CN111460378A - Power distribution network accurate investment project optimization method considering risk measure - Google Patents

Abstract

The invention relates to a risk measure considered power distribution network accurate investment project optimization method, and provides a risk measure considered power distribution network accurate investment combination optimization model for realizing power distribution network project optimization with controllable risk and optimal input and output benefits. The model takes projects as decision objects, can take account of the confidence coefficient of achievement of benefits of a single project, well represents the cost performance attributes of planned projects in the investment decisions of the power distribution network by constructing actual constraints such as production performance, investment limitation, project association and the like on the basis of introducing the development diagnosis of the power distribution network and the performance evaluation index of the projects, and finally calculates the optimization strategy of the project combination of the power distribution network under certain investment risk preference based on the fuzzy satisfaction degree of the profitability and the unit output benefits.

Description

Power distribution network accurate investment project optimization method considering risk measure

Technical Field

The invention belongs to the technical field of investment construction of power distribution networks, and particularly relates to a power distribution network accurate investment project optimization method considering risk measure.

Background

The power distribution network construction transformation action plan (2015-2020) proposes that the power distribution network construction transformation investment is not less than 2 trillion yuan in recent years, along with the development of social requirements, the power distribution network planning development faces a series of new challenges that ① distributed energy, electric vehicles and other controllable equipment are connected to lead to diversification of power distribution network operation scenes and increase of uncertainty and complexity in planning, ② diversification of power distribution network investment targets are not limited to only meeting load growth, improvement of power utilization reliability, enhancement of flexible operation capacity, full consumption of clean energy and the like also become important targets, ③ social investment participates in incremental power distribution network investment, construction and operation, power grid enterprises are no longer single main bodies of power distribution network investment and compete with social capital on investment benefit and operation efficiency, scientificity and rationality of investment play a key role in forming of effective assets of power grids, meanwhile, the future development target of power grid enterprises must be improved, accurate investment and decision making are realized, the current ④ investment is still improved, the expansion of power grid enterprises is needed to be improved, the new industrial investment needs to be developed, the theoretical investment is needed to be improved, the new investment and the new investment is difficult to be developed, the high-investment risk of the power distribution network construction and the high-oriented industrial investment management and control project is difficult to be developed.

The invention provides a power distribution network accurate investment project optimization method considering risk measure to overcome the defects.

Disclosure of Invention

The invention aims to provide a method for optimizing an accurate investment project of a power distribution network in consideration of risk measure, which is used for solving one of the technical problems in the prior art, such as: at present, a power distribution network still needs to be transformed, expanded and newly built in a large quantity to adapt to the new development situation of the power industry and meet the increasing high-quality energy utilization requirement. Different from planning and designing one or more feeders, the investment of the power distribution network is usually in a medium-long period, a systematic engineering project group is taken as an object, the problems of large scale of development and construction, numerous related factors, difficulty in quantification of monomer benefits, large structural difference of project description and the like exist, a method for guiding accurate investment and project optimization of the power distribution network is lacked in practice, and the investment risk of power grid management and control is difficult to assist. Therefore, an accurate investment analysis theory suitable for the development characteristics of the power distribution network needs to be deeply researched, project optimization basis is quantized, an investment combination system considering risk management and control is introduced, and the development investment decision benefits of the power distribution network are improved.

In order to achieve the purpose, the technical scheme of the invention is as follows:

a power distribution network accurate investment project optimization method considering risk measure is characterized by comprising the following steps:

s1, collecting historical data of the regional distribution network needing to arrange investment plans, comprising the following steps: five categories of power distribution network equipment operation information, the current investment budget amount, declaration project collection, power load increase information and urgent problem solving are provided, and the declaration projects of the power distribution network are numbered;

s2, combining the characteristics of the urban medium-high voltage distribution network, constructing a distribution network project investment benefit evaluation double-layer index system, wherein the double-layer index system comprises microscopic indexes and macroscopic indexes, the index system covers five categories of net rack robustness, power supply safety and quality, operation economy and efficiency, power supply harmony, interactivity and social friendliness, and is further subdivided into various sub indexes;

s3, developing and diagnosing the power distribution network in the area where the investment plan is arranged by using the macroscopic indexes, determining the current investment thought and the corresponding investment constraint conditions, and determining the investment concern;

s4, quantifying investment efficiency of a single investment project unit by using microscopic indexes, quantifying benefit output dimensions which cannot be directly quantified by money amount by using scores, such as net rack strength degree improvement, heavy overload equipment reduction and the like, and scoring each alternative project according to index weight and scoring standards so as to assist accurate investment;

s5, based on a conditional risk measurement method, the distribution characteristics of the profitability of a single project are taken into the investment risk, the profitability and the risk evaluation are carried out on the single project, and an investment risk minimum objective function containing a risk preference confidence coefficient is constructed in the dimension;

s6, taking the total investment, the power grid development requirement, the investment portfolio profitability and the coupling relation among projects as constraints, establishing a power distribution network investment portfolio optimization model taking efficiency and risk as targets, and considering both investment cost and investment risk under the boundary condition of investment income and income; let x^T＝(x₁,x₂,x₃,…,x_n)^TAs investment decision vector, x_i1 represents that the ith project is determined to be put into operation, if the ith project is 0, the ith project does not enter the investment plan, and n is the total number of the projects; let V be (V)₁,v₂,v₃,…,v_n) As a project profitability vector, v_iExpressing the yield of the ith project, conforming to normal distribution, wherein the expectation is the average yield, and the variance is the estimate of the risk degree of the project;

s7, converting the fuzzy satisfaction into a single-target mixed integer linear programming problem easy to solve by using the fuzzy satisfaction so as to calculate a power distribution network accurate investment strategy;

and S8, judging whether the investment target and the feasibility are met, if not, returning to S3, otherwise, ending the method.

Preferably, in step S1, the regional distribution network historical data includes: five categories of power distribution network equipment operation information, the current investment budget amount, declaration project collection, power load increase information and urgent problem solving are provided, and the declaration projects of the power distribution network are numbered;

the operation information of the power distribution network equipment comprises daily, monthly and annual load rate statistics, equipment operation age, line loss and transformer loss, power supply reliability, voltage qualification rate, capacity-to-load ratio, N-1 passage rate, line connection rate, line section qualification rate, line power supply radius qualification rate, line insulation rate, distribution automation coverage rate, high-loss distribution-to-occupation ratio, line overload rate, distribution-to-transformation overload rate, comprehensive line loss rate, distributed energy grid-connection rate, dynamic electricity price and electricity consumption proportion and electric quantity proportion for electric vehicles;

the budget amount of the investment comprises the budget amount of the investment;

the declaration project collection comprises investment construction amount, newly-added capacity, newly-added load, newly-added power supply load, estimated power supply reliability (RS-3) improvement degree, estimated voltage qualification rate improvement degree, N-1 line lifting number, distribution line connecting line lifting number, segmented reasonable line lifting number, power supply radius qualified line lifting number, distribution line insulation, lifting kilometers number, distribution automation station number, high-loss distribution transformer descending station number, line heavy-load descending number, distribution transformer heavy-load descending number, estimated line loss rate improvement degree, newly-accessed distributed power supply capacity, dynamic electricity price increase capacity and newly-added charging pile number;

the electricity load increase information comprises annual maximum load increase rate and local shortage load;

the problems to be solved urgently include key special construction plans, outstanding power supply contradictions and the like.

Preferably, in step S2, a power distribution network project investment benefit evaluation double-layer index system is constructed, which includes a microscopic index and a macroscopic index, and specifically includes:

macroscopic indexes are as follows: the method comprises the following steps of (N-1) passing rate (unit%), line contact rate (unit%), line subsection qualification rate (unit%), line power supply radius qualification rate (unit%), power supply reliability rate (RS-3) (unit%), voltage qualification rate (unit%), line insulation rate (unit%), distribution automation coverage rate (unit%), capacity-load ratio, high-loss distribution transformation ratio (unit%), line overload rate (unit%), distribution transformation overload rate (unit%), comprehensive line loss rate (unit%), distributed energy grid-connected rate (unit%), dynamic electricity price and electricity consumption proportion (unit%) and electric automobile electricity consumption proportion (unit%);

microscopic indexes are as follows: the method comprises the steps of increasing the number of N-1 lines (unit: one/ten thousand), increasing the number of interconnection lines of distribution lines (unit: one/ten thousand), increasing the number of sectionally reasonable lines (unit: one/ten thousand) by unit investment, increasing the number of qualified lines (unit: one/ten thousand) by unit investment power supply radius, predicting and improving the local comprehensive voltage qualification rate (unit:%), increasing the insulation degree of the distribution lines (unit: km/ten thousand), predicting and improving the local power supply reliability rate (RS-3) (unit:%), automatically increasing the number of distribution lines (unit: one/ten thousand), decreasing the number of distribution lines (unit: one/ten thousand) by unit investment and heavily-loaded decreasing number of the distribution lines (unit: km/ten thousand) by unit investment, The unit investment distribution transformer heavy load reduction number (unit: station/ten thousand yuan), the local line loss rate predicted improvement degree (unit: percent), the unit investment power increase amount (unit: kWh/yuan), the unit investment power increase load (unit: kW/yuan), the unit investment newly-accessed distributed power capacity (unit: kWh/yuan), the unit investment dynamic power price increase amount (unit: kW/yuan), and the unit investment newly-added charging pile number (unit: pieces/ten thousand yuan).

Preferably, in the step S3, the macro step refers to developing a power distribution network development diagnosis for an area where an investment plan is arranged, determining the current investment idea and part of investment constraint conditions, and determining an investment concern, specifically:

the "N-1" pass rate (unit%) is × 100% of the number of distribution lines/total number of distribution lines satisfying the "N-1" safety criterion;

the line connection rate (unit%) is × 100% which satisfies the number of the distribution lines of the main line interconnection/the total number of the distribution lines;

the line segmentation qualified rate (unit%) is equal to the reasonable segmentation distribution line number/total distribution line number × 100%;

the qualification rate (unit%) of the power supply radius of the line is × 100% (the area of A +, A, B is not more than 3 km, the area of C is not more than 5 km, and the area of D is not more than 15 km) of the number of lines/total number of distribution lines with qualified power supply radius;

power supply reliability (RS-3) (unit%) [1- (average user outage time-average user outage time/statistical period time ] × 100%;

the voltage qualification rate (unit%) is × 100% of the sum of the voltage qualification rates of all the power grid monitoring points/the number of the power grid monitoring points;

the insulation rate (unit%) of the line is × 100% of the length of the insulated line/the total length of the line;

distribution automation coverage (unit%) -total number of automation transformers/distribution transformers × 100;

capacity-to-load ratio is × 100% as distribution capacity/maximum load;

the high-loss distribution ratio (unit%) is × 100% of the number of high-loss distribution/total distribution;

the line heavy load rate (unit%) is × 100% of the number of distribution lines/total number of distribution lines with the annual maximum load rate of more than 70%;

the distribution transformer overloading rate (unit%) is × 100% of distribution transformer station number/total distribution transformer station number, the urban annual maximum overloading rate is greater than 80%;

the integrated line loss rate (unit%) — line loss power/head end transmission power × 100%;

the grid connection rate (unit%) of the distributed energy is distributed energy internet access capacity/total distributed energy capacity;

the proportion (unit%) of the electricity consumption of the dynamic electricity price is the dynamic electricity price load/the total load of the power distribution users;

the electric automobile power consumption proportion (unit percent) is the electric automobile power consumption proportion/distribution network ground power consumption;

through the indexes, the overall operation condition of the power distribution game is known, five power distribution network project emphasis points which meet newly added loads, strengthen the grid structure, eliminate potential safety hazards, match the transformer substation and send out projects and solve heavy overload equipment are determined, and the capacity-to-load ratio, the line heavy load rate, the distribution transformer heavy load rate and the N-1 passing rate target value are used as boundary conditions for model optimization in the step S6.

Preferably, in step S4, scoring each candidate item according to the index weight and the scoring criteria to assist accurate investment is performed by using a benefit output dimension that cannot be directly quantified by money amount quantified by scoring quantification, such as net rack robustness improvement, heavy overload equipment reduction, and the like. The method specifically comprises the following steps:

the unit investment of N-1 line lifting number (unit: number of lines/ten thousand yuan) is equal to the number of lines/total investment of the project which satisfies the increase of the number of N-1 distribution lines after the project is implemented;

the number of the contact lines of the unit investment distribution line is increased (unit: one/ten thousand yuan), namely the number of the single radiation line is reduced or the number of the contact lines is increased/total investment of the project after the project is implemented;

the reasonable line lifting number (unit: one/ten thousand yuan) of the unit investment segmentation is equal to the lifting number of the reasonable line/the total investment of the project after the project is implemented;

the number of the lifting lines (unit: one/ten thousand yuan) of the power supply radius qualified lines in unit investment is equal to the number of the power supply radius qualified lines/total investment of the project after the project is implemented;

the predicted improvement degree (unit:%) of the local integrated voltage yield (voltage yield of the line (or the block line) after the implementation of the item-voltage yield of the line (or the block line) before the implementation of the item);

the insulated lifting kilometer number (unit: km/ten thousand yuan) of the distribution line is equal to the kilometer number/total investment of the project increased by the insulated line after the project is implemented;

the estimated improvement degree of local power supply reliability (RS-3) (unit:%) (power supply reliability of the line (or the block line) after the item is implemented-power supply reliability of the line (or the block line) before the item is implemented);

the number of automatic units (unit: unit/ten thousand yuan) of unit investment power distribution is the number of automatic transformers which are changed into the number of ascending units/total investment of the project after the project is implemented;

the number of the descending units of the unit investment high loss distribution transformer (unit: unit/ten thousand yuan) is equal to the number of the descending units of the high loss distribution transformer/total investment of the project after the project is implemented;

the heavy load of the line of unit investment is decreased by the number (unit: km/ten thousand yuan), namely the number of the distribution lines reduced by the distribution line with the load rate of more than 70% after the project is implemented/the total investment of the project;

the unit investment distribution transformer heavy load descending number (unit: station/ten thousand yuan) is the number of stations/total project investment of distribution transformer reduction with the load rate of more than 80% after the project is implemented;

the predicted improvement degree (unit:%) of the local line loss rate is (theoretical loss of the station distribution transformer (or the block distribution transformer) before the implementation of the item-theoretical loss of the station distribution transformer (or the block distribution transformer) after the implementation of the item;

the unit investment increased power supply amount (unit: kWh/yuan) is the theoretically increased power supply amount/total investment of the distribution transformer related to the project;

the unit investment increase load (unit: kW/yuan) is the theoretically increased load/total investment of the distribution transformer related to the project;

unit investment newly-accessed distributed power capacity (unit: kWh/yuan) is newly-accessed distributed power capacity/total project investment;

the dynamic electricity price increase capacity (unit: kW/yuan) of unit investment is equal to the dynamic electricity price increase capacity/total investment of project;

the unit investment of the newly increased charging pile number (unit: one/ten thousand yuan) is equal to the number of the newly increased charging piles/total project investment;

aiming at different evaluation targets, different relative importance degrees can be reflected in a weighting mode; considering that the investment evaluation dimensionality of the project is multiple, and finally determining the weight through an expert; the net rack strength degree, the power supply safety and quality, the operation economy and efficiency, the power supply coordination, the interactivity and the social friendliness are respectively entitled to 0.2, 0.25, 0.2 and 0.1, and can be flexibly adjusted according to the experience or the work key point requirement;

the single project investment performance index reflects unit investment income and is a benefit index; at present, no mature application experience exists, and no regulation guide rule indicates the range; because the indexes have different dimensions and different attributes, the data needs to be initialized; according to the principle of preferred investment, the larger the unit investment output value is, the more ideal the unit investment output value is; the maximum value is set to 100 points, the minimum value is set to 0 points, and each index score is calculated as follows:

in the formula: r is_iScoring the jth index of the ith item, wherein y_i,jIs the jth of the ith itemIndex data, y_{j max}Is the jth finger maximum, y_{j min}Is the jth index minimum;

converting each index into a score in a non-dimensionalization way, and combining the index weight, wherein the investment benefit evaluation model of a single project is as follows:

in the formula: r is the comprehensive evaluation value of a certain investment project, y_jIs the score of the jth index, w_jThe weight of the jth index is n indexes; considering the index system of the hierarchical structure, the scores of each target layer can be calculated successively:

in the formula: the superscript t +1 denotes the t +1 th layer index.

Preferably, in step S5, specifically, the method includes:

value at Risk (VaR) refers to the worst expected loss Value for a given confidence level for a given portfolio over a future holding period; CVaR refers to the condition mean value of investment loss exceeding the VaR value, represents the average level of excess loss, and can reflect the potential risk value of the tail part better than the VaR;

let X be the feasible set of investment portfolio,

let f (X, y) be a loss function, wherein X ∈ X is an n-dimensional power distribution network project investment portfolio scheme vector, n is the number of projects, y ∈ R^mSetting the joint probability density function of y as p (y), and for determined x, the loss f (x, y) caused by y is a random variable obeying a certain distribution on R, and the distribution function of the random variable not exceeding the confidence α is as follows:

for any fixed x, ψ (x, a) as a function of α is the cumulative probability distribution function of loss under portfolio x, with β representing confidence, α_β(x) And the VaR value corresponding to the loss f (x, y) when the investment combination is x is expressed, and the calculation formula is as follows:

α_β(x)＝min{α∈R；ψ(x，α)≥β}

calculating the maximum possible loss at a given probability level of β∈ (0,1), and further, calculating by definition the asset loss f (x, y) to be no less than α_β(x) CVaR value of (a):

the VaR function α is contained in the above formula_β(x) Term, and α_β(x) The analytical expression of (A) is difficult to solve, and usually an analyzable function F is introduced_β(x, α) instead of phi_β(x) Calculating CVaR:

wherein [ f (x, y) - α]⁺Represents max {0, f (x, y) - α }, α is VaR value, β is confidence coefficient;

the analytical expression of the probability density function p (y) is difficult to obtain, and the Monte Carlo method is often used for extracting sample data to give integral estimation in the formula above; is provided with y¹,y²,…,y^qQ samples of y, the function F_βThe calculated value of (x, α) is:

in actual calculations, the optimal combination coefficient vector X of the asset and the corresponding CVaR and VaR values are usually determined based on the above formula, and according to the CVaR method, the minimum α -CVaR value is found in the feasible region X, which can be defined as the following optimization model:

min CVaR_α(x^Tw)

in the formula: w is an investment income vector; when discrete samples are used to estimate the asset weights x and CVaR in the model combination, the optimization model can be converted into:

in the formula z_k＝[f(x,y^k)-a)]⁺Converting the original equation into an analytically solvable linear programming model by adding the linear constraints shown in the following equation:

the income of the power distribution network project is generally normally distributed, and the expected income rate can be regarded as the profit level of the project and can represent sigma²Can represent an estimate of the degree of risk, σ²A larger value means more extreme cases and higher investment risk. From the investment risk perspective, each project has a certain decision risk of uncertainty in revenue. For example, the price of electricity for sale is adjusted, and the amount of electricity used is influenced by industrial structure and economic development, etc., which may lead to uncertainty in price difference, amount of electricity sold, standard discount rate and unit power supply cost, and finally lead to difference in profitability and fluctuation range of each project, and even loss of part of the projects, resulting in loss of operation of the power grid company. This distribution is given by an expert integrating the above factors.

Preferably, in step S6, specifically, the method includes:

let x^T＝(x₁,x₂,x₃,…,x_n)^TAs investment decision vector, x_i1 represents that the ith project is determined to be put into operation, if the ith project is 0, the ith project does not enter the investment plan, and n is the total number of the projects; let V be (V)₁,v₂,v₃,…,v_n) As a project profitability vector, v_iExpressing the yield of the ith project, conforming to normal distribution, wherein the expectation is the average yield, and the variance is the estimate of the risk degree of the project; defining a combination of itemsThe return on investment function is R (x, y) ═ x^Tw, wherein w ═ v₁U₁,v₂U₂,v₃U₃,…,v_nU_n]I th project expected profit w_i＝v₁U_i，U_iAn amount of investment for the ith project; defining the portfolio investment loss function as f (x, w) — R (x, w), R (x, w) being the portfolio investment yield, the loss function is given by:

1) investment risk minimization objective function

Under the influence of a plurality of uncertain factors, the profitability and the fluctuation range of each project are different, loss is easy to occur in part of projects, the operation loss of a power grid company is caused, the investment risk is reduced as much as possible, and the method specifically comprises the following steps:

wherein f (X, w) is formula (12), (X, α) ∈ (X, R); α is the VaR value;

2) unit investment output efficiency maximization objective function

The investment performance indexes of the single lower-layer project need to calculate unit investment output, and the higher the project score is, the higher the unit output is represented; in order to avoid excessive investment, the average score of the project should be improved as follows:

in the formula: r is the item score vector, R ═ R₁,r₂,r₃,…,r_n)^TWherein r is_iScoring the ith item;

the investment of the power distribution network should preferably consider to meet the power demand constraint and the power utilization reliability constraint, and in addition, the constraints of equipment bearing capacity, total investment, expected income and the like are also considered; meanwhile, coupling constraints such as sequence and the like are established among a plurality of power distribution network projects, and if a certain access supporting project needs to be subjected to capacity expansion at a certain interval (the two projects may not be listed in the same project), the relevance of the two projects also needs to be considered;

power demand constraints:

load prediction is a precondition for power distribution network development, and influence factors such as economic development, electric quantity level and the like of each region need to be integrated; the investment construction of the power distribution network should meet the power demand of the region, and the constraint conditions are as follows:

in the formula: d represents the newly increased power demand; c. C_nRepresenting the newly added power supply capacity of the nth item;

and (3) power utilization reliability constraint:

the passing rate of the N-1 of the power distribution network is directly related to the average annual power failure time of a user and the power supply reliability (RS-3); the power utilization reliability can be remarkably increased when the N-1 passing rate is increased, and the constraint expression is as follows:

wherein gamma is the passing rate of a target N-1 after investment and is given out through upper layer index system diagnosis and analysis, L is the total number of the power distribution network lines, and l is the number of the lines passing the N-1 verification;

power supply coordination constraint:

the power supply coordination constraint is represented by a capacity-to-load ratio and is one of important indexes reflecting power supply capacity; the high capacity-to-load ratio can increase the operation cost of the power grid and cause excessive investment; otherwise, the load increase requirement may not be met, heavy equipment overload is easily caused, and the operation reliability is reduced; according to the relevant guide rule, the capacity-load ratio is controlled in a reasonable interval:

in the formula:_maxand_minis the upper and lower limits of the capacity-to-load ratioThe specific value is determined by the load increase speed; p_maxThe annual maximum peak load;

heavy overload equipment constraints

The load of the power distribution network is increased rapidly, the problem of heavy load of equipment becomes an increasingly prominent contradiction, and the problem of heavy load of the equipment can be relieved to a certain extent by capacity increase of a transformer and newly added lines; therefore, the target of which the heavy overload ratio should be reduced should be determined by upper layer diagnostic indexes as shown in the following formula:

in the formula: s_overIs the number of heavy-load devices, S is the total number of devices, N_overResolving the number of heavy-duty devices for a single project; total investment constraint

When an investment plan in a planning period is made, a power grid company needs to estimate the investment capacity in the whole planning period, and the total investment amount cannot exceed the investment capacity of the company;

in the formula: u shape_maxRepresents the maximum investment capacity of the investment, U_nInvesting construction costs for the nth project;

prospective return of investment constraint

Selecting the net present value of the investment expectation to represent the investment income; the investment income is restricted as follows:

in the formula: mu represents the expected minimum yield of the current investment of the power grid enterprise;

inter-item coupling constraints

Distribution network projects often have a mutual constraint relationship in implementation, and the distribution network projects are as follows:

2) and (4) item mutual exclusion constraint:

x_i+x_j≤1

i.e. two projects cannot exist simultaneously; aiming at the same problem, a plurality of reconstruction measures are provided, and investment can not be simultaneously or repeatedly invested;

2) item dependency constraints:

x_i-x_j≥0

item j may exist only if item i exists; for example, in the extension project, the low voltage level needs to be equal to the high voltage level in the distribution network project.

3) Strict complementary constraint:

x_i-x_j＝0

two projects must be built simultaneously or not, for example, the construction of the transformer substation matching project and the transformer substation are interdependent.

Preferably, in the step S7,

because the CVaR minimization is different from the maximization target of unit investment benefit, a decreasing half gradient membership function and an increasing half linear membership function are respectively introduced, and the specific steps are as follows:

1) inputting model original data, respectively solving objective functions according to a single-objective optimization model to obtain target values under different optimization objectives, and forming a decision attribute table of the objective functions, specifically F₁*、F₁ ^min、F₁ ⁽²⁾、F₂*、F₂ ⁽¹⁾、F₂ ^maxWhere the band values represent the model solved with the objective function only:

2) the multi-target function is expanded to different degrees according to the will and preference of a decision maker, and then F is determined₁ ^max、F₂ ^minAnd further determining a value range, wherein the specific calculation result is as follows:

0≤F₁ ^max-F₁ ^min≤F₁ ⁽²⁾-F₁ ^min

F₂ ^max-F₂ ^min≥F₂ ^max-F₂ ⁽¹⁾

3) and (3) processing a CVaR minimum objective function by applying a decreasing semigradient membership function, and processing an investment score maximization objective function by applying an increasing semilinear membership function to obtain an objective function membership function as follows:

in the formula: f_i ^maxAnd F_i ^minAs an objective function f_iMinimum and maximum values of (·), π (f)_i) As an objective function f_i(ii) a membership function of;

the membership degree of each objective function is obtained by applying a fuzzy satisfaction degree theory, and the membership degree can be converted into a multi-objective CVaR model according to a max-min principle of a fuzzy set theory, which is as follows:

min λπ₁(f₁)+(1-λ)π₂(f₂)

in the formula: pi ═ min { pi ═_k(f_k),k＝1,2}；π_k(f_k) Membership degrees for each objective function; λ is the weight; the model is a linear model and is convenient to solve.

Preferably, in the step S8, it is determined whether the investment objective and the feasibility are met, if not, the method returns to S3, otherwise, the method is ended, specifically: if the investment is in accordance with the actual situation, the investment is brought into the current-year investment construction plan.

Compared with the prior art, the invention has the beneficial effects that: the scheme provides a power distribution network accurate investment combination optimization model considering risk measure; the model takes projects as decision objects, can take account of the confidence coefficient of achievement of benefits of a single project, well represents the cost performance attributes of planned projects in the investment decisions of the power distribution network by constructing actual constraints such as production performance, investment limitation, project association and the like on the basis of introducing the development diagnosis of the power distribution network and the performance evaluation index of the projects, and finally calculates the optimization strategy of the project combination of the power distribution network under certain investment risk preference based on the fuzzy satisfaction degree of the profitability and the unit output benefits.

Drawings

FIG. 1 is a schematic diagram of the basic steps of an embodiment of the present invention.

Fig. 2 is a schematic diagram of an investment decision flow of an embodiment of the present invention.

Fig. 3 is a schematic diagram of an investment benefit assessment index system for a double-layer power distribution network project according to an embodiment of the present invention.

Fig. 4 is a schematic diagram of an investment application flow of a power distribution network portfolio based on CVaR according to an embodiment of the present invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to fig. 1 to 4 of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

Example (b):

as shown in fig. 1 to 4, a method for optimizing a precision investment project of a power distribution network considering risk measure is characterized by comprising the following steps:

s1, collecting historical data of the regional distribution network needing to arrange investment plans, comprising the following steps: five categories of power distribution network equipment operation information, the current investment budget amount, declaration project collection, power load increase information and urgent problem solving are provided, and the declaration projects of the power distribution network are numbered;

s2, combining the characteristics of the urban medium-high voltage distribution network, constructing a distribution network project investment benefit evaluation double-layer index system, wherein the double-layer index system comprises microscopic indexes and macroscopic indexes, the index system covers five categories of net rack robustness, power supply safety and quality, operation economy and efficiency, power supply harmony, interactivity and social friendliness, and is further subdivided into various sub indexes;

s3, developing and diagnosing the power distribution network in the area where the investment plan is arranged by using the macroscopic indexes, determining the current investment thought and the corresponding investment constraint conditions, and determining the investment concern;

s4, quantifying investment efficiency of a single investment project unit by using microscopic indexes, quantifying benefit output dimensions which cannot be directly quantified by money amount by using scores, such as net rack strength degree improvement, heavy overload equipment reduction and the like, and scoring each alternative project according to index weight and scoring standards so as to assist accurate investment;

s5, based on a conditional risk measurement method, the distribution characteristics of the profitability of a single project are taken into the investment risk, the profitability and the risk evaluation are carried out on the single project, and an investment risk minimum objective function containing a risk preference confidence coefficient is constructed in the dimension;

s6, taking the total investment, the power grid development requirement, the investment portfolio profitability and the coupling relation among projects as constraints, establishing a power distribution network investment portfolio optimization model taking efficiency and risk as targets, and considering both investment cost and investment risk under the boundary condition of investment income and income; let x^T＝(x₁,x₂,x₃,…,x_n)^TAs investment decision vector, x_i1 represents that the ith project is determined to be put into operation, if the ith project is 0, the ith project does not enter the investment plan, and n is the total number of the projects; let V be (V)₁,v₂,v₃,…,v_n) As a project profitability vector, v_iExpressing the yield of the ith project, conforming to normal distribution, wherein the expectation is the average yield, and the variance is the estimate of the risk degree of the project;

s7, converting the fuzzy satisfaction into a single-target mixed integer linear programming problem easy to solve by using the fuzzy satisfaction so as to calculate a power distribution network accurate investment strategy;

and S8, judging whether the investment target and the feasibility are met, if not, returning to S3, otherwise, ending the method.

Preferably, in step S1, the regional distribution network historical data includes: five categories of power distribution network equipment operation information, the current investment budget amount, declaration project collection, power load increase information and urgent problem solving are provided, and the declaration projects of the power distribution network are numbered;

the operation information of the power distribution network equipment comprises daily, monthly and annual load rate statistics, equipment operation age, line loss and transformer loss, power supply reliability, voltage qualification rate, capacity-to-load ratio, N-1 passage rate, line connection rate, line section qualification rate, line power supply radius qualification rate, line insulation rate, distribution automation coverage rate, high-loss distribution-to-occupation ratio, line overload rate, distribution-to-transformation overload rate, comprehensive line loss rate, distributed energy grid-connection rate, dynamic electricity price and electricity consumption proportion and electric quantity proportion for electric vehicles;

the budget amount of the investment comprises the budget amount of the investment;

the declaration project collection comprises investment construction amount, newly-added capacity, newly-added load, newly-added power supply load, estimated power supply reliability (RS-3) improvement degree, estimated voltage qualification rate improvement degree, N-1 line lifting number, distribution line connecting line lifting number, segmented reasonable line lifting number, power supply radius qualified line lifting number, distribution line insulation, lifting kilometers number, distribution automation station number, high-loss distribution transformer descending station number, line heavy-load descending number, distribution transformer heavy-load descending number, estimated line loss rate improvement degree, newly-accessed distributed power supply capacity, dynamic electricity price increase capacity and newly-added charging pile number;

the electricity load increase information comprises annual maximum load increase rate and local shortage load;

the problems to be solved urgently include key special construction plans, outstanding power supply contradictions and the like.

Preferably, in step S2, a power distribution network project investment benefit evaluation double-layer index system is constructed, which includes a microscopic index and a macroscopic index, and specifically includes:

macroscopic indexes are as follows: the method comprises the following steps of (N-1) passing rate (unit%), line contact rate (unit%), line subsection qualification rate (unit%), line power supply radius qualification rate (unit%), power supply reliability rate (RS-3) (unit%), voltage qualification rate (unit%), line insulation rate (unit%), distribution automation coverage rate (unit%), capacity-load ratio, high-loss distribution transformation ratio (unit%), line overload rate (unit%), distribution transformation overload rate (unit%), comprehensive line loss rate (unit%), distributed energy grid-connected rate (unit%), dynamic electricity price and electricity consumption proportion (unit%) and electric automobile electricity consumption proportion (unit%);

microscopic indexes are as follows: the method comprises the steps of increasing the number of N-1 lines (unit: one/ten thousand), increasing the number of interconnection lines of distribution lines (unit: one/ten thousand), increasing the number of sectionally reasonable lines (unit: one/ten thousand) by unit investment, increasing the number of qualified lines (unit: one/ten thousand) by unit investment power supply radius, predicting and improving the local comprehensive voltage qualification rate (unit:%), increasing the insulation degree of the distribution lines (unit: km/ten thousand), predicting and improving the local power supply reliability rate (RS-3) (unit:%), automatically increasing the number of distribution lines (unit: one/ten thousand), decreasing the number of distribution lines (unit: one/ten thousand) by unit investment and heavily-loaded decreasing number of the distribution lines (unit: km/ten thousand) by unit investment, The unit investment distribution transformer heavy load reduction number (unit: station/ten thousand yuan), the local line loss rate predicted improvement degree (unit: percent), the unit investment power increase amount (unit: kWh/yuan), the unit investment power increase load (unit: kW/yuan), the unit investment newly-accessed distributed power capacity (unit: kWh/yuan), the unit investment dynamic power price increase amount (unit: kW/yuan), and the unit investment newly-added charging pile number (unit: pieces/ten thousand yuan).

Preferably, in the step S3, the macro step refers to developing a power distribution network development diagnosis for an area where an investment plan is arranged, determining the current investment idea and part of investment constraint conditions, and determining an investment concern, specifically:

the "N-1" pass rate (unit%) is × 100% of the number of distribution lines/total number of distribution lines satisfying the "N-1" safety criterion;

the line connection rate (unit%) is × 100% which satisfies the number of the distribution lines of the main line interconnection/the total number of the distribution lines;

the line segmentation qualified rate (unit%) is equal to the reasonable segmentation distribution line number/total distribution line number × 100%;

the qualification rate (unit%) of the power supply radius of the line is × 100% (the area of A +, A, B is not more than 3 km, the area of C is not more than 5 km, and the area of D is not more than 15 km) of the number of lines/total number of distribution lines with qualified power supply radius;

power supply reliability (RS-3) (unit%) [1- (average user outage time-average user outage time/statistical period time ] × 100%;

the voltage qualification rate (unit%) is × 100% of the sum of the voltage qualification rates of all the power grid monitoring points/the number of the power grid monitoring points;

the insulation rate (unit%) of the line is × 100% of the length of the insulated line/the total length of the line;

distribution automation coverage (unit%) -total number of automation transformers/distribution transformers × 100;

capacity-to-load ratio is × 100% as distribution capacity/maximum load;

the high-loss distribution ratio (unit%) is × 100% of the number of high-loss distribution/total distribution;

the line heavy load rate (unit%) is × 100% of the number of distribution lines/total number of distribution lines with the annual maximum load rate of more than 70%;

the distribution transformer overloading rate (unit%) is × 100% of distribution transformer station number/total distribution transformer station number, the urban annual maximum overloading rate is greater than 80%;

the integrated line loss rate (unit%) — line loss power/head end transmission power × 100%;

the grid connection rate (unit%) of the distributed energy is distributed energy internet access capacity/total distributed energy capacity;

the proportion (unit%) of the electricity consumption of the dynamic electricity price is the dynamic electricity price load/the total load of the power distribution users;

the electric automobile power consumption proportion (unit percent) is the electric automobile power consumption proportion/distribution network ground power consumption;

through the indexes, the overall operation condition of the power distribution game is known, five power distribution network project emphasis points which meet newly added loads, strengthen the grid structure, eliminate potential safety hazards, match the transformer substation and send out projects and solve heavy overload equipment are determined, and the capacity-to-load ratio, the line heavy load rate, the distribution transformer heavy load rate and the N-1 passing rate target value are used as boundary conditions for model optimization in the step S6.

Preferably, in step S4, scoring each candidate item according to the index weight and the scoring criteria to assist accurate investment is performed by using a benefit output dimension that cannot be directly quantified by money amount quantified by scoring quantification, such as net rack robustness improvement, heavy overload equipment reduction, and the like. The method specifically comprises the following steps:

the unit investment of N-1 line lifting number (unit: number of lines/ten thousand yuan) is equal to the number of lines/total investment of the project which satisfies the increase of the number of N-1 distribution lines after the project is implemented;

the number of the contact lines of the unit investment distribution line is increased (unit: one/ten thousand yuan), namely the number of the single radiation line is reduced or the number of the contact lines is increased/total investment of the project after the project is implemented;

the reasonable line lifting number (unit: one/ten thousand yuan) of the unit investment segmentation is equal to the lifting number of the reasonable line/the total investment of the project after the project is implemented;

the number of the lifting lines (unit: one/ten thousand yuan) of the power supply radius qualified lines in unit investment is equal to the number of the power supply radius qualified lines/total investment of the project after the project is implemented;

the predicted improvement degree (unit:%) of the local integrated voltage yield (voltage yield of the line (or the block line) after the implementation of the item-voltage yield of the line (or the block line) before the implementation of the item);

the insulated lifting kilometer number (unit: km/ten thousand yuan) of the distribution line is equal to the kilometer number/total investment of the project increased by the insulated line after the project is implemented;

the estimated improvement degree of local power supply reliability (RS-3) (unit:%) (power supply reliability of the line (or the block line) after the item is implemented-power supply reliability of the line (or the block line) before the item is implemented);

the number of automatic units (unit: unit/ten thousand yuan) of unit investment power distribution is the number of automatic transformers which are changed into the number of ascending units/total investment of the project after the project is implemented;

the number of the descending units of the unit investment high loss distribution transformer (unit: unit/ten thousand yuan) is equal to the number of the descending units of the high loss distribution transformer/total investment of the project after the project is implemented;

the heavy load of the line of unit investment is decreased by the number (unit: km/ten thousand yuan), namely the number of the distribution lines reduced by the distribution line with the load rate of more than 70% after the project is implemented/the total investment of the project;

the unit investment distribution transformer heavy load descending number (unit: station/ten thousand yuan) is the number of stations/total project investment of distribution transformer reduction with the load rate of more than 80% after the project is implemented;

the predicted improvement degree (unit:%) of the local line loss rate is (theoretical loss of the station distribution transformer (or the block distribution transformer) before the implementation of the item-theoretical loss of the station distribution transformer (or the block distribution transformer) after the implementation of the item;

the unit investment increased power supply amount (unit: kWh/yuan) is the theoretically increased power supply amount/total investment of the distribution transformer related to the project;

the unit investment increase load (unit: kW/yuan) is the theoretically increased load/total investment of the distribution transformer related to the project;

unit investment newly-accessed distributed power capacity (unit: kWh/yuan) is newly-accessed distributed power capacity/total project investment;

the dynamic electricity price increase capacity (unit: kW/yuan) of unit investment is equal to the dynamic electricity price increase capacity/total investment of project;

the unit investment of the newly increased charging pile number (unit: one/ten thousand yuan) is equal to the number of the newly increased charging piles/total project investment;

aiming at different evaluation targets, different relative importance degrees can be reflected in a weighting mode; considering that the investment evaluation dimensionality of the project is multiple, and finally determining the weight through an expert; the net rack strength degree, the power supply safety and quality, the operation economy and efficiency, the power supply coordination, the interactivity and the social friendliness are respectively entitled to 0.2, 0.25, 0.2 and 0.1, and can be flexibly adjusted according to the experience or the work key point requirement;

the single project investment performance index reflects unit investment income and is a benefit index; at present, no mature application experience exists, and no regulation guide rule indicates the range; because the indexes have different dimensions and different attributes, the data needs to be initialized; according to the principle of preferred investment, the larger the unit investment output value is, the more ideal the unit investment output value is; the maximum value is set to 100 points, the minimum value is set to 0 points, and each index score is calculated as follows:

in the formula: r is_iScoring the jth index of the ith item, wherein y_i,jJ-th index data, y, for the i-th item_{j max}Is the jth finger maximum, y_{j min}Is the jth index minimum;

converting each index into a score in a non-dimensionalization way, and combining the index weight, wherein the investment benefit evaluation model of a single project is as follows:

in the formula: r is the comprehensive evaluation value of a certain investment project, y_jIs the score of the jth index, w_jThe weight of the jth index is n indexes; considering the index system of the hierarchical structure, the scores of each target layer can be calculated successively:

in the formula: the superscript t +1 denotes the t +1 th layer index.

Preferably, in step S5, specifically, the method includes:

value at Risk (VaR) refers to the worst expected loss Value for a given confidence level for a given portfolio over a future holding period; CVaR refers to the condition mean value of investment loss exceeding the VaR value, represents the average level of excess loss, and can reflect the potential risk value of the tail part better than the VaR;

let X be the feasible set of investment portfolio,

let f (X, y) be a loss function, wherein X ∈ X is an n-dimensional power distribution network project investment portfolio scheme vector, n is the number of projects, y ∈ R^mM-dimensional random variables and m random factors (such as price difference of electricity purchased and sold, electricity sold amount, unit cost and the like); let the joint probability density function of y be p (y), and for a certain x, the loss f (x, y) caused by y is RFrom a certain distributed random variable, the distribution function that does not exceed the confidence α is:

for any fixed x, ψ (x, a) as a function of α is the cumulative probability distribution function of loss under portfolio x, with β representing confidence, α_β(x) And the VaR value corresponding to the loss f (x, y) when the investment combination is x is expressed, and the calculation formula is as follows:

α_β(x)＝min{α∈R；ψ(x，α)≥β}

calculating the maximum possible loss at a given probability level of β∈ (0,1), and further, calculating by definition the asset loss f (x, y) to be no less than α_β(x) CVaR value of (a):

the VaR function α is contained in the above formula_β(x) Term, and α_β(x) The analytical expression of (A) is difficult to solve, and usually an analyzable function F is introduced_β(x, α) instead of phi_β(x) Calculating CVaR:

wherein [ f (x, y) - α]⁺Represents max {0, f (x, y) - α }, α is VaR value, β is confidence coefficient;

the analytical expression of the probability density function p (y) is difficult to obtain, and the Monte Carlo method is often used for extracting sample data to give integral estimation in the formula above; is provided with y¹,y²,…,y^qQ samples of y, the function F_βThe calculated value of (x, α) is:

in actual calculations, the optimal combination coefficient vector X of the asset and the corresponding CVaR and VaR values are usually determined based on the above formula, and according to the CVaR method, the minimum α -CVaR value is found in the feasible region X, which can be defined as the following optimization model:

min CVaR_α(x^Tw)

in the formula: w is an investment income vector; when discrete samples are used to estimate the asset weights x and CVaR in the model combination, the optimization model can be converted into:

in the formula z_k＝[f(x,y^k)-a)]⁺Converting the original equation into an analytically solvable linear programming model by adding the linear constraints shown in the following equation:

the income of the power distribution network project is generally normally distributed, and the expected income rate can be regarded as the profit level of the project and can represent sigma²Can represent an estimate of the degree of risk, σ²A larger value means more extreme cases and higher investment risk. From the investment risk perspective, each project has a certain decision risk of uncertainty in revenue. For example, the price of electricity for sale is adjusted, and the amount of electricity used is influenced by industrial structure and economic development, etc., which may lead to uncertainty in price difference, amount of electricity sold, standard discount rate and unit power supply cost, and finally lead to difference in profitability and fluctuation range of each project, and even loss of part of the projects, resulting in loss of operation of the power grid company. This distribution is given by an expert integrating the above factors.

Preferably, in step S6, specifically, the method includes:

let x^T＝(x₁,x₂,x₃,…,x_n)^TAs investment decision vector, x_i1 represents that the ith project is determined to be put into operation, if the ith project is 0, the ith project does not enter the investment plan, and n is the total number of the projects; let V be (V)₁,v₂,v₃,…,v_n) As a project profitability vector, v_iExpressing the yield of the ith project, conforming to normal distribution, wherein the expectation is the average yield, and the variance is the estimate of the risk degree of the project; defining a project portfolio investment return function as R (x, y) ═ x^Tw, wherein w ═ v₁U₁,v₂U₂,v₃U₃,…,v_nU_n]I th project expected profit w_i＝v₁U_i，U_iAn amount of investment for the ith project; defining the portfolio investment loss function as f (x, w) — R (x, w), R (x, w) being the portfolio investment yield, the loss function is given by:

1) investment risk minimization objective function

Under the influence of a plurality of uncertain factors, the profitability and the fluctuation range of each project are different, loss is easy to occur in part of projects, the operation loss of a power grid company is caused, the investment risk is reduced as much as possible, and the method specifically comprises the following steps:

wherein f (X, w) is formula (12), (X, α) ∈ (X, R); α is the VaR value;

2) unit investment output efficiency maximization objective function

The investment performance indexes of the single lower-layer project need to calculate unit investment output, and the higher the project score is, the higher the unit output is represented; in order to avoid excessive investment, the average score of the project should be improved as follows:

in the formula: r is the item score vector, R ═ R₁,r₂,r₃,…,r_n)^TWherein r is_iScoring the ith item;

the investment of the power distribution network should preferably consider to meet the power demand constraint and the power utilization reliability constraint, and in addition, the constraints of equipment bearing capacity, total investment, expected income and the like are also considered; meanwhile, coupling constraints such as sequence and the like are established among a plurality of power distribution network projects, and if a certain access supporting project needs to be subjected to capacity expansion at a certain interval (the two projects may not be listed in the same project), the relevance of the two projects also needs to be considered;

power demand constraints:

load prediction is a precondition for power distribution network development, and influence factors such as economic development, electric quantity level and the like of each region need to be integrated; the investment construction of the power distribution network should meet the power demand of the region, and the constraint conditions are as follows:

in the formula: d represents the newly increased power demand; c. C_nRepresenting the newly added power supply capacity of the nth item;

and (3) power utilization reliability constraint:

the passing rate of the N-1 of the power distribution network is directly related to the average annual power failure time of a user and the power supply reliability (RS-3); the power utilization reliability can be remarkably increased when the N-1 passing rate is increased, and the constraint expression is as follows:

wherein gamma is the passing rate of a target N-1 after investment and is given out through upper layer index system diagnosis and analysis, L is the total number of the power distribution network lines, and l is the number of the lines passing the N-1 verification;

power supply coordination constraint:

the power supply coordination constraint is represented by a capacity-to-load ratio and is one of important indexes reflecting power supply capacity; the high capacity-to-load ratio can increase the operation cost of the power grid and cause excessive investment; otherwise, the load increase requirement may not be met, heavy equipment overload is easily caused, and the operation reliability is reduced; according to the relevant guide rule, the capacity-load ratio is controlled in a reasonable interval:

in the formula:_maxand_minthe specific value is determined by the load increase speed for the upper limit and the lower limit of the capacity-load ratio; p_maxThe annual maximum peak load;

heavy overload equipment constraints

The load of the power distribution network is increased rapidly, the problem of heavy load of equipment becomes an increasingly prominent contradiction, and the problem of heavy load of the equipment can be relieved to a certain extent by capacity increase of a transformer and newly added lines; therefore, the target of which the heavy overload ratio should be reduced should be determined by upper layer diagnostic indexes as shown in the following formula:

in the formula: s_overIs the number of heavy-load devices, S is the total number of devices, N_overResolving the number of heavy-duty devices for a single project; total investment constraint

When an investment plan in a planning period is made, a power grid company needs to estimate the investment capacity in the whole planning period, and the total investment amount cannot exceed the investment capacity of the company;

in the formula: u shape_maxRepresents the maximum investment capacity of the investment, U_nInvesting construction costs for the nth project;

prospective return of investment constraint

Selecting the net present value of the investment expectation to represent the investment income; the investment income is restricted as follows:

in the formula: mu represents the expected minimum yield of the current investment of the power grid enterprise;

inter-item coupling constraints

Distribution network projects often have a mutual constraint relationship in implementation, and the distribution network projects are as follows:

3) and (4) item mutual exclusion constraint:

x_i+x_j≤1

i.e. two projects cannot exist simultaneously; aiming at the same problem, a plurality of reconstruction measures are provided, and investment can not be simultaneously or repeatedly invested;

2) item dependency constraints:

x_i-x_j≥0

item j may exist only if item i exists; for example, in the extension project, the low voltage level needs to be equal to the high voltage level in the distribution network project.

3) Strict complementary constraint:

x_i-x_j＝0

two projects must be built simultaneously or not, for example, the construction of the transformer substation matching project and the transformer substation are interdependent.

Preferably, in the step S7,

because the CVaR minimization is different from the maximization target of unit investment benefit, a decreasing half gradient membership function and an increasing half linear membership function are respectively introduced, and the specific steps are as follows:

1) inputting model original data, respectively solving objective functions according to a single-objective optimization model to obtain target values under different optimization objectives, and forming a decision attribute table of the objective functions, specifically F₁*、F₁ ^min、F₁ ⁽²⁾、F₂*、F₂ ⁽¹⁾、F₂ ^maxWhere the band values represent the model solved with the objective function only:

2) the multi-target function is expanded to different degrees according to the will and preference of a decision maker, and then F is determined₁ ^max、F₂ ^minAnd further determining a value range, wherein the specific calculation result is as follows:

0≤F₁ ^max-F₁ ^min≤F₁ ⁽²⁾-F₁ ^min

F₂ ^max-F₂ ^min≥F₂ ^max-F₂ ⁽¹⁾

3) and (3) processing a CVaR minimum objective function by applying a decreasing semigradient membership function, and processing an investment score maximization objective function by applying an increasing semilinear membership function to obtain an objective function membership function as follows:

in the formula: f_i ^maxAnd F_i ^minAs an objective function f_iMinimum and maximum values of (·), π (f)_i) As an objective function f_i(ii) a membership function of;

the membership degree of each objective function is obtained by applying a fuzzy satisfaction degree theory, and the membership degree can be converted into a multi-objective CVaR model according to a max-min principle of a fuzzy set theory, which is as follows:

min λπ₁(f₁)+(1-λ)π₂(f₂)

in the formula: pi ═ min { pi ═_k(f_k),k＝1,2}；π_k(f_k) Membership degrees for each objective function; λ is the weight; the model is a linear model and is convenient to solve.

Preferably, in the step S8, it is determined whether the investment objective and the feasibility are met, if not, the method returns to S3, otherwise, the method is ended, specifically: if the investment is in accordance with the actual situation, the investment is brought into the current-year investment construction plan. .

The above are preferred embodiments of the present invention, and all changes made according to the technical scheme of the present invention that produce functional effects do not exceed the scope of the technical scheme of the present invention belong to the protection scope of the present invention.

Claims (9)

1. A power distribution network accurate investment project optimization method considering risk measure is characterized by comprising the following steps:

s1, collecting historical data of a regional power distribution network in which an investment plan needs to be arranged;

s2, constructing a power distribution network project investment benefit evaluation double-layer index system by combining the characteristics of a medium-high voltage power distribution network in a city, wherein the double-layer index system comprises microscopic indexes and macroscopic indexes, and the index system covers five categories of net rack robustness, power supply safety and quality, operation economy and efficiency, power supply harmony, interactivity and social friendliness;

s3, developing and diagnosing the power distribution network in the area where the investment plan is arranged by using the macroscopic indexes, determining the current investment thought and the corresponding investment constraint conditions, and determining the investment concern;

s4, quantifying investment efficiency of a single investment project unit by using microscopic indexes, quantifying benefit output dimensions which cannot be directly quantified by using scores, and scoring each alternative project according to index weight and scoring standards so as to assist accurate investment;

s5, based on a conditional risk measurement method, the distribution characteristics of the profitability of a single project are taken into the investment risk, the profitability and the risk evaluation are carried out on the single project, and an investment risk minimum objective function containing a risk preference confidence coefficient is constructed in the dimension;

s6, taking the total investment, the power grid development requirement, the investment portfolio profitability and the coupling relation among projects as constraints, establishing a power distribution network investment portfolio optimization model taking efficiency and risk as targets, and considering both investment cost and investment risk under the boundary condition of investment income and income;

s7, converting the fuzzy satisfaction into a single-target mixed integer linear programming problem easy to solve by using the fuzzy satisfaction so as to calculate a power distribution network accurate investment strategy;

and S8, judging whether the investment target and the feasibility are met, if not, returning to S3, otherwise, ending the method.

2. The method for optimizing the investment project of the power distribution network based on the risk measure according to claim 1,

in step S1, the regional distribution network historical data includes: five categories of power distribution network equipment operation information, the current investment budget amount, declaration project collection, power load increase information and urgent problem solving are provided, and the declaration projects of the power distribution network are numbered;

the operation information of the power distribution network equipment comprises daily, monthly and annual load rate statistics, equipment operation age, line loss and transformer loss, power supply reliability, voltage qualification rate, capacity-to-load ratio, N-1 passage rate, line connection rate, line section qualification rate, line power supply radius qualification rate, line insulation rate, distribution automation coverage rate, high-loss distribution-to-occupation ratio, line overload rate, distribution-to-transformation overload rate, comprehensive line loss rate, distributed energy grid-connection rate, dynamic electricity price and electricity consumption proportion and electric quantity proportion for electric vehicles;

the budget amount of the investment comprises the budget amount of the investment;

the declaration project collection comprises investment construction amount, newly-added capacity, newly-added load, newly-added power supply load, predicted power supply reliability improvement degree, predicted voltage qualification rate improvement degree, N-1 line lifting number, distribution line connecting line lifting number, segmented reasonable line lifting number, power supply radius qualification line lifting number, distribution line insulation, lifting kilometer number, distribution automation station number, high-loss distribution transformer descending station number, line heavy-load descending number, distribution transformer heavy-load descending number, predicted line loss rate improvement degree, newly-accessed distributed power capacity, dynamic electricity price increase capacity and newly-added charging pile number;

the electricity load increase information comprises annual maximum load increase rate and local shortage load;

the problems to be solved urgently comprise key special construction plans and outstanding power supply contradictions.

3. The method for optimizing the accurate investment project of the power distribution network in consideration of the risk measure according to claim 1, wherein in the step S2, a power distribution network project investment benefit assessment double-layer index system is constructed, which comprises micro indexes and macro indexes, and specifically comprises:

macroscopic indexes are as follows: the system comprises an N-1 passage rate, a line interconnection rate, a line section qualification rate, a line power supply radius qualification rate, a power supply reliability rate, a voltage qualification rate, a line insulation rate, a distribution automation coverage rate, a capacity-load ratio, a high-loss distribution duty ratio, a line heavy load rate, a distribution change heavy load rate, a comprehensive line loss rate, a distributed energy grid-connection rate, a dynamic electricity price and electricity consumption proportion and an electric quantity proportion for electric vehicles;

microscopic indexes are as follows: the method comprises the steps of increasing the number of N-1 line hoisting pieces per unit investment, increasing the number of interconnection lines of distribution lines per unit investment, increasing the number of reasonable line hoisting pieces per unit investment by sections, increasing the number of line hoisting pieces per unit investment, increasing the number of qualified line hoisting pieces per unit investment power supply radius, improving the degree of predicted qualification rate of local comprehensive voltage, improving the number of insulated hoisting kilometers of distribution lines per unit investment, improving the degree of predicted reliability rate of local power supply, increasing the number of automatic distribution pieces per unit investment, decreasing the number of high-loss distribution transformer descending pieces per unit investment, decreasing the number of heavy load per unit investment line heavy load per unit investment, improving the degree of loss rate of local lines, increasing the power supply amount per unit investment, increasing the load per unit investment, increasing the capacity of a newly-accessed distributed power supply per unit investment, increasing the.

4. The method for optimizing the precision investment project of the power distribution network based on the risk measure of the claim 1, wherein in the step S3, the macro means to perform development diagnosis of the power distribution network for the area where the investment plan is arranged, determine the current investment idea and part of the investment constraint conditions, and determine the investment concern, specifically:

the "N-1" pass rate is × 100% of the number of distribution lines/total number of distribution lines meeting the "N-1" safety criterion;

the line connection rate is × 100% which satisfies the number of the distribution lines of the main line interconnection/the total number of the distribution lines;

the line segmentation yield is × 100% of reasonable segmentation distribution line number/total distribution line number;

the line power supply radius qualification rate is × 100% of the number of lines with qualified power supply radius/the total number of distribution lines;

power supply reliability is [1- (user average power failure time-user average power limit power failure time/statistical period time ] × 100%;

the voltage qualification rate is × 100% of the sum of the voltage qualification rates of all the power grid monitoring points/the number of the power grid monitoring points;

the insulation rate of the line is × 100% of the length of the insulated line/the total length of the line;

the distribution automation coverage rate is × 100% of the total number of the automation transformers/distribution transformers;

capacity-to-load ratio is × 100% as distribution capacity/maximum load;

the ratio of the high-loss distribution transformer is × 100% of the number of the high-loss distribution transformers/the total number of the distribution transformers;

the line overloading rate is × 100% of the number of distribution lines/total number of distribution lines with the annual maximum load rate more than 70%;

the distribution transformer overloading rate is × 100% of the number of distribution transformer stations/total number of distribution transformer stations with the urban annual maximum overloading rate being greater than 80%;

the integrated line loss rate is × 100% of the line loss power/the head end transmission power;

the grid connection rate of the distributed energy is distributed energy internet access capacity/total distributed energy capacity;

the proportion of the electricity consumption of the dynamic electricity price is the dynamic electricity price load/the total load of the power distribution users;

the electric quantity proportion for the electric vehicle is equal to the electric quantity proportion for the electric vehicle/the ground electric quantity of the power distribution network;

through the indexes, the overall operation condition of the power distribution game is known, five power distribution network project emphasis points which meet newly added loads, strengthen the grid structure, eliminate potential safety hazards, match the transformer substation and send out projects and solve heavy overload equipment are determined, and the capacity-to-load ratio, the line heavy load rate, the distribution transformer heavy load rate and the N-1 passing rate target value are used as boundary conditions for model optimization in the step S6.

5. The method for optimizing the precision investment project of the power distribution network based on the risk measure according to claim 1, wherein the step S4 specifically comprises:

the unit investment of N-1 line lifting number is equal to the number of the N-1 distribution lines increased after the project is implemented/total project investment;

the number of the contact lines of the unit investment distribution line is equal to the number of the single radiation lines reduced or the number of the contact lines increased/total investment of the project after the project is implemented;

the number of lifting lines of the segmental reasonable line of unit investment is equal to the number of lifting lines of the segmental reasonable line/total investment of the project after the project is implemented;

the number of the lifting lines of the power supply radius qualified line in unit investment is equal to the number of the power supply radius qualified lines/total investment of the project after the project is implemented;

the predicted improvement degree of the local comprehensive voltage qualification rate is equal to the qualification rate of the line voltage after the project is implemented, namely the qualification rate of the line voltage before the project is implemented;

the number of insulated and lifted kilometers of the distribution line is equal to the number of kilometers/total investment of the project increased by the insulated line after the project is implemented;

the predicted improvement degree of the local power supply reliability is the reliability of the power supply of the line after the project is implemented, and the reliability of the power supply of the line before the project is implemented;

the number of automatic units for unit investment and power distribution is equal to the number of automatic transformers which rise after the project is implemented/the total investment of the project;

the number of the descending units of the unit investment high loss distribution transformer is equal to the number of the descending units of the high loss distribution transformer/total investment of the project after the project is implemented;

the number of heavy load descending pieces of the unit investment line is equal to the number of reduced distribution lines with the load rate of more than 70% after the project is implemented/the total investment of the project;

the unit investment distribution transformer heavy load reduction number is the number of distribution transformer reduced units/total project investment with the load rate of more than 80% after the project is implemented;

the predicted improvement degree of the local line loss rate is the theoretical loss of the distribution transformer before the project is implemented, and the theoretical loss of the distribution transformer after the project is implemented;

the unit investment increased power supply amount is the power supply amount theoretically increased by the distribution transformer related to the project/total investment of the project;

the unit investment increase load is the load/total investment of the project which is increased theoretically by the distribution transformer related to the project;

the unit investment and the capacity of the newly accessed distributed power supply are equal to the capacity of the newly accessed distributed power supply/the total investment of a project;

the unit investment dynamic electricity price increase capacity is the dynamic electricity price increase capacity/total investment of the project;

the unit investment newly increased charging pile number is equal to the newly increased charging pile number/total project investment;

aiming at different evaluation targets, different relative importance degrees can be reflected in a weighting mode; considering that the investment evaluation dimensionality of the project is multiple, and finally determining the weight through an expert; the net rack strength degree, the power supply safety and quality, the operation economy and efficiency, the power supply coordination, the interactivity and the social friendliness are respectively entitled to 0.2, 0.25, 0.2 and 0.1, and can be flexibly adjusted according to the experience or the work key point requirement;

the single project investment performance index reflects unit investment income and is a benefit index; at present, no mature application experience exists, and no regulation guide rule indicates the range; because the indexes have different dimensions and different attributes, the data needs to be initialized; according to the principle of preferred investment, the larger the unit investment output value is, the more ideal the unit investment output value is; the maximum value is set to 100 points, the minimum value is set to 0 points, and each index score is calculated as follows:

in the formula: r is_iScoring the jth index of the ith item, wherein y_i,jJ-th index data, y, for the i-th item_{j max}Is the jth finger maximum, y_{j min}Is the jth index minimum;

converting each index into a score in a non-dimensionalization way, and combining the index weight, wherein the investment benefit evaluation model of a single project is as follows:

in the formula: r is the comprehensive evaluation value of a certain investment project, y_jIs the jth fingerTarget score, w_jThe weight of the jth index is n indexes; considering the index system of the hierarchical structure, the scores of each target layer can be calculated successively:

in the formula: the superscript t +1 denotes the t +1 th layer index.

6. The method for optimizing the precision investment project of the power distribution network based on the risk measure according to claim 5, wherein the step S5 specifically comprises:

the risk value VaR refers to the worst expected loss value of a certain investment portfolio in a certain future holding period at a given confidence level; CVaR refers to the condition mean value of investment loss exceeding the VaR value, represents the average level of excess loss, and can reflect the potential risk value of the tail part better than the VaR;

let X be the feasible set of investment portfolio,

let f (X, y) be a loss function, wherein X ∈ X is an n-dimensional power distribution network project investment portfolio scheme vector, n is the number of projects, y ∈ R^mAnd setting the joint probability density function of y as p (y), wherein for the determined x, the loss f (x, y) caused by y is a random variable obeying a certain distribution on R, and the distribution function of the random variable not exceeding the confidence α is as follows:

Ψ(x,α)＝∫_{f(x,y)，，a}p(y)dy

for any fixed x, ψ (x, a) as a function of α is the cumulative probability distribution function of loss under portfolio x, with β representing confidence, α_β(x) And the VaR value corresponding to the loss f (x, y) when the investment combination is x is expressed, and the calculation formula is as follows:

α_β(x)＝min{α∈R；ψ(x，α)≥β}

calculating the maximum possible loss at a given probability level of β∈ (0,1), and calculating the asset loss f (x, y) by definition to be no less than α_β(x) CVaR value of (a):

the VaR function α is contained in the above formula_β(x) Term, and α_β(x) The analytical expression of (A) is difficult to solve, and usually an analyzable function F is introduced_β(x, α) instead of phi_β(x) Calculating CVaR:

wherein [ f (x, y) - α]⁺Represents max {0, f (x, y) - α }, α is VaR value, β is confidence coefficient;

the analytical expression of the probability density function p (y) is difficult to obtain, and the Monte Carlo method is often used for extracting sample data to give integral estimation in the formula above; is provided with y¹,y²,…,y^qQ samples of y, the function F_βThe calculated value of (x, α) is:

in actual calculations, the optimal combination coefficient vector X of the asset and the corresponding CVaR and VaR values are usually determined based on the above formula, and according to the CVaR method, the minimum α -CVaR value is found in the feasible region X, which can be defined as the following optimization model:

min CVaR_α(x^Tw)

in the formula: w is an investment income vector; when discrete samples are used to estimate the asset weights x and CVaR in the model combination, the optimization model can be converted into:

in the formula z_k＝[f(x,y^k)-a)]⁺By adding the linear constraint shown in the following formula, the original equation can be converted into analytically solvingLinear programming model of solution:

the income of the power distribution network project is generally normally distributed, and the expected income rate can be regarded as the profit level of the project and can represent sigma²Can represent an estimate of the degree of risk, σ²A larger value means more extreme cases and higher investment risk.

7. The method for optimizing the precision investment project of the power distribution network based on the risk measure according to claim 6, wherein the step S6 specifically comprises:

let x^T＝(x₁,x₂,x₃,…,x_n)^TAs investment decision vector, x_i1 represents that the ith project is determined to be put into operation, if the ith project is 0, the ith project does not enter the investment plan, and n is the total number of the projects; let V be (V)₁,v₂,v₃,…,v_n) As a project profitability vector, v_iExpressing the yield of the ith project, conforming to normal distribution, wherein the expectation is the average yield, and the variance is the estimate of the risk degree of the project; defining a project portfolio investment return function as R (x, y) ═ x^Tw, wherein w ═ v₁U₁,v₂U₂,v₃U₃,…,v_nU_n]I th project expected profit w_i＝v₁U_i，U_iAn amount of investment for the ith project; defining the portfolio investment loss function as f (x, w) — R (x, w), R (x, w) being the portfolio investment yield, the loss function is given by:

1) investment risk minimization objective function

Under the influence of a plurality of uncertain factors, the profitability and the fluctuation range of each project are different, loss is easy to occur in part of projects, the operation loss of a power grid company is caused, the investment risk is reduced as much as possible, and the method specifically comprises the following steps:

wherein f (X, w) is formula (12), (X, α) ∈ (X, R); α is the VaR value;

2) unit investment output efficiency maximization objective function

The investment performance indexes of the single lower-layer project need to calculate unit investment output, and the higher the project score is, the higher the unit output is represented; in order to avoid excessive investment, the average score of the project should be improved as follows:

in the formula: r is the item score vector, R ═ R₁,r₂,r₃,…,r_n)^TWherein r is_iScoring the ith item;

the investment of the power distribution network is preferably considered to meet the power demand constraint and the power utilization reliability constraint, and in addition, the constraints of equipment bearing capacity, total investment, expected income and the like are also considered; meanwhile, coupling constraints such as sequence and the like are established among a plurality of power distribution network projects, and the relevance of the coupling constraints is also considered;

power demand constraints:

load prediction is a precondition for power distribution network development, and influence factors such as economic development, electric quantity level and the like of each region need to be integrated; the investment construction of the power distribution network should meet the power demand of the region, and the constraint conditions are as follows:

in the formula: d represents the newly increased power demand; c. C_nRepresenting the newly added power supply capacity of the nth item;

and (3) power utilization reliability constraint:

the N-1 passing rate of the power distribution network is directly related to the average annual power failure time and the power supply reliability of a user; the power utilization reliability can be remarkably increased when the N-1 passing rate is increased, and the constraint expression is as follows:

wherein gamma is the passing rate of a target N-1 after investment and is given out through upper layer index system diagnosis and analysis, L is the total number of the power distribution network lines, and l is the number of the lines passing the N-1 verification;

power supply coordination constraint:

the power supply coordination constraint is represented by a capacity-to-load ratio and is one of important indexes reflecting power supply capacity; the high capacity-to-load ratio can increase the operation cost of the power grid and cause excessive investment; otherwise, the load increase requirement may not be met, heavy equipment overload is easily caused, and the operation reliability is reduced; according to the relevant guide rule, the capacity-load ratio is controlled in a reasonable interval:

in the formula:_maxand_minthe specific value is determined by the load increase speed for the upper limit and the lower limit of the capacity-load ratio; p_maxThe annual maximum peak load;

heavy overload equipment constraints

The load of the power distribution network is increased rapidly, the problem of heavy load of equipment becomes an increasingly prominent contradiction, and the problem of heavy load of the equipment can be relieved to a certain extent by capacity increase of a transformer and newly added lines; therefore, the target of which the heavy overload ratio should be reduced should be determined by upper layer diagnostic indexes as shown in the following formula:

in the formula: s_overIs the number of heavy-load devices, S is the total number of devices, N_overResolving the number of heavy-duty devices for a single project; total investment constraint

When an investment plan in a planning period is made, a power grid company needs to estimate the investment capacity in the whole planning period, and the total investment amount cannot exceed the investment capacity of the company;

in the formula: u shape_maxRepresents the maximum investment capacity of the investment, U_nInvesting construction costs for the nth project;

prospective return of investment constraint

Selecting the net present value of the investment expectation to represent the investment income; the investment income is restricted as follows:

in the formula: mu represents the expected minimum yield of the current investment of the power grid enterprise;

inter-item coupling constraints

Distribution network projects often have a mutual constraint relationship in implementation, and the distribution network projects are as follows:

1) and (4) item mutual exclusion constraint:

x_i+x_j≤1

i.e. two projects cannot exist simultaneously; aiming at the same problem, a plurality of reconstruction measures are provided, and investment can not be simultaneously or repeatedly invested;

2) item dependency constraints:

x_i-x_j≥0

item j may exist only if item i exists;

3) strict complementary constraint:

x_i-x_j＝0

both projects must be built at the same time or not.

8. The method for optimizing the investment project of the power distribution network based on the risk measure according to claim 7, wherein in the step S7,

because the CVaR minimization is different from the maximization target of unit investment benefit, a decreasing half gradient membership function and an increasing half linear membership function are respectively introduced, and the specific steps are as follows:

1) inputting model original data, respectively solving objective functions according to a single-objective optimization model to obtain target values under different optimization objectives, and forming a decision attribute table of the objective functions, specifically F₁*、F₁ ^min、F₁ ⁽²⁾、F₂*、F₂ ⁽¹⁾、F₂ ^maxWhere the band values represent the model solved with the objective function only:

2) the multi-target function is expanded to different degrees according to the will and preference of a decision maker, and then F is determined₁ ^max、F₂ ^minAnd further determining a value range, wherein the specific calculation result is as follows:

0≤F₁ ^max-F₁ ^min≤F₁ ⁽²⁾-F₁ ^min

F₂ ^max-F₂ ^min≥F₂ ^max-F₂ ⁽¹⁾

3) and (3) processing a CVaR minimum objective function by applying a decreasing semigradient membership function, and processing an investment score maximization objective function by applying an increasing semilinear membership function to obtain an objective function membership function as follows:

in the formula: f_i ^maxAnd F_i ^minAs an objective function f_iMinimum and maximum values of (·), π (f)_i) As an objective function f_i(ii) a membership function of;

the membership degree of each objective function is obtained by applying a fuzzy satisfaction degree theory, and the membership degree can be converted into a multi-objective CVaR model according to a max-min principle of a fuzzy set theory, which is as follows:

min λπ₁(f₁)+(1-λ)π₂(f₂)

in the formula: pi ═ min { pi ═_k(f_k),k＝1,2}；π_k(f_k) Membership degrees for each objective function; λ is the weight; the model is a linear model and is convenient to solve.

9. The method for optimizing the accurate investment project of the power distribution network based on the risk measure of the claim 8, wherein the step S8 is performed to determine whether the investment objective and the feasibility are met, if not, the method returns to S3, otherwise, the method is ended, specifically: if the investment is in accordance with the actual situation, the investment is brought into the current-year investment construction plan.

Images