CN106709625A - Electricity market demand response planning evaluation method - Google Patents

Abstract

The invention discloses an electricity market demand response planning evaluation method, which comprises the steps of: calculating user response characteristics in different demand response projects according to an improved user response characteristic model based on a price elasticity matrix; constructing a demand response project evaluation index and establishing a decision-making matrix; and determining evaluation index weight by adopting an entropy weight method, grading each demand response alternative scheme by adopting a gray comprehensive evaluation method, and ranking quality of the demand response schemes according to scores of the schemes. The electricity market demand response planning evaluation method provides technical support for the demand response planning decision-making and project popularization.

Description

A kind of electric power market demand responds planning appraisal method

Technical field

The present invention relates to a kind of electric power market demand response planning appraisal method, category electricity needs planning technology field.

Background technology

With the development of Electricity Market Competition, demand response electricity marketization has been increasingly becoming inexorable trend, but due to lacking The demand response theoretical system of few science and the support of method, comment during Project there are still demand response effect Valency system is not perfect, and effect efficiency assessment missing, the object of planning deviates the problems such as actual conditions.Therefore, development is needed badly Demand response fundamental research and demand response recruitment evaluation correlative study, set up demand response planning appraisal method.

Scholar both domestic and external and mechanism have carried out substantial amounts of research to the performance evaluation of demand response, but research lay particular emphasis on it is right The assessment of demand response implementation result and implementation benefit, preferably studies for demand response scheme and its resource distribution and is related to It is less.Accordingly, it would be desirable to set up set of system, comprehensive demand response implementation result assessment technology system, evaluation requirement response exists Reduce energy resource consumption, reduce load peak-valley difference, improve efficiency in terms of impact effect, excavate demand response scheme in advantage and Weak link, for the planning appraisal of demand response project provides technical support and guiding opinion.

The content of the invention

The purpose of the present invention is to reduce energy resource consumption to improve evaluation requirement response, reduces load peak-valley difference, is improved Influence in terms of efficiency, the present invention proposes a kind of electric power market demand response planning appraisal method.

Realize that technical scheme is as follows：A kind of electric power market demand responds planning appraisal method, and specific steps are such as Under：

(1) identify project the target of planning appraisal；

(2) determine different demands response item user response characteristic now, rung by the user based on price elasticity matrix Improved properties model is answered, the user response characteristic in different demands response item now is calculated, is demand response Indexes of Evaluation Effect And the calculating of economic indicator lays the foundation；

(3) build demand response Output Ratio and set up decision matrix, it is thus necessary to determine that adapt under different market environments Demand response project evaluation index and quantization method, with this build evaluate demand response scheme decision matrix.

(4) determine optimal demand response project, the weight of each index in decision matrix is determined using entropy assessment, according to assessment Application scenarios, are modified by the weight modifying factor for introducing index to index, on this basis, use Grey Comprehensive Evaluation Method is given a mark to each demand response alternative, and the quality of demand response scheme is entered by the height of each scheme fraction Row sequence.

The present invention introduces the participation rate of different type user in existing user response characteristic model, with more accurately anti- Reflect implement demand response before and after user power utilization situation of change.

The present invention obtains final user response characteristic model：

In formula：η_ωRepresent the percentage of all types of user；q₀I () and q (i) are to implement demand response in i-th hour Front and rear workload demand, p₀J () and p (j) are the electricity price after the initial electricity price and demand response of jth hour；A (j) is represented in jth Unit excitation in hour；p_cJ () represents the unit punishment in jth hour.

A (j) and p_cJ () two amount is on the occasion of being zero in other periods in the demand response effect period.

In demand response project, the participation of user is different under different project rule, project mechanism；As i=j, ε_ω(i, j) is ω class user's self-elasticity coefficients；As i ≠ j, ε_ω(i, j) is ω class user's coefficients of cross elasticity.

Electric power market demand response evaluation index of the present invention can be divided into Indexes of Evaluation Effect and economic evaluation index,

Indexes of Evaluation Effect includes peak load reduction e₁, peak-valley difference decrement e₂, energy curtlaiment amount e₃, rate of load condensate incrementss e₄, wholesale electricity price reduction amount e₅With wholesale Electricity price fluctuation rate variable quantity e₆；Economic evaluation index is divided according to benefited main body, can be divided It is user's economic benefit b₁, electrical network economy benefit b₂With environmental benefit b₃。

Electric power market demand response of the present invention based on Grey Comprehensive Evaluation Method is evaluated and includes that entropy assessment determines evaluation index Weight and Grey Comprehensive Evaluation Method evaluate demand response.

The method that evaluation criterion weight is determined using Information Entropy：

(1) assessment indicator system is determined according to market environment and demand response project planning target, determines evaluation index Data.

If n achievement data sequence forms following matrix：

In formula：M is demand response scheme number, and n is the number of evaluation index.

(2) linear function method for normalizing is processed index.

(3) contribution degree of Calculation Estimation index

U-th demand response scheme is calculated to the v contribution degree P of index_uv：

In formula：P_uvRepresent u-th scheme to the v contribution degree of evaluation index；x'_uvIt is u-th demand response scheme institute Corresponding v-th index is through the dimensionless number after treatment.

(4) v evaluation index entropy is determined

The computing formula of v evaluation index entropy is shown below：

In formula：K=1/ln (m)>0,0≤e_v≤ 1, m are the quantity of scheme.When each demand response scheme refers to a certain evaluation When target contribution degree reaches unanimity, e_vTend to 1.

(5) Calculation Estimation index x'_vValue of utility (i.e. difference property coefficient)

The size of evaluation index value coefficient is determined by all schema differences sizes, therefore, can introduce that value of utility is defined (i.e. Difference property coefficient) d_v, d_v=1-e_v, d_vRepresent the degree of consistency of each demand response scheme contribution degree under v-th evaluation index. For evaluation index x'_v, x'_uvDifference is smaller, e_vIt is bigger, d_vIt is smaller, x'_uvDifference is bigger, e_vIt is smaller, d_vBigger, index is for being The comparing of system is used as greatly.

(6) entropy assessment determines the weight of each evaluation index

As the following formula, the weight of each evaluation index is calculated：

In formula：W_vIt is the weight for having normalized, especially, works as d_vWhen=0, v-th effect of index can be rejected, its power 0 is equal to again.

(7) amendment of index weights

Demand response policymaker rule of thumb can introduce subjective estimation weight modifying factor λ_v, then can be repaiied by weight Positive divisor λ_vTo W_vIt is modified.

In formula：W_v ⁰To add the revised normalization evaluation criterion weight of subjective experience.

Grey Comprehensive Evaluation Method：

The step of carrying out overall merit using grey correlation analysis be：

(1) reference data sequence is determined

Reference data sequence is a preferable standard of comparison.In the present invention, the x ' of the optimum value of each index is chosen_0v As reference data sequence X₀′：

X₀'=(x '₀₁,x′₀₂,…,x′_0v,…,x′_0n)；

In formula：x′_0v=Optimum (x '_uv), u=1,2 ..., m；V=1,2 ..., n.

(2) calculate correlation coefficient

By following formula, the incidence coefficient size of each comparative sequences and reference data sequence corresponding element is calculated.

In formula：U=1,2 ..., m；V=1,2 ..., n；ρ is resolution ratio, ρ ∈ [0,1], and ρ is smaller, poor between incidence coefficient Different bigger, separating capacity is stronger, and ρ values are usually 0.5.

(3) calculate correlation coefficient matrix

By above formula calculate correlation coefficient ζ_uv, and then obtain corresponding incidence coefficient matrix：

In formula：ζ_uvIt is u-th v of scheme index and the v incidence coefficient of optimal parameter in all schemes.

(4) calculating correlation

The significance level acted on played in overall merit in view of each index is different, and reply incidence coefficient is multiplied by weight, It is by the upper priority weight for saving each evaluation index being calculated：

R=(r_u)_1×m=(r₁,r₂,…,r_m)=W⁰E^T；

In formula：W⁰To introduce each index weights of modifying factor.

(5) evaluation result is drawn according to inteerelated order.

According to the degree of association r being calculated_u(u=1,2 ..., size order m) is arranged the priority of scheme Sequence, the size order of the degree of association is the priority of each demand response project alternative.

The beneficial effects of the invention are as follows the present invention proposes the electricity needs response project rule based on Grey Comprehensive Evaluation Method Draw appraisal procedure, establish the user response improved properties model based on price elasticity matrix, construct meter and effectiveness indicator with And the demand response planning appraisal index of economic indicator, using the assessment technology of entropy assessment and Grey Comprehensive Evaluation Method in multiple The demand response project most matched with planning appraisal target is selected in demand response scheme.It is different to many scenes by typical case The effect efficiency of the demand response project under the object of planning carries out comprehensive assessment, and integration project rule and planning appraisal mesh Mark, deep interpretation has been carried out to assessment result, it was demonstrated that the validity and applicability of assessment models.The appraisal procedure is demand Response programmed decision-making and the popularization of project provide technical support.

Brief description of the drawings

Fig. 1 is demand response project planning estimation flow figure；

Fig. 2 is demand response Output Ratio system；

Fig. 3 is the demand response load curve based on electricity price；

Fig. 4 is the demand response load curve based on excitation；

Fig. 5 is mixed type demand response load curve.

Specific embodiment

The concrete application of demand response project planning appraisal procedure of the present invention is illustrated by the present embodiment.According to being System running status, the several scenes of the present embodiment initialization system planning appraisal use the user response based on price elasticity matrix Improved properties model calculates the response characteristic of user under different demands response scheme, on this basis, under different application scene The response effect of each scheme carries out comprehensive assessment, determines quality of each response scheme under different application scene.

Table 1 is certain electricity market typical case's day power load, according to load curve and different types of demand response project Feature, devises several demand response scheme, and project alternative rule is as shown in table 2.

Scene settings

Scene one：System lacks capacity not short of electricity amount, and in the case where power system capacity vacancy is larger, ISO assume responsibility for maintaining electricity The most basic responsibility of software safety, the reduction of peak load can increase the spare capacity of system, greatly increase the security of system.

Scene two：System not only lacks capacity but also short of electricity amount, to take into account the reduction of peak load simultaneously in demand response planning and designing And the reduction of electricity.

Scene three：Under the scene that wholesale electricity price is higher, Electricity price fluctuation is larger, the main target of demand response planning and designing is Wholesale market electricity price is reduced, wholesale market Electricity price fluctuation rate is reduced.

Interpretation of result：

(1) user response specificity analysis

Using the improved model of the user response characteristic based on price elasticity matrix, region load curve before and after response is calculated Change, result of calculation is as shown in Fig. 3, Fig. 4, Fig. 5.

From Fig. 3, Fig. 4, Fig. 5, under rational project rule, the demand response based on electricity price, the need based on excitation Ask response and mixed type demand response that certain peak load shifting ability, but different demand responses are respectively provided with to region load Purpose features of response is different, such as from the figure 3, it may be seen that because CPP peak electricity tariff advance notification times are short, electricity price between peak and valley is big, Peak clipping effect is more obvious, and the influence to Pinggu period load is smaller, and TOU tou power prices are long due to advance notification times, phase Become apparent compared with peak CPP electricity price Fill valley effects.

(2) index result

By the result of user response property calculation, with reference to quantification of targets computational methods, demand under many scenes is calculated Response item indices such as peak load reduction, peak-valley difference reduction amount now, result of calculation is as shown in table 3 below.

Certain electricity market typical case's day power load of table 1

Hour	1	2	3	4	5	6	7	8
									Load (GW)	47.536	48.171	49.109	50.665	58.762	65.014	80.493	90.444
Hour	9	10	11	12	13	14	15	16
									Load (GW)	85.268	70.615	68.145	65.932	65.746	65.168	65.382	67.597
Hour	17	18	19	20	21	22	23	24
									Load (GW)	69.875	76.406	91.43	95.043	90.695	70.071	60.341	50.271

The demand response project of table 2 rule

The demand response scheme index Comparative result of table 3

TOU：Tou power price；CPP：Peak electricity tariff；RTP：Spot Price；EDRP：Urgent need is responded；CAP：Capacity city Field/assistant service plan；IL：Interruptible load.

Claims (3)

1. a kind of target of electric power market demand response planning appraisal method, including planning appraisal of identifying project, it is characterised in that Methods described is calculated in different demands response item now by the improved user response characteristic model based on price elasticity matrix User response characteristic；Build demand response Output Ratio and set up decision matrix；Determine evaluation index using entropy assessment Weight, is given a mark, by the height of each scheme fraction using Grey Comprehensive Evaluation Method to each demand response alternative Quality to demand response scheme is ranked up；

The user response characteristic model is：

$q\left(i\right)=q_0\left(i\right)\{1+\underset{\mathrm{\ω}=1}{\overset{4}{\mathrm{\Σ}}}{\mathrm{\η}}_{\mathrm{\ω}}\underset{j=1}{\overset{24}{\mathrm{\Σ}}}{\mathrm{\ϵ}}_{\mathrm{\ω}}(i,j)\frac{\[p\left(j\right)-p_0\left(j\right)+A\left(j\right)+p_c\left(j\right)\]}{p_0\left(j\right)}\},i=1,2,\mathrm{..},24$

In formula：η_ωRepresent the percentage of all types of user；q₀I () and q (i) are before and after implementing within i-th hour demand response Workload demand, p₀J () and p (j) are the electricity price after the initial electricity price and demand response of jth hour；A (j) is represented in jth hour Unit excitation；p_cJ () represents the unit punishment in jth hour.

2. a kind of electric power market demand responds planning appraisal method according to claim 1, it is characterised in that the use entropy Value method determines that the method for evaluation criterion weight is as follows：

(1) assessment indicator system is determined according to market environment and demand response project planning target, determines evaluation index data；

If n achievement data sequence forms following matrix：

$A=\left(\begin{array}{cccc}{x}_{11}& x_{12}& ...& x_{1n}\\ x_{21}& x_{22}& ...& x_{2n}\\ .& .& .& .\\ .& .& .& .\\ .& .& .& .\\ x_{m1}& x_{m2}& ...& x_{mn}\end{array}\right)---\left(6\right)$

In formula：M is demand response scheme number, and n is the number of evaluation index；

(2) linear function method for normalizing is processed index；

(3) contribution degree of Calculation Estimation index

U-th demand response scheme is calculated to the v contribution degree P of index_uv：

$P_{uv}=\frac{x_{uv}^{\′}}{\underset{u=1}{\overset{m}{\mathrm{\Σ}}}{x}_{uv}^{\′}}$

In formula：P_uvRepresent u-th scheme to the v contribution degree of evaluation index；x'_uvCorresponding to u-th demand response scheme V-th index through the dimensionless number after treatment；

(4) v evaluation index entropy is determined

The computing formula of v evaluation index entropy is shown below：

$e_v=-K\underset{u=1}{\overset{m}{\mathrm{\Σ}}}{P}_{uv}ln\left(P_{uv}\right)$

In formula：K=1/ln (m)>0,0≤e_v≤ 1, m are the quantity of scheme；When each demand response scheme is to a certain evaluation index When contribution degree reaches unanimity, e_vTend to 1；

(5) Calculation Estimation index x'_vValue of utility

The size of evaluation index value coefficient is determined by all schema differences sizes, therefore, can introduce that value of utility d is defined_v, d_v= 1-e_v, d_vRepresent the degree of consistency of each demand response scheme contribution degree under v-th evaluation index；For evaluation index x'_v, x'_uv Difference is smaller, e_vIt is bigger, d_vIt is smaller, x'_uvDifference is bigger, e_vIt is smaller, d_vBigger, index is used as greatly for the comparing of system；

(6) entropy assessment determines the weight of each evaluation index

As the following formula, the weight of each evaluation index is calculated：

$W_v=\frac{d_v}{\underset{v=1}{\overset{n}{\mathrm{\Σ}}}{d}_v}$

In formula：W_vIt is the weight for having normalized, especially, works as d_vWhen=0, v-th effect of index can be rejected, its weight etc. In 0；

(7) amendment of index weights

Demand response policymaker rule of thumb can introduce subjective estimation weight modifying factor λ_v, then can be by weight modifying factor λ_vTo W_vIt is modified；

$W_v^0=\frac{{\mathrm{\λ}}_v{W}_v}{\underset{v=1}{\overset{n}{\mathrm{\Σ}}}{\mathrm{\λ}}_v{W}_v}$

In formula：W_v ⁰To add the revised normalization evaluation criterion weight of subjective experience.

3. a kind of electric power market demand responds planning appraisal method according to claim 1, it is characterised in that the grey is closed Join analysis is the step of carrying out overall merit：

(1) reference data sequence is determined

Choose the x ' of the optimum value of each index_0vAs reference data sequence X '₀：

X′₀=(x '₀₁,x′₀₂,…,x′_0v,…,x′_0n)

In formula：x′_0v=Optimum (x '_uv), u=1,2 ..., m；V=1,2 ..., n；

(2) calculate correlation coefficient

By following formula, the incidence coefficient size of each comparative sequences and reference data sequence corresponding element is calculated：

${\mathrm{\ζ}}_{uv}=\frac{\underset{u}{min}\underset{v}{min}\left|x_{0v}-x_{uv}^{\′}\right|+\mathrm{\ρ}\underset{u}{max}\underset{v}{max}\left|x_{0v}-x_{uv}^{\′}\right|}{\left|x_{0v}-x_{uv}^{\′}\right|+\mathrm{\ρ}\underset{u}{max}\underset{v}{max}\left|x_{0v}-x_{uv}^{\′}\right|}$

In formula：U=1,2 ..., m；V=1,2 ..., n；ρ is resolution ratio, ρ ∈ [0,1], and ρ is smaller, and difference is got between incidence coefficient Greatly, separating capacity is stronger, and ρ values are usually 0.5；

(3) calculate correlation coefficient matrix

By above formula calculate correlation coefficient ζ_uv, and then obtain corresponding incidence coefficient matrix：

$E={\left({\mathrm{\ζ}}_{uv}\right)}_{m\×n}=\left(\begin{array}{cccc}{\mathrm{\ζ}}_{11}& {\mathrm{\ζ}}_{12}& ...& {\mathrm{\ζ}}_{1n}\\ {\mathrm{\ζ}}_{21}& {\mathrm{\ζ}}_{22}& ...& {\mathrm{\ζ}}_{2n}\\ .& .& .& .\\ .& .& .& .\\ .& .& .& .\\ {\mathrm{\ζ}}_{m1}& {\mathrm{\ζ}}_{m2}& ...& {\mathrm{\ζ}}_{mn}\end{array}\right)$

In formula：ζ_uvIt is u-th v of scheme index and the v incidence coefficient of optimal parameter in all schemes；

(4) calculating correlation

The significance level acted on played in overall merit in view of each index is different, and reply incidence coefficient is multiplied by weight, by upper The priority weight of each evaluation index that section is calculated is：

$W^0=(W_1^0,W_2^0,\mathrm{...},W_n^0)$

R=(r_u)_1×m=(r₁,r₂,…,r_m)=W⁰E^T

In formula：W⁰To introduce each index weights of modifying factor；

(5) evaluation result is drawn according to inteerelated order：

According to the degree of association r being calculated_u(u=1,2 ..., size order m) is ranked up to the priority of scheme, association The size order of degree is the priority of each demand response project alternative.