year basis;

(4) calculate time span of forecast and the power prediction consumer confidence index I of adjacent with described time span of forecast M historical phase _t, wherein M=N-a _imax, a _imaxthe maximal value representing.

The building method of described power prediction consumer confidence index, described step (2) is determined each economic leading indicators weight of speedup on year-on-year basis, specifically comprises the following steps:

(21), according to tracking power consumption speedup data and the each economic leading indicators speedup data on year-on-year basis on year-on-year basis of the N adjacent with time span of forecast that obtains historical phase, in Excel, utilize correlation coefficient function CORREL to calculate i economic leading indicators speedup and its a that lags behind on year-on-year basis in described N historical phase _ithe tracking power consumption of the phase related coefficient C of speedup on year-on-year basis _i, utilize standard deviation function STDEV.S to calculate i the economic leading indicators standard deviation S of speedup on year-on-year basis in described N historical phase _i;

(22) calculate i the economic leading indicators weights omega of speedup on year-on-year basis _i:

${\mathrm{\ω}}_i={\mathrm{\ω}}_{\mathrm{ic}}{*\mathrm{\ω}}_{\mathrm{is}}/{\mathrm{\Σ}}_{i=1}^n{\mathrm{\ω}}_{\mathrm{ic}}*{\mathrm{\ω}}_{\mathrm{is}},$

Wherein, ${\mathrm{\ω}}_{\mathrm{ic}}=C_i/{\mathrm{\Σ}}_{i=1}^n{C}_i,{\mathrm{\ω}}_{\mathrm{is}}=S_i/{\mathrm{\Σ}}_{i=1}^n{S}_i.$

As shown from the above technical solution, the present invention for the economic target of constructing power prediction consumer confidence index on year-on-year basis speedup on tracking power consumption on year-on-year basis the impact of speedup all there is hysteresis quality, therefore, can utilize the data of speedup on year-on-year basis of these economic targets of historical phase to synthesize the power prediction consumer confidence index of time span of forecast, and then obtain the tracking power consumption predicted value of speedup on year-on-year basis of time span of forecast, to contribute to carry out power planning.

Brief description of the drawings

Fig. 1 is method flow diagram of the present invention;

Fig. 2 is the tracking power consumption tendency situation map of speedup and power prediction consumer confidence index on year-on-year basis.

Embodiment

Further illustrate in conjunction with specific embodiments the present invention below.

Take Anhui Province as example, first, the economic leading indicators that qualitative analysis and tracking power consumption are closely related, according to two of economic leading indicators necessary conditions, relevant to tracking power consumption and the impact of tracking power consumption is had to hysteresis quality, the following economic target of initial option: area, Anhui Province's crude steel output, Anhui Province's cement output and national crude steel output are sold in Anhui Province's investment in fixed assets, Anhui Province's end of term loan balance, Anhui Province's medium-term and long-term credit remaining sum, V (M2), market, Anhui Province.

Secondly, each economic target of quantitative test initial option is speedup and the tracking power consumption correlativity between speedup and hysteresis quality on year-on-year basis on year-on-year basis, concrete grammar is the historical data (2005 first quarter～2013 year third quarter) of issuing according to authoritative institution, in Excel, utilize correlation coefficient function CORREL to calculate each economic target its speedup and tracking power consumption related coefficient between speedup on year-on-year basis on year-on-year basis of initial option, from without the lag period to 5 phases of hysteresis; For each economic target, the result calculating is sorted, find hysteresis issue corresponding to related coefficient maximal value, if related coefficient maximal value appeared at without the lag period, illustrate initial option this economic target its on year-on-year basis speedup on tracking power consumption on year-on-year basis the impact of speedup do not there is hysteresis quality, this economic target is got rid of.By said method, the index that obtains synthetic Anhui Province power prediction consumer confidence index is: area accumulative total speedup (impact lagged behind for 2 phases), national crude steel output speedup (impact lagged behind for 1 phase), Anhui Province's end of term loan balance speedup (impact lagged behind for 3 phases), Anhui Province's cement output speedup (impact lagged behind for 3 phases) on year-on-year basis on year-on-year basis on year-on-year basis is on year-on-year basis sold in market, Anhui Province.

As shown in Figure 1, a kind of building method of power prediction consumer confidence index, comprises the following steps:

S1, obtain tracking power consumption speedup data and the each economic leading indicators speedup data on year-on-year basis on year-on-year basis of N history phase adjacent with time span of forecast before time span of forecast:

In the present embodiment, adopt season data, taking the fourth quarter in 2013 as time span of forecast, taking each season in the first quarter～2013 year third quarter in 2005 as the historical phase, the tracking power consumption of historical phase on year-on-year basis speedup and each economic leading indicators speedup data are as shown in table 1 on year-on-year basis:

Table 1

S2, determine each economic leading indicators weight of speedup on year-on-year basis:

(1) correlation coefficient process: economic leading indicators on year-on-year basis speedup and the tracking power consumption weight that the correlativity of speedup is given more greatly on year-on-year basis larger.

In Excel, utilize correlation coefficient function CORREL calculate respectively four economic leading indicators on year-on-year basis speedup be that area accumulative total speedup, national crude steel output speedup, Anhui Province's end of term loan balance speedup, Anhui Province's cement output speedup and the tracking power consumption related coefficient C of speedup on year-on-year basis on year-on-year basis on year-on-year basis on year-on-year basis on year-on-year basis sold in market, Anhui Province ₁, C ₂, C ₃, C ₄.

Consider the hysteresis quality of impact, calculate C ₁time, adopt the market, Anhui Province in the first quarter～2013 year first quarter in 2005 to sell area accumulative total speedup data and the tracking power consumption speedup data on year-on-year basis in the third quarter～2013 year third quarter in 2005 on year-on-year basis; Calculate C ₂time, the national crude steel output that adopts the first quarter～2013 year second quarter in 2005 is speedup data and the tracking power consumption speedup data on year-on-year basis in the second quarter～2013 year third quarter in 2005 on year-on-year basis; Calculate C ₃time, the Anhui Province's end of term loan balance that adopts the first quarter～2012 year fourth quarter in 2005 is speedup data and the tracking power consumption speedup data on year-on-year basis in the fourth quarter～2013 year third quarter in 2005 on year-on-year basis; Calculate C ₄time, the Anhui Province's cement output that adopts the first quarter～2012 year fourth quarter in 2005 is speedup data and the tracking power consumption speedup data on year-on-year basis in the fourth quarter～2013 year third quarter in 2005 on year-on-year basis.

Note C=C ₁+ C ₂+ C ₃+ C ₄, four economic leading indicators on year-on-year basis the weight of speedup be respectively:

ω _1c＝C _１／C，ω _2c＝C ₂／C，ω _3c＝C ₃／C，ω _4c＝C ₄／C。

(2) undulatory property method: the economic leading indicators weight that the undulatory property of speedup is given more greatly is on year-on-year basis less, and the size of undulatory property is weighed by standard deviation.

In Excel, utilize standard deviation function STDEV.S calculate respectively four economic leading indicators on year-on-year basis speedup be that area accumulative total speedup, national crude steel output speedup, Anhui Province's end of term loan balance speedup, the Anhui Province's cement output standard deviation S of speedup on year-on-year basis on year-on-year basis on year-on-year basis is on year-on-year basis sold in market, Anhui Province ₁, S ₂, S ₃, S ₄.

When calculating, do not need to consider the hysteresis quality of impact, each economic leading indicators on year-on-year basis speedup all adopts the data in the first quarter～2013 year third quarter in 2005.

Note S=S ₁+ S ₂+ S ₃+ S ₄, four economic leading indicators on year-on-year basis the weight of speedup be respectively:

ω _１s＝S _１／S，ω _2s＝S ₂／S，ω _3s＝S ₃／S，ω _4s＝S ₄／S。

(3) overall approach: overall approach is to combine correlation coefficient process and undulatory property method, comprehensively provides weight according to the rule of correlation coefficient process and undulatory property method.

Note γ ₁=ω _1c* ω _1s, γ ₂=ω _2c* ω _2s, γ ₃=ω _3c* ω _3s, γ ₄=ω _4c* ω _4s, γ=γ ₁+ γ ₂+ γ ₃+ γ ₄, four economic leading indicators on year-on-year basis the weight of speedup be respectively:

ω ₁＝γ ₁／γ，ω ₂＝γ ₂／γ，ω ₃＝γ ₃／γ，ω ₄＝γ ₄／γ。

S3, set up the computation model (weight providing taking above-mentioned overall approach is example) of power prediction consumer confidence index:

I _t＝ω ₁*R _1，t-2+ω ₂*R _2，t-1+ω ₃*R _3，t-3+ω ₄*R _4，t-3

Wherein, I _trepresent current power prediction consumer confidence index, R _{1, t-2}area accumulative total speedup on year-on-year basis, R are sold in the market, Anhui Province that represents leading current 2 phases _{2, t-1}represent the national crude steel output speedup on year-on-year basis of leading current l phase, R _{3, t-3}represent Anhui Province's end of term loan balance speedup, R on year-on-year basis of leading current 3 phases _{4, t-3}represent Anhui Province's cement output speedup on year-on-year basis of leading current 3 phases.

S4, consider and affect hysteresis quality, utilize the computation model of above-mentioned power prediction consumer confidence index to calculate the power prediction consumer confidence index in the fourth quarter～2013 year each season in the third quarter in 2005, and the power prediction consumer confidence index in 2013 year fourth quarter of time span of forecast.

The power prediction consumer confidence index that utilization calculates, predicts the tracking power consumption in the fourth quarter in 2013, comprises the following steps:

S5, set up tracking power consumption speedup forecast model on year-on-year basis, its regression equation is:

Y _t＝A+B*I _t

Wherein, Y _trepresent tracking power consumption speedup on year-on-year basis, A is that constant term, B are I _tcoefficient, A, B be by the power prediction consumer confidence index in historical phase in the fourth quarter～2013 year each season in the third quarter in 2005 and corresponding tracking power consumption on year-on-year basis in the above-mentioned regression equation of speedup data substitution matching obtain.

S6, by the i.e. above-mentioned regression equation of power prediction consumer confidence index substitution in the fourth quarter in 2013 of the time span of forecast calculating, can calculate the tracking power consumption predicted value of speedup on year-on-year basis in the fourth quarter in 2013, according to this predicted value and the tracking power consumption numerical value in the fourth quarter in 2012, just can obtain the tracking electricity demand forecasting value in the fourth quarter in 2013 again.

As shown in Figure 2, on long terms, power prediction consumer confidence index and tracking the power consumption on year-on-year basis tendency of speedup are basic identical, thus can utilize the power prediction consumer confidence index before the excess of export of the current known numerical evaluation of speedup on year-on-year basis of each economic leading indicators, to tracking power consumption on year-on-year basis speedup predict.

The above embodiment is only that the preferred embodiment of the present invention is described; not scope of the present invention is limited; design under the prerequisite of spirit not departing from the present invention; various distortion and improvement that those of ordinary skill in the art make technical scheme of the present invention, all should fall in the definite protection domain of claims of the present invention.

Claims (2)

1. a building method for power prediction consumer confidence index, is characterized in that, the method comprises the following steps:

(1) obtain tracking power consumption speedup data and the each economic leading indicators speedup data on year-on-year basis on year-on-year basis of the N adjacent with time span of forecast historical phase;

(2) determine each economic leading indicators weight of speedup on year-on-year basis;

(3) set up the computation model of constructing power prediction consumer confidence index:

$I_t={\mathrm{\Σ}}_{i=1}^n{\mathrm{\ω}}_i*R_{i,t-a_i},$

Wherein, I _trepresent historical phase or the current power prediction consumer confidence index of time span of forecast, n represents the number of economic leading indicators, ω _irepresent i the economic leading indicators weight of speedup on year-on-year basis, a _irepresent i economic leading indicators on year-on-year basis speedup on the tracking power consumption hysteresis issue of speedup impact on year-on-year basis, represent leading current a _ithe i of a phase economic leading indicators numerical value of speedup on year-on-year basis;

(4) calculate time span of forecast and the power prediction consumer confidence index I of adjacent with described time span of forecast M historical phase _t, wherein M=N-a _imax, a _imaxrepresent a _imaximal value.

2. the building method of power prediction consumer confidence index according to claim 1, is characterized in that, described step (2) is determined each economic leading indicators weight of speedup on year-on-year basis, specifically comprises the following steps:

(21), according to tracking power consumption speedup data and the each economic leading indicators speedup data on year-on-year basis on year-on-year basis of the N adjacent with time span of forecast that obtains historical phase, in Excel, utilize correlation coefficient function CORREL to calculate i economic leading indicators speedup and its a that lags behind on year-on-year basis in described N historical phase _ithe tracking power consumption of the phase related coefficient C of speedup on year-on-year basis _i, utilize standard deviation function STDEV.S to calculate i the economic leading indicators standard deviation S of speedup on year-on-year basis in described N historical phase _j;

(22) calculate i the economic leading indicators weights omega of speedup on year-on-year basis _i:

${\mathrm{\ω}}_i={\mathrm{\ω}}_{\mathrm{ic}}{*\mathrm{\ω}}_{\mathrm{is}}/{\mathrm{\Σ}}_{i=1}^n{\mathrm{\ω}}_{\mathrm{ic}}*{\mathrm{\ω}}_{\mathrm{is}},$

Wherein, ${\mathrm{\ω}}_{\mathrm{ic}}=C_i/{\mathrm{\Σ}}_{i=1}^n{C}_i,{\mathrm{\ω}}_{\mathrm{is}}=S_i/{\mathrm{\Σ}}_{i=1}^n{S}_i.$