Two-Layer Optimization Method for Prosumer Distributed Transactions Based on Distribution Network Operation Constraints

Abstract

The present invention belongs to the field of power markets, and specifically relates to a two-layer optimization method for prosumer distributed transactions based on distribution network operation constraints, including: Step 1: Construct a two-layer optimization model for prosumer distributed transactions, specifically: Step 101: Establish an upper-layer prosumer distributed transaction model; Step 102: Establish a lower-layer distribution network optimization model; Step 2: Perform distributed solution on the upper-layer prosumer distributed transaction model to obtain the distributed transaction volumes of each prosumer; input the distributed transaction volumes of each prosumer into the lower-layer distribution network optimization model to update the bilateral transaction coefficients between prosumers, and return the bilateral transaction coefficients to the upper-layer prosumer distributed transaction model; iteratively solve the two-layer optimization model to minimize the transaction costs of each prosumer and the operation cost of the distribution network, and obtain the distributed transaction strategy among multiple prosumers under the distribution network operation constraints. The present invention updates the bilateral transaction coefficients between prosumers and reduces the costs of prosumers.

Description

Technical Field

[0001] The present invention belongs to the field of power markets, and particularly relates to a two-layer optimization method for prosumer distributed transactions based on distribution network operation constraints. Background Art

[0002] At present, the world's energy structure is mainly based on fossil energy. However, against the backdrop of the shortage of fossil energy and the continuous exacerbation of environmental pollution problems, distributed resources such as photovoltaic power generation, energy storage batteries, and electric vehicles have become a new direction for countries around the world to develop low-carbon energy. Distributed resources have the characteristics of low pollution, safe and reliable operation, high energy conversion rate, and small and easy-to-install equipment, greatly improving the load flexibility on the user side. As the penetration rate of distributed resources in the distribution network continues to increase, the user side has gradually changed from traditional electricity consumers with uncontrollable loads to prosumers with controllable loads, with the ability to generate electricity independently, manage power generation and consumption, and energy storage, realizing self-generation and self-consumption of electric energy.

[0003] Traditional centralized scheduling methods have the disadvantages of a large amount of information, large computational complexity, poor stability, and inability to protect user privacy, and cannot support multiple prosumers to participate in distributed peer-to-peer transactions. Distributed transactions enable prosumers to conduct electricity transactions through information interaction, realize the local consumption of new energy, can tap the flexibility of resources on the prosumer side, and improve the operation stability of the power system. However, most of the research rarely involves the impact of distribution network operation constraints on prosumer distributed transactions. In view of this, providing a two-layer optimization method for prosumer distributed transactions based on distribution network operation constraints has become an urgent problem to be solved in this field. Summary of the Invention

[0004] The purpose of the present invention is to provide a two-layer optimization method for prosumer distributed transactions based on distribution network operation constraints, update the bilateral transaction coefficients between prosumers, and reduce the costs of prosumers.

[0005] To solve the above technical problems, the technical solution of the present invention is: a two-layer optimization method for prosumer distributed transactions based on distribution network operation constraints, including the following steps:

[0006] Step 1: Construct a two-layer optimization model for prosumer distributed transactions, specifically:

[0007] Step 101: Establish an upper-layer prosumer distributed transaction model;

[0008] Step 102: Establish a lower-layer distribution network optimization model;

[0009] Step 2: Solve the two-layer optimization model to obtain the distributed transaction strategy between prosumers, specifically:

[0010] Step 201: Conduct distributed solution for the upper-layer prosumer distributed trading model to obtain the distributed trading volumes of each prosumer; input the distributed trading volumes of each prosumer into the lower-layer distribution network optimization model to update the bilateral trading coefficients between prosumers, and return the bilateral trading coefficients to the upper-layer prosumer distributed trading model;

[0011] Step 202: Iteratively solve the two-layer optimization model to minimize the trading costs of each prosumer and the operation cost of the distribution network, and obtain the distributed trading strategy among multiple prosumers under the operation constraints of the distribution network.

[0012] The present invention has the following beneficial effects:

[0013] The present invention proposes a two-layer optimization method for prosumer distributed trading based on the operation constraints of the distribution network, which broadens the trading channels between prosumers, realizes the balance of internal supply and demand and the consumption of new energy, reduces its own costs, and promotes more prosumers to actively enter the market to participate in trading. The present invention is based on the optimization of the distribution system, and the distribution network optimization model can optimize and update the bilateral trading coefficients between prosumers, reducing the costs of prosumers. The alternating direction multiplier method is used to solve the upper-layer prosumer distributed trading model. Compared with the centralized solution method, it can protect the privacy of prosumers and reduce the communication pressure. Description of the Drawings

[0014] Figure 1 is the flowchart of the method of the present invention;

[0015] Figure 2 is the schematic diagram of the two-layer model structure;

[0016] Figure 3 is the topological diagram of the improved IEEE33-node distribution system;

[0017] Figure 4(1) shows the internal energy management of prosumers under the photovoltaic low output scenario 2;

[0018] Figure 4(2) shows the internal energy management of prosumers under the photovoltaic high output scenario 7;

[0019] Figure 5(1) shows the energy trading situation of prosumer 1;

[0020] Figure 5(2) shows the energy trading situation of prosumer 2;

[0021] Figure 5(3) shows the energy trading situation of prosumer 3;

[0022] Figure 6 are the grid purchase and sale electricity prices and the trading prices between prosumers. Detailed Embodiment

[0023] To make the objectives, technical solutions, and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0024] Please refer to Figure 1 and Figure 2 The present invention is a two - layer optimization method for prosumer distributed transactions based on distribution network operation constraints, including the following steps:

[0025] Step 1: Construct a two - layer optimization model for prosumer distributed transactions, specifically:

[0026] Step 101: Establish an upper - layer prosumer distributed transaction model;

[0027] Step 102: Establish a lower - layer distribution network optimization model;

[0028] Step 2: Solve the two - layer optimization model to obtain the distributed transaction strategy among prosumers, specifically:

[0029] Step 201: Conduct distributed solution for the upper - layer prosumer distributed transaction model to obtain the distributed transaction volume of each prosumer; input the distributed transaction volume of each prosumer into the lower - layer distribution network optimization model to update the bilateral transaction coefficient among prosumers, and return the bilateral transaction coefficient to the upper - layer prosumer distributed transaction model;

[0030] Step 202: Iteratively solve the two - layer optimization model to minimize the transaction cost of each prosumer and the operation cost of the distribution network, and obtain the distributed transaction strategy among multiple prosumers under the distribution network operation constraints.

[0031] The input data of the present invention are the purchase and sale electricity prices, the basic load of each prosumer, the internal photovoltaic output, the fuel cell parameters, the energy storage parameters, etc. First, for different distributed resources, the present invention constructs an upper - layer model for prosumer distributed transactions; secondly, based on the influence of distribution network operation on prosumer distributed transactions, a two - layer optimization model is constructed, where the upper - layer model is the prosumer distributed transaction model and the lower - layer model is the distribution network optimization model; finally, by iteratively solving the two - layer optimization model, the transaction cost of each prosumer and the operation cost of the distribution network are minimized to realize the distributed transactions among multiple prosumers under the distribution network operation constraints. The present invention fully exploits the scheduling potential of the internal flexible resources of prosumers, suppresses the uncertainty of new - energy output, promotes the local consumption of new energy, reduces the transaction cost of prosumers, encourages more prosumers to participate in market transactions, and improves the economic efficiency and security of system operation.

[0032] The following will specifically describe each of the above steps.

[0033] Step 101 includes the following steps:

[0034] Step 1011: Taking the minimum cost of prosumers' participation in the electricity market for distributed transactions as the goal, establish the objective function of the prosumer distributed transaction model, which is expressed by the formula as follows:

[0035]

[0036] In the formula: the subscript x represents the prosumer; the subscript t represents the trading period; the subscript s represents the photovoltaic output scenario; ρ s represents the probability of each scenario; C all represents the total cost of the prosumer; represents the cost of the prosumer x trading with the power grid at time t; represents the energy storage cost of the prosumer x in scenario s at time t; represents the air-conditioning load cost of the prosumer x in scenario s at time t; represents the flexible load cost of the prosumer x in scenario s at time t; represents the fuel cell cost of the prosumer x in scenario s at time t; C x,t represents the distributed transaction cost of the prosumer x with other prosumers at time t;

[0037] Step 1012: Establish the constraint conditions of the prosumer distributed transaction model, including energy storage constraints, central air-conditioning constraints, flexible load constraints, fuel cell constraints, and distributed transaction constraints.

[0038] Specifically, in Step 1011,

[0039] The formula for the cost of trading with the power grid is expressed as follows:

[0040]

[0041] In the formula: represents the cost of the prosumer x trading with the power grid at time t; represents the electricity purchase price in the electricity market during period t; represents the electricity purchase and sale prices in the electricity market during period t; represents the electricity quantity purchased by the virtual power plant x from the electricity market during period t; represents the electricity quantity sold by the virtual power plant x to the electricity market during period t;

[0042] The formula for the energy storage cost is expressed as follows:

[0043]

[0044] In the formula: represents the energy storage cost of the prosumer x in scenario s at time t; represents the charging quantity of the energy storage of the prosumer x during period t in scenario s; Represents the discharge amount of the energy storage of prosumer x during time period t in scenario s; ε c Represents the charging dissipation coefficient of the energy storage; ε d Represents the discharge dissipation coefficient of the energy storage;

[0045] The cost formulas for central air conditioners and flexible loads are expressed as follows:

[0046]

[0047]

[0048] In the formula: Represents the air-conditioning load cost of prosumer x at time t in scenario s; Represents the flexible load cost of prosumer x at time t in scenario s; m and α are user discomfort coefficients; Is the indoor temperature of prosumer x during time period t in scenario s; Is the most comfortable temperature for the user's body feeling inside prosumer x; Is the flexible load value of the user of prosumer x during time period t in scenario s; Is the load reference value of the user of prosumer x during time period t;

[0049] The cost formula of the fuel cell is expressed as follows:

[0050]

[0051] In the formula: Represents the fuel cell cost of prosumer x at time t in scenario s; c on Is the operating cost coefficient of the fuel cell; Is the power generation power of the fuel cell of prosumer x during time period t in scenario s;

[0052] The distributed transaction cost formula is expressed as follows:

[0053]

[0054] In the formula: C x,t Represents the distributed transaction cost of prosumer x with other prosumers at time t; c x,y Represents the bilateral transaction coefficient between prosumer x and prosumer y; P x,y,t Represents the energy transaction volume between prosumer x and prosumer y at time t.

[0055] Specifically, in step 1012,

[0056] The fuel cell constraint formula is expressed as follows:

[0057]

[0058]

[0059] In the formula: represents the minimum value of the fuel cell power within prosumer x; represents the maximum value of the fuel cell power within prosumer x; is the power generation power of the fuel cell of prosumer x during time period t in scenario s; represents the upward ramp rate of the fuel cell within prosumer x; represents the downward ramp rate of the fuel cell within prosumer x;

[0060] The energy storage constraint formula is expressed as follows:

[0061]

[0062]

[0063]

[0064] S min ≤ S x,s,t ≤ S max

[0065] In the formula: P c,max represents the maximum value of the energy storage charging power; P d,max represents the maximum value of the energy storage discharging power; represents the charging amount of the energy storage of prosumer x during time period t in scenario s; represents the discharging amount of the energy storage of prosumer x during time period t in scenario s; S x,s,t is the state of charge of the energy storage of prosumer x during time period t in scenario s; represents the minimum stored electricity of the energy storage system within prosumer x; represents the maximum stored electricity of the energy storage system within prosumer x; represents the charging efficiency of the energy storage within prosumer x; represents the discharging efficiency of the energy storage within prosumer x;

[0066] The central air - conditioning constraint formula is expressed as follows:

[0067]

[0068] In the formula: is the indoor temperature of prosumer x during time period t in scenario s; represents the outdoor temperature of prosumer x during time period t in scenario s; Δt represents the time interval; C x 、R x represent the physical parameters of the air - conditioner; η represents the energy efficiency ratio of the refrigeration unit; Indicates the operating power of the internal building air conditioner of prosumer x at time t in scenario s;

[0069] Based on user comfort, the indoor temperature is restricted within an adjustable temperature range:

[0070]

[0071] In the formula: T in,min is the minimum indoor temperature, and T in,max is the maximum indoor temperature;

[0072] The flexible load constraint formula is expressed as follows:

[0073]

[0074]

[0075] In the formula: is the flexible load value of the user of prosumer x at time t in scenario s; is the load reference value of the user of prosumer x within time t; represents the minimum flexible load of prosumer x at time t; represents the maximum flexible load of prosumer x at time t;

[0076] The distributed transaction balance constraint formula is expressed as follows:

[0077] P x,y,t +P y,x,t =0 y≠x

[0078]

[0079] In the formula: P x,y,t is the energy trading volume between prosumer x and prosumer y within time t; P y,x,t is the energy trading volume between prosumer y and prosumer x within time t; g x,s,t is the PV output of prosumer x at time t in scenario s; represents the electricity purchased from the power grid by prosumer x within time t in scenario s; represents the electricity sold to the power grid by prosumer x within time t in scenario s.

[0080] Step 102 includes the following steps:

[0081] Step 1021: With the goal of minimizing the operating cost of the distribution network, establish the objective function of the distribution network optimization model, which is expressed as follows:

[0082]

[0083] Where: subscript t represents the trading period; subscript i represents the distribution network node; is the power purchase cost from the superior power grid in period t; is the load shedding cost of node i in period t;

[0084] Step 1022: Establish the model constraints of the distribution network optimization model, including the branch power flow model, upper and lower limits of voltage amplitude constraints, branch capacity constraints, and power balance constraints;

[0085] Step 1023: Define the node marginal price according to the active power balance equation of the distribution network node.

[0086] Specifically, in step 1021,

[0087] Power purchase and sale cost and distributed transaction cost:

[0088]

[0089] Where: represents the cost of the prosumer x trading with the power grid at time t in the upper layer model; C x,t represents the distributed transaction cost of the prosumer x trading with other prosumers at time t in the upper layer model;

[0090] Power purchase cost from the superior power grid:

[0091]

[0092] Where: represents the wholesale price in period t; P t sub is the active power purchased from the superior power grid in period t;

[0093] Load shedding cost:

[0094]

[0095] Where: represents the unit load shedding cost coefficient in period t; represents the demand power of node i in period t; represents the curtailed power of node i in period t.

[0096] Specifically, in step 1022,

[0097] Branch power flow model:

[0098]

[0099]

[0100] Where: subscripts i, j, k represent the distribution network nodes; Pi,t Denote the active power injection of node i at time period t; Q i,t Denote the reactive power injection of node i at time period t; P ij,t Denote the active power on branch ij at time period t; Q ij,t Denote the reactive power on branch ij at time period t; P ki,t Denote the active power on branch ki at time period t; Q ki,t Denote the reactive power on branch ki at time period t; F(i) represents the set of end nodes of the branches with node i as the starting end point; T(i) represents the set of starting end nodes of the branches with node i as the end node; l ij,t Denote the square of the current amplitude on branch ij at time period t; l ki,t Denote the square of the current amplitude on branch ki at time period t; r ij Denote the resistance of branch ij; x ij Denote the reactance of branch ij; r ki Denote the resistance of branch ki; x ki Denote the reactance of branch ki; V i,t Denote the square of the voltage amplitude at node i at time period t; V j,t Denote the square of the voltage amplitude at node j at time period t; N s Denote the set of system nodes; B s Denote the set of system branches;

[0101] Voltage amplitude upper and lower limit constraint:

[0102] V i min ≤V i,t ≤V i max

[0103] In the formula: V i min 、V i max respectively represent the minimum and maximum values of the voltage amplitude of node i;

[0104] Branch capacity constraint:

[0105]

[0106] Power balance constraint:

[0107]

[0108]

[0109]

[0110]

[0111] Where: P i,t represents the active power injection of node i at time period t; represents the electricity sold by the prosumer connected to node i at time period t; represents the electricity purchased by the prosumer connected to node i at time period t; represents the electricity for distributed transactions between prosumers connected to node i and other prosumers at time period t; Q i,t represents the reactive power injection of node i at time period t; represents the reactive power purchased from the superior power grid at time period t; tanψ represents the tangent value of the load power factor angle.

[0112] Specifically, in step 1023,

[0113] Active power balance equation of distribution network nodes:

[0114]

[0115] Define the Lagrangian dual multiplier of this formula as the marginal price of distribution network nodes, and obtain the marginal price of distribution network nodes of node i at time period t as λ i,t ;

[0116] If prosumer x is connected to node i and prosumer y is connected to node j, then λ x,t = λ i,t , λ y,t = λ j,t , then the bilateral transaction coefficient between prosumer x and prosumer y is expressed as:

[0117] c ij,t = (λ i,t - λ j,t ) / 2.

[0118] Step 201 includes the following steps:

[0119] Step 2011: Initialize the number of iterations k = 0 for solving the upper-layer prosumer distributed transaction model, the number of iterations n = 0 for solving the prosumer distributed transaction two-layer optimization model, and the bilateral transaction coefficient between prosumers

[0120] Step 2012: Use the ADMM algorithm to perform distributed solution on the upper-layer prosumer distributed transaction model to obtain the distributed transaction volume of each prosumer; In this embodiment, step 2012 includes the following steps:

[0121] Step 20121: Use the ADMM distributed algorithm to solve the upper-layer prosumer distributed transaction model, and the Lagrangian function is:

[0122]

[0123] Where: C all is the total cost of prosumers participating in distributed energy trading; u i,j,t is the dual variable, defined as the price of the energy trading P x,y,t between prosumer x and prosumer y at time period t; X x,y,t is the auxiliary variable; ρ is the penalty factor, i.e., the step size;

[0124] Step 20122: Iterate on the energy trading P i,j,t , the auxiliary variable X i,j,t and the dual variable, and the formula is expressed as follows:

[0125]

[0126]

[0127]

[0128] Where: k is the number of iterations; the superscripts k and k + 1 represent the k-th and (k + 1)-th iterations respectively;

[0129] Step 20123: Calculate the primal residual and the dual residual after each iteration, and the formula is expressed as follows:

[0130]

[0131] Where: are the primal residual and the dual residual of the energy of prosumer x and prosumer y at time period t in the distributed transaction of the (k + 1)-th iteration respectively;

[0132] Step 20124: Determine whether the ADMM algorithm converges through the iteration stop condition until the iteration stop condition is satisfied, and the upper-layer prosumer distributed transaction model stops iterating; the iteration stop condition formula is expressed as follows:

[0133]

[0134] Where: ε pri and ε dual are the upper tolerance limits of the primal residual and the dual residual respectively.

[0135] Step 2013: Obtain the electricity purchase and sale volumes of each prosumer and input the electricity purchase and sale volumes of each prosumer and the distributed trading volume P x,y,t into the lower-layer distribution network optimization model to obtain the nodal marginal price λ i,t of each node, so as to obtain the updated bilateral trading coefficient And return the bilateral trading coefficient to the upper-layer prosumer distributed trading model.

[0136] Step 202 specifically includes: repeating the iterative solution of steps 2012 - 2013 until the bilateral trading coefficient reaches the iterative condition, that is, obtaining the prosumer distributed trading strategy output by the current upper-layer prosumer distributed trading model; in this step, the iterative condition is: Wherein: respectively represent the bilateral trading coefficients of the nth and (n + 1)th iterations between prosumer x and prosumer y, and ε c represents the tolerance upper limit. In this embodiment, the GAMS software is used to solve the above model to obtain the prosumer distributed trading strategy under the two-layer optimization model.

[0137] In this embodiment, a simulation example of three prosumers incorporated into the distribution network is used to verify the effectiveness of the proposed method. Among them, all 3 prosumers include photovoltaic, energy storage system, air-conditioning load and flexible load, and prosumer 1 also includes a fuel cell. The parameters of the fuel cell, energy storage, central air-conditioning, and photovoltaic are shown in Table 1. The time-of-use average purchase and sale electricity prices are shown in Table 2.

[0138] Table 1 Parameters of each device of the prosumer

[0139]

[0140] Table 2 Purchase and sale electricity prices from the power grid

[0141]

[0142] Taking prosumer 1 as an example, the internal energy management situation based on the uncertain output of photovoltaic is shown in Figures 4(1) and 4(2). Comparing the low photovoltaic output scenario 2 and the high photovoltaic output scenario 7, the multi-load scheduling inside the prosumer under uncertain photovoltaic output is analyzed. When the electricity purchase price is low (1:00 - 7:00), the energy storage system charges; when the electricity purchase price is high (12:00 - 19:00), the energy storage system discharges, which can reduce the electricity purchased from outside and lower its own cost. It can be seen from the figure that at the moments of low photovoltaic output (1:00 - 7:00) and low electricity consumption valleys (15:00 - 17:00), the air-conditioning load and flexible load are low, and at the moments of high photovoltaic output (12:00 - 14:00) and high electricity consumption peaks (18:00 - 21:00), the air-conditioning load and flexible load are high. Therefore, the air-conditioning load and flexible load can be adjusted and the flexibility of the energy storage can be fully utilized to smooth the uncertain photovoltaic output. Under scenarios 2 and 7, the net load of the prosumer is the same. Therefore, the load can be adjusted inside the prosumer to achieve the consumption of uncertain photovoltaic and obtain the deterministic trading volume.

[0143] The specific trading situations of prosumers are shown in Figures 5(1), (2), and (3). In the figures, the positive direction indicates that the prosumer purchases electric energy, and the negative direction indicates that the prosumer sells electric energy. The trading volume of the prosumer is positive and negative, indicating that the prosumer can participate in the transaction as a buyer or a seller, and the buying and selling identities are endogenous. By comparing Figures 5(1), (2), and (3), it can be seen that due to the large photovoltaic output and the inclusion of fuel cells of prosumer 1, it mainly participates in energy trading as a "power-dominated" seller, while prosumers 2 and 3 mainly participate in energy trading as "load-dominated" buyers. During the low photovoltaic output period, the output of the three prosumers cannot meet the internal load demand, and they need to purchase electricity from the power grid. While during the high photovoltaic output period, the three prosumers only need to conduct distributed trading to meet the load demand and do not need to purchase electricity from the power grid anymore, and the distributed trading volume increases. From Figure 6 It can be seen that the distributed trading price is between the power grid trading purchase and sale prices. The distributed trading of prosumers can reduce the trading cost, so it encourages more prosumers to participate in distributed trading and realizes the sharing of electric energy.

[0144] The trading costs of each prosumer are shown in Table 3. Compared with directly trading with the power grid, although the cost of the "power-dominated" prosumer 1 increases to a certain extent when the prosumer participates in distributed trading, the costs of the "load-dominated" prosumers 2 and 3 decrease significantly, and the total cost of participating in distributed trading is lower. And under the two-layer model, by updating the bilateral trading coefficients between prosumers, since the bilateral trading coefficients decrease, the distributed trading volume increases and the cost decreases, which proves the economy of the two-layer optimization model.

[0145] Table 3 Comparison of trading costs

[0146]

[0147] The parts not involved in the present invention are the same as or implemented by the prior art.

[0148] The above content is a further detailed description of the present invention in combination with specific implementation manners. It cannot be determined that the specific implementation of the present invention is only limited to these descriptions. For those of ordinary skill in the technical field to which the present invention belongs, without departing from the concept of the present invention, several simple deductions or substitutions can still be made, which should all be regarded as belonging to the protection scope of the present invention.

Claims

1. A two - layer optimization method for prosumer distributed trading based on distribution network operation constraints, Characterized in that: It includes the following steps: Step 1: Construct a two - layer optimization model for prosumer distributed trading, specifically: Step 101: Establish an upper - layer prosumer distributed trading model, specifically including: Step 1011: Taking the minimum cost of prosumers participating in the electricity market for distributed trading as the goal, establish the objective function of the prosumer distributed trading model, and the formula is expressed as follows: Where: the subscript x represents prosumers; the subscript t represents trading periods; the subscript s represents PV output scenarios; ρ s represents the probability of each scenario; C all represents the total cost of prosumers; represents the cost of prosumer x's transaction with the power grid at time t; represents the energy storage cost of prosumer x in scenario s at time t; represents the air conditioning load cost of prosumer x in scenario s at time t; represents the flexible load cost of prosumer x in scenario s at time t; represents the fuel cell cost of prosumer x in scenario s at time t; C x,t represents the distributed transaction cost of prosumer x with other prosumers at time t; Step 1012: Establish the constraint conditions of the prosumer distributed trading model, including energy storage constraints, central air - conditioning constraints, flexible load constraints, fuel cell constraints, and distributed trading constraints Step 102: Establish a lower - layer distribution network optimization model, specifically including: Step 1021: Taking the lowest operation cost of the distribution network as the goal, establish the objective function of the distribution network optimization model, and the formula is expressed as follows: In the formula: the subscript t represents the trading period; the subscript i represents the distribution network node; is the power purchase cost from the superior power grid in period t; is the load shedding cost of node i in period t; Step 1022: Establish the model constraints of the distribution network optimization model, including branch power flow model, upper and lower limits of voltage amplitude constraints, branch capacity constraints, and power balance constraints; Step 1023: Define the nodal marginal price according to the active power balance equation of the distribution network nodes; Step 2: Solve the two - layer optimization model to obtain the distributed trading strategy among prosumers; specifically: Step 201: Conduct distributed solution on the upper - layer prosumer distributed trading model to obtain the distributed trading volume of each prosumer; input the distributed trading volume of each prosumer into the lower - layer distribution network optimization model to update the bilateral trading coefficient among prosumers, and return the bilateral trading coefficient to the upper - layer prosumer distributed trading model; Step 202: Iteratively solve the two - layer optimization model to minimize the trading costs of each prosumer and the operation cost of the distribution network, and obtain the distributed trading strategy among multiple prosumers under the distribution network operation constraints.

2. The two - layer optimization method for prosumer distributed trading based on distribution network operation constraints according to claim 1, Characterized in that: In step 1011, The formula for the trading cost with the power grid is expressed as follows: In the formula: Represents the cost of prosumer x trading with the power grid at time t; Represents the electricity purchase price in the power market during period t; Represents the electricity purchase and sale prices in the power market during period t; Represents the electricity quantity purchased by virtual power plant x from the power market during period t; Represents the electricity quantity sold by virtual power plant x to the power market during period t; The formula for the energy storage cost is expressed as follows: In the formula: represents the energy storage cost of prosumer x at time t in scenario s; represents the charging amount of the energy storage of prosumer x during time period t in scenario s; represents the discharging amount of the energy storage of prosumer x during time period t in scenario s; ε c represents the charging dissipation coefficient of the energy storage; ε d represents the discharging dissipation coefficient of the energy storage; The formula for the calling cost of central air - conditioning and flexible load is expressed as follows: Wherein: represents the air-conditioning load cost of prosumer x at time t in scenario s; represents the flexible load cost of prosumer x at time t in scenario s; m and α are user discomfort coefficients; T in x,s,t is the indoor temperature of prosumer x during time period t in scenario s; is the most comfortable temperature for the user's body feeling inside prosumer x; is the flexible load value of the user of prosumer x during time period t in scenario s; is the load reference value of the user of prosumer x during time period t; The formula for the fuel cell cost is expressed as follows: In the formula: represents the fuel cell cost of prosumer x in scenario s at time t; c on is the operating cost coefficient of the fuel cell; is the power generation of the fuel cell of prosumer x during time period t in scenario s; The formula for the distributed trading cost is expressed as follows: Where: C x,t represents the distributed transaction cost of prosumer x with other prosumers at time t; c x,y represents the bilateral transaction coefficient between prosumer x and prosumer y; P x,y,t represents the energy transaction volume between prosumer x and prosumer y at time t.

3. The two - layer optimization method for prosumer distributed trading based on distribution network operation constraints according to claim 1, Characterized in that: In step 1012, The fuel cell constraint formula is expressed as follows: In the formula: represents the minimum value of the fuel cell power within prosumer x; represents the maximum value of the fuel cell power within prosumer x; is the power generation of the fuel cell within prosumer x during time period t in scenario s; represents the upward ramp rate of the fuel cell within prosumer x; represents the downward ramp rate of the fuel cell within prosumer x; The energy storage constraint formula is expressed as follows: S min ≤ S x,s,t ≤ S max ; Where: P c,max represents the maximum value of the energy storage charging power; P d,max represents the maximum value of the energy storage discharging power; represents the charging amount of the energy storage of prosumer x during period t in scenario s; represents the discharging amount of the energy storage of prosumer x during period t in scenario s; S x,s,t is the state of charge of the energy storage of prosumer x during period t in scenario s; represents the minimum stored electricity of the energy storage system within prosumer x; represents the maximum stored electricity of the energy storage system within prosumer x; represents the charging efficiency of the energy storage within prosumer x; represents the discharging efficiency of the energy storage within prosumer x; The central air - conditioning constraint formula is expressed as follows: In the formula: is the indoor temperature of prosumer x during time period t in scenario s; represents the outdoor temperature of prosumer x under scenario s at time period t; Δt represents the time interval; C x , R x represent the physical parameters of the air conditioner; η represents the energy efficiency ratio of the refrigeration unit; represents the operating power of the internal building air conditioner of prosumer x under scenario s at time period t; Based on user comfort, limit the indoor temperature within the adjustable temperature range: where: T in,min is the minimum indoor temperature, and T in,max is the maximum indoor temperature; The flexible load constraint formula is expressed as follows: In the formula: is the flexible load value of prosumer x for the user during the period t in the internal scenario s; is the load reference value of the user for prosumer x during the period t; represents the minimum flexible load of prosumer x during the period t; represents the maximum flexible load of prosumer x during the period t; The distributed trading balance constraint formula is expressed as follows: P x,y,t +P y,x,t = 0, y ≠ x; Where: P x,y,t is the energy trading volume between prosumer x and prosumer y during time period t; P y,x,t represents the energy trading volume between prosumer y and prosumer x at time t; g x,s,t is the photovoltaic output of prosumer x during time period t in scenario s; represents the electricity purchased by prosumer x from the power grid during time period t in scenario s; represents the electricity sold by prosumer x to the power grid during time period t in scenario s.

4. The two - layer optimization method for prosumer distributed trading based on distribution network operation constraints according to claim 1, Characterized in that: In step 1021, The cost of buying and selling electricity and the distributed trading cost: In the formula: represents the cost of the prosumer x trading with the power grid at time t in the upper-level model; C x,t represents the distributed trading cost of the prosumer x trading with other prosumers at time t in the upper-level model; The cost of buying electricity from the superior power grid: Wherein: represents the wholesale price at time period t; is the active power of power purchase from the superior power grid at time period t; The cost of cutting off load: In the formula: represents the unit cost coefficient of cutting electricity load at time period t; represents the demand power of node i at time period t; represents the power curtailment of node i at time period t.

5. The two - layer optimization method for prosumer distributed trading based on distribution network operation constraints according to claim 1, Characterized in that: In step 1022, The branch power flow model: In the formula: subscripts i, j, k represent the nodes of the distribution network; P i,t represents the active power injection of node i at time t; Q i,t represents the reactive power injection of node i at time t; P ij,t represents the active power of branch ij at time t; Q ij,t represents the reactive power of branch ij at time t; P ki,t represents the active power of branch ki at time t; Q ki,t represents the reactive power of branch ki at time t; F(i) represents the set of end nodes of the branch with node i as the starting end point; $T(i)$ represents the set of the head nodes of the branches with node $i$ as the end node; $l$ ij,t represents the square of the current amplitude at time period $t$ on branch $ij$; $l$ ki,t represents the square of the current amplitude at time period $t$ on branch $ki$; $r$ ij represents the resistance of branch $ij$; $x$ ij represents the reactance of branch $ij$; $r$ ki represents the resistance of branch $ki$; $x$ ki represents the reactance of branch $ki$; $V$ i,t represents the square of the voltage amplitude at node $i$ at time period $t$; $V$ j,t represents the square of the voltage amplitude at node $j$ at time period $t$; $N$ s represents the set of system nodes; $B$ s represents the set of system branches; The upper and lower limits of voltage amplitude constraints: Where: respectively represent the minimum and maximum voltage magnitudes of node i; Branch capacity constraint: Power balance constraint: Where: P i,t represents the active power injection of node i at time period t; represents the electricity sold by the prosumer connected to node i at time period t; represents the electricity purchased by the prosumer connected to node i at time period t; represents the electricity for distributed transactions between the prosumer connected to node i and other prosumers at time period t; Q i,t represents the reactive power injection of node i at time period t; represents the reactive power purchased from the superior power grid at time period t; tanψ represents the tangent value of the load power factor angle.

6. The prosumer distributed trading two - layer optimization method based on distribution network operation constraints according to claim 1, characterized in that: in step 1023, Active power balance equation of distribution network nodes: Define the Lagrangian dual multiplier of this formula as the nodal marginal price of the distribution network. The nodal marginal price of node i at time t in the distribution network is obtained as λ i,t ; If prosumer x connects to node i and prosumer y connects to node j, then λ x,t = λ i,t , λ y,t = λ j,t , then the bilateral trading coefficient between prosumer x and prosumer y is expressed as: c ij,t = (λ i,t - λ j,t ) / 2。 7. The prosumer distributed trading two - layer optimization method based on distribution network operation constraints according to claim 1, characterized in that: step 201 is specifically as follows: Step 2011: Initialize the number of iterations k = 0 for solving the upper-layer prosumer distributed transaction model, the number of iterations n = 0 for solving the two-layer optimization model of the prosumer distributed transaction, and the bilateral transaction coefficient between prosumers step 2012: Use the ADMM algorithm to perform distributed solution on the upper - layer prosumer distributed trading model to obtain the distributed trading volumes of each prosumer; specifically including the following steps: step 20121: Use the ADMM distributed algorithm to solve the upper - layer prosumer distributed trading model, and the Lagrangian function is: Where: C all is the total cost of prosumers participating in distributed energy trading; u x,y,t is the dual variable, defined as the price of the energy trading P x,y,t between prosumer x and prosumer y at time period t; X x,y,t is the auxiliary variable; ρ is the penalty factor, i.e., the step size; Step 20122: Iterate on the energy transaction P x,y,t , the auxiliary variable X x,y,t and the dual variable, and the formula is expressed as follows: where: k is the number of iterations; the superscripts k and k + 1 represent the k - th and (k + 1)-th iterations respectively; step 20123: Calculate the primal residual and dual residual after each iteration, and the formula is expressed as follows: In the formula: are respectively the original residual and dual residual of the energy of prosumer x and prosumer y in time period t in the (k + 1)-th iterative distributed transaction; step 20124: Judge whether the ADMM algorithm converges through the iteration stop condition until the iteration stop condition is met, and the upper - layer prosumer distributed trading model stops iterating; the iteration stop condition formula is expressed as follows: where: ε pri and ε dual are the upper tolerance limits of the original residual and the dual residual, respectively; step 2013: Obtain the power purchase and sale volumes of each prosumer, and input the power purchase and sale volumes and distributed trading volumes of each prosumer into the lower - layer distribution network optimization model to obtain the nodal marginal price of each node, thereby obtaining the updated bilateral trading coefficient between prosumers, and return the bilateral trading coefficient to the upper - layer prosumer distributed trading model.

8. The prosumer distributed trading two - layer optimization method based on distribution network operation constraints according to claim 7, characterized in that: Step 202 is specifically as follows: Repeat the iterative solution of steps 2012 - 2013 until the bilateral trading coefficient reaches the iterative condition, that is, the difference in the bilateral trading coefficients between the nth and (n + 1)th iterations between prosumer x and prosumer y. When it is less than or equal to the preset tolerance upper limit, obtain the prosumer distributed trading strategy output by the current upper-layer prosumer distributed trading model.

Images