CN115733178A

CN115733178A - Optical storage charging station capacity configuration method and system based on cost and risk multiple targets

Info

Publication number: CN115733178A
Application number: CN202211362402.7A
Authority: CN
Inventors: 黄婧杰; 李金成; 袁亮; 周任军; 杨洪明; 禹海峰
Original assignee: Changsha University of Science and Technology
Current assignee: Changsha University of Science and Technology
Priority date: 2022-11-02
Filing date: 2022-11-02
Publication date: 2023-03-03

Abstract

The invention belongs to the technical field of planning of optical storage charging stations, and discloses a method and a system for configuring the capacity of an optical storage charging station based on cost and risk multiple targets, wherein a cost and risk multiple-target capacity optimal configuration model of the optical storage charging station is established, a risk function in the model is investment maintenance and operation cost risk quantified by condition risk value, and a cost function is the sum of expected investment maintenance and operation cost under a typical scene considering charging demand and photovoltaic output; combining an amplification epsilon-constraint method, taking cost as a main target, and taking a risk secondary target as a constraint; and solving the model to obtain pareto fronts of cost and risk under different risk preferences and corresponding configuration capacity, and screening out an objective decision scheme by adopting an entropy weight-TOPSIS method. Compared with a linear weighting method in the traditional risk management, the method has the advantages that after the targets are processed by the epsilon-constraint method, a front edge with better distribution and boundary optimality can be obtained, and an investment scheme with more detailed cost and risk division is provided.

Description

Optical storage charging station capacity configuration method and system based on cost and risk multiple targets

Technical Field

The invention belongs to the technical field of planning of optical storage charging stations, and particularly relates to a cost and risk multi-objective-based capacity configuration method and system for an optical storage charging station.

Background

At present, under the drive of a double-carbon target, industries of Electric Vehicles (EVs) and charging stations develop rapidly, and light storage charging stations coupling photovoltaic and energy storage are widely concerned due to the advantages of on-site photovoltaic consumption, direct reduction of carbon emission and the like.

When planning and constructing the optical storage charging station, investors pay attention to the cost of the optical storage charging station in the operation process. Reasonable photovoltaic and energy storage capacity size and ratio not only can reduce the equipment investment or the running cost of light storage charging station, can also reduce the running cost through promoting photovoltaic absorption space, reducing the peak valley difference of purchasing the electric power curve etc.. However, the uncertainty in photovoltaic output and EV charging demand can affect the capacity configuration of the optical storage charging station and further affect the cost of commissioning process for the optical storage charging station investors. Therefore, the uncertainty plan is converted into the certainty plan by utilizing the opportunity constraint plan, the investment cost is reduced, the investment recovery period is shortened, and the economy of investment operation of the optical storage charging station is realized. But in most planning operations investors also face the risk of high costs due to uncertainty, i.e. exceeding the expected costs. This risk in the planning of a light storage charging station is caused by uncertainties in photovoltaic output and EV charging demand, ignoring the risk will affect the investor's estimate of the cost of the commissioning process, and therefore it is necessary to quantify the risk and measure how much it affects the cost.

The main methods for dealing with risks are: the method comprises the following steps of a variance method, a value at risk (VaR) value, a conditional value at risk (CVaR), and the like, wherein the CVaR has good mathematical properties, and under the condition of meeting a sub-additivity and consistency axiom, the tail risk of the uncertain quantity under the given probability distribution can be accurately quantified. A great deal of applications have been made in power system planning and optimal scheduling, such as: when the energy system is planned and the capacity of the virtual power plant is configured comprehensively, the CVaR is adopted to measure the operation cost risk caused by the uncertainty of the output of renewable energy, the load change and the electricity price fluctuation; and when the wind power system rotates for standby optimization scheduling, the risk brought to the safe operation of the system by using the CVaR to measure the uncertainty factor is adopted.

In the research including risk management, the method is implemented by optimizing a composite objective function. Among them, the traditional approach to deal with the risk term is to linearly weight it into the objective function with one weighting factor and provide a solution set and scheme under different weighting factors set subjectively. However, when the multi-target problem is processed by adopting the linear weighting method, the pareto solution set with optimal distribution cannot be guaranteed, and the enhanced epsilon-constraint method can guarantee that when the multi-target optimization problem is solved, the obtained pareto solution set has better distribution and boundary optimality, can map the actual pareto front edge of the multi-target problem, and provides an investment scheme with easier control on cost and gradient of risk solution sets for investors. Therefore, it is desirable to design a new capacity allocation method for optical storage and charging stations.

Through the above analysis, the problems and defects of the prior art are as follows:

(1) In the prior art, uncertainty of photovoltaic output and EV charging demand affects capacity allocation of the optical storage charging station, and further affects cost of commissioning process of optical storage charging station investors.

(2) Investors also face a high cost risk of uncertainty in most planning runs, i.e., a risk of exceeding the expected cost, and ignoring this risk will affect the investor's estimate of the commissioning process cost.

(3) In the conventional method for processing the risk items, when a linear weighting method is adopted to process a multi-target problem, a pareto solution set with optimal distribution cannot be obtained.

Disclosure of Invention

Aiming at the problems in the prior art, the invention provides a cost and risk multi-objective optical storage charging station capacity configuration method and system, and particularly relates to a cost and risk amplification epsilon-constraint multi-objective optical storage charging station capacity optimal configuration method, system, medium, equipment and terminal.

The invention is realized in such a way, and provides a capacity configuration method of an optical storage charging station based on cost and risk multiple targets, which comprises the following steps: quantifying risks brought by uncertainty of photovoltaic output and charging requirements of the electric automobile by adopting CVaR, and quantifying the risks into economic indexes, namely risk values; establishing a risk and cost multi-target planning operation model by combining an augmentation epsilon-constraint method and taking cost as a main target and taking a risk secondary target as a constraint; the pareto frontier of cost and risk under different risk preferences and corresponding configuration capacity are obtained by the solving model, and an objective decision scheme is screened out by adopting an entropy weight-TOPSIS method while providing a subjective decision basis for investors.

Further, the method for configuring the capacity of the optical storage charging station based on cost and risk multiple targets comprises the following steps:

the method comprises the following steps that firstly, a Monte Carlo sampling method is adopted to obtain an electric automobile charging demand scene, and a four-season typical photovoltaic output is utilized to obtain a photovoltaic power generation scene; an uncertainty function of the charging requirement and the photovoltaic output of the electric automobile is constructed by combining a scene method;

establishing an investment maintenance cost function, an operation cost function and a CVaR risk measurement function of the optical storage system; the investment and maintenance cost of the optical storage system comprises equal-annual-value initial investment cost and annual operation and maintenance cost; the cost of the whole planning operation stage is the sum of the investment maintenance cost and the operation cost of the light storage system, the value of the risk represents the high-cost risk caused by uncertainty, and the value is quantified by utilizing a CVaR theory;

thirdly, processing multiple targets by combining an amplification epsilon-constraint method according to the target of minimizing cost and risk, and establishing a cost and risk multiple-target optical storage charging station capacity optimal configuration model;

and step four, solving the converted mixed integer linear programming model by using a Gurobi solver, and obtaining a capacity programming result and a corresponding optimized operation strategy of the optical storage charging station under different risk values after the solution is finished.

Further, in the step one, a process of converting the charging requirement of the electric vehicle and the uncertainty problem of the photovoltaic output into a scene to be researched is as follows:

the charging requirement of the electric automobile is determined by initial charging time and initial charging SOC (state of charge), the initial charging SOC of the electric automobile approximately obeys log-normal distribution, and the initial charging time approximately obeys normal distribution;

in the formula, S _OC1 Charging the electric automobile with an initial SOC; t is t ₁ Is the initial time of charging;

and

charging the average value and the standard deviation of the initial SOC variable logarithm for the electric automobile;

and

and charging the average value and the standard deviation of the logarithm of the initial SOC variable of the electric automobile. Sampling and sampling the initial charging time and the initial SOC state of the electric automobile by using a Monte Carlo method to obtain the charging requirement of the electric automobile, and obtaining the number of charging scenes required by the electric automobile by combining with kmeans clustering;

P _PV (t)＝p _er (t)P _PV ；

in the formula, P _PV (t)、P _PV 、p _er (t) photovoltaic output power at the time t, photovoltaic output percentage at the time t and photovoltaic configuration capacity. The uncertainty of photovoltaic output and the charging requirement of the electric automobile is processed by adopting a scene method, various possible conditions are simulated through a large number of scenes, and the random planning problem is converted into the deterministic planning problem. Setting photovoltaic output scene set s = { s = } _i ,i＝1,2,...,n _s Electric vehicle charging demand scene set e = { e = } _j ,i＝1,2,...,n _e }; wherein n is _s And e _j A total field Jing Geshu respectively for photovoltaic output and electric vehicle charging requirements; superscripts se all indicate values at s _i Individual photovoltaic output scene, e _j In the scene of charging requirement of the electric automobile.

Further, in the second step, the process of describing the cost function and the risk metric function when the optical storage charging station optical storage system capacity is configured is as follows:

the investment and maintenance cost of the optical storage system comprises two aspects of equal-annual-value initial investment cost and annual operation and maintenance cost;

C _inv ＝(C _PV P _PV +C _ESS,W W _ESS +C _ESS,P P _ESS )C _RF ；

C _cost ＝C _inv +C _OM ；

in the formula, C _inv 、C _PV 、C _ESS,W 、C _ESS,P 、C _RF Respectively representing the equal annual investment cost of the photovoltaic and energy storage system, the unit capacity investment cost of the photovoltaic and energy storage system and the unit power investment cost of the energy storage systemThe annual value investment coefficient; r and m respectively represent the discount rate and the service life of a corresponding system; c _OM 、

C _cost Respectively representing the annual maintenance cost of the optical storage system, the annual maintenance cost of the unit capacity of the photovoltaic system and the energy storage system and the investment maintenance cost of the optical storage system; r ^se 、

The annual running cost of the scene se, the power purchased and sold from the optical storage charging station to the power grid and the charging power of the electric automobile are respectively calculated; a. b and c are the electricity purchase and sale price of the light storage charging station to the power grid and the charging price of the electric vehicle respectively; t is the number of operating hours, and 24 hours are taken; pi(s) _i )、π(e _j ) Are respectively the s _i Individual photovoltaic output scene, e _j Probability of individual electric vehicle charging demand scenarios;

wherein alpha is the confidence level, and the potential maximum loss risk at the confidence level alpha is C _VaR ；C _CVaR Is in excess of C _VaR Average loss of parts, representing cost risk; z is a radical of ^se Is a virtual variable; introducing a CVaR quantitative cost risk value characterizing risks associated with expected investment maintenance and operating costs.

Further, in the third step, an augmented epsilon-constraint method is constructed to process a cost and risk multi-target model, and the process of solving the multi-target planning problem of the optical storage charging station is as follows:

the objective function for the lowest cost is:

the objective function for the individual minimum risk is:

min{F ₂ (x)＝C _CVaR }；

an objective function of the CVaR programming problem introduced with the traditional linear weighting method:

min{(1-β)F ₁ (x)+βF ₂ (x)}；

and manually regulating and controlling the risk weight through the weighting factor to obtain different planning schemes. Where β is a weighting factor in the range of [0,1] used to achieve a tradeoff between cost and risk, representing a risk preference factor. Different investment schemes are obtained by changing the parameter beta, and an effective front edge of cost and risk is constructed; the larger beta is, the more serious risk is shown, and an investor is of a risk evasion type; the smaller beta indicates the more neglect of risk, and the investor is of the risk pursuit type.

The process of processing cost and risk multiple targets by the epsilon-constraint method is as follows:

the method comprises the steps that an augmentation epsilon-constraint method is used for optimizing another main target by taking a secondary target as a constraint condition, adjusting the value of an auxiliary variable epsilon within a certain range for solving, and calculating the value range of each target;

F ₁₁ ＝min{F ₁ (x):x∈S}；

F ₂₂ ＝min{F ₂ (x):x∈S}；

F ₁₂ ＝min{F ₂ (x):F ₁ (x)＝F ₁₁ ，x∈S}；

F ₂₁ ＝min{F ₁ (x):F ₂ (x)＝F ₂₂ ，x∈S}；

in the formula, F ₁₁ And F ₂₂ All minimum values under a single target, only considering F ₁ (x) Or F ₂ (x) The minimum value of time; f ₁₂ To minimize the risk with minimal cost per target, F ₂₁ To minimize costs with minimal risk single objectives.

Selecting a cost target as a main target, a risk target as a secondary target and a constraint to divide the range into p equal intervals, and converting the multi-objective optimization problem into a single-objective optimization problem by combining an auxiliary variable epsilon and a relaxation variable s;

ε＝lb+(k+r)/p，k＝0,1,...,p；

s.t.F ₂ (x)+s＝ε,s∈R ⁺ ；

where lb is the minimum value of the risk target, p is the number of intervals the risk target is divided into, r is the range of the risk target, α is a sufficiently small number, and s is the non-negative slack variable corresponding to the risk target.

Capacity power function of energy storage battery:

0.2W _ESS ≤P _ESS ≤W _ESS ；

in the formula, W _ESS 、P _ESS Indicating the configured capacity and rated power of the energy storage battery.

And (3) restraining the electric quantity and the charge-discharge power of the energy storage system:

the relationship between the energy storage capacity and the charge and discharge power is as follows:

the energy storage electric quantity range at the moment t is as follows:

the electric quantity is equal at the beginning and the end in the operation period:

in the formula (I), the compound is shown in the specification,

η、

D _max 、D _min the electric quantity, the charge-discharge efficiency, the charge-discharge power and the maximum charge-discharge depth of the energy storage system at the t moment and the t-1 moment are respectively.

In the formula, P _ESS Rated power, u, allocated to the energy storage system _t The variable is 0-1, and only charging can be carried out when the value is 1, and only discharging can be carried out when the value is 0.

Decoupling by a big-M method:

in the formula, M is a positive number which is set to be large enough, and the decoupling of the nonlinear constraint is realized.

And power balance constraint:

and (3) power exchange constraint with a power grid:

in the formula, P _max For the maximum value of the power exchanged between the optical storage charging station and the power grid, u _e The variable is 0-1, and the electricity can be purchased only from the power grid when the value is 1, and can be sold only from the power grid when the value is 0.

CVaR risk constraint:

z ^se ≥0；

the risk value of the CVaR risk constraint metric is considered for the expected cost of each scenario, the high cost risk value that the investment maintenance and running costs describing the light storage capacity configuration present in the face of deterministic scenarios consisting of uncertainty. And after the first step to the third step, converting the planning model into a mixed integer linear planning model, and solving the model by calling a Gurobi solver.

Further, in the fourth step, in the conventional linear weighting method, the confidence level α =0.9, and the risk preference coefficient is increased by a value of 0.05; in the augmented epsilon-constraint method, if the interval p =20, the number of solution sets is 21.

Another object of the present invention is to provide a cost and risk multiple objective based optical storage charging station capacity allocation system applying the cost and risk multiple objective based optical storage charging station capacity allocation method, the cost and risk multiple objective based optical storage charging station capacity allocation system comprising:

the optimal configuration model building module is used for building a multi-target capacity optimal configuration model of the optical storage charging station with cost and risk; the risk function is investment maintenance and operation cost risk quantified by condition risk value, and the cost function is the sum of expected investment maintenance and operation cost under a typical scene considering charging demand and photovoltaic output;

the system comprises an amplification epsilon-constraint module, a risk secondary target and a risk analysis module, wherein the amplification epsilon-constraint module is used for combining an amplification epsilon-constraint method, taking cost as a main target and taking a risk secondary target as a constraint;

and the objective decision scheme deleting module is used for solving the model to obtain pareto fronts of cost and risk under different risk preferences and corresponding configuration capacity, and screening the objective decision scheme by adopting an entropy weight-TOPSIS method.

Another object of the present invention is to provide a computer device comprising a memory and a processor, the memory storing a computer program which, when executed by the processor, causes the processor to perform the steps of the cost and risk multi-objective based optical storage charging station capacity allocation method.

Another object of the present invention is to provide a computer readable storage medium, storing a computer program which, when executed by a processor, causes the processor to perform the steps of the cost and risk multi-objective based optical storage charging station capacity allocation method.

Another object of the present invention is to provide an information data processing terminal for implementing the cost and risk multi-objective based optical storage charging station capacity configuration system.

By combining the technical scheme and the technical problem to be solved, the technical scheme to be protected by the invention has the advantages and positive effects that:

the uncertainty of the charging requirement and the photovoltaic output of the electric automobile causes certain uncertainty of the investment maintenance and operation cost of the optical storage charging station. In order to quantify the risk brought by the uncertainty to the optical storage charging station, namely the risk exceeding the expected cost, the invention establishes an optical storage charging station multi-target capacity optimization configuration model containing cost and risk, wherein a risk function in the model is investment maintenance and operation cost risk quantified by a conditional risk value, a cost function is the sum of the expected investment maintenance and operation cost under a typical scene considering the charging demand and the photovoltaic output, an augmentation epsilon-constraint method is combined to take the cost as a main target, and a risk secondary target is taken as a constraint; and (4) solving the model to obtain pareto fronts of cost and risk under different risk preferences and corresponding configuration capacity, and screening out an objective decision scheme by adopting an entropy weight-TOPSIS method. Compared with a linear weighting method in traditional risk management, the method has the advantages that after the targets are processed by the epsilon-constraint method, the front edge with better distribution and boundary optimality can be obtained, the actual pareto front edge of a multi-target problem can be mapped, and an investment scheme with more detailed cost and risk division is provided.

The invention provides a cost and risk multi-objective-based capacity allocation method for an optical storage charging station. The invention adopts an amplification epsilon-constraint method to process the multi-objective optimization problem, provides investment schemes with different costs and risk combinations for investors, and screens out objective decision schemes for the investors through an entropy weight-TOPSIS method. In order to quantify the cost risk, the invention establishes a multi-target capacity configuration model for CVaR to evaluate the cost risk, provides an augmentation epsilon-constraint method to process the multi-target model containing the cost and the risk, optimizes the capacity configuration of the optical storage charging station, quantifies the risk of the investor exceeding the expected cost, and provides an objective planning scheme and an operation strategy for the investor.

The method comprises the steps of measuring investment maintenance and operation cost risks brought to the photovoltaic output and EV charging demand uncertainty, visually displaying the relation between the risks and the costs, and providing investment and operation schemes for investors with different risk preferences, wherein the higher the cost is, the smaller the risk is. The invention quantifies the cost risk brought by uncertainty to the optical storage charging station, establishes a multi-target capacity optimization configuration model of cost and risk, and establishes a method for processing multiple targets by an augmented epsilon-constraint method. Compared with the traditional linear weighting method for processing the risk items, the pareto effective front edge obtained by the method of the invention by adopting the amplification epsilon-constraint method is more uniform in distribution and more excellent in boundary points; investment and operation schemes with different risk preferences are divided more carefully, and investment traders can conveniently control risks and cost. In different investment operation schemes, investors can subjectively select decision schemes according to risk preference; or comprehensively evaluating each scheme by adopting an entropy weight-TOPSIS method, screening out objective decision schemes, and obtaining the optimal balance of cost and risk.

The expected income and commercial value after the technical scheme of the invention is converted are as follows: the influence of uncertainty of photovoltaic output and charging requirement of the electric vehicle on capacity configuration and operation of the optical storage system is considered, cost risk values caused by the influence are quantized, investment operation schemes with different risks and costs are presented for optical storage charging station investors, and investment schemes of different types of investors are convenient to select.

The technical scheme of the invention fills the technical blank in the industry at home and abroad: and establishing a cost and risk multi-target model by adopting the CVaR quantitative risk. Compared with a traditional CVaR risk planning operation model, a solution set obtained by processing a multi-target problem by combining an amplification epsilon-constraint method has a more selective gradient, namely the obtained planning operation scheme is more finely regulated; and the solution set boundary is optimal, namely the cost of the operation plan is lower when only the risk single target is considered, and the risk of the operation plan is lower when only the cost single target is considered.

The technical scheme of the invention solves the technical problem that people are eagerly to solve but can not be successfully solved all the time: firstly, the specific risk brought by uncertainty is quantified for investors, and the risk is quantified into economic indexes. Second, different investment operation schemes are provided for investors with different investment types, and the investors can be enabled to negotiate a combination scheme of cost and risk. Thirdly, a model containing a risk target is optimized, and the problems that the solution set obtained by a traditional CVaR risk planning operation model is poor in distributivity and the boundary point risk is too high are solved, namely the obtained planning operation scheme is still the same when different risk preferences are met, or the cost is too high when only a risk single target is considered and the risk is too high when only the cost single target is considered.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the embodiments of the present invention will be briefly described below, and it is obvious that the drawings described below are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without creative efforts.

Fig. 1 is a flowchart of a capacity allocation method for an optical storage charging station based on cost and risk multiple targets according to an embodiment of the present invention;

fig. 2 is a schematic diagram of a capacity allocation method for an optical storage charging station based on cost and risk multiple targets according to an embodiment of the present invention;

FIG. 3 is a graph illustrating typical four season photovoltaic output percentages provided by embodiments of the present invention;

FIG. 4 is a flow chart of a method for processing cost and risk multiple targets by the augmented ε -constraint method provided by an embodiment of the present invention;

FIG. 5 is a graph of an effective leading edge comparison of a linear weighting method and an augmented ε -constrained method provided by an embodiment of the present invention;

FIG. 6 is a schematic diagram of the photovoltaic capacity and energy storage capacity power configuration results provided by embodiments of the present invention;

FIG. 7 is a schematic diagram of the total EV charging power after clustering according to an embodiment of the present invention;

fig. 8 is a schematic diagram of EV charging electricity price and purchasing and selling electricity price provided by an embodiment of the present invention;

fig. 9 is a schematic diagram of a pareto solution set comprehensive evaluation value provided by the embodiment of the present invention;

FIG. 10 is a schematic diagram of photovoltaic output and EV charging requirements for a 13 th scenario provided by an embodiment of the present invention;

FIG. 11 is a diagram illustrating the operation of scenario 13 in an objective decision making scheme according to an embodiment of the present invention;

FIG. 12 is a schematic diagram of the operation of the 13 th scenario in an investment scenario considering only cost according to an embodiment of the present invention;

fig. 13 is a schematic diagram of the operation of the 13 th scenario in the investment scenario only considering the risk according to the embodiment of the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail with reference to the following embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

In order to solve the problems in the prior art, the invention provides a method and a system for configuring capacity of an optical storage charging station based on cost and risk multiple targets, and the invention is described in detail below with reference to the accompanying drawings.

This section is an explanatory embodiment expanding on the claims so as to fully understand how the present invention is embodied by those skilled in the art.

As shown in fig. 1, the method for configuring capacity of an optical storage charging station based on cost and risk multiple targets provided by the embodiment of the present invention includes the following steps:

s101, establishing a multi-target capacity optimal configuration model of the optical storage charging station with cost and risk;

s102, combining an amplification epsilon-constraint method, taking cost as a main target, and taking a risk secondary target as a constraint;

s103, solving the model to obtain pareto fronts of cost and risk under different risk preferences and corresponding configuration capacity, and screening out an objective decision scheme by adopting an entropy weight-TOPSIS method.

The risk function in the cost and risk-containing optical storage charging station multi-target capacity optimal configuration model is investment maintenance and operation cost risk quantified by condition risk value, and the cost function is the sum of expected investment maintenance and operation cost under a typical scene considering charging demand and photovoltaic output.

As a preferred embodiment, as shown in fig. 2, the method for configuring capacity of an optical storage charging station based on multiple objectives of cost and risk provided in the embodiment of the present invention specifically includes the following steps:

the method comprises the following steps of S1, obtaining a charging demand scene of the electric automobile by adopting a Monte Carlo sampling method, and obtaining a photovoltaic power generation scene by utilizing a four-season typical photovoltaic output; and (4) constructing an uncertainty function of the charging requirement and the photovoltaic output of the electric automobile by combining a scene method.

S2, establishing an investment maintenance cost function, an operation cost function and a CVaR risk measurement function of the optical storage system; the investment and maintenance cost of the optical storage system comprises equal-annual-value initial investment cost and annual operation and maintenance cost. The cost of the whole planning operation stage is the sum of the investment maintenance cost and the operation cost of the light storage system, the value of the risk represents the high-cost risk caused by uncertainty, and the value is quantified by using a CVaR theory.

And S3, processing multiple targets by combining an amplification epsilon-constraint method according to the target of minimizing cost and risk, and establishing a cost and risk multiple-target optical storage charging station capacity optimal configuration model.

And S4, solving the converted mixed integer linear programming model by using a Gurobi solver, and obtaining the capacity programming result of the optical storage charging station and the corresponding optimized operation strategy under different risk values after the solution is finished.

And step S1, converting the uncertainty problem into the certainty problem.

And S2, quantifying cost risks brought to planned operation of the optical storage charging station by the charging requirements of the electric vehicle and the uncertainty of the photovoltaic output.

Step S3, the function is: and establishing a planning operation model of the augmentation epsilon-constraint method for processing cost and risk multiple targets, wherein the planning operation scheme provided for investors can have a more detailed risk regulation gradient.

Step S4, the function is: and solving the converted mixed integer linear programming model by adopting a solver to obtain a programming operation scheme.

In step S1 provided by the embodiment of the present invention, a process of converting the uncertainty problem of the charging demand and the photovoltaic output of the electric vehicle into a scene to be researched is as follows:

the charging requirement of the electric automobile is determined by initial charging time and initial charging SOC, the initial charging SOC of the electric automobile approximately obeys the lognormal distribution, and the initial charging time approximately obeys the normal distribution.

and

initial SO for charging electric vehicleThe mean and standard deviation of the logarithm of the C variable;

and

and charging the average value and the standard deviation of the logarithm of the initial SOC variable of the electric automobile. Sampling and sampling the initial charging time and the initial SOC state of the electric automobile by using a Monte Carlo method to obtain the charging requirement of the electric automobile, and obtaining the number of charging scenes required by the electric automobile by combining with kmeans clustering.

P _PV (t)＝p _er (t)P _PV (3)

In the formula, P _PV (t)、P _PV 、p _er (t) photovoltaic output power at the time t, photovoltaic output percentage at the time t and photovoltaic configuration capacity. The typical four season photovoltaic output percentage is shown in fig. 3. The uncertainty of photovoltaic output and the charging requirement of the electric automobile is processed by adopting a scene method, various possible conditions are simulated through a large number of scenes, and finally the random planning problem is converted into the deterministic planning problem. Setting photovoltaic output scene set s = { s = } _i ,i＝1,2,...,n _s Electric vehicle charging demand scene set e = { e = } _j ,i＝1,2,...,n _e In which n is _s And e _j The total scene number of the photovoltaic output and the electric automobile charging requirement is respectively. In the analysis that follows, se in all alphabetical superscripts indicates that the value is at the s-th _i Individual photovoltaic output scene, e _j In the scene of charging requirement of the electric automobile.

In step S2 provided in the embodiment of the present invention, a process of describing a cost function and a risk measurement function when capacity of an optical storage system of an optical storage charging station is configured is as follows:

the investment and maintenance cost of the optical storage system comprises two aspects of equal-annual-value initial investment cost and annual operation and maintenance cost.

C _inv ＝(C _PV P _PV +C _ESS,W W _ESS +C _ESS,P P _ESS )C _RF (4)

C _cost ＝C _inv +C _OM (7)

In the formula, C _inv 、C _PV 、C _ESS,W 、C _ESS,P 、C _RF Respectively representing the equal-annual-value investment cost of the photovoltaic and energy storage system, the unit capacity investment cost of the photovoltaic and energy storage system, the unit power investment cost of the energy storage system and the equal-annual-value investment coefficient; and r and m respectively represent the discount rate and the service life of the corresponding system. C _OM 、

C _cost And respectively representing the annual maintenance cost of the light storage system, the annual maintenance cost of the unit capacity of the photovoltaic system and the energy storage system and the investment maintenance cost of the light storage system. R is ^se 、

The annual running cost of the scene se, the power purchased and sold from the optical storage charging station to the power grid and the charging power of the electric automobile are respectively calculated; a. b and c are the electricity purchase and sale price of the light storage charging station to the power grid and the charging price of the electric vehicle respectively; t is the number of operating hours (24 hours is taken in the invention); pi(s) _i )、π(e _j ) Are respectively the s _i Individual photovoltaic output scene, e _j Probability of individual electric vehicle charging demand scenarios.

Wherein alpha is the confidence level, and the potential maximum loss risk at the confidence level alpha is C _VaR And C is _CVaR Is in excess of C _VaR Average loss of part, i.e. cost risk in the present invention, z ^se Are virtual variables. To overcome this ambiguity, a CVaR quantified cost risk value is introduced, and equation (9) characterizes the risks associated with the expected investment maintenance and operating costs, the specific risk items of which are the CVaR risk constraints in step S3.

In step S3 provided in the embodiment of the present invention, an augmented epsilon-constraint method is constructed to process a cost and risk multi-target model, and a process of solving a multi-target planning problem of an optical storage charging station is as follows:

min{F ₂ (x)＝C _CVaR } (11)

equations (10) and (11) represent the objective functions for the minimum cost and the minimum risk individually. The objective function of the CVaR programming problem of the conventional linear weighting method is introduced as follows:

min{(1-β)F ₁ (x)+βF ₂ (x)} (12)

the formula (12) is an objective function form for researching the CVaR planning problem containing the risk items at present, and different planning schemes are finally obtained by artificially regulating and controlling the risk weight through a weighting factor. Wherein: β is a weighting factor in the range of [0,1] that is used to achieve a tradeoff between cost and risk, i.e., a risk preference factor. Different investment schemes can be obtained by changing the parameter beta, and effective leading edges of cost and risk are constructed. The larger beta indicates a more significant risk, and such investors are risk evasive, i.e. wish to minimize the risk as much as possible; the smaller beta indicates a more neglect of risk, i.e. such investors are risk-chasing, i.e. wish to minimize costs as much as possible. However, pareto front edge distribution and boundary optimality obtained by a multi-target model constructed by a traditional linear weighting method are poor. That is, when an investor makes a selection balance based on the cost and risk value of each investment scheme, the gradient of the investment scheme is not easy to be regulated, the risk is too large when only the cost is considered, and the cost is too large when only the risk is considered.

The process of processing cost and risk multiple targets by the amplification epsilon-constraint method provided by the invention comprises the following steps:

the idea of the augmented epsilon-constraint method is to optimize another main target by taking a secondary target as a constraint condition, adjust the value of an auxiliary variable epsilon within a certain range and solve the problem, and a flow chart is shown in fig. 4. First, the value range of each target is calculated.

F ₁₁ ＝min{F ₁ (x):x∈S} (13)

F ₂₂ ＝min{F ₂ (x):x∈S} (14)

F ₁₂ ＝min{F ₂ (x):F ₁ (x)＝F ₁₁ ，x∈S} (15)

F ₂₁ ＝min{F ₁ (x):F ₂ (x)＝F ₂₂ ，x∈S} (16)

In the formula, F ₁₁ And F ₂₂ All minimum values under a single target, i.e. considering only F ₁ (x) Or F ₂ (x) Minimum value of time, F ₁₂ To minimize the risk with minimal cost per target, F ₂₁ To minimize costs with minimal risk single objectives. And at the moment, selecting a cost target as a main target, taking a risk target as a secondary target and constraining the range of the risk target into p equal intervals, and converting the multi-objective optimization problem into a single-objective optimization problem by combining an auxiliary variable epsilon and a slack variable s.

ε＝lb+(k+r)/p，k＝0,1,...,p (17)

s.t.F ₂ (x)+s＝ε,s∈R ⁺ (19)

Where lb is the minimum value of the risk target, p is the number of intervals the risk target is divided into, r is the range of the risk target, α is a sufficiently small number, and s is the non-negative slack variable corresponding to the risk target. The above is the construction of cost and risk multi-objective functions, followed by the listing of relevant constraints and variables.

Capacity power function of energy storage battery:

0.2W _ESS ≤P _ESS ≤W _ESS (20)

in the formula, W _ESS 、P _ESS The capacity and rated power of the configured energy storage battery are shown, and the relation between the energy storage capacity and the power is limited by the formula.

And (3) restraining the electric quantity and the charge and discharge power of the energy storage system:

in the formula (I), the compound is shown in the specification,

η、

D _max 、D _min the energy storage system comprises the following components of t and t-1 time electric quantity, charge and discharge efficiency, charge and discharge power and maximum charge and discharge depth of the energy storage system respectively; the formula (21) is the relation between the energy storage capacity and the charge and discharge power, the formula (22) limits the range of the energy storage capacity at the time t, and the formula (23) ensures that the starting and ending capacities are equal in one operation period (24 h in the invention).

In the formula, P _ESS Rated power, u, allocated to the energy storage system _t Is a variable of 0 to 1, and is only chargeable when the value is 1When the value is 0, only discharge can be realized; equations (24) and (25) define the range of the charging and discharging power of the energy storage system and ensure that the energy storage system is not charged and discharged simultaneously.

Due to the introduction of the variable u of 0 to 1 in the formulae (24) and (25) _t And P is _ESS Is also a decision variable, leading to the appearance of non-linear constraints. Thus, the big-M method is used to decouple formulae (24) and (25).

In the formula, M is a positive number which is set to be large enough, and the decoupling of the nonlinear constraint is realized. At this time, the constraints (24) to (25) are converted into constraints (26) to (29).

And power balance constraint:

and power exchange constraint with the power grid:

in the formula, P _max For the maximum value of the power exchanged between the optical storage charging station and the power grid, u _e Is a variable of 0 to 1, and only can purchase electricity from a power grid when the value is 1And when the value is 0, electricity can be sold to the power grid only, so that the condition that the light storage charging station does not purchase electricity to the power grid at the same time is ensured.

CVaR risk constraint:

z ^se ≥0 (33)

the risk value of the metric of equation (34) is considered for the expected cost of each scenario, describing the high cost risk value that the investment maintenance and operating costs of the light storage capacity configuration presents in the face of a deterministic scenario consisting of uncertainty. Through the steps S1, S2 and S3, the planning model provided by the invention is converted into a mixed integer linear planning model, so that a Gurobi solver solving model can be called.

In step S4 provided in the embodiment of the present invention, in the conventional linear weighting method, the confidence level α =0.9, and the risk preference coefficient is increased by a value of 0.05, while in the augmented epsilon-constraint method, at an interval p =20, the solution sets are all 21, and the simulation result obtained after the solution is as follows.

A comparison of the linear weighting method and the augmented epsilon-constraint solution set is shown in fig. 5.

As can be seen from the pareto effective frontier in fig. 5, as the solution centralization investment maintenance and operation costs decrease, the corresponding risk will increase. The cost is negative, the light storage charging station is in a profit state in actual investment operation, and the risk is negative, and the minimum income which can be obtained after operation is indicated. The actual physical significance of cost, risk, benefit and sign is explained as follows: the cost corresponding to point A of FIG. 5 is-281085 yuan, which indicates that the expected profit of the investor is 281085 yuan; risk-141492 Yuan, and gain not less than 141492 Yuan at 90% confidence level.

On the effective frontier, along with the increase of risks and the reduction of cost, the investment trend of risk pursuing type investors is shown to pay attention to the cost neglect risks; with the increase of the cost and the reduction of the risk, the method indicates that the risk neglect cost is emphasized and is the investment trend of a risk-evading investor. And the uniformly distributed weights can be seen from the effective front edge obtained by the linear weighting methodThe set of weights beta does not guarantee a valid solution set { F } ₁ ,F ₂ The mapping of pareto active sets is therefore not sufficient and different weight combinations will yield the same active solution, for example: when β =0.9,0.95 or β = 0.05. In this case, it can be seen that: the pareto effective frontier obtained by the method of the augmentation epsilon-constraint is better in distribution, and an investment scheme which is easier to regulate and control cost and risk solution set gradient can be provided.

The effective solution sets found by the two methods are not comparable, since the result of the two methods is two different mappings of the same pareto boundary. However, at the boundary points B, β =1 and a, β =0, it is possible to compare, and at the boundary between the two, the schemes B and a dominate the schemes β =1 and β =0, respectively. For the upper left border, this time equivalent to the same risk, while solution B has a lower cost compared to the solution β = 1; for the lower right border, this time corresponding to the same cost, scheme a has a lower risk than the scheme β =0. In conclusion, the superiority of the extended ε -constraint method was demonstrated.

Cost and risk and different capacity configuration analysis: the capacity configuration of the solution set solved based on the augmented epsilon-constraint method is analyzed, and the photovoltaic capacity and energy storage capacity power configuration results are shown in fig. 6.

Because the uncertainty of photovoltaic output and the charging requirement of the electric automobile influences the planning result, after the uncertainty is processed by adopting a scene method, the optimization result with the minimum investment maintenance and operation cost is actually a probability sum of all scene costs, namely an expected cost value. In order to overcome the ambiguity, the CVaR introduced by the invention can quantify the risk of the cost, and the investor can make subjective decision by autonomously balancing the relationship between the cost and the risk during investment.

Uncertainty of photovoltaic output and electric vehicle charging requirements can bring fluctuation to the operation cost of each scene, operation cost risks are generated, and capacity configuration of photovoltaic energy storage is further influenced, so that investment and maintenance cost risks are brought. From the figure, it can be seen that: in the process of making more attention to risks and investing, investors can reduce the capacity allocation of photovoltaic and energy storage, but the proportion of the energy storage to the photovoltaic capacity can be increased. Reducing the capacity allocation can reduce the investment maintenance cost risk to a certain extent, but can improve the operating cost to improve the operating cost risk, so the change of the energy storage capacity is not large, thus it can be ensured that the system cost risk reduced by reducing the investment can be larger than the operating cost risk improved by reducing the investment, i.e. the total risk can be in a downward trend. In the process, the electricity is purchased from the power grid to meet the charging requirement of the electric automobile, the photovoltaic capacity stabilizing fluctuation is reduced, and the electricity selling to the power grid is reduced.

On the contrary, when the investors pay attention to the risk of cost neglect, the investors can invest a large amount of photovoltaic and energy storage capacity, although the investment and maintenance cost is improved, the operation cost is reduced by reducing the electricity purchasing and increasing the electricity selling to the power grid after the charging requirement of the electric automobile is met, and the reduced operation cost is more than the improved investment and maintenance cost, so the cost is reduced. However, due to the increase of investment and the large photovoltaic fluctuation, the risk of uncertainty of investors is further increased.

The invention quantifies the cost risk brought by uncertainty to the optical storage charging station, establishes a multi-target capacity optimization configuration model of cost and risk, and establishes a method for processing multiple targets by an augmented epsilon-constraint method.

The method comprises the steps of measuring investment maintenance and operation cost risks brought to a light storage charging station by uncertainty of photovoltaic output and electric vehicle charging requirements, visually displaying the relation between the risks and the costs, and providing investment and operation schemes for investors with different risk preferences as the costs are higher and the risks are smaller. The pareto effective front edge obtained by the method of the invention by adopting the amplification epsilon-constraint method is more uniform in distribution and more excellent in boundary point. Investment and operation schemes with different risk preferences are divided more carefully, and investment traders can conveniently control risks and costs.

The capacity configuration system of the optical storage charging station based on cost and risk multiple targets provided by the embodiment of the invention comprises:

In order to prove the creativity and the technical value of the technical scheme of the invention, the part is the application example of the technical scheme of the claims on specific products or related technologies.

The technical scheme of the invention carries out example simulation analysis.

1 example parameters

Calling a Gurobi solver in Matlab to solve the models established in 3.1 and 3.2. The typical photovoltaic four-season output scene probabilities are all 0.25, charging requirements of the EVs are obtained by adopting Monte Carlo sampling, the EV charging requirements are grouped into four types by using kmeans, as shown in FIG. 7, and the probabilities of the four types of charging power after clustering are 0.238, 0.262 and 0.238 respectively. From a time perspective, the optical storage charging stations all use the fast charging piles to provide charging services for the EVs, and it is assumed that the situation of queuing up for charging does not occur, that is, the EVs are charged immediately after arriving at the optical storage charging stations. Fig. 8 shows EV charging power rates a, light storage charging stations purchasing power from the grid, and power rates b, c. The maximum power exchange between the optical storage charging station and the power grid is 200kW. The initial charge of the energy storage battery is 50% of the total capacity of the energy storage battery. In the setting of the three prices, the charging price of the EV is larger than the electricity purchasing price and is far larger than the electricity selling price, so that the light storage charging station is ensured to provide charging service for the EV as much as possible, and the electric quantity is not sold to the power grid.

2 example results and analysis

2.1 comparison of Linear weighting with augmented ε -constraint solution sets

To compare the advantages and disadvantages of the multi-objective optimization problem processing method with risk management, the respective effective fronts are drawn as shown in fig. 5. In the conventional linear weighting method, the confidence level α =0.9, the risk preference coefficient is incremented by a value of 0.05, whereas in the augmented epsilon-constraint method, the interval p =20, the solution set number is 21.

As can be seen from the pareto effective frontier of fig. 5, as the solution centralization investment maintenance and operation costs decrease, the corresponding risk will increase. The cost is negative, the light storage charging station is in a profit state in actual investment operation, and the risk is negative, and the minimum income which can be obtained after operation is indicated. The actual physical significance of cost, risk, benefit and sign is explained as follows: the cost corresponding to point A in FIG. 5 is-281085 yuan, which indicates that the expected profit of the investor is 281085 yuan; risk-141492 Yuan, and yield not less than 141492 Yuan at 90% confidence level.

On the effective frontier, along with the increase of risks and the reduction of cost, the investment trend of risk pursuing type investors is shown to pay attention to the cost neglect risks; with the increase of the cost and the reduction of the risk, the investment trend of risk-evasive investors is shown to pay attention to the risk neglect cost. And as can be seen from the effective frontier obtained by the linear weighting method, the uniformly distributed weight coefficient set beta can not ensure the effective solution set { F } ₁ ,F ₂ The mapping of pareto active sets is therefore not sufficient and different weight combinations will yield the same active solution, for example: when β =0.9,0.95 or β = 0.05. Here, it can be known that: the pareto effective frontier obtained by the method of the enlargement epsilon-constraint has better distribution, and an investment scheme which is easier to regulate and control cost and risk solution set gradient can be provided.

The effective solution sets found by the two methods are not comparable, since the result of the two methods is two different mappings of the same pareto boundary. However, at the boundary points B, β =1 and a, β =0, it is possible to compare, and at the boundary between the two, the schemes B and a dominate the schemes β =1 and β =0, respectively. For the upper left border, this time corresponding to the same risk, solution B has a lower cost than the solution with β = 1; for the lower right border, this time corresponding to the same cost, scheme a has a lower risk than the scheme β =0. In conclusion, the superiority of the extended ε -constraint method was demonstrated.

2.2 cost and risk and different Capacity configuration analysis

The invention analyzes the capacity configuration of the solution set based on the augmentation epsilon-constraint method, and the photovoltaic capacity and energy storage capacity power configuration results are shown in figure 6.

Because the uncertainty of photovoltaic output and EV charging requirements influences the planning result, after the uncertainty is processed by adopting a scene method, the optimization result with the minimum investment maintenance and operation cost is actually a probability sum of all scene costs, namely an expected cost value. In order to overcome the ambiguity, the CVaR introduced by the invention can quantify the risk of the cost, and the investor can make subjective decision by autonomously balancing the relationship between the cost and the risk during investment.

Uncertainty of photovoltaic output and EV charging requirements can bring fluctuation to the operation cost of each scene, generate operation cost risks, and further influence capacity allocation of photovoltaic energy storage, so that investment and maintenance cost risks are brought. From fig. 6, it can be seen that: in the process of making more attention to risks and investing, investors can reduce the capacity allocation of photovoltaic and energy storage, but the proportion of the energy storage to the photovoltaic capacity can be increased. Reducing the capacity allocation can reduce the investment maintenance cost risk to a certain extent, but can improve the operating cost to improve the operating cost risk, so the change of the energy storage capacity is not large, thus it can be ensured that the system cost risk reduced by reducing the investment can be larger than the operating cost risk improved by reducing the investment, i.e. the total risk can be in a downward trend. In the process, the electricity is purchased from the power grid to meet the charging requirement of the EV, the photovoltaic capacity stabilizing fluctuation is reduced, and the electricity selling to the power grid is reduced.

On the contrary, when the investors pay attention to the risk of cost neglect, the investors can invest a large amount of photovoltaic and energy storage capacity, although the investment and maintenance cost is improved, after the EV charging requirement is met, the number of electricity purchasing and electricity selling to the power grid is reduced, the operation cost is reduced, and the reduced operation cost is larger than the improved investment and maintenance cost, so that the cost is reduced. However, due to the increase of investment and the large photovoltaic fluctuation, the risk of uncertainty of investors is further increased.

2.3 Objective decision making scheme based on entropy weight-TOPSIS method screening

The weights of the cost and risk targets in the solution set of the augmented epsilon-constraint method in fig. 5 are first calculated by the entropy weight method: w ₁ ＝0.5461、W ₂ =0.4539. The overall evaluation value of each solution set obtained by the toposis method is shown in fig. 9.

The solution set sequence numbers from small to large in fig. 6 correspond one-to-one to the solution sets from right to left of the risk values in fig. 5, and correspond one-to-one to the capacity configurations in which the risk values gradually decrease from left to right in fig. 6. The larger the comprehensive evaluation value is, the better the solution set is, the maximum comprehensive evaluation value is 0.6176 is in the 7 th solution set in fig. 9, and the solution set is an objective decision scheme; referring to fig. 9, the result of configuring the capacity of each device in the objective decision scheme is as follows: p _PV ＝265kW、W _ESS ＝352kWh、P _ESS =145kW, cost-271521 Yuan, risk-148031 Yuan.

2.4 comparison of Objective decision scheme with typical scheme operating results

In actual investment, there are two typical schemes that are only minimum with a single target of cost or risk according to the subjective intention of the investor, and the 1 st solution set and the 21 st solution set in fig. 9 represent the two typical schemes respectively. The 1 st solution set investment results are: p _PV ＝315kW、W _ESS ＝354kWh、P _ESS =146kW, cost-281085 Yuan, risk-141492 Yuan, investment result for the 21 st solution set is: p _PV ＝112kW、W _ESS ＝272kWh、P _ESS =112kW, cost-234071 Yuan, risk-163290 Yuan. The cost is the desired cost, including the running cost, which is made up of individual scenarios, while the actual cost is related to the scenario probability, and the quantified risk value characterizes the maximum cost that may be encountered. The invention has 16 scenes, the cost of the 13 th scene of each scheme in the simulation result exceeds the risk value, and the cost of the scene exceeds the expected cost. Photovoltaic output of 13 th sceneThe force and EV charging demand are shown in fig. 10, and the operation of the objective decision scheme, the investment scheme considering only the cost, and the investment scheme considering only the risk in this scenario are shown in fig. 11, fig. 12, and fig. 13.

Since the risk is excessively avoided by the investment scheme considering only the risk, the capacity configuration of the photovoltaic and the energy storage is far lower than that of the objective decision scheme and the investment scheme considering only the cost, as can be seen from comparison of fig. 11 and fig. 12 and 13: in the investment scenario where only the risk is considered, the capacity due to the deployed photovoltaic and energy storage is low. When the photovoltaic power generation capacity is weak (1 to 7 periods 00), the scheme can only meet the charging demand of the EV by purchasing power to the grid. When the photovoltaic power generation capacity is strong (11-00 period, 15), the scheme can not only obtain the income for selling the electricity to the power grid, but also can only store a small amount of electric energy. Thus, excessive conservative investments in this approach, while reducing risks, greatly increase operating costs.

Objective decision schemes compare to investment schemes that only consider cost: the photovoltaic capacity configuration is reduced, and the energy storage capacity is basically unchanged. The objective decision scheme reduces the photovoltaic capacity allocation, so it can be seen in the comparison of fig. 11 and 12 that the objective decision scheme reduces the electricity sales to the grid, while increasing the operating cost, reducing the photovoltaic capacity allocation reduces the investment cost. The investment maintenance and operating cost risk in the risk item at this time can be interpreted as: the reduction in risk value resulting from reduced investment and maintenance costs is greater than the increase in risk value resulting from increased operating costs. Thus, the present invention reduces the overall cost risk.

In conclusion, compared with an investment scheme only considering cost, the objective decision scheme reduces the risk by 4.6% on the premise of only improving the cost by 3.4%; compared with the investment scheme only considering the risk, the optimal investment scheme has the advantage that the cost is reduced by 16.0% on the premise of only improving the risk by 9.3%.

The embodiment of the invention achieves some positive effects in the process of research and development or use, and has great advantages compared with the prior art, and the following contents are described by combining data, diagrams and the like in the test process.

EV charging demand and optical storage charging station function

1.1EV uncertainty charging demand function

The charging requirement of the EV is determined by the initial charging time and the initial charging SOC, the initial charging SOC of the EV approximately follows the lognormal distribution, and the initial charging time approximately follows the normal distribution.

In the formula, S _OC1 Charging the EV with an initial SOC; t is t ₁ Is the initial time of charging;

and

the mean value and the standard deviation of the logarithm of the initial SOC variable of the EV charge are obtained;

and

the mean and standard deviation of the EV charge initial SOC variable logarithm are obtained.

1.2 photovoltaic typical output function

P _PV (t)＝p _er (t)P _PV (3)

In the formula, P _PV (t)、P _PV 、p _er (t) photovoltaic output power at the time t, photovoltaic output percentage at the time t and photovoltaic configuration capacity. The photovoltaic four season typical output percentage is shown in figure 3.

1.3 energy storage Battery Capacity Power function

The energy storage battery can store electric energy in the low-price valley period or when the photovoltaic power generation power is greater than the EV charging power; and when the photovoltaic power generation power is lower than the EV charging power, electric energy is released to meet the charging requirement of the EV.

0.2W _ESS ≤P _ESS ≤W _ESS (4)

In the formula, W _ESS 、P _ESS The configured capacity and rated power of the energy storage battery are shown, and the relation between the energy storage capacity and the power is limited by the formula.

1.4 optical storage System investment maintenance cost function

C _inv ＝(C _PV P _PV +C _ESS,W W _ESS +C _ESS,P P _ESS )C _RF (5)

C _cost ＝C _inv +C _OM (8)

C _cost And respectively representing the annual maintenance cost of the light storage system, the annual maintenance cost of the unit capacity of the photovoltaic system and the energy storage system and the investment maintenance cost of the light storage system.

2. Light storage charging station multi-target capacity optimal configuration model considering CVaR measurement risk

2.1 Scenario Explanation

Multi-target capacity optimization configuration of optical storage charging station containing uncertain quantities such as photovoltaic output, EV charging demand and the likeThe problem belongs to a stochastic programming problem, uncertainty of photovoltaic output and EV charging requirements is processed by a scene method, various possible conditions are simulated through a large number of scenes, and the stochastic programming problem is finally converted into a deterministic programming problem. Setting photovoltaic output scene set s = { s = } _i ,i＝1,2,...,n _s }, EV charging demand scenario set e = { e _j ,i＝1,2,...,n _e }; wherein n is _s And e _j The total scene number of the photovoltaic output and the EV charging requirement is respectively. In the analysis that follows, se in all alphabetical superscripts indicates that the value is at the s-th _i Individual photovoltaic output scene, e _j In one EV charging demand scenario.

2.2 investment maintenance and operating cost objective function

Minimizing the investment maintenance and operating costs of optical storage charging stations:

in the formula, R ^se 、

The annual running cost of the scene se, the power purchased and sold from the optical storage charging station to the power grid and the charging power of the EV are respectively; a. b and c are the electricity purchase and sale price and the EV charging price of the light storage charging station to the power grid respectively; t is the number of operating hours (24 hours is taken in the invention); pi(s) _i )、π(e _j ) Are respectively the s _i Individual photovoltaic output scene, e _j Probability of individual EV charging demand scenarios.

2.3 CVaR risk metric objective function

Minimizing the risk faced by light storage charging stations:

min{F ₂ (x)＝C _CVaR } (12)

wherein alpha is the confidence level, and the potential maximum loss risk at the confidence level alpha is C _VaR And C is _CVaR Is in excess of C _VaR Average loss of part, i.e. cost risk in the present invention, z ^se Is a virtual variable. To overcome this ambiguity, introducing a CVaR quantified cost risk value, equation (11) characterizes the risk associated with the expected investment maintenance and operating costs, the specific risk terms of which are given in the CVaR risk constraint of 2.4.

2.4 constraints

2.4.1 energy storage system electric quantity and charge-discharge power constraint

In the formula (I), the compound is shown in the specification,

η、

D _max 、D _min the energy storage system comprises the following components of t and t-1 time electric quantity, charge and discharge efficiency, charge and discharge power and maximum charge and discharge depth of the energy storage system respectively; the formula (13) is the relation between the energy storage capacity and the charge and discharge power, the formula (14) limits the range of the energy storage capacity at the time t, and the formula (15) ensures that the starting and ending capacities are equal in one operation period (24 h in the invention).

In the formula, P _ESS Rated power, u, allocated to the energy storage system _t The variable is 0-1, the variable can be charged only when the value is 1, and can be discharged only when the value is 0; the formula (16) and the formula (17) limit the range of the charging and discharging power of the energy storage system and ensure that the energy storage system is not charged and discharged at the same time.

Due to the introduction of the variable u of 0 to 1 into the formulas (16) and (17) _t And P is _ESS Are also decision variables, leading to the appearance of non-linear constraints. Thus, the big-M method is used to decouple formulae (16) and (17).

In the equation, M is a sufficiently large positive number set to achieve decoupling of the nonlinear constraints, and in this case, the constraints (16) to (17) are converted into constraints (18) to (21).

2.4.2 Power balance constraints

2.4.3 Power exchange constraints with the grid

In the formula, P _max For the maximum value of the power exchanged between the optical storage charging station and the power grid, u _e The variable is 0-1, electricity can only be purchased from the power grid when the value is 1, and electricity can only be sold to the power grid when the value is 0, so that the condition that the optical storage charging station cannot purchase and sell electricity to the power grid at the same time is ensured.

2.4.4CVaR Risk constraint

For ease of solution, virtual variables have been introduced, and for ease of computation, the virtual variables are relaxed into the following two constraints.

z ^se ≥0 (25)

The risk value of the equation (26) metric is considered for the expected cost of each scenario, describing the high cost risk value that the investment maintenance and operating costs of the light storage capacity configuration presents in the face of a deterministic scenario consisting of uncertainty.

3. Multi-objective combination and solution with risk management

3.1 Linear weighting method

The model established by the invention not only considers the economic target of minimizing the total cost of the optical storage charging station, but also considers the risk target of minimizing the CVaR quantization risk. When the multi-objective optimization problem with risk management is solved, a linear weighting method is used for multiple times, risk preference coefficients are introduced to process risk items, and the multi-objective optimization problem is converted into a single-objective optimization problem.

min{(1-β)F ₁ (x)+βF ₂ (x)} (27)

s.t.x∈S (28)

The linear weighting method constructs a composite objective function (27) containing cost and risk. In the formula: β is a weighting factor in the range of [0,1] that is used to achieve a tradeoff between cost and risk, i.e., a risk preference factor. Different investment schemes can be obtained by changing the parameter beta, and effective leading edges of cost and risk are constructed. The larger beta indicates a more significant risk, and such investors are risk evasive, i.e. wish to minimize the risk as much as possible; smaller β indicates a more neglect of risk, i.e. such investors are risk-chasing, i.e. it is desirable to minimize costs as much as possible. S is a feasible set of decision variables x that satisfy equations (4), (13) - (26).

Solving the multi-target model constructed by the linear weighting method to obtain poor pareto leading edge distribution and boundary optimality. That is, when an investor makes a selection and balance based on the cost and risk value of each investment scheme, the gradient of the investment schemes is not easy to regulate, the risk is too large when only the cost is considered, and the cost is too large when only the risk is considered.

3.2 method of increasing Epsilon-restraint

In order to improve the distribution of the multi-target combination problem solution set constructed by the linear weighting method and the values of boundary points, an augmented epsilon-constraint method is provided to construct a multi-target model containing risk management, the actual pareto frontier of the multi-target optimization problem is mapped, and an investment scheme which is easier to regulate and control cost and gradient of the risk solution set and a solution set with more optimal boundary points are provided.

The idea of the augmentation epsilon-constraint method is to optimize another main target by taking a secondary target as a constraint condition, and adjust the value of an auxiliary variable epsilon within a certain range to solve. First, the value range of each objective, i.e. the value range of the cost and risk objectives in the present invention, is calculated, as shown in table 1.

TABLE 1 cost and Risk target value ranges

F in Table 1 ₁₁ And F ₂₂ All minimum under a single target, i.e. considering only F ₁ (x) Or F ₂ (x) Minimum value of time, F ₁₂ To minimize the risk with minimal cost per target, F ₂₁ To minimize the cost with the minimum risk single target, the calculation process is as follows:

F ₁₁ ＝min{F ₁ (x):x∈S} (29)

F ₂₂ ＝min{F ₂ (x):x∈S} (30)

F ₁₂ ＝min{F ₂ (x):F ₁ (x)＝F ₁₁ ，x∈S} (31)

F ₂₁ ＝min{F ₁ (x):F ₂ (x)＝F ₂₂ ，x∈S} (32)

the maximum and minimum values in each column of table 1 may determine the range of each target on the pareto frontier. And at the moment, selecting a cost target as a main target, taking a risk target as a secondary target and constraining the range of the risk target into p equal intervals, and converting the multi-objective optimization problem into a single-objective optimization problem by combining an auxiliary variable epsilon and a slack variable s.

ε＝lb+(k+r)/p，k＝0,1,...,p (33)

s.t.F ₂ (x)+s＝ε,s∈R ⁺ ,x∈S (35)

3.3 screening Objective optimal solution based on entropy weight-TOPSIS method

Solving the multi-target models constructed by 3.1 and 3.2 to obtain pareto frontiers, wherein investors can subjectively select a decision scheme according to the balance of the investors on cost and risk, and can also objectively determine the decision scheme by adopting an entropy weight-TOPSIS method. The core idea is to standardize the cost and the risk indexes, objectively weight the cost and the risk indexes based on the information entropy of index data, and quantify the relative distance between each solution set and a positive ideal solution and a negative ideal solution to serve as a comprehensive evaluation value.

First, using a range method to measure each index X _ij A normalization process is performed to eliminate the magnitude and dimension effects.

In the formula, i represents the serial number of the solution, and j represents the measure index, namely the cost or risk index; x _ij And Y _ij Respectively representing the cost value, the risk value and the normalized cost value and risk value; max (X) _ij ) And min (X) _ij ) Each represents the most significant value of cost or risk.

Second, each index Y is calculated _ij Information entropy E of _j 。

Thirdly, calculating each index Y _ij Weight W of _j 。

And fourthly, constructing a weighted evaluation matrix S of each index.

S＝(s _ij ) _n×m (39)

In the formula, n is the solution set number in pareto solution set, namely p +1 in the invention; m is the index number, s _ij ＝W _j ×Y _ij 。

Fifthly, determining the positive ideal solution k of each index according to the weighted evaluation matrix S ^* And negative ideal solution k ⁰ . Since both cost and risk are cost-type indicators, i.e., the smaller the better, the:

sixthly, calculating each solution set and the positive ideal solution

And negative ideal solution

Euclidean distance of

And

and seventhly, calculating a comprehensive evaluation value:

comprehensive evaluation value R _i The largest solution set is optimal.

It should be noted that the embodiments of the present invention can be realized by hardware, software, or a combination of software and hardware. The hardware portion may be implemented using dedicated logic; the software portions may be stored in a memory and executed by a suitable instruction execution system, such as a microprocessor or specially designed hardware. Those skilled in the art will appreciate that the apparatus and methods described above may be implemented using computer executable instructions and/or embodied in processor control code, such code being provided on a carrier medium such as a disk, CD-or DVD-ROM, programmable memory such as read only memory (firmware), or a data carrier such as an optical or electronic signal carrier, for example. The apparatus and its modules of the present invention may be implemented by hardware circuits such as very large scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., or by software executed by various types of processors, or by a combination of hardware circuits and software, e.g., firmware.

The above description is only for the purpose of illustrating the present invention and the appended claims are not to be construed as limiting the scope of the invention, which is intended to cover all modifications, equivalents and improvements that are within the spirit and scope of the invention as defined by the appended claims.

Claims

1. The capacity configuration method of the optical storage charging station based on the cost and risk multiple targets is characterized by comprising the following steps of: establishing a multi-target capacity optimal configuration model of the light storage charging station with cost and risk, wherein a risk function in the model is investment maintenance and operation cost risk quantified by condition risk value, and a cost function is the sum of expected investment maintenance and operation cost under a typical scene considering charging demand and photovoltaic output; combining an amplification epsilon-constraint method, taking cost as a main target, and taking a risk secondary target as a constraint; and solving the model to obtain pareto fronts of cost and risk under different risk preferences and corresponding configuration capacity, and screening out an objective decision scheme by adopting an entropy weight-TOPSIS method.

2. The cost and risk multi-objective based optical storage charging station capacity allocation method according to claim 1, wherein the cost and risk multi-objective based optical storage charging station capacity allocation method comprises the steps of:

establishing an investment maintenance cost function, an operation cost function and a CVaR risk measurement function of the optical storage system; the investment and maintenance cost of the optical storage system comprises an equal-annual-value initial investment cost and an annual operation and maintenance cost; the cost of the whole planning operation stage is the sum of the investment maintenance cost and the operation cost of the light storage system, the value of the risk represents the high-cost risk caused by uncertainty, and the value is quantized by utilizing a CVaR theory;

3. The cost and risk multi-objective based capacity allocation method for the optical storage charging station according to claim 2, wherein in the step one, the process of converting the uncertainty problems of the charging demands and the photovoltaic outputs of the electric vehicles into the required research scenes is as follows:

in the formula, S _OC1 Charging an initial SOC for the electric vehicle; t is t ₁ Is the initial time of charging;

and

and

charging the average value and the standard deviation of the initial SOC variable logarithm for the electric automobile; sampling and sampling the initial charging time and the initial SOC state of the electric automobile by using a Monte Carlo method to obtain the charging time and the initial SOC state of the electric automobileThe method comprises the steps of obtaining the number of charging scenes required by the electric automobile by combining electricity demand and kmeans clustering;

P _PV (t)＝p _er (t)P _PV ；

in the formula, P _PV (t)、P _PV 、p _er (t) photovoltaic output power at time t, photovoltaic output percentage at time t, and photovoltaic configuration capacity; the uncertainty of photovoltaic output and the charging requirement of the electric automobile is processed by adopting a scene method, various possible conditions are simulated through a large number of scenes, and the random planning problem is converted into a deterministic planning problem; setting photovoltaic output scene set s = { s = _i ,i＝1,2,...,n _s Electric vehicle charging demand scene set e = { e = } _j ,i＝1,2,...,n _e }; wherein n is _s And e _j A total field Jing Geshu respectively for photovoltaic output and electric vehicle charging requirements; superscripts se all indicate values at s _i Individual photovoltaic output scene, e _j In the scene of charging requirement of the electric automobile.

4. The method for configuring capacity of an optical storage and charging station based on cost and risk multiple targets as claimed in claim 2, wherein in the second step, the process of describing the cost function and the risk metric function when the capacity of the optical storage and charging station optical storage system is configured is as follows:

the investment and maintenance cost of the optical storage system comprises two aspects of equal annual value initial investment cost and annual operation and maintenance cost;

C _inv ＝(C _PV P _PV +C _ESS,W W _ESS +C _ESS,P P _ESS )C _RF ；

C _cost ＝C _inv +C _OM ；

in the formula, C _inv 、C _PV 、C _ESS,W 、C _ESS,P 、C _RF Respectively representing the equal-annual-value investment cost of the photovoltaic and energy storage system, the unit capacity investment cost of the photovoltaic and energy storage system, the unit power investment cost of the energy storage system and the equal-annual-value investment coefficient; r and m respectively represent the discount rate and the service life of a corresponding system; c _OM 、

C _cost Respectively representing the annual maintenance cost of the light storage system, the annual maintenance cost of the unit capacity of the photovoltaic system and the energy storage system and the investment maintenance cost of the light storage system; r is ^se 、

P _s ^se 、

The annual running cost of the scene se, the power purchased and sold from the optical storage charging station to the power grid and the charging power of the electric automobile are respectively calculated; a. b and c are the electricity purchase and sale price of the light storage charging station to the power grid and the charging price of the electric vehicle respectively; t is the number of operating hours, and 24 hours are taken; pi(s) _i )、π(e _j ) Are respectively the s _i Individual photovoltaic output scene, e _j Probability of each electric vehicle charging demand scenario;

wherein alpha is the confidence level, and the potential maximum loss risk at the confidence level alpha is C _VaR ；C _CVaR Is in excess of C _VaR Average loss of parts, representing a cost risk; z is a radical of ^se Is a virtual variable; introduction ofCVaR quantifies cost risk values characterizing risks associated with expected investment maintenance and operating costs.

5. The capacity allocation method for the optical storage charging station based on the cost and risk multiple targets as claimed in claim 2, wherein in the third step, an augmented epsilon-constraint method is constructed to process a cost and risk multiple target model, and the process of solving the problem of the multiple target planning of the optical storage charging station is as follows:

the objective function for the lowest cost is:

the objective function for the individual minimum risk is:

min{F ₂ (x)＝C _CVaR }；

min{(1-β)F ₁ (x)+βF ₂ (x)}；

manually regulating and controlling the risk weight through a weighting factor to obtain different planning schemes; wherein β is a weighting factor in the range of [0,1] for achieving a tradeoff between cost and risk, representing a risk preference coefficient; different investment schemes are obtained by changing the parameter beta, and an effective front edge of cost and risk is constructed; the larger beta is, the more serious risk is shown, and an investor is of a risk evasion type; the smaller beta is, the more neglect the risk is indicated, and the investors are risk pursuit type;

F ₁₁ ＝min{F ₁ (x):x∈S}；

F ₂₂ ＝min{F ₂ (x):x∈S}；

F ₁₂ ＝min{F ₂ (x):F ₁ (x)＝F ₁₁ ，x∈S}；

F ₂₁ ＝min{F ₁ (x):F ₂ (x)＝F ₂₂ ，x∈S}；

in the formula, F ₁₁ And F ₂₂ All minimum values under a single target, only considering F ₁ (x) Or F ₂ (x) The minimum value of time; f ₁₂ To minimize the risk with minimal cost per target, F ₂₁ The cost is minimized on the premise that the risk single target is minimum; selecting a cost target as a main target, a risk target as a secondary target and a constraint to divide the range into p equal intervals, and converting the multi-objective optimization problem into a single-objective optimization problem by combining an auxiliary variable epsilon and a relaxation variable s;

ε＝lb+(k+r)/p，k＝0,1,...,p；

s.t.F ₂ (x)+s＝ε,s∈R ⁺ ；

where lb is the minimum value of the risk target, p is the number of intervals over which the risk target is divided, r is the range of the risk target, α is a sufficiently small number, and s is the non-negative slack variable corresponding to the risk target;

capacity power function of energy storage battery:

0.2W _ESS ≤P _ESS ≤W _ESS ；

in the formula, W _ESS 、P _ESS The configured capacity and rated power of the energy storage battery are represented;

the energy storage electric quantity range at the moment t is as follows:

the electric quantity at the beginning and the end in the operation period is equal:

in the formula (I), the compound is shown in the specification,

η、

D _max 、D _min the electric quantity, the charge-discharge efficiency, the charge-discharge power and the maximum charge-discharge depth of the energy storage system at the t moment and the t-1 moment respectively;

in the formula, P _ESS Rated power, u, allocated to the energy storage system _t The variable is 0-1, the variable can be charged only when the value is 1, and can be discharged only when the value is 0;

decoupling by a big-M method:

in the formula, M is a positive number which is set to be large enough, and the decoupling of nonlinear constraint is realized;

and power balance constraint:

and power exchange constraint with the power grid:

0≤P _s ^se (t)≤(1-u _e )P _max ；

in the formula, P _max For the maximum value of the power exchanged between the optical storage charging station and the power grid, u _e The variable is 0-1, and the variable can only buy electricity from the power grid when the value is 1 and can only sell electricity from the power grid when the value is 0;

CVaR risk constraint:

z ^se ≥0；

the risk value of the CVaR risk constraint metric is considered for the expected cost of each scenario, the high cost risk value that the investment maintenance and running costs describing the configuration of the optical storage capacity exist in the face of deterministic scenarios consisting of uncertainty; and after the first step to the third step, converting the planning model into a mixed integer linear planning model, and solving the model by calling a Gurobi solver.

6. The cost and risk multi-objective based optical storage charging station capacity allocation method according to claim 2, wherein in the fourth step, in the conventional linear weighting method, the confidence level α =0.9 and the risk preference coefficient is increased by a value of 0.05; in the augmented epsilon-constraint method, if the interval p =20, the number of solution sets is 21.

7. A cost and risk multi-objective based optical storage charging station capacity allocation system applying the cost and risk multi-objective based optical storage charging station capacity allocation method according to any one of claims 1 to 6, wherein the cost and risk multi-objective based optical storage charging station capacity allocation system comprises:

8. A computer arrangement, characterized in that the computer arrangement comprises a memory and a processor, the memory storing a computer program which, when executed by the processor, causes the processor to carry out the steps of the cost and risk multi-objective based optical storage charging station capacity configuration method according to any one of claims 1-6.

9. A computer-readable storage medium, storing a computer program which, when executed by a processor, causes the processor to carry out the steps of the cost and risk multi-objective based optical storage charging station capacity configuration method according to any one of claims 1 to 6.

10. An information data processing terminal, characterized in that the information data processing terminal is used for realizing the cost and risk multi-objective based optical storage charging station capacity configuration system according to claim 7.