CN105654197A - Grey theory-based comprehensive plan investment total prediction method - Google Patents

Abstract

The invention relates to a grey theory-based comprehensive plan investment total prediction method. The method includes the following steps that: a) key influence factors are selected: possible influence factors which influence comprehensive plan investment total are analyzed from the three aspects of economic development, company management and investment performance; b) a grey prediction method is utilized to predict the change trend of the key influence factors: corresponding program codes are written by means of tools and according to the algorithm steps of grey prediction, and the future trend calculation values of the key influence factors such as regional GDP, the maximum load of power consumption of the whole society, power sale quantity and total profits can be obtained after programs are operated; c) the degree of influence of the key influence factors is determined: the qualitative sequence of the calculation factors is converted into a decision coefficient by means of analytic hierarchy process and based on qualitative analysis and quantitative analysis, and the degree of influence of the change of the factors on total investment is determined according to the decision coefficient; and d) comprehensive plan investment scale is predicted: the annual prediction result of comprehensive plan investment is obtained according to a calculation model.

Description

A kind of overall planning based on gray theory drops into total dish Forecasting Methodology

Technical field

What the present invention relates to is that a kind of overall planning based on gray theory drops into total dish Forecasting Methodology, belongs to the investment decision-making technical field in Enterprise Integrated planning management.

Background technology

Existing Forecasting Methodology mainly comprises qualitative forecasting method (market survey prediction procedure, scholarly forecast method, subjective probability method, omen prediction procedure etc.), time-series flat forecast (exponential smoothing, difference index smoothing method, adaptive filtering method etc.), regression model (Linear Regression Model in One Unknown, multiple linear regression model, nonlinear regression model (NLRM) etc.), trend extrapolation prediction procedure (Exponential Curve Method, Prediction by Modified Index Curve method, growth curve method, enveloping curve method etc.), Markov prediction procedure (is used for merchandise sales status predication, share of market is predicted, Expectation Returns Forecast) etc.

The above prediction analysis method of Integrated comparative, although regression model is comparatively common Forecasting Methodology, but it needs bigger size of a sample as support. Once size of a sample is excessively little, predicts the outcome and it is easy to that deviation occurs.

Grey prediction is taught by Central China Polytechnics Deng Julong after proposing in nineteen eighty-two, it is widely applied at the numerous areas such as social, economic, scientific and technical, become that relevant field carries out predicting, decision-making, assessment, planning control, system analysis and modeling one of important method, it has advantages such as " modeling information are few; computing is convenient, modeling accuracy height ".

Summary of the invention

It is an object of the invention to overcome the deficiency of prior art existence, and provide a kind of by determining to affect the key influence factor that overall planning drops into total dish, and in conjunction with gray prediction method and analytical hierarchy process, reduce the human factor in investment decision-making management process, guaranteeing prediction process science, simplicity, efficient more, the overall planning based on gray theory more consistent with practical situation that predict the outcome drops into total dish Forecasting Methodology.

It is an object of the invention to complete by following technical solution, a kind of overall planning based on gray theory drops into total dish Forecasting Methodology, and described Forecasting Methodology comprises the steps:

A) key influence factor is selected: drop into which influence factor total dish may exist from Economic development class factor, company management class factor, portfolio performance class factor three aspect analyzing influence overall plannings: and the computing function eventually through Excel form obtains five candidate's factors: operation revenue, area GDP, total profit, electricity sales amount and electricity consumption maximum load;

B) gray prediction method is used to predict the variation tendency of key influence factor: by MATLAB instrument, according to the algorithm steps of grey prediction, write corresponding program code, after working procedure, just can obtain the Trend value measuring and calculating value in future such as key influence factor such as area GDP, whole society's electricity consumption maximum load, electricity sales amount, total profit etc.;

C) determine the influence degree of key influence factor: by analytical hierarchy process, comprehensive qualitative analysis and quantitative analysis, the qualitative sequence of measuring and calculating factor is converted into decision-making coefficient, then determines, according to decision-making coefficient, the influence degree that total dish is dropped into by each factors vary.

Key step is as follows: (1) Judgement Matricies, determines the importance of each influence factor according to " Method of nine marks ", and carries out consistency check; (2) calculate decision-making coefficient, find out the proper vector that the maximum eigenwert of judgment matrix is corresponding, and be normalized; (3) calculating influence coefficient: by above-mentioned decision-making coefficient, calculates the size that overall planning is dropped into influence degree by each factor;

D) predict overall planning scale of input: according to measuring and calculating model, overall planning can be obtained according to following formula and drop into annual prediction result:

$z=\underset{i=1}{\overset{n}{\Σ}}{u}_i\×\left(x_i\right(j+1)-x_i(j\left)\right)$

In above-mentioned formula, Z represents that the overall planning of program year drops into measuring and calculating value; u_iIt is the influence coefficient (influence degree size) of i-th key influence factor, the variable quantity that namely overall planning corresponding to factor i unit change drops into; x_i(j+1) it is the target value of jth+1 year (program year) factor i; x_iJ () is the actual value of jth year factor i.

As preferably: described: step a) in, the selection of influence factor (1) Economic development class factor is: tentatively determine area 3 factors such as GDP, whole society's power consumption, whole society's electricity consumption maximum load. Wherein, area GDP is the important indicator of reflection local economic development trend, and what whole society's power consumption, whole society's electricity consumption maximum load reflected is the alteration trend of region electricity needs, and these are all the important factor in order that electric power enterprise carries out Investment decision-making;

The selection of influence factor (2) company management class factor is: tentatively determine 4 factors such as electricity sales amount, operation revenue, gross assets, total profit. Wherein, electricity sales amount associates with " whole society's power consumption " greatly, the alteration trend of reflection electricity needs; Operation revenue, gross assets, total profit are then the important indicators of reflection company management present situation and the investment ability;

The selection of influence factor (3) portfolio performance class factor is: tentatively determine 3 factors such as net assets income ratio, unit electric grid investment sale of electricity increment, unit electric grid investment load increment. These indexs, in order to weigh the benefit of company's input-output, should be included in when screening influence factor, and power grid enterprises also to be considered gain on investments after all;

By the relevant historical data of Comprehensive Plan Management is carried out correlation analysis, directly reject the relatively low factor of dependency, reduce scope further, obtain five candidate's factors: operation revenue, area GDP, total profit, electricity sales amount and electricity consumption maximum load.

Prediction overall planning drops into total dish, as long as possessing the basic data of 5 years before this regional GDP, electricity consumption maximum load, electricity sales amount and total profit, by the utility software of self-developing, it is not necessary to the predicted data that next annual overall planning drops into total dish within 10 minutes, can be obtained. As being necessary, in the utility software of self-developing, relevant data can be adjusted (such as the numerical value in change judgment matrix, fine setting relevant historical data etc.), obtain new predicting the outcome fast, very convenient and practical.

Generally, the present invention is on the basis reducing person works's amount, it is to increase overall planning drops into accuracy and the science of total dish prediction.

Embodiment

Below in conjunction with specific embodiment, the present invention will be described in detail: the overall planning based on gray theory of the present invention drops into total dish Forecasting Methodology, and it mainly comprises the steps:

A) key influence factor is selected:

First, drop into which influence factor total dish may exist from three aspect analyzing influence overall plannings:

(1) Economic development class factor: tentatively determine 3 factors such as area GDP, whole society's power consumption, whole society's electricity consumption maximum load. Wherein, area GDP is the important indicator of reflection local economic development trend, and what whole society's power consumption, whole society's electricity consumption maximum load reflected is the alteration trend of region electricity needs, and these are all the important factor in order that electric power enterprise carries out Investment decision-making.

(2) company management class factor: tentatively determine 4 factors such as electricity sales amount, operation revenue, gross assets, total profit. Wherein, electricity sales amount associates with " whole society's power consumption " greatly, the alteration trend of reflection electricity needs; Operation revenue, gross assets, total profit are then the important indicators of reflection company management present situation and the investment ability.

(3) portfolio performance class factor: tentatively determine 3 factors such as net assets income ratio, unit electric grid investment sale of electricity increment, unit electric grid investment load increment. These indexs, in order to weigh the benefit of company's input-output, should be included in when screening influence factor, and power grid enterprises also to be considered gain on investments after all.

Secondly, by the relevant historical data of Comprehensive Plan Management is carried out correlation analysis, directly reject the relatively low factor of dependency, reduce scope further, obtain five influence factors: operation revenue, area GDP, total profit, electricity sales amount and electricity consumption maximum load. This step realizes mainly through the computing function of Excel form. Concrete operation: newly-built Excel list, input raw data associated, clicks " Data analysis instrument relation conefficient ", can carry out overall planning and drop into total dish and the correlation calculations of candidate's influence factor.

An alternative embodiment of the invention can also on the basis of above-mentioned five influence factors, candidate's influence factor that 5 dependencys drawn through correlation calculations are higher is analysed in depth, therefrom finally determine that key influence factor is four, i.e. area GDP, electricity consumption maximum load, electricity sales amount and total profit. These four key factors in fact from meeting socio-economic development, to ensure electrical network strong, represent company management achievement and manage four different sides such as performance, the scale of input of common restriction overall planning.

B) gray theory is used to predict:

Actually operating needs by MATLAB instrument, according to the algorithm steps of grey prediction, write corresponding program code, after working procedure, just can obtain the Trend value measuring and calculating value in future such as key influence factor such as area GDP, whole society's electricity consumption maximum load, electricity sales amount, total profit etc.

C) influence degree of key influence factor is determined:

In order to determine that overall planning is dropped into the influence degree of total dish by five or four key influence factors, by analytical hierarchy process, comprehensive qualitative analysis and quantitative analysis, by the qualitative sequence of measuring and calculating factor (according to expertise, the importance of five or four key influence factors is sorted) it is converted into decision-making coefficient, then determine, according to decision-making coefficient, the influence degree that total dish is dropped into by each factors vary.

Key step is as follows: (1) Judgement Matricies, determines the importance of each influence factor according to " Method of nine marks ", and carries out consistency check;(2) calculate decision-making coefficient, find out the proper vector that the maximum eigenwert of judgment matrix is corresponding, and be normalized; (3) calculating influence coefficient: by above-mentioned decision-making coefficient, calculates the size that overall planning is dropped into influence degree by each factor.

D) overall planning scale of input is predicted:

According to measuring and calculating model, overall planning can be obtained according to following formula and drop into annual prediction result:

$z=\underset{i=1}{\overset{n}{\Σ}}{u}_i\×\left(x_i\right(j+1)-x_i(j\left)\right)$

In above-mentioned formula, Z represents that the overall planning of program year drops into measuring and calculating value; u_iIt is the influence coefficient (influence degree size) of i-th key influence factor, the variable quantity that namely overall planning corresponding to factor i unit change drops into; x_i(j+1) it is the target value of jth+1 year (program year) factor i; x_iJ () is the actual value of jth year factor i.

Embodiment: predict GM (1 according to grey, 1) model, by MATLAB program, the predictor of Zhejiang area GDP in 2016 is 46,566 hundred million yuan, whole society's electricity consumption maximum load predictor is 6,617 ten thousand kilowatts, electricity sales amount predictor is 2,004 hundred million kilowatt-hours, and total profit predictor is 38.2 hundred million. The influence coefficient simultaneously extrapolating key influence factor is respectively: area GDP is 0.053, and whole society's electricity consumption maximum load is 0.39, and electricity sales amount is 0.37, and total profit is 5.28. Dropping into measure formula according to total dish, knowing Zhejiang company by inference, within 2016, to drop into total dish measuring and calculating value be 32,800,000,000 yuan, and the investment amount of assign 30,000,000,000 yuan of this and higher level company is closely.

Claims (2)

1. the overall planning based on gray theory drops into total dish Forecasting Methodology, it is characterised in that described Forecasting Methodology comprises the steps:

A) key influence factor is selected: drop into which influence factor total dish may exist from Economic development class factor, company management class factor, portfolio performance class factor three aspect analyzing influence overall plannings, and the computing function eventually through Excel form obtains five candidate's factors: operation revenue, area GDP, total profit, electricity sales amount and electricity consumption maximum load;

B) gray prediction method is used to predict the variation tendency of key influence factor: by MATLAB instrument, according to the algorithm steps of grey prediction, write corresponding program code, after working procedure, just can obtain the trend measuring and calculating value in future such as key influence factor such as area GDP, whole society's electricity consumption maximum load, electricity sales amount, total profit etc.;

C) determine the influence degree of key influence factor: by analytical hierarchy process, comprehensive qualitative analysis and quantitative analysis, the qualitative sequence of measuring and calculating factor is converted into decision-making coefficient, then determines, according to decision-making coefficient, the influence degree that total dish is dropped into by each factors vary.

Key step is as follows: (1) Judgement Matricies, determines the importance of each influence factor according to " Method of nine marks ", and carries out consistency check; (2) calculate decision-making coefficient, find out the proper vector that the maximum eigenwert of judgment matrix is corresponding, and be normalized; (3) calculating influence coefficient: by above-mentioned decision-making coefficient, calculates the size that overall planning is dropped into influence degree by each factor;

D) predict overall planning scale of input: according to measuring and calculating model, overall planning can be obtained according to following formula and drop into annual prediction result:

$z=\underset{i=1}{\overset{n}{\Σ}}{u}_i\×\left(x_i\right(j+1)-x_i(j\left)\right)$

In above-mentioned formula, Z represents that the overall planning of program year drops into measuring and calculating value; u_iIt is the influence coefficient (influence degree size) of i-th key influence factor, the variable quantity that namely overall planning corresponding to factor i unit change drops into; x_i(j+1) it is the target value of jth+1 year (program year) factor i;X_iJ () is the actual value of jth year factor i.

2. the overall planning based on gray theory according to claim 1 drops into total dish Forecasting Methodology, it is characterised in that described:

Step a) in, the selection of influence factor (1) Economic development class factor is: tentatively determine area 3 factors such as GDP, whole society's power consumption, whole society's electricity consumption maximum load. Wherein, area GDP is the important indicator of reflection local economic development trend, and what whole society's power consumption, whole society's electricity consumption maximum load reflected is the alteration trend of region electricity needs, and these are all the important factor in order that electric power enterprise carries out Investment decision-making;

The selection of influence factor (2) company management class factor is: tentatively determine 4 factors such as electricity sales amount, operation revenue, gross assets, total profit. Wherein, electricity sales amount associates with " whole society's power consumption " greatly, the alteration trend of reflection electricity needs; Operation revenue, gross assets, total profit are then the important indicators of reflection company management present situation and the investment ability;

The selection of influence factor (3) portfolio performance class factor is: tentatively determine 3 factors such as net assets income ratio, unit electric grid investment sale of electricity increment, unit electric grid investment load increment. These indexs, in order to weigh the benefit of company's input-output, should be included in when screening influence factor, and power grid enterprises also to be considered gain on investments after all;

By the relevant historical data of Comprehensive Plan Management is carried out correlation analysis, directly reject the relatively low factor of dependency, reduce scope further, obtain five candidate's factors: operation revenue, area GDP, total profit, electricity sales amount and electricity consumption maximum load.