CN106779205A - A kind of purchase sale of electricity optimization method of meter and the market share and value-added service cost of investment - Google Patents

Abstract

The invention discloses a kind of meter and the purchase sale of electricity policy optimization method of the market share and value-added service cost of investment.It is determined that all kinds of factors of influence sale of electricity houses market fixed portion size, the weight of each index of the influence market share is calculated using analytic hierarchy process (AHP), the assessment result of user utility is obtained；With reference to Logistic regression models, market share model is set up；Finally according to Conditional Lyapunov ExponentP method（CVaR）Set up the purchase sale of electricity model of meter and market share size and value-added service cost of investment.The sale of electricity company that the present invention sets up after being opened for sale of electricity side provides the purchase sale of electricity model of meter and market share influence, sale of electricity company can be according to the risk dynamics that oneself can be born, the marketing strategy of adjustment company, formulate optimal purchase sale of electricity strategy and power marketing value-added service capital project, to a certain extent, the market share shared by lifting itself, while avoiding risk, improves income.

Description

A kind of purchase sale of electricity optimization method of meter and the market share and value-added service cost of investment

Technical field

The present invention relates to electricity market field, and in particular to the purchase of a kind of meter and the market share and value-added service cost of investment Sale of electricity policy optimization method.

Background technology

With deepening continuously for electric system reform, sales market main body starts that diversification form, sale of electricity is presented Market Competition general layout is just gradually formed.According to up to hundreds of families of sale of electricity company that relevant data display, the whole nation have been set up, in market Under competitive environment, sale of electricity enterprise meets the business risk such as face customer churn, lack of capital at any time.To win competitive advantage, if only Consider price competition, user is attracted by reducing electricity price, such competitive method is being come to user, enterprise and entirely sold in the long term All lose more than gain in electric market.So, sale of electricity side open market situation under, sale of electricity main body how to improve it is non-in price Competitiveness, attract more users, extend volume growth, while avoid risk again, improve income, be asking for a urgent need to resolve Topic.

At present, to sale of electricity enterprise income, risk investigation are generally concentrated in the purchase sale of electricity decision in the face of risk of sale of electricity enterprise, and The yield risk research that sale of electricity Company Establishment value-added service etc. is extended volume growth is not directed to, research before is more focused on Sale of electricity main body how to distribute power purchase ratio and how sale of electricity bigger income and bears smaller risk to obtain.To user's After demand is not studied specifically with selection, and sale of electricity side opens, user can freely select sale of electricity main body, and tradition to sell The method of the electric state monopoly for purchase and marketing is different, and in the long run/term, sale of electricity company is more contemplated that the big of the market share shared by company It is small, power marketing value-added service is set up to attract more clients, influence of the market share to sale of electricity corporate income and risk is added, Set up and more comprehensively purchase sale of electricity model.

The content of the invention

Goal of the invention：In order to lift the market share shared by sale of electricity company itself, raising income of avoiding risk, the present invention is carried For a kind of meter and the purchase sale of electricity policy optimization method of the market share and value-added service cost of investment.

Technical scheme：The purchase sale of electricity policy optimization method of a kind of meter and the market share and value-added service cost of investment, the party Method is comprised the following steps：

(1) determine all kinds of factors of influence sale of electricity houses market fixed portion size, influence market is calculated using analytic hierarchy process (AHP) The weight coefficient of all kinds of indexs of share, obtains the assessment result of user utility；

(2) Logistic regression models are combined, according to the result of calculation of user utility, sale of electricity houses market share mould is set up Type；

(3) sale of electricity company meter and the market share and value-added service cost of investment are set up with Conditional Lyapunov ExponentP method (CVaR) Purchase sale of electricity model, with sale of electricity company maximization of utility as target, the market share shared by sale of electricity company is added in the model Size, and the cost of investment of power marketing value-added service is added in the cost payout of sale of electricity company.

Beneficial effect：Compare prior art, shared by the sale of electricity company that the purchase sale of electricity model set up in the present invention is included The market share and value-added service cost of investment, more comprehensively, objectively can purchase sale of electricity and value-added service throwing for sale of electricity company provides Financial strength degree etc. is referred to；The risk dynamics that sale of electricity company be able to can bear according to oneself is analyzed to model, so as to formulate phase The purchase sale of electricity strategy and capital project answered, to a greater extent, lift the market share shared by itself, and raising of avoiding risk is received Benefit, while also having ensured the power quality of user.

Brief description of the drawings

Accompanying drawing 1 is meter of the present invention and the purchase sale of electricity policy optimization method of the market share and value-added service cost of investment Flow chart.

Specific embodiment

With reference to the accompanying drawings and detailed description, the present invention will be further described.

As shown in figure 1, the purchase sale of electricity policy optimization method of meter and the market share and value-added service cost of investment is including following Step：

(1) determine all kinds of factors of influence sale of electricity houses market fixed portion size, influence market is calculated using analytic hierarchy process (AHP) The weight coefficient of all kinds of indexs of share, obtains the assessment result of user utility；

1) hierarchical structure is set up

Influence all kinds of indexs of market share size to be determined by the marketing strategy of sale of electricity company, stress comprising electric power in index The every value-added service of marketing, after determining rear index, sets up destination layer, rule layer and the indicator layer of appraisement system, wherein, due to User utility corresponding to sale of electricity company is higher, and the market share shared by sale of electricity company is bigger, so the present invention is from user's effect With being destination layer；Rule layer is all kinds of performance indications for influenceing user utility, and indicator layer is the specific city under corresponding rule layer Market share influence factor.

2) weight valuation

After having set up hierarchical structure, by Judgement Matricies, by a certain of more than policymaker or relevant expert one level Factor is criterion, and this level factor associated therewith is compared two-by-two, determines its relative importance.The present invention takes 1 ~9 proportion quotiety methods, are respectively compared the relative importance of each level index, and clearly quantified with numerical value.For example, setting up Each rule layer be A for the judgment matrix of destination layer, the element of judgment matrix determines as shown in table 1.

The judgment matrix element of table 1 determines

Judgment matrix to establishing carries out individual layer sequence and consistency check, and individual layer sequence is according to judgment matrix, meter Calculate for certain element in last layer time, the weights of the importance of the associated element of this level.It judges square by this level The component that characteristic vector corresponding to the eigenvalue of maximum of battle array is done after normalized draws：

CW=λ_maxW (2)

Wherein, λ_max, W represent the eigenvalue of maximum and corresponding characteristic vector of judgment matrix respectively；

Each component of characteristic vector W is normalized, you can obtain the weight vectors of individual layer sequence：

Calculate the uniformity that CR checks judgment matrix：

Wherein,N is the exponent number of judgment matrix；RI is Aver-age Random Consistency Index, with judgment matrix rank Number is related；If CR<0.1, then judgment matrix there is uniformity, otherwise need to modify judgment matrix.

3) total hierarchial sorting

Based on same level all levels list weight order, the combining weights weighting of a hierarchical elements is used, you can Calculate the weights of this level all elements importance for the whole level of last layer time.Total hierarchial sorting need on to Under successively carry out, for top, its Mode of Level Simple Sequence is total sequence.By total hierarchial sorting, each index is calculated The total evaluation result of weight, i.e. user utility, the i.e. market share, is represented with following formula：

U=β₁C₁+β₂C₂+β₃C₃+β₄C₄+…+β_iC_i (5)

Wherein, β_iRepresent the corresponding weight coefficient of each two-level index, C_iRepresent i-th index in appraisement system.

(2) Logistic regression models are combined, according to the result of calculation of user utility, sale of electricity houses market share mould is set up Type；

Logistic regression models, are based on the maximum a kind of discrete selection mould set up with stochastic utility theory of effectiveness Type, emphasizes the housing choice behavior of individuality, and after sale of electricity side opens, customer just decides public affairs for the housing choice behavior of sale of electricity company The size of the shared market share of department, the probability that company is easily selected by a user is bigger, and the market share shared by it is bigger, so according to The assessment result of user utility can set up the market share model of sale of electricity company.

Market share model based on Logistic regression models is expressed as follows：

In above formula, P represents market share size, is the real number in 0~1；x₁~x_pThe construction of each two-level index is corresponded to respectively The construction degree of each two-level index is quantified as degree, the present invention 1~9 real number, and two-level index its construction level is higher, x_i's Value is bigger；β₁~β_pValue is respectively the corresponding weight coefficient of each two-level index, wherein β₀Value can be calculated by reasonable assumption Draw (such as：When the construction degree of all two-level index is mean level, i.e. numerical value 5, the market share of occupancy is 0.5, is brought into β can be tried to achieve in market share model₀Value).

(3) sale of electricity company meter and the market share and value-added service cost of investment are set up with Conditional Lyapunov ExponentP method (CVaR) Purchase sale of electricity model, with sale of electricity company maximization of utility as target, the market share shared by sale of electricity company is added in the model Size, and the cost of investment of power marketing value-added service is added in the cost payout of sale of electricity company.

Wherein, the process of setting up of purchase sale of electricity model includes：

1) sale of electricity company purchases strategies C

The purchases strategies of sale of electricity company include two parts, a part be bilateral contract market, ahead market and in real time The purchases strategies in market, another part is the cost of investment of market share size shared by sale of electricity company and its value-added service, by , it is necessary to be important to notice that by the construction of value-added service to extend volume growth after sale of electricity side opens, therefore all two In level index, the overall cost of ownership γ of value-added service is only added, and the quantification gradation of index is higher, γ is bigger, i.e.,：

C=P (C_C+C_DA+C_RT)+γ (7)

C_C=q_C,t·p_C (8)

Constraints：

q_C,t+q_DA,t+q_RT,t=q_all,t (11)

In formula：C is the purchases strategies of sale of electricity company；C_C、C_DAAnd C_RTRespectively sale of electricity company in bilateral contract market, a few days ago Market and the purchases strategies of Real-time markets；T is research period sum；p_C、p_DA,tAnd p_RT,tRespectively two day market power purchase price with And the purchase electricity price of t periods ahead market and Real-time markets, wherein p_CIt is bilateral transaction price with q_C,tThe function of change；q_C,t、 q_DA,t、q_RT,tAnd q_all,tRespectively t period bilateral contract market power purchase electricity, Day-ahead Electricity Purchase electricity, Real-time markets power purchase Electricity and total power purchase electricity；P is market share size shared by sale of electricity company；γ is the overall cost of ownership of value-added service.

2) sale of electricity company power selling income R

Sale of electricity company is that can provide the user with different sale of electricity contract type in sale of electricity, and the power selling income of sale of electricity company is exactly The income obtained by all kinds of sale of electricity contracts, i.e.,：

Constraints：

I=1,2 ..., h (14)

In formula, h is the species of sale of electricity contract；R_iIt is i-th kind of power selling income of sale of electricity contract；N is total number of users；N_iIt is purchase Buy i-th kind of number of users of sale of electricity contract；f_iAnd p_iRespectively i-th kind cost function and pricing structure of sale of electricity contract；q_j,tIt is jth Load of the individual user in the t periods.

3) the purchase sale of electricity model of sale of electricity company meter and the market share

The purchase sale of electricity model of sale of electricity company is in tradition purchase sale of electricity model, to add market share size and increment clothes The cost of investment of business, is modeled by Conditional Lyapunov ExponentP method (CVaR) to the uncertainty that electricity price and demand are brought.Root According to the definition of CVaR, the present invention purchases the N of dynamoelectric benefit P ' using the generation of Monte Carlo simulations method_MIndividual sample is sold obtaining purchase The CVaR values of electric profit.

Object function：

Wherein, β is the risk averse coefficient of sale of electricity company；N_MIt is Monte Carlo simulation number of times；α is the confidence of CVaR Level；η_xIt is the decision variable for introducing；ξ₁It is sale of electricity company in the power purchase allocation proportion of ahead market Yu bilateral contract market；ξ₂ It is sale of electricity company in the power purchase allocation proportion of Real-time markets Yu bilateral contract market.

Constraints：

C_C=q_C,t·p_C (16)

q_C,t+q_DA,t+q_RT,t=q_all,t (19)

I=1,2 ..., n (21)

X=1,2 ..., N_M (22)

P '=R-C (23)

η_x≥E(P′)-P_i′-ξ₁-ξ₂ (24)

η_x≥0 (25)

Wherein, P_i' P ' the values drawn by i ＆ lt Monte Carlo simulations；E (P ') is that sale of electricity company purchases dynamoelectric benefit Desired value.

Formula (15) is the optimization aim with sale of electricity company maximization of utility as model, wherein comprising two parts：1) sale of electricity is public Department's profit value, profit value is equal to sale of electricity and subtracts each market purchases strategies and value-added service throwing to the income that user obtains in the present invention The difference of cost is provided, multiplied by with the percentage of the market share shared by sale of electricity company；2) Conditional Lyapunov ExponentP (CVaR) and risk averse The product of factor beta.

Claims (7)

1. it is a kind of count and the market share and value-added service cost of investment purchase sale of electricity optimization method, it is characterised in that the method bag Include following steps：

(1) determine all kinds of factors of influence sale of electricity houses market fixed portion size, the influence market share is calculated using analytic hierarchy process (AHP) All kinds of indexs weight coefficient, obtain the assessment result of user utility；

(2) Logistic regression models are combined, according to the result of calculation of user utility, sale of electricity houses market share model is set up；

(3) purchase of sale of electricity company meter and the market share and value-added service cost of investment is set up with Conditional Lyapunov ExponentP method (CVaR) Sale of electricity model, with sale of electricity company maximization of utility as target, adds the market share size shared by sale of electricity company in the model, And the cost of investment of power marketing value-added service is added in the cost payout of sale of electricity company.

2. the purchase sale of electricity optimization method of meter according to claim 1 and the market share and value-added service cost of investment, it is special Levy and be, in step (1), after influenceing all kinds of indexs of market share size to determine, set up destination layer, the criterion of appraisement system Layer and indicator layer, wherein destination layer are user utility, and rule layer is all kinds of performance indications for influenceing user utility, and indicator layer is phase Answer the specific market share influence factor under rule layer.

3. the purchase sale of electricity optimization method of meter according to claim 1 and the market share and value-added service cost of investment, it is special Levy and be, in step (2), the market share model of sale of electricity company is calculated with equation below：

$P=\frac{e^{{\mathrm{\&beta;}}_0+{\mathrm{\&beta;}}_1{x}_1+...+{\mathrm{\&beta;}}_p{x}_p}}{1+e^{{\mathrm{\&beta;}}_0+{\mathrm{\&beta;}}_1{x}_1+...+{\mathrm{\&beta;}}_p{x}_p}}$

In above formula, P represents market share size, is the real number in 0~1；x₁~x_pThe construction journey of each two-level index is corresponded to respectively Degree；β₁~β_pValue is respectively the corresponding weight coefficient of each two-level index, wherein β₀Value calculated by reasonable assumption.

4. the purchase sale of electricity optimization method of meter according to claim 1 and the market share and value-added service cost of investment, it is special Levy and be, in step (3), set up sale of electricity company meter and the market share purchase sale of electricity model the step of include：1) sale of electricity is set up public Department's purchases strategies model；2) sale of electricity company power selling income model is set up；3) sale of electricity company meter and the market share is set up to be sought with electric power Sell the purchase sale of electricity model of value-added service cost of investment.

5. the purchase sale of electricity optimization method of the meter and the market share according to claim 1 or 4 and value-added service cost of investment, its It is characterised by, in step (3), the sale of electricity company purchases strategies model of structure is：

C=P (C_C+C_DA+C_RT)+γ

Constraints：

C_C=q_C,t·p_C

$C_{DA}=\underset{t=1}{\overset{T}{\&Sigma;}}{p}_{DA,t}{q}_{DA,t}$

$C_{RT}=\underset{t=1}{\overset{T}{\&Sigma;}}{p}_{RT,t}{q}_{RT,t}=\underset{t=1}{\overset{T}{\&Sigma;}}{p}_{RT,t}(q_{all,t}-q_{C,t}-q_{DA,t})$

q_C,t+q_DA,t+q_RT,t=q_all,t

In formula：C is the purchases strategies of sale of electricity company；C_C、C_DAAnd C_RTRespectively sale of electricity company is in bilateral contract market, ahead market With the purchases strategies of Real-time markets；T is research period sum；p_C、p_DA,tAnd p_RT,tRespectively two day market power purchase price and t The purchase electricity price of period ahead market and Real-time markets, wherein p_CIt is bilateral transaction price with q_C,tThe function of change；q_C,t、q_DA,t、 q_RT,tAnd q_all,tRespectively t period bilateral contract market power purchase electricity, Day-ahead Electricity Purchase electricity, Real-time markets power purchase electricity with And total power purchase electricity；P is market share size shared by sale of electricity company；γ is the overall cost of ownership of value-added service.

6. the purchase sale of electricity optimization method of meter according to claim 5 and the market share and value-added service cost of investment, it is special Levy and be, in step (3), the sale of electricity company power selling income model of structure is：

$R=P\underset{i=1}{\overset{h}{\&Sigma;}}{R}_i=P\underset{j=1}{\overset{N_i}{\&Sigma;}}\left\{(1+f_i)\underset{t=1}{\overset{T}{\&Sigma;}}{p}_{i,l}{q}_{j,t}\right\}$

Constraints：

$N=\underset{i=1}{\overset{h}{\&Sigma;}}{N}_i$

I=1,2 ..., h

In formula, h is the species of sale of electricity contract；R_iIt is i-th kind of power selling income of sale of electricity contract；N is total number of users；N_iIt is purchase i-th Plant the number of users of sale of electricity contract；f_iAnd p_iRespectively i-th kind cost function and pricing structure of sale of electricity contract；q_j,tIt is j-th use Load of the family in the t periods.

7. the purchase sale of electricity optimization method of meter according to claim 6 and the market share and value-added service cost of investment, it is special Levy and be, in step (3), the purchase sale of electricity of the sale of electricity company of structure meter and the market share and power marketing value-added service cost of investment Model purchases sale of electricity maximization of utility as target with electric company, and the object function built is：

$\max U=P\underset{j=1}{\overset{N_i}{\&Sigma;}}\left\{(1+f_i)\underset{t=1}{\overset{T}{\&Sigma;}}{p}_{i,l}{q}_{j,t}\right\}-P(C_C+C_{DA}+C_{RT})+\mathrm{\&gamma;}-\mathrm{\&beta;}\&lsqb;{\mathrm{\&xi;}}_1+{\mathrm{\&xi;}}_2+\frac{1}{N_M(1-\mathrm{\&alpha;})}\underset{x=1}{\overset{N_M}{\&Sigma;}}{\mathrm{\&eta;}}_x\&rsqb;$

Wherein, β is the risk averse coefficient of sale of electricity company；N_MIt is Monte Carlo simulation number of times；α is the confidence level of CVaR； η_xIt is the decision variable for introducing；ξ₁It is sale of electricity company in the power purchase allocation proportion of ahead market Yu bilateral contract market；ξ₂It is sale of electricity Power purchase allocation proportion of the company in Real-time markets Yu bilateral contract market；

Constraints：

C_C=q_C,t·p_C

$C_{DA}=\underset{t=1}{\overset{T}{\&Sigma;}}{p}_{DA,t}{q}_{DA,t}$

$C_{RT}=\underset{t=1}{\overset{T}{\&Sigma;}}{p}_{RT,t}{q}_{RT,t}=\underset{t=1}{\overset{T}{\&Sigma;}}{p}_{RT,t}(q_{all,t}-q_{C,t}-q_{DA,t})$

q_C,t+q_DA,t+q_RT,t=q_all,t

$N=\underset{i=1}{\overset{n}{\&Sigma;}}{N}_i$

I=1,2 ..., n

X=1,2 ..., N_M

P '=R-C

η_x≥E(P′)-P_i′-ξ₁-ξ₂

η_x≥0

Wherein, P_i' P ' the values drawn by i ＆ lt Monte Carlo simulations；E (P ') is the expectation that sale of electricity company purchases dynamoelectric benefit Value.