Modeling method for distributed power generation and supporting power grid investment decision-making based on multi-agent game

Abstract

The present invention discloses a method for modeling investment decisions for distributed power sources and supporting power grids based on multi-agent game theory, comprising the following steps: constructing a distributed power source investment decision model and a supporting power grid investment decision model to obtain raw data; generating corresponding strategy sets based on a set of candidate power generation equipment and a set of candidate distributed power source investment plans in the raw data; iteratively conducting a game based on the elements in the strategy set and the flow information of each game, and obtaining the investment return of this game based on the distributed power source investment decision model and the supporting power grid investment decision model; and outputting the equilibrium solution and the final investment return if the game reaches equilibrium. During the joint game process between the distributed power source and the supporting power grid, the present invention continuously optimizes the investment decisions of the two entities by measuring the mixed flow and based on the flow to achieve the optimal result.

Description

Technical Field

[0001] The present invention relates to the field of power grid investment decision modeling, and in particular to a distributed power source and supporting power grid investment decision modeling method based on multi-agent game. Background Art

[0002] With the advancement of energy reform and marketization, my country's energy landscape has gradually undergone major changes, and distributed power generation has become an increasingly important part of the energy system. Figure 1 As shown in the figure, distributed power sources (distributed wind and photovoltaic power) and traditional thermal power plants are boosted by step-up substations and then connected to the grid at substations (nodes), where they transmit electricity to load nodes through the grid. The distributed operating company invests in the distributed power sources, the step-up substations, and the lines connecting the step-up substations to the grid nodes; the grid company invests in the substations and the transmission lines between them and the users. The grid company purchases electricity from the distributed power sources and resells it to load users through the grid. In this model, if the distributed power sources are connected to a 10kV grid or above, self-generation and self-consumption, or the grid connection of surplus power, is not considered. If the distributed power sources are connected to a 380V / 220V grid, the load includes self-generation and self-consumption, or the grid connection of surplus power.

[0003] In recent years, research on distributed generation (DG) and distribution network investment decisions has often incorporated methods for analyzing the correlation between distribution network investment and benefits. These methods also incorporate variables such as distribution network construction and renovation measures, affected operating parameters, and investment benefit indicators. These methods then propose distribution network planning and investment decision-making methods with the goal of optimizing the overall benefits of the distribution network. Alternatively, multi-objective investment optimization models are established, using functions such as DG investment and operating costs, network losses, and transaction costs as objectives. Through optimization algorithms and case studies, investment optimization plans suitable for DG projects are derived. However, existing research on DG investment considers distribution network transaction costs, time-of-use electricity prices, operating benefits, wind and solar timing characteristics, power quality, and environmental performance, but rarely considers distribution network trends. This results in existing models being less realistic and inaccurate. Summary of the Invention

[0004] The technical problem to be solved by the present invention is: in response to the technical problems existing in the prior art, the present invention provides a distributed power supply and supporting power grid investment decision modeling method based on multi-agent game. In the joint game process of distributed power supply and supporting power grid, the investment decisions of the two subjects are continuously optimized by measuring the mixed flow and based on the flow to achieve the optimal result.

[0005] In order to solve the above technical problems, the technical solution proposed by the present invention is:

[0006] A distributed power generation and supporting power grid investment decision modeling method based on multi-agent game includes the following steps:

[0007] Build a distributed power generation investment decision model and a supporting power grid investment decision model to obtain raw data;

[0008] Generate a first strategy set and a second strategy set respectively according to a candidate set of power generation equipment and a candidate set of distributed power generation investment plans in the original data;

[0009] The game is iterated based on the elements in the first strategy set and the second strategy set. In each round of the game, a target element is selected from the first strategy set based on the element selected from the second strategy set in the previous round of the game, and a target element is selected from the second strategy set based on the element selected from the first strategy set in the current round of the game. The power system network structure is changed according to the target element, and then the power flow information of the current round of the game is calculated based on the network structure and the verification is performed. If the verification passes, the distributed power generation investment decision model and the supporting power grid investment decision model are solved based on the network structure and original data of the current round of the game to obtain the investment return of the current round of the game;

[0010] If the game reaches equilibrium, the equilibrium solution and the final investment return are output. If the elements in the first strategy set and the second strategy set are selected and the game has not reached equilibrium, the power system network structure is changed according to the selected power generation equipment and the selected power source investment plan in the preset plan, and the flow information of this round of game is calculated according to the network structure and verified. If the verification passes, the distributed power source investment decision model and the supporting power grid investment decision model are solved according to the network structure and original data of this round of game to obtain the investment return of this round of game, and the game is iterated again according to the elements in the first strategy set and the second strategy set until the game reaches equilibrium.

[0011] Furthermore, the objective profit function of the distributed power investment decision model is the electricity sales revenue , Renewable Energy Quota Income and government subsidies The sum of the total minus the investment cost of the distributed generation units , operating costs of distributed power generation units .

[0012] Furthermore, electricity sales revenue The expression is as follows:

[0013]

[0014] Where n is the life cycle level year, N is the total number of life cycles, is the electricity sales of distributed generation in the nth year, is the average on-grid electricity price in the n-level year;

[0015] Renewable Energy Portfolio Revenue The expression is as follows:

[0016]

[0017] in, is the transaction price of green certificates, k is the coefficient for quantifying renewable energy quotas into green certificates, θs is the number of typical days s, is the power generation of the distributed generation at time t on the sth typical day, A collection of power generation equipment to be selected;

[0018] government subsidies The expression is as follows:

[0019]

[0020] in, is the active power of distributed generation equipment i, num is the number of distributed generation units, is the unit price of government subsidy in the nth level year;

[0021] Investment cost of distributed power generation units The expression is as follows:

[0022]

[0023] in, is the investment variable of distributed generation equipment i, is the investment cost of distributed generation equipment i, is the service life of distributed power generation equipment, ω is the capital discount rate;

[0024] Operating costs of distributed power generation units The expression is as follows:

[0025]

[0026] Where i is the number of the distributed power generation equipment, is the operating time of distributed generation equipment i in n horizontal years, is the operating cost of power generation equipment i per unit power in the nth year, is the active power of distributed generation equipment i.

[0027] Furthermore, the constraints of the distributed power investment decision model include:

[0028] The power constraint is expressed as follows:

[0029]

[0030] Where x is the confidence capacity factor of the distributed generation unit, The maximum load of the year. is the capacity reserve factor, is the active power of distributed generation equipment i, is the investment variable of distributed generation equipment i, A collection of power generation equipment to be selected;

[0031] The installed capacity constraint is expressed as follows:

[0032]

[0033] in, is the active power of distributed generation equipment i, is the investment variable of distributed generation equipment i, is a collection of power generation equipment to be selected, It is the maximum installed capacity of distributed power generation units.

[0034] Furthermore, the target income of the investment decision model of the supporting power grid is the income from electricity sales. Minus electricity sales revenue , Grid investment cost , network loss cost and the penalty costs for curtailing renewable energy .

[0035] Furthermore, the revenue from electricity sales The expression is as follows:

[0036]

[0037] in, is the annual load for the n-level year, is the electricity sales price of the power grid company in the n-level year;

[0038] Grid investment cost The expression is as follows:

[0039]

[0040] in, is a collection of candidate power grid investment projects; is the investment variable of project j; is the investment cost of project j; is the useful life of the asset;

[0041] Network loss cost The expression is as follows:

[0042]

[0043] in, l Number the project; Supporting lines for distributed power generation l Network loss in the n-level year; is the network loss cost of the unit line in the nth horizontal year;

[0044] Penalty costs for curtailing renewable energy The expression is as follows:

[0045]

[0046] in, represents the predicted output of the distributed generation unit in period t, is a collection of power generation equipment to be selected, Indicates actual output. Represents the penalty coefficient.

[0047] Furthermore, the constraints of the supporting power grid investment decision model include:

[0048] The power network constraints are expressed as follows:

[0049]

[0050] Where H, J and K represent the correlation matrices of transmission lines, generators, loads and power network nodes respectively; Indicates the n-level year line l the tide that flows past; represents the output of generator m in horizontal year n; represents the load of node k in the n-level year; S1, S2, S3 and S4 represent the set of transmission lines, the set of generators, the set of power loads and the set of power network nodes respectively;

[0051] The power flow constraint is expressed as follows:

[0052]

[0053] in, 、 are the injected active power and injected reactive power at node q respectively; 、 are the voltage amplitudes at nodes q and r, respectively; 、 are the conductance and susceptance of branch qr respectively; is the voltage phase angle difference between nodes q and r;

[0054] The line transmission capacity constraint is expressed as follows:

[0055]

[0056] in, is the power flow of line qr between nodes q and r; The maximum capacity allowed for transmission on the line qr between nodes q and r.

[0057] Furthermore, calculating and verifying the current game flow information based on the network structure includes the following steps:

[0058] Calculating the power system flow under the network structure and transmitting it to the coupling node after passing the first check;

[0059] Calculate the energy flow of the coupling node, substitute it into the power flow model, and then perform a second check on the calculation results.

[0060] Furthermore, calculating the power system flow under the network structure includes the following steps:

[0061] According to the initial voltage value of each node, the imbalance of the node voltage squared with injected power is calculated;

[0062] Calculate the voltage change value of each node according to the unbalance amount;

[0063] According to the change value of each node voltage, the step of calculating the imbalance amount of the node voltage square of the injected power is performed until the corrected voltage value of each node meets the accuracy requirement.

[0064] Furthermore, the unbalanced expression of the square of the node voltage for calculating the injected power is as follows:

[0065]

[0066] in, e i 、 f i are the real and imaginary parts of the node voltage obtained during the iteration process, P i for PQ Node and PV The injected active power of the node, is the network conductance matrix, is the network susceptance matrix;

[0067]

[0068] in, e i 、 f iare the real and imaginary parts of the node voltage obtained during the iteration process, Q i for PQ The injected reactive power of the node, is the network conductance matrix, is the network susceptance matrix.

[0069] Compared with the prior art, the advantages of the present invention are:

[0070] The present invention constructs investment decision-making models for distributed power sources and supporting power grids respectively, and establishes a joint game mechanism for the two subjects. Through continuous iteration, the optimal result is guaranteed. The selection of the game plan in each round depends on whether the current information verification of this round is passed, so as to ensure that the result conforms to the actual situation of the power system. In addition to involving multiple income and costs, the constructed investment decision-making model also sets corresponding constraints based on the power system network structure, thereby ensuring that the result takes into account both economy and accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 This is a physical structure diagram of a typical distributed power system.

[0072] Figure 2 This is a diagram of the game relationship between the distributed power source and the supporting power grid according to an embodiment of the present invention.

[0073] Figure 3 Schematic diagram of gaming behavior according to an embodiment of the present invention.

[0074] Figure 4 A simplified diagram of the steps of an embodiment of the present invention.

[0075] Figure 5 This is a flowchart of the game behavior of an embodiment of the present invention.

[0076] Figure 6 This is the initial scenario designed in the embodiment of the present invention.

[0077] Figure 7 Detailed flowchart of the method corresponding to the scenario of the embodiment of the present invention.

[0078] Figure 8 The figure is a comparison chart of the results of the embodiment of the present invention and the results of the other two scenarios. DETAILED DESCRIPTION

[0079] The present invention will be further described below in conjunction with the accompanying drawings and specific preferred embodiments, but the scope of protection of the present invention is not limited thereby.

[0080] Before describing the specific embodiments of the present invention in detail, it is necessary to state in advance the relevant concepts and assumptions involved in the specific embodiments as follows:

[0081] For distributed generation investors, joint planning through game theory can help reduce investment waste. Modeling analysis can clarify the grid structure and load, thereby determining whether new network lines are necessary. This can reduce the cost of distributed generation electricity and fundamentally save on line investment.

[0082] For power grid companies, model building and planning can clarify whether the investment in distributed power sources can adapt to the grid structure, further determine the optimal power flow of the grid, and minimize line losses. On the one hand, it reduces the cost of power transmission and distribution of the grid, and on the other hand, it reduces unnecessary investment in the grid, thereby reducing the electricity price transmitted to users and reducing social electricity costs.

[0083] Transfer of game relationship between distributed power generation and supporting power grid investment:

[0084] Distributed generation (DG) investors (or DG operators) make decisions about new DG construction plans, while power grid companies make decisions about supporting DG grid construction plans. These two entities influence each other through a hybrid power flow model, which is implemented during the safety verification process. During this decision-making process, DG operators can determine their own generation equipment investment, power consumption plans, grid access plans, and trading models, thereby influencing the power grid company's investment decisions and electricity sales revenue. Power grid companies can also decide on grid construction investment plans, striving for optimal investment, thereby influencing the investment decisions and trading models of DG operators.

[0085] The main body of distributed power generation investment first formulates its power capacity investment strategy. On this basis, the power grid company formulates the grid construction plan strategy. The distributed power generation and the power grid company both aim to maximize their respective benefits and optimize and adjust the game strategy to the best. When the other party chooses the optimal strategy, the other party also chooses the optimal strategy. That is, under this strategy, the entire system achieves the highest benefit in the Nash equilibrium state. The game relationship transmission diagram is as follows Figure 2 The DG investment candidate set includes the candidate decisions for DG generator sets throughout their life cycle. The grid connection plan and power consumption plan refer to the plans and schedules for each time period. The grid construction investment includes the main and distribution network construction plans throughout their life cycle.

[0086] Distributed power generation and supporting power grid joint game investment decision-making model:

[0087] During the construction of new distributed power sources, it is necessary to comprehensively consider the direct connection between the grid and the distribution network load nodes, so that the original network flow will change accordingly. When the integrated production and consumption users and distributed energy storage are connected to the grid, the flow will flow from the bus to the load end in one direction, thereby changing the size and direction of the flow, making it no longer fixed. Based on this, the present invention will use a mixed flow method for calculation and analysis, and research from the perspective of flow will help the research results to be close to the actual energy consumption situation. The game process is based on the idea of ​​sequence. The distributed power source first gives the initial investment plan, and the power grid company then gives the initial decision plan, which in turn affects the investment strategy of the distributed power source. Figure 6 As shown in the figure, the solid line part is the generator set and the transmission line, and the dotted line part is the candidate generator set and the candidate transmission line. When the distributed power source selects the candidate generator set, the power grid company selects the corresponding candidate transmission line. The subsequent distributed power source continues to select from the remaining candidate generator sets according to the candidate transmission line selected by the power grid company and the original line according to actual needs. The above process is repeated, and a game pattern is formed between the distributed power source and the power grid company. The specific game behavior is as follows Figure 3 It should be noted that the specific process of selecting a new generator set from the candidate generator sets according to the lines in the power system is well known to those skilled in the art. This solution does not involve any improvement to this process and will not be described in detail here.

[0088] First, the power network flow is calculated. The distributed generation decides on the initial investment plan based on its own situation. The grid company also determines the initial investment plan for grid construction based on the actual situation. Then, the information is transmitted to the coupling node. The coupling node feeds back the flow status to the distributed generation and the grid company. Each entity changes the network structure, and the game continues until the strategies of the distributed generation and the grid company are optimal and there is no room for improvement. The game behavior of each entity reaches the Nash equilibrium:

[0089] (1)

[0090] Where, 、 Each strategy is the optimal strategy for one party when the other party chooses the optimal strategy. Under this strategy combination, both the distributed generation and the power grid company can achieve the maximum benefits in the equilibrium sense. argmax() is the set of variables that maximizes the value of the objective function.

[0091] like Figure 4 As shown, this embodiment proposes a distributed power supply and supporting power grid investment decision modeling method based on multi-agent game, including the following steps:

[0092] S1) Build a distributed power generation investment decision model and a supporting power grid investment decision model to obtain raw data;

[0093] S2) generating a first strategy set and a second strategy set respectively based on a set of candidate power generation equipment and a set of candidate distributed generation investment plans in the original data;

[0094] S3) iteratively playing the game based on the elements in the first strategy set and the second strategy set, wherein in each round of the game, a target element is selected from the first strategy set based on the element selected from the second strategy set in the previous round of the game, and a target element is selected from the second strategy set based on the element selected from the first strategy set in the current round of the game, and the power system network structure is changed based on the target elements selected from the first strategy set and the second strategy set, and then the power flow information of the current round of the game is calculated based on the network structure and the original data and verified. If the verification passes, the distributed power generation investment decision model and the supporting power grid investment decision model are solved based on the network structure and the original data of the current round of the game to obtain the investment return of the current round of the game;

[0095] S4) If the game reaches an equilibrium state, the equilibrium solution and the final investment return are output. If the elements in the first strategy set and the second strategy set are selected and the game has not reached an equilibrium state, the power system network structure is changed according to the selected power generation equipment and the selected power source investment plan in the preset plan, and the flow information of this round of game is calculated based on the network structure and verified. If the verification passes, the distributed power source investment decision model and the supporting power grid investment decision model are solved according to the network structure and original data of this round of game to obtain the investment return of this round of game, and the game is iterated again according to the elements in the first strategy set and the second strategy set until the game reaches an equilibrium state.

[0096] In step S1 of this embodiment, the original data includes the necessary parameters such as user load information, parameters of selected power generation equipment, electricity price, cost of distributed power generation equipment, investment cost of power grid structure, and original network topology parameters required for establishing the model in this embodiment. Based on these parameters, the target profit function of the distributed power investment decision model in this embodiment is the electricity sales revenue. , Renewable Energy Quota Income and government subsidies The sum of the total minus the investment cost of the distributed generation units , operating costs of distributed power generation units , the expression is as follows:

[0097] (2)

[0098] in:

[0099] Electricity sales revenue The expression is as follows:

[0100] (3)

[0101] Where n is the life cycle level year, N is the total number of life cycles, is the electricity sales of distributed generation in the nth year, is the average on-grid electricity price in the n-level year;

[0102] Renewable Energy Portfolio Revenue The expression is as follows:

[0103] (4)

[0104] in, is the transaction price of green certificates, k is the coefficient for quantifying renewable energy quotas into green certificates, θs is the number of typical days s, is the power generation of the distributed generation at time t on the sth typical day, A collection of power generation equipment to be selected;

[0105] government subsidies The expression is as follows:

[0106] (5)

[0107] in, is the active power of distributed generation equipment i, num is the number of distributed generation units, is the unit price of government subsidy in the nth level year;

[0108] Investment cost of distributed power generation units The expression is as follows:

[0109] (6)

[0110] in, is the investment variable of distributed generation equipment i, is the investment cost of distributed generation equipment i, is the service life of distributed power generation equipment, ω is the capital discount rate;

[0111] Operating costs of distributed power generation units The expression is as follows:

[0112] (7)

[0113] Where i is the number of the distributed power generation equipment, is the operating time of distributed generation equipment i in n horizontal years, is the operating cost of power generation equipment i per unit power in the nth year, is the active power of distributed generation equipment i.

[0114] In addition, the constraints of the distributed power investment decision model include:

[0115] The power constraint is expressed as follows:

[0116] (8)

[0117] Where x is the confidence capacity factor of the distributed generation unit, The maximum load of the year. is the capacity reserve factor, is the active power of distributed generation equipment i, is the investment variable of distributed generation equipment i, A collection of power generation equipment to be selected;

[0118] The installed capacity constraint is expressed as follows:

[0119] (9)

[0120] in, is the active power of distributed generation equipment i, is the investment variable of distributed generation equipment i, is a collection of power generation equipment to be selected, It is the maximum installed capacity of distributed power generation units.

[0121] Based on the parameters of the original data, in this embodiment, the target income of the investment decision model of the supporting power grid is the income from electricity sales. Minus electricity sales revenue , Grid investment cost , network loss cost and the penalty costs for curtailing renewable energy , the expression is as follows:

[0122] (10)

[0123] in:

[0124] Revenue from electricity sales The expression is as follows:

[0125] (11)

[0126] in, is the annual load for the n-level year, is the electricity sales price of the power grid company in the n-level year;

[0127] Grid investment cost The expression is as follows:

[0128] (12)

[0129] in, is a collection of candidate power grid investment projects; is the investment variable of project j; is the investment cost of project j; is the useful life of the asset;

[0130] Network loss cost The expression is as follows:

[0131] (13)

[0132] in, l Number the project; Supporting lines for distributed power generation l Network loss in the n-level year; is the network loss cost of the unit line in the nth horizontal year;

[0133] Penalty costs for curtailing renewable energy The expression is as follows:

[0134] (14)

[0135] in, represents the predicted output of the distributed generation unit in period t, is a collection of power generation equipment to be selected, Indicates actual output. Represents the penalty coefficient.

[0136] In addition, the constraints of the supporting power grid investment decision model include:

[0137] The power network constraints are expressed as follows:

[0138] (15)

[0139] Where H, J and K represent the correlation matrices of transmission lines, generators, loads and power network nodes respectively; represents the current flowing on line l in horizontal year n; represents the output of generator m in horizontal year n; represents the load of node k in the n-level year; S1, S2, S3 and S4 represent the set of transmission lines, the set of generators, the set of power loads and the set of power network nodes respectively;

[0140] The power flow constraint is expressed as follows:

[0141] (16)

[0142] in, 、 are the injected active power and injected reactive power at node q respectively; 、 are the voltage amplitudes at nodes q and r, respectively; 、 are the conductance and susceptance of branch qr respectively; is the voltage phase angle difference between nodes q and r;

[0143] The line transmission capacity constraint is expressed as follows:

[0144] (17)

[0145] in, is the power flow of line qr between nodes q and r; The maximum capacity allowed for transmission on the line qr between nodes q and r.

[0146] In step S2 of this embodiment, the DG investment entity generates an investment plan set (hereinafter referred to as the first strategy set) f(i) = {MF1, MF2, …, MFnF} based on the candidate set of power generation equipment in the original data. The power grid company generates a grid strategy set (hereinafter referred to as the second strategy set) y(i) = {ME1, ME2, …, MEnE} based on the candidate set of DG investment plans in the original data. nF and nE are the total number of elements in the first strategy set and the second strategy set, respectively.

[0147] Step S3) of this embodiment is as follows Figure 5 Shown, including:

[0148] Randomly select a set of solutions f0 and y0 from the first strategy set and the second strategy set as the initial value of the iteration;

[0149] Set the initial iteration value δ=2;

[0150] According to the selected power generation equipment corresponding to f0 and the selected investment plan corresponding to y0, the power system network structure is changed. The power flow information of the current network structure in the initial round of the game is calculated and verified. If the verification passes, the current network structure and original data are input into the distributed power generation investment decision model and the supporting power grid investment decision model to solve the corresponding benefits;

[0151] In each subsequent round of game, the corresponding plan fn is selected from the first strategy set based on the plan yn-1 selected from the second strategy set of the previous round of game, and then the corresponding plan yn is selected from the second strategy set based on the plan fn, and the mixed flow calculation is performed to obtain the corresponding flow information. After the flow information is verified, the corresponding benefits can be obtained according to the above steps, and the benefits corresponding to the current round of game are compared with the benefits corresponding to the previous round of game. If the benefits of the two game rounds are the same, that is, (fn, yn) = (fn-1, yn-1) = (MF*, ME*) is satisfied, it is considered that the equilibrium state is reached and the process goes to step S4). Otherwise, let δ=δ+1 and execute this step.

[0152] In this embodiment, when performing hybrid power flow measurement, based on the information from the previous round, the distributed power source and the power grid company make another decision, check, and calculate their own plans, and obtain the final benefits of this round of game after the hybrid power flow calculation. Specifically, the distributed power source is generally connected at the end of the distribution line. Due to the large fluctuation range of light and wind power, the active and reactive flow directions of the line undergo complex changes. The interactive behavior of each subject is achieved through safety verification. During this process, it is necessary to calculate the hybrid power flow to provide a guarantee for the stable operation of the distributed power system. In the game process, the mutual influence between the decision-making plans is indirectly achieved through the transmission and conversion of power flow parameters during the hybrid power flow calculation process. The specific process of hybrid power flow safety verification is as follows:

[0153] (1) After the distributed generation and the power grid company give a decision plan, the network structure of the power system is updated according to the elements selected from the first strategy set and the second strategy set, and the power system flow under the grid structure is calculated. The plan is then checked to see whether the power system flow at this time exceeds a preset first threshold. If the check is satisfied, the flow information is transmitted to the coupling node.

[0154] (2) Using the power balance relationship of the coupling node, the energy flow of the coupling node is calculated and substituted into the power flow model for calculation. The calculated result is then compared with the preset second threshold to achieve safety verification of the decision-making scheme.

[0155] (3) The final decision-making scheme of this game round is determined according to the verification results. That is, when the verification is passed, the final decision-making scheme of this game round is the decision-making scheme given by the distributed power source and the power grid company.

[0156] In this embodiment, the process of calculating the power system flow is as follows:

[0157] According to the initial voltage value of each node, the unbalanced amount of the square of the node voltage of the injected power is calculated as follows:

[0158] (18)

[0159] Where, I B and U B are the current and voltage of each node respectively, Y B is the node admittance matrix.

[0160] (19)

[0161] Where, j is the Jacobian matrix, is the network conductance matrix, is the network susceptance matrix.

[0162] (20)

[0163] Where, e i (0) 、 f i (0) is the initial value of each node voltage.

[0164] Equations (21) and (22) are the squared imbalances of node voltages used to calculate injected power.

[0165] (twenty one)

[0166] Where, e i 、 f i are the real and imaginary parts of the node voltage obtained during the iteration process. P i for PQ Node and PV The injected active power of the node.

[0167] (twenty two)

[0168] Where, Q i for PQ The injected reactive power of the node. U i for PV The voltage magnitude of the node. Substitute the initial voltage value of each node into the equation to find the unbalance Δ in the correction equation. P i (0) , Δ Q i (0)Etc., as shown in Equations (23) to (25).

[0169] (twenty three)

[0170] (twenty four)

[0171] (25)

[0172] After the convergence condition is met, the solution is continued to calculate the value of each node voltage change, that is, the corrected value, as shown in Equation (26) and Equation (27).

[0173] (26)

[0174] (27)

[0175] Where, e i (1) 、 f i (1) is the corrected value of each node voltage.

[0176] The corrected value is used to continue the next iteration from Equation (21) until the accuracy requirement is met and the loop is exited. Finally, the balanced node power and line power are calculated.

[0177] Therefore, in step S3 of this embodiment, calculating and verifying the current game flow information based on the network structure includes the following steps:

[0178] Calculating the power system flow under the network structure and transmitting it to the coupling node after passing the first check;

[0179] Calculate the energy flow of the coupling node, substitute it into the power flow model, and then perform a second check on the calculation results.

[0180] Calculating the power system flow under the network structure includes the following steps:

[0181] According to the initial voltage value of each node, the unbalanced amount of the node voltage squared with injected power is calculated according to equations (21) and (22);

[0182] Calculate the voltage change value of each node according to the unbalance amount;

[0183] According to the change value of each node voltage, the step of calculating the imbalance amount of the node voltage square of the injected power is performed until the corrected voltage value of each node meets the accuracy requirement.

[0184] In summary, the method of this embodiment takes into account the investment and operation stages to establish an investment decision-making model for distributed power sources, supporting power grids, and the joint game between the two entities, which not only ensures the economy of the decision-making plan, but also takes into account the accuracy of investment, realizes the optimal allocation of resources, meets the needs of distributed power development and regional economic development, and improves the social and economic benefits of investment. In the joint game process between distributed power sources and supporting power grids, by measuring the mixed flow and transmitting and converting the flow parameters, the decision-making plans of distributed power sources and supporting power grids can be mutually influenced, thereby continuously optimizing the investment decisions of the two entities to achieve the optimal

[0185] The following describes the experimental process for verifying the method of this embodiment:

[0186] First, determine the investment strategies of various stakeholders. Distributed power generation (DG) operators aim to improve the return on investment of their power generation equipment and reduce operating costs. Furthermore, DGs primarily utilize renewable clean energy, and their consumption will bring environmental benefits. Multi-energy systems effectively leverage the complementary nature of distributed wind and photovoltaic power generation, integrating power generation and sales. Grid companies aim to reduce line investment and improve investment returns. While the grid company and DGs generate power, deployed distributed energy storage can control power absorption or output to smooth grid power fluctuations, provide peak shaving and valley shifting, and participate in primary frequency regulation. Consumer electricity is supplied by these three parties.

[0187] Then, the initial scenario of the IEEE24-node model is designed, such as Figure 6 As shown, the solid line part represents the generator sets and transmission lines; there are 13 generator sets, whose parameters are detailed in Table 1, and 41 transmission lines, whose parameters are detailed in Table 2.

[0188] Table 1 Capacity of existing generator sets

[0189]

[0190] Table 2 Parameters of existing transmission lines

[0191]

[0192] like Figure 6 As shown in the figure, the dotted part represents the candidate power generation equipment and the candidate transmission line in the distributed power system; there are 6 candidate power generation equipment and 9 candidate transmission lines, and the maximum expansion number of each line is 1. The parameters are detailed in Table 3 and Table 4.

[0193] Table 3 Distributed power generation equipment parameters

[0194]

[0195] Table 4 Parameters of transmission lines to be built

[0196]

[0197] Based on a survey of transmission line construction costs in a certain province over the past two years, we set the investment costs for 500kV and 220kV transmission lines at 2.5 million yuan / km and 1 million yuan / km, respectively. The operating costs of distributed generators are calculated as 2% of the investment cost. The load parameters for the power network are shown in Table 5.

[0198] Table 5 Maximum load distribution of each node

[0199]

[0200] Because the investment decision primarily examines the potential challenges of near-term distributed generation (DG) investment, the scenario design references Beijing's 2022 on-grid electricity price and residential electricity sales price standards. The average on-grid electricity price for distributed users is assumed to be 0.5 yuan / kW·h, and the grid company's sales price is 0.58 yuan / kW·h. The renewable energy quota factor is 20%, the purchase price of green certificates is 100 yuan per certificate, and the penalty fee for unit wind and solar curtailment is 400 yuan / MWh. Typical daily load and wind and solar output data are taken from a provincial power grid.

[0201] The distributed power generation and supporting power grid investment decision-making under three different scenarios are simulated: scenario 1: independent decision-making of distributed power generation and power grid without considering game; scenario 2: joint decision-making of distributed power generation and power grid without considering game; scenario 3: joint decision-making of distributed power generation and power grid with considering game, which are the scenarios corresponding to the method of this embodiment.

[0202] like Figure 7 As shown in the figure, scenario 3 generates a power flow diagram based on the initial situation, derives the power flow information, and uses Matlab software to perform game calculations. If Nash equilibrium is reached, the results are generated, forming a power flow diagram and obtaining the profit results. If Nash equilibrium is not reached, PSASP (Power System Analysis Synthesis Program) is used to generate a new power flow diagram based on the investment plan, derive the power flow information, and use Matlab software again to perform game calculations.

[0203] The distributed user load demands in the three scenarios are the same. The comparison of power network decision scenarios is as follows: Figure 8As shown in the figure, in the distributed generation investment decision, scenario 1 selects units at nodes 1, 2, 7, and 22; scenario 2 selects units at nodes 2, 7, and 22; and scenario 3 selects units at nodes 2, 7, 15, and 16. Regarding the transmission line decision, the three scenarios differ in that scenario 1 builds new transmission lines on all candidate branches, scenario 2 does not build new transmission lines on branches 1-5, and scenario 3 does not build new lines on branches 1-5, 17-22, and 16-17. The remaining transmission line construction remains the same.

[0204] The table below compares the benefits of Scenario 1 and Scenario 2. Overall, Scenario 2 generates 1,087.29 million yuan more than Scenario 1. This is because in Scenario 1, the distributed generation and the power grid company make independent decisions. Furthermore, in Scenario 1, DG Company builds new power generation equipment at both Nodes 1 and 2, totaling 320.4 MW. The power grid company also builds new transmission lines. However, based on the grid load, only 171.2 MW of new generation equipment at Node 2 is needed to meet the grid load. This results in excess capacity, reduced generation equipment utilization, and a waste of investment. Therefore, considering joint decision-making can optimize investment overall and increase the system's total benefits.

[0205] Table 6 Comparison of multi-agent benefits under scenario 1 and scenario 2

[0206] Unit: Ten thousand yuan

[0207]

[0208] The table below compares the benefits of Scenario 1 and Scenario 2. Compared to Scenario 2, Scenario 3 reduces the grid company's transmission line investment by 455 million yuan, while increasing its revenue by 455 million yuan. This is because, in Scenario 2, the DG operating company built a new 415MW power generation facility at node 22. However, this node lacked off-grid load to absorb the power, requiring it to flow through the 22-17 line to node 17 and transmit it to other load nodes on the grid. This resulted in long-distance transmission of power and load, increasing line losses and the number of transmission lines required by the grid company, significantly increasing the grid company's investment and network loss costs. After the introduction of multi-agent game, in order to achieve a balanced state between the power grid company and the DG operating company, in Scenario 3, the DG operating company built new 258MW and 157MW power generation equipment at nodes 15 and 16, respectively, close to the load requirements. This optimized the grid flow. The power grid company only needed to build three new lines, 16-19, 16-15, and 15-24, to meet the grid load requirements and flow constraints. Although the DG operating company's investment increased, the investment plan of Scenario 3 was superior to that of Scenario 2 because it greatly reduced the investment and network loss costs of the power grid company.

[0209] Table 7 Comparison of total benefits of distributed power generation in scenarios 2 and 3

[0210] Unit: Ten thousand yuan

[0211]

[0212] The above description is merely a preferred embodiment of the present invention and does not constitute any form of limitation to the present invention. Although the present invention has been disclosed above with reference to the preferred embodiment, it is not intended to limit the present invention. Therefore, any simple modifications, equivalent variations, and modifications to the above embodiment that do not depart from the technical solution of the present invention and are based on the technical essence of the present invention shall fall within the scope of protection of the technical solution of the present invention.

Claims

1. A distributed power generation and supporting power grid investment decision modeling method based on multi-agent game, characterized by: The following steps are involved: Build a distributed power generation investment decision model and a supporting power grid investment decision model to obtain raw data; Generate a first strategy set and a second strategy set respectively according to a candidate set of power generation equipment and a candidate set of distributed power generation investment plans in the original data; The game is iterated based on the elements in the first strategy set and the second strategy set. In each round of the game, a target element is selected from the first strategy set based on the element selected from the second strategy set in the previous round of the game, and a target element is selected from the second strategy set based on the element selected from the first strategy set in the current round of the game. The power system network structure is changed according to the target element, and then the power flow information of the current round of the game is calculated based on the network structure and the verification is performed. If the verification passes, the distributed power generation investment decision model and the supporting power grid investment decision model are solved based on the network structure and original data of the current round of the game to obtain the investment return of the current round of the game; If the game reaches equilibrium, the equilibrium solution and the final investment return are output. If the elements in the first strategy set and the second strategy set are selected and the game has not reached equilibrium, the power system network structure is changed according to the selected power generation equipment and the selected power source investment plan in the preset plan, and the power flow information of this round of game is calculated based on the network structure and verified. If the verification passes, the distributed power generation investment decision model and the supporting power grid investment decision model are solved based on the network structure and original data of this round of game to obtain the investment return of this round of game, and the game is iterated again according to the elements in the first strategy set and the second strategy set until the game reaches equilibrium. The objective profit function of the distributed power investment decision model is the electricity sales revenue , Renewable Energy Quota Income and government subsidies The sum of the total minus the investment cost of the distributed generation units , operating costs of distributed power generation units ; Electricity sales revenue The expression is as follows: Where n is the life cycle level year, N is the total number of life cycles, is the electricity sales of distributed generation in the nth year, is the average on-grid electricity price in the n-level year; Renewable Energy Portfolio Revenue The expression is as follows: in, is the transaction price of green certificates, k is the coefficient for quantifying renewable energy quotas into green certificates, θs is the number of typical days s, is the power generation of the distributed generation at time t on the sth typical day, A collection of power generation equipment to be selected; government subsidies The expression is as follows: in, is the active power of distributed generation equipment i, num is the number of distributed generation units, is the unit price of government subsidy in the nth level year; Investment cost of distributed power generation units The expression is as follows: in, is the investment variable of distributed generation equipment i, is the investment cost of distributed generation equipment i, is the service life of distributed power generation equipment, ω is the capital discount rate; Operating costs of distributed power generation units The expression is as follows: Where i is the number of the distributed power generation equipment, is the operating time of distributed generation equipment i in n horizontal years, is the operating cost of power generation equipment i per unit power in the nth year, is the active power of distributed generation equipment i; The constraints of the distributed power investment decision model include: The power constraint is expressed as follows: Where x is the confidence capacity factor of the distributed generation unit, The maximum load of the year. is the capacity reserve factor, is the active power of distributed generation equipment i, is the investment variable of distributed generation equipment i, A collection of power generation equipment to be selected; The installed capacity constraint is expressed as follows: in, is the active power of distributed generation equipment i, is the investment variable of distributed generation equipment i, is a collection of power generation equipment to be selected, It is the maximum installed capacity of distributed power generation units.

2. The distributed power generation and supporting power grid investment decision modeling method based on multi-agent game according to claim 1 is characterized in that: The target income of the supporting power grid investment decision model is the income from electricity sales. Minus electricity sales revenue , Grid investment cost , network loss cost and the penalty costs for curtailing renewable energy .

3. The distributed power generation and supporting power grid investment decision modeling method based on multi-agent game according to claim 2 is characterized in that: Revenue from electricity sales The expression is as follows: in, is the annual load for the n-level year, is the electricity sales price of the power grid company in the n-level year; Grid investment cost The expression is as follows: in, is a collection of candidate power grid investment projects; is the investment variable of project j; is the investment cost of project j; is the useful life of the asset; Network loss cost The expression is as follows: in, l Number the project; Supporting lines for distributed power generation l Network loss in the n-level year; is the network loss cost of the unit line in the nth horizontal year; Penalty costs for curtailing renewable energy The expression is as follows: in, represents the predicted output of the distributed generation unit in period t, is a collection of power generation equipment to be selected, Indicates actual output. Represents the penalty coefficient.

4. The distributed power generation and supporting power grid investment decision modeling method based on multi-agent game according to claim 2 is characterized in that: The constraints of the supporting power grid investment decision model include: The power network constraints are expressed as follows: Where H, J and K represent the correlation matrices of transmission lines, generators, loads and power network nodes respectively; Indicates the n-level year line l the tide that flows past; represents the output of generator m in horizontal year n; represents the load of node k in the n-level year; S1, S2, S3 and S4 represent the set of transmission lines, the set of generators, the set of power loads and the set of power network nodes respectively; The power flow constraint is expressed as follows: in, 、 are the injected active power and injected reactive power at node q respectively; 、 are the voltage amplitudes at nodes q and r, respectively; 、 are the conductance and susceptance of branch qr respectively; is the voltage phase angle difference between nodes q and r; The line transmission capacity constraint is expressed as follows: in, is the power flow of line qr between nodes q and r; The maximum capacity allowed for transmission on the line qr between nodes q and r.

5. The distributed power generation and supporting power grid investment decision modeling method based on multi-agent game according to claim 1 is characterized in that: Calculating and verifying the current round of game flow information based on the network structure includes the following steps: Calculating the power system flow under the network structure and transmitting it to the coupling node after passing the first check; Calculate the energy flow of the coupling node, substitute it into the power flow model, and then perform a second check on the calculation results.

6. The distributed power generation and supporting power grid investment decision modeling method based on multi-agent game according to claim 5 is characterized in that: Calculating the power system flow under the network structure includes the following steps: According to the initial voltage value of each node, the imbalance of the node voltage squared with injected power is calculated; Calculate the voltage change value of each node according to the unbalance amount; According to the change value of each node voltage, the step of calculating the imbalance amount of the node voltage square of the injected power is performed until the corrected voltage value of each node meets the accuracy requirement.

7. The distributed power generation and supporting power grid investment decision modeling method based on multi-agent game according to claim 6 is characterized in that: The expression for the unbalanced amount of the square of the node voltage for calculating the injected power is as follows: in, e i 、 f i are the real and imaginary parts of the node voltage obtained during the iteration process, P i for PQ Node and PV The injected active power of the node, is the network conductance matrix, is the network susceptance matrix; in, e i 、 f i are the real and imaginary parts of the node voltage obtained during the iteration process, Q i for PQ The injected reactive power of the node, is the network conductance matrix, is the network susceptance matrix.

Images