ADN distributed optimization method and system based on enhanced price update P2P transaction

Abstract

The invention is suitable for the technical field of active power distribution network operation optimization, and provides an ADN distributed optimization method and system based on enhanced price update P2P transaction, and the method comprises the following steps: constructing a double-layer distributed optimization model comprising a producer and consumer P2P transaction layer and a physical optimization operation layer; the method comprises the following steps: iteratively optimizing an electric energy transaction price and a transaction volume between adjacent consumer consumers and / or between a consumer consumer and a superior main network based on an enhanced price updating mechanism by taking minimization of the operating cost of the consumer consumers as a target, and outputting an optimized active power transaction plan; on the basis of the optimized active power transaction plan, node voltage is calculated, and reactive power output of the producer and the consumer is optimized; judging whether all node voltages meet safe operation constraints or not; if yes, outputting a final optimization result; and if not, adjusting the active power transaction upper limit of each producer, and returning to the P2P transaction layer of the producer to carry out iterative optimization again.

Description

Technical Field

[0001] This invention belongs to the field of active distribution network operation optimization technology, and in particular relates to an ADN decentralized optimization method and system based on enhanced price update P2P transactions. Background Technology

[0002] With the large-scale integration of distributed generation (DG), represented by distributed photovoltaic (DPV), and battery storage devices (BSD), the traditional distribution network has undergone a transformation from passive to active power. On the one hand, this transformation can effectively integrate and utilize different types of renewable energy, promoting the construction and development of low-carbon, green power systems. On the other hand, the high proportion of DG integration, characterized by strong randomness and uncertainty, also brings technical challenges to the operation of the distribution network, such as frequent power flow reversals and power congestion, and voltage fluctuations exceeding limits.

[0003] To address the aforementioned issues, existing research has explored energy management and operational optimization of distribution networks through the control and scheduling of different types of flexible resources. However, when the ownership of controllable equipment differs or conflicts of interest exist, the control measures implemented by the distribution system operator (DSO) with the goal of system safety and economic operation may become ineffective to some extent. Therefore, for systems containing multiple entities that combine power production and consumption functions (prosumers), a relatively relaxed peer-to-peer (P2P) power trading model can enable each prosumer to actively participate in distribution network control. Combined with a reasonable market clearing mechanism, this can achieve economically optimized operation of the active distribution network (ADN) while ensuring the benefits for all parties. Most P2P trading models only aim at the economic benefits of prosumers and do not fully consider the physical operational constraints of the distribution network (such as voltage safety and power flow), which can easily lead to problems such as voltage exceeding limits and line congestion after the transaction, failing to guarantee operational safety. Some schemes that consider physical constraints rely on centralized control by the DSO, which has drawbacks such as user privacy leakage, single points of failure, and high communication complexity.

[0004] Traditional P2P transaction prices are updated using a fixed iteration step size, which can easily lead to price oscillations when supply and demand fluctuations are small, and make it difficult to quickly guide the market to restore balance when supply and demand are severely imbalanced. In addition, the lack of a price benchmark means that transaction prices can easily deviate from actual value in extreme scenarios, failing to provide reliable price protection for producers and consumers and weakening the macro-control capabilities of the electricity market. Summary of the Invention

[0005] The purpose of this invention is to provide an ADN decentralized optimization method for P2P transactions based on enhanced price updates, in order to solve the aforementioned technical problems.

[0006] This invention is implemented as follows: an ADN decentralized optimization method based on enhanced price update P2P transactions, comprising the following steps:

[0007] Construct a two-layer decentralized optimization model that includes a producer-consumer P2P transaction layer and a physical optimization operation layer;

[0008] In the producer-consumer P2P transaction layer, with the goal of minimizing the operating costs of producers and consumers, an enhanced price update mechanism is used to iteratively optimize the electricity transaction price and transaction volume between adjacent producers and consumers and / or between producers and consumers and the upper-level main grid, and output the optimized active power transaction plan. The enhanced price update mechanism realizes the adaptive adjustment of the iteration step size of the ADMM algorithm through the tanh function, and introduces a transaction price anchoring term with the upper-level main grid retail electricity price and the producer-consumer on-grid electricity price as the boundary.

[0009] In the physical optimization operation layer, based on the optimized active power trading plan, the ADMM algorithm is used to solve the fully decentralized reactive power / voltage optimization model, calculate the node voltage, and optimize the reactive power output of producers and consumers.

[0010] Determine whether the voltage of all nodes meets the safe operation constraints; if it does, output the final optimization result; if it does not, adjust the active power trading limit of each producer-consumer according to the voltage limit violation situation, and return to the producer-consumer P2P transaction layer for iterative optimization again.

[0011] Furthermore, the prosumers include distributed photovoltaic (DPV), power generation equipment, and battery storage devices (BSD). In the prosumer P2P transaction layer, the method for minimizing the operating cost of prosumers is as follows: For any prosumer i, construct its cost minimization objective function within the scheduling period T, as follows:

[0012] ;

[0013] ;

[0014] ;

[0015] In the formula, , These represent the total electricity sold by producer-consumer i at time t and the revenue per unit of electricity sold, respectively. , These represent the electricity purchased by producer-consumer i from producer-consumer j at time t, and the unit electricity purchase cost, respectively. , These represent the power purchased by producer-consumer i from the upper-level power grid at time t and the unit power purchase cost, respectively. , These represent the electricity sold by producer-consumer i to the upper-level power grid at time t and the unit electricity sales cost, respectively. , and , These represent the output cost and active power output of the power generation equipment and BSD in producer-consumer i at time t, respectively. The unit energy transaction cost charged by the distribution network operator (DSO) to the buyer at time t; , and This refers to the cost coefficient for the corresponding item; The total number of prosumers in the system; Let i be the set of neighboring nodes of node i; This indicates the cost of electricity transactions between producers and consumers; This represents the cost of electricity transactions between producers / consumers and the upper-level power grid.

[0016] Furthermore, the transaction price update formula for the enhanced price update mechanism is as follows:

[0017] ;

[0018] In the formula, This indicates the ratio of local power shortage or surplus among producers and consumers at time t; The hyperbolic tangent function is the adaptive gain factor. Anchor to the transaction price; is the iteration step size of the ADMM algorithm; k is the current iteration number.

[0019] Furthermore, in the transaction price update formula:

[0020] ;

[0021] ;

[0022] .

[0023] Furthermore, in the physical optimization operation layer, the reactive power / voltage optimization model aims to minimize the voltage deviation and active power loss at each producer-consumer node, specifically expressed as follows:

[0024] ;

[0025] In the formula, U is the square of the voltage at consumer node i at time t; max U min These are the upper and lower limits of the allowable node voltage, respectively; UN This is the system's rated voltage; , These represent the active and reactive power injected by the upstream producer-consumer into producer-consumer i, respectively; r i The line resistance between producer-consumer i and the upstream producer-consumer; This is the penalty coefficient for voltage exceeding the limit.

[0026] Furthermore, the constraints of the prosumer-consumer P2P transaction layer include power balance constraints, power generation equipment output constraints, BSD output constraints, and prosumer-consumer transaction constraints.

[0027] Furthermore, the constraints of the physical optimization operation layer include linear Dist-Flow power flow constraints, capacity and limit constraints of power generation equipment and BSD.

[0028] Furthermore, the adjustment formula for the active power trading ceiling is as follows:

[0029] ;

[0030] ;

[0031] In the formula, The upper limit of the total electricity sold by producer-consumer i at time t; Let t be the upper limit of the electricity sold by producer-consumer j to producer-consumer i at time t; The trading power is adjusted for each iteration; U is the square of the voltage at consumer node i at time t; max U min These are the upper and lower limits of the allowable node voltage, respectively. This represents the allowed operating voltage range; k is the current iteration number.

[0032] Furthermore, the ADN distributed optimization method is applicable to improved IEEE 33-node distribution networks that include distributed photovoltaic (DPV), power generation equipment, and battery storage devices (BSD), with an allowable operating voltage range of 0.95-1.05 pu.

[0033] Another objective of this invention is to provide an ADN decentralized optimization system based on enhanced price update P2P transactions, for implementing the aforementioned ADN decentralized optimization method, comprising:

[0034] The producer-consumer (P2P) transaction layer aims to minimize the operating costs of producers and consumers. Based on an enhanced price update mechanism, it uses the ADMM algorithm to iteratively optimize the electricity transaction price and volume between adjacent producers and consumers and / or between producers and consumers and the upstream main grid, outputting an optimized active power transaction plan. The enhanced price update mechanism uses the tanh function to adaptively adjust the iteration step size of the ADMM algorithm and introduces a transaction price anchoring term with the upstream main grid retail electricity price and the producer-consumer on-grid electricity price as the boundary.

[0035] The physical optimization operation layer is used to solve the fully decentralized reactive power / voltage optimization model based on the optimized active power trading plan using the ADMM algorithm, calculate node voltage, and optimize the reactive power output of producers and consumers.

[0036] The safety constraint module is used to determine whether the voltage of all nodes meets the safety operation constraints. If it does, the final optimization result is output. If it does not, the active power trading limit of each producer-consumer is adjusted according to the voltage limit violation, and the process is returned to the producer-consumer P2P transaction layer for iterative optimization.

[0037] This invention provides a decentralized optimization method for ADN based on enhanced price update P2P transactions. This method ensures the profitability of each trading entity and the safe and economical operation of the ADN in a P2P electricity trading environment. It mainly consists of two parts: a producer-consumer P2P trading layer and a physical optimization operation layer. The producer-consumer P2P trading layer minimizes the operating costs of producers by iteratively optimizing the electricity trading prices and volumes between adjacent producers or between producers and the upper-level main grid. Its trading prices are adaptively adjusted based on a preset enhanced price update mechanism to ensure profitability and transaction security under different supply-demand imbalance scenarios. The physical optimization operation layer performs fully decentralized reactive power / voltage optimization of the ADN based on the electricity prices optimized by the producer-consumer P2P trading layer. When voltage exceedances occur, feedback is given to the producer-consumer P2P trading layer, which adjusts the active power trading limits and actual trading volumes between producers and consumers to meet the system's voltage safety constraints as much as possible, ensuring the safe operation of the ADN. This method is effective and superior when applied to an improved IEEE 33-node distribution network containing multiple types of producers and consumers. Attached Figure Description

[0038] Figure 1 This is an overall framework diagram of the ADN decentralized optimization method based on enhanced price update P2P transactions provided in the embodiments of the present invention.

[0039] Figure 2 A schematic diagram illustrating the transaction logic of producers and consumers.

[0040] Figure 3 Schematic diagram for adaptive iterative step size update.

[0041] Figure 4 This is a diagram of an improved IEEE 33-node distribution network structure.

[0042] Figure 5 Net load diagram for producer-consumer nodes.

[0043] Figure 6 A diagram illustrating the P2P transaction process for different consumer schemes.

[0044] Figure 7 The diagram shows the system voltage distribution and reactive power optimization process for different schemes.

[0045] Figure 8 A price chart for P2P transactions between producers and consumers.

[0046] Figure 9 A graph showing the iterative process of producer-consumer transaction prices at different times.

[0047] Figure 10 The diagram shows the adaptive update process for the iteration step size.

[0048] Figure 11 A diagram showing the prosumer operating costs for different scenarios.

[0049] Figure 12 The graph shows the convergence residual curve of the algorithm. Detailed Implementation

[0050] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0051] In one embodiment of the present invention, an ADN decentralized optimization method based on enhanced price update P2P transactions is provided, comprising the following steps:

[0052] Construct a two-layer decentralized optimization model that includes a producer-consumer P2P transaction layer and a physical optimization operation layer;

[0053] In the producer-consumer P2P transaction layer, with the goal of minimizing the operating costs of producers and consumers, an enhanced price update mechanism is used to iteratively optimize the electricity transaction price and transaction volume between adjacent producers and consumers and / or between producers and consumers and the upper-level main grid, and output the optimized active power transaction plan. The enhanced price update mechanism realizes the adaptive adjustment of the iteration step size of the ADMM algorithm through the tanh function, and introduces a transaction price anchoring term with the upper-level main grid retail electricity price and the producer-consumer on-grid electricity price as the boundary.

[0054] In the physical optimization operation layer, based on the optimized active power trading plan, the ADMM algorithm is used to solve the fully decentralized reactive power / voltage optimization model, calculate the node voltage, and optimize the reactive power output of producers and consumers.

[0055] Determine whether the voltage of all nodes meets the safe operation constraints; if it does, output the final optimization result; if it does not, adjust the active power trading limit of each producer-consumer according to the voltage limit violation situation, and return to the producer-consumer P2P transaction layer for iterative optimization again.

[0056] Specifically, the distributed optimization method for ADN provided in this embodiment of the invention enables electricity producers and consumers to actively participate in voltage regulation of the ADN, thereby improving system operational safety and economy. The logical framework of this distributed optimization method is as follows: Figure 1 As shown, the constructed two-layer distributed optimization model mainly includes two parts: the producer-consumer P2P transaction layer and the physical optimization operation (i.e., ADN operation optimization) layer. The optimization process can be divided into the following steps: ① The producer-consumer P2P transaction layer optimizes the transaction volume and price between producers and consumers to achieve the goal of minimizing the operating cost of each producer and consumer; ② Based on the optimized active power of each producer and consumer transaction, the system is optimized in a distributed manner for reactive power / voltage to improve the safety and economic performance of the system operation; ③ Each producer and consumer monitors the voltage safety of its own grid-connected node. If the allowable operating voltage range is met, the transaction and operation optimization results are directly output. Otherwise, the producer and consumer actively adjusts the upper limit of active power transaction according to the following formulas (1) and (2), and iterates and optimizes the P2P transaction and reactive power / voltage regulation again until the voltage safety requirements are met, and the optimal ADN optimization operation result is obtained. Steps ① and ② are executed in the P2P multi-agent environment, and step ③ is executed locally based on the optimization results of steps ① and ②.

[0057] The formula for adjusting the active power trading limit is as follows:

[0058] (1);

[0059] (2);

[0060] In the formula, The upper limit of the total electricity sold by producer-consumer i at time t; Let t be the upper limit of the electricity sold by producer-consumer j to producer-consumer i at time t; The trading power is adjusted for each iteration. If its value is too large, it may reduce the stability of the trading and increase the difficulty of convergence. If its value is too small, it may cause the iteration to be too slow. U is the square of the voltage at consumer node i at time t; max U min These are the upper and lower limits of the allowable node voltage, respectively. This refers to the allowed operating voltage range; k is the current iteration number, i.e., the superscript (k) indicates the kth iteration.

[0061] As can be seen from the above analysis, the two-layer distributed optimization model constructed in this embodiment of the invention transfers the responsibility for the operation safety of the distribution network from the DSO to the producers and consumers to a certain extent. As can be seen from equations (1) and (2), when the voltage of producer-consumer node i exceeds the upper limit, the corresponding producer-consumer reduces the amount of electricity sold, suppresses the voltage rise of adjacent producer-consumer nodes, and thus controls its own voltage drop; when the voltage of producer-consumer node i exceeds the lower limit, the corresponding producer-consumer reduces the amount of electricity purchased from adjacent producer-consumers, and thus increases its own node voltage. It should be noted that since this step is a local control, each producer-consumer can only adjust the voltage of the connected nodes by adjusting the amount of electricity traded, and cannot optimize the voltage of nodes in the system that have no producers-consumers connected.

[0062] In a preferred embodiment of the present invention, the defined prosumers include DPVs, power generation equipment, and BSDs, serving a certain amount of inflexible loads. The power generation equipment mainly includes micro gas turbines (MGs). A schematic diagram of the ADN structure and the transaction logic between prosumers is shown below. Figure 2 As shown in the figure. This embodiment of the invention stipulates that only physically adjacent prosumers can engage in electricity trading. That is, prosumers can purchase electricity from neighboring prosumers or the upstream main grid to meet their load demands and to charge energy storage devices, and can sell excess electricity back to neighboring prosumers or the upstream grid to generate revenue. In the prosumer P2P transaction layer, the method for minimizing the operating cost of prosumers is as follows: For any prosumer i, a cost minimization objective function is constructed within the scheduling period T, as shown in the following equation:

[0063] (3);

[0064] (4);

[0065] (5);

[0066] In the formula, , These represent the total electricity sold by producer-consumer i at time t and the revenue per unit of electricity sold, respectively. , These represent the electricity purchased by producer-consumer i from producer-consumer j at time t, and the unit electricity purchase cost, respectively. , These represent the power purchased by producer-consumer i from the upper-level power grid at time t and the unit power purchase cost, respectively. , These represent the electricity sold by producer-consumer i to the upper-level power grid at time t and the unit electricity sales cost, respectively. , and , These represent the output cost and active power output of the power generation equipment (mainly MG) and BSD in producer-consumer i at time t, respectively. The unit energy transaction cost charged by the distribution network operator (DSO) to the buyer at time t; , and This refers to the cost coefficient for the corresponding item; The total number of prosumers in the system; Let i be the set of neighboring nodes of node i; This indicates the cost of electricity transactions between producers and consumers; This represents the cost of electricity transactions between producers / consumers and the upper-level power grid.

[0067] In a preferred embodiment of the present invention, the constraints of the prosumer-consumer P2P transaction layer include power balance constraints, power generation equipment output constraints, BSD output constraints, and prosumer-consumer transaction constraints.

[0068] Specifically, the power balance constraint is shown in the following equation:

[0069] (6);

[0070] (7);

[0071] In the formula, Let be the overflow power of producer i at time t. Equation (7) is used to ensure the power balance of producer i at any time. Let r be the net load power of producer i at time t; ij The resistance of the power distribution line between producers i and j; U N This is the system's rated voltage.

[0072] The output constraint of the power generation equipment is shown in the following formula:

[0073] (8);

[0074] (9);

[0075] Equations (8) and (9) represent the output limit constraint and ramp constraint of the power generation equipment MG, respectively; where, , These represent the upper and lower limits of MG's output, respectively. , These represent the upper and lower limits of the MG's ramp rate, respectively.

[0076] The BSD output constraint is shown in the following formula:

[0077] (10);

[0078] (11);

[0079] (12);

[0080] (13);

[0081] (14);

[0082] Equations (10) and (11) represent the output configuration and charge / discharge limit constraints of the BSD, respectively; Equation (12) describes the state of charge (SOC) change process of the BSD; Equations (13) and (14) describe the usable SOC range of the BSD and its initial and final states; where, , These represent the BSD charging and discharging power of consumer i at time t; The upper limit of the charging and discharging power of BSD in consumer i; Let SOC be the BSD value of consumer i at time t; These are the BSD charge / discharge coefficients; E is the total energy of the BSD. , These are the upper and lower limits of SOC, respectively.

[0083] Prosumer transaction constraints are shown in the following formula:

[0084] (15);

[0085] (16);

[0086] (17);

[0087] (18);

[0088] Equations (15) to (17) are used to limit the amount of electricity traded between the two parties; Equation (18) describes the electricity trading capacity of the producers and consumers. In the equations, The upper limit of the electricity sales power from the main grid to consumer i; The upper limit of the electricity sales power from the upstream power grid to producer-consumer i; The set of neighboring producer-consumer nodes of producer-consumer i is represented by j, where j represents the neighboring nodes of node i. This represents the total electrical energy that producer i can output. This represents the electrical energy purchased by neighboring producers from point i.

[0089] In a preferred embodiment of the present invention, the producer-consumer P2P transaction layer aims to minimize the cost of each producer-consumer, which is a decentralized transaction model with strong scalability, elasticity and privacy, and can be plug-and-play, and can be solved in a multi-agent environment. P2P transactions are implemented by multiple producers-consumers calling various flexible resources. Each local transaction market contains one seller and several buyers. The transaction price of producers-consumers is not determined by a single buyer or seller. Both buyers and sellers will affect the transaction price. In order to make the transaction price independent of the influence of buyers and sellers, the embodiments of the present invention consider the constraints of the electricity market mechanism and propose an electricity transaction price update strategy between producers-consumers. Specifically, when the provisional transaction volume (the electricity that producer-consumer i can sell) exceeds the provisional electricity demand (the electricity that producer-consumer i's neighboring producers-consumers decide to buy), the electricity selling price of producer-consumer i will decrease, and vice versa. Finally, when the provisional electricity transaction volume reaches a balance, the transaction price is output. The transaction between producers-consumers can be mainly divided into two steps: (1) transaction volume update. According to the P2P transaction price, each producer-consumer conducts optimal electricity trading based on its own interests through self-scheduling; (2) Transaction price update. Producers-consumers adjust the transaction price based on the electricity purchase demand of their neighboring producers-consumers and their own interests.

[0090] In practical applications, this embodiment of the invention uses the ADMM algorithm to solve the producer-consumer P2P transaction layer model. The proposed model can be characterized by the standard ADMM algorithm model. The augmented Lagrangian function of the original objective function shown in equation (3) can be defined as:

[0091] (19);

[0092] In the formula, p is the decision variable of the model, i.e. ;F tr The objective function value of expression (3) is expressed; The auxiliary variables introduced to decouple the coupling constraints of equations (15) and (16), and have ; For Lagrange multipliers; This represents the iteration step size of the ADMM algorithm.

[0093] In the ADMM algorithm, the decision variables, auxiliary variables, and Lagrange multipliers need to be updated alternately. Their update formulas are shown in equations (20) to (22) below:

[0094] (20);

[0095] (twenty one);

[0096] (twenty two);

[0097] In the formula: The values ​​of the decision variables corresponding to the objective function shown in equation (3); Represents the Lagrange multiplier terms; This represents the penalty function term for supply and demand imbalance.

[0098] Based on this, taking the partial derivative of equation (21), we obtain the simplified formula for updating the auxiliary variable as follows:

[0099] (twenty three);

[0100] By passing all prosumers i in equation (23) to all adjacent prosumers ( Summing the power of ) yields:

[0101] (twenty four);

[0102] Furthermore, by uniformly defining the Lagrange multipliers, which are equivalent to shadow prices, as producer-consumer transaction prices, we can integrate equations (22) and (23) to obtain:

[0103] (25);

[0104] Finally, by combining equations (22), (24), and (25), we can obtain the following traditional trading pricing mechanism based on the imbalance between electricity supply and demand:

[0105] (26);

[0106] The specific expression for the penalty function in equation (20) is:

[0107] (27);

[0108] In a preferred embodiment of the present invention, the traditional price update strategy can achieve P2P transaction optimization with the goal of minimizing producer-consumer costs, but it may have the following two problems in practical applications:

[0109] (1) Fixed iteration step size. When it is too large, price updates will overreact to small supply and demand fluctuations, causing market prices to fluctuate violently and lack stability; when it is too small, the system response is slow and it is difficult to effectively guide supply and demand to quickly restore balance, resulting in low market efficiency.

[0110] (2) Lack of price benchmark checks and balances. When there is a serious supply and demand imbalance in the system, transaction prices may rise or fall excessively, deviating significantly from the actual value, causing market regulation to fail and failing to provide reliable price protection for producers and consumers.

[0111] Therefore, in order to address the above problems, this embodiment of the invention introduces the tanh function to construct an enhanced price update mechanism for an adaptive iterative step size collaborative price anchoring strategy, and its transaction price update formula is shown in equation (28).

[0112] (28);

[0113] In the formula, The ratio of local power shortage or surplus of producer-consumer i at time t can be specifically described by equation (29); The hyperbolic tangent function is the adaptive gain factor, which can be specifically described by equation (30); The anchor term for the transaction price is as shown in equation (31); is the iteration step size of the ADMM algorithm; k is the current iteration number.

[0114] Furthermore, in the transaction price update formula:

[0115] (29);

[0116] (30);

[0117] (31);

[0118] Equation (29) is used to quantify the supply and demand balance of the system. The range of is [-1, 1], and its physical meaning can be interpreted as: when When the supply and demand relationship between producers and consumers is such that supply exceeds demand, and when... When, it means that supply is less than demand; when When the value approaches 0, it indicates that the electricity available for sale by producers and consumers matches the demand of their neighbors, meaning that supply and demand are in equilibrium; when... When this occurs, it indicates that neighboring producers and consumers have no intention to purchase electricity, resulting in an electricity surplus; when When this occurs, it indicates that producers and consumers have no electricity to sell, and there is a severe shortage of electricity. The function shown in equation (30) is used to implement the adaptive adjustment of the iteration step size of the ADMM algorithm. According to basic economic principles, prices decrease as supply increases or demand decreases, and increase as demand increases or supply decreases. For the P2P trading market, the trend of trading prices follows... The function (with a range of (0,1)) increases as the degree of supply-demand imbalance between producers and consumers increases, i.e., the value of increases, thus increasing the iteration step size of the ADMM algorithm; when the degree of supply-demand imbalance between producers and consumers decreases, the value of decreases accordingly; when supply and demand are balanced, i.e., it approaches 0, ... The value of tanh will also tend to 0. This demonstrates that the tanh function can adaptively adjust the iteration step size of the ADMM algorithm based on the supply and demand relationship in the P2P market. The principle is as follows: Figure 3 As shown in Equation (31), the anchor term can provide a price reference for P2P transactions.

[0119] In summary, this invention, through the introduction of the tanh function and the transaction price anchoring term, forms an enhanced real-time price update mechanism for decentralized P2P electricity trading. Its main advantages are: (1) when supply and demand are balanced, the adaptive iteration step size tends to 0, effectively suppressing price fluctuations and improving the stability of P2P transactions; (2) when supply and demand are unbalanced, the transaction price is driven to make rapid adjustments to the supply and demand difference, improving the responsiveness of the P2P market; (3) under extreme supply and demand conditions, the algorithm intelligently anchors the transaction price to the main grid retail electricity price and the producer-consumer on-grid electricity price, ensuring the robustness of the market.

[0120] In a preferred embodiment of the present invention, to ensure the safe and economical operation of the system, this embodiment proposes a decentralized optimization operation model for an ADN (Active Data Network) based on the P2P transaction results of each producer-consumer, namely, a reactive power / voltage optimization model for the physical optimization operation layer. In the physical optimization operation layer, the reactive power / voltage optimization model aims to minimize the voltage deviation and active power loss of each producer-consumer node, specifically expressed as follows:

[0121] (32);

[0122] In the formula, U is the square of the voltage at consumer node i at time t; max U min These are the upper and lower limits of the allowable node voltage, respectively; U N This is the system's rated voltage; , These represent the active and reactive power injected by the upstream producer-consumer into producer-consumer i, respectively; r i The line resistance between producer-consumer i and the upstream producer-consumer; This is the voltage over-limit penalty coefficient, which is a sufficiently large positive number.

[0123] In addition, the constraints of the decentralized optimization operation model of each producer and consumer in the above-mentioned physical optimization operation layer include linear Dist-Flow power flow constraints, capacity and limit constraints of power generation equipment and BSD.

[0124] Specifically, the linear Dist-Flow power flow constraint is shown in the following equation:

[0125] (33);

[0126] (34);

[0127] (35);

[0128] In the formula, These represent the net injected active and reactive power of producer node i at time t, respectively. , and These represent the reactive power output by MG and BSD connected to producer node i at time t, and the net reactive load, respectively. The active and reactive power transmitted between nodes i and k at time t are respectively, where node k represents the downstream node of node i. Let be the square of the voltage at node k at time t; Resistance and reactance of the power distribution lines between nodes i and k; This represents the set of end nodes of a branch with node i as the first end node.

[0129] The capacity and limit constraints of the power generation equipment MG and BSD are shown in the following formula:

[0130] (36);

[0131] (37);

[0132] In the formula: These represent the maximum reactive power output of the MG and BSD connected to the producer-consumer node i at time t, respectively. These represent the maximum apparent capacity of the MG and BSD connected to the producer-consumer node i, respectively.

[0133] In a preferred embodiment of the present invention, the objective function shown in equation (32) above is transformed into the following form:

[0134] (38);

[0135] The coupling constraint is shown in the following equation:

[0136] (39);

[0137] In the formula: These are the local variables of node i at time t after decoupling, representing the reactive power absorbed from the upstream node and injected into the downstream node. Let be the voltage of the upstream node of node i at time t; This is a local variable representing the voltage of the upstream node of node i at time t after decoupling; This represents the set of downstream nodes of node i.

[0138] The reactive power constraint is shown in the following formula:

[0139] (40);

[0140] (41);

[0141] (42);

[0142] Where st represents the constraint condition.

[0143] Furthermore, the boundary conditions of the distributed optimization operation model are defined as follows: if the upstream node of node i is the balanced node of the network, then let If node i is the last node of each branch, then let The decision variables of the proposed decentralized optimization model after decoupling can be expressed as:

[0144] .

[0145] Based on this, the decentralized optimization operation model can also be solved using the ADMM algorithm. The process is similar to the decentralized solution of the above-mentioned producer-consumer P2P transaction layer model. The update formula of the decision variables in the algorithm is shown in equation (43):

[0146] (43);

[0147] In the formula, The values ​​of the decision variables corresponding to the objective function shown in equation (38); This represents the iteration step size of the ADMM algorithm; Let Lagrange multipliers be the variables corresponding to the coupled variables shown in equation (39), and their update formulas are as follows:

[0148] (44);

[0149] In the formula, Let represent the set of downstream nodes of node 'l', where node 'l' represents the upstream node of node 'i'. It should be noted that, according to the defined boundary conditions, when the upstream node of node 'i' is a balancing node in the network, the squared value of its upstream node voltage... ,but No update is needed; when node i is the last node of a branch, it injects power into downstream nodes. ,but No updates are required. The models provided in the embodiments of this invention are all solved using the ADMM algorithm.

[0150] In another embodiment of the present invention, an ADN decentralized optimization system based on enhanced price update P2P transactions is also provided to implement the above-mentioned ADN decentralized optimization method, comprising:

[0151] The producer-consumer (P2P) transaction layer aims to minimize the operating costs of producers and consumers. Based on an enhanced price update mechanism, it uses the ADMM algorithm to iteratively optimize the electricity transaction price and volume between adjacent producers and consumers and / or between producers and consumers and the upstream main grid, outputting an optimized active power transaction plan. The enhanced price update mechanism uses the tanh function to adaptively adjust the iteration step size of the ADMM algorithm and introduces a transaction price anchoring term with the upstream main grid retail electricity price and the producer-consumer on-grid electricity price as the boundary.

[0152] The physical optimization operation layer is used to solve the fully decentralized reactive power / voltage optimization model based on the optimized active power trading plan using the ADMM algorithm, calculate node voltage, and optimize the reactive power output of producers and consumers.

[0153] The safety constraint module is used to determine whether the voltage of all nodes meets the safety operation constraints. If it does, the final optimization result is output. If it does not, the active power trading limit of each producer-consumer is adjusted according to the voltage limit violation, and the process is returned to the producer-consumer P2P transaction layer for iterative optimization.

[0154] It should be noted that each of the above modules can be implemented as a computer program, which can run on a computer device. The computer device's memory can store the computer program that makes up each module, enabling the processor to execute each step of the above method.

[0155] Example 1: Establish a test system for a modified IEEE 33-node system, including distributed photovoltaic (DPV), power generation equipment (MG), and battery energy storage devices (BSD). The network structure and the access locations and configurations of the power producers and consumers are as follows: Figure 4 As shown. The system reference voltage and reference capacity are taken as 12.66kV and 100MVA, respectively. The allowable operating voltage range is... The impedance of each line is determined according to the standard IEEE 33 test system parameters; the unit energy trading cost is taken as... =0.04 yuan / kW, the retail and on-grid usage (ToU) of the upstream power grid unit are shown in Table 1, where the valley period is 0:00-6:00, the normal period is 7:00-9:00, 13:00-16:00 and 22:00-23:00, and the peak period is 10:00-13:00 and 17:00-21:00; the transaction power adjustment value is taken as; the initial iteration step size of the ADMM algorithm in the P2P transaction model is. The convergence threshold is set to 0.1; the iteration step size in the running optimization model is... The convergence threshold is set to 0.001, and the voltage over-limit penalty coefficient is set to a certain value. Except for producer-consumer nodes, this paper does not consider the load of other nodes. The net load of each producer-consumer node is as follows: Figure 5 As shown, the configuration parameters for prosumers are as follows:

[0156] MG1: 0.012 yuan / kW 2 , =0.3 yuan / kW 2 , =500kW, =0, =500kVA =125kW, =-125kW;

[0157] MG2: 0.001 yuan / kW 2 , =0.65 yuan / kW 2 , =300kW =0, =500kVA =100kW =-100kW;

[0158] BSD: =0.005 yuan / kW 2 , =150kW, =150kVA, E=300kW∙h, =0.95, =0.85, =0.15, =0.5.

[0159] Table 1 Time-of-use electricity pricing for purchasing and selling electricity

[0160]

[0161] To verify the effectiveness of the model constructed in the embodiments of the present invention, four schemes were designed, and comparative analyses were conducted from three aspects: model optimization effect, the role of the price update mechanism, and prosumer costs.

[0162] Option 1: Only consider traditional P2P transactions in the producer-consumer P2P transaction layer, without considering ADN optimization in the physical optimization operation layer;

[0163] Option 2: Only consider P2P transactions based on the enhanced price update mechanism in the producer-consumer P2P transaction layer, without considering the ADN optimization operation of the physical optimization operation layer;

[0164] Option 3: Based on global centralized optimization, the producer-consumer P2P transaction layer and the physical optimization operation layer are operated and optimized collaboratively;

[0165] Scheme 4 (i.e., the method provided in the above embodiments of the present invention): Decentralized collaborative operation optimization of the producer-consumer P2P transaction layer and the physical optimization operation layer based on the enhanced price update mechanism.

[0166] (I) Comparative Analysis of ADN Operation Optimization Effects; During the entire optimization period T, the MG output and BSD SOC change curves of the above four comparative schemes are as follows: Figure 6 As shown. By Figure 6 It can be seen that during peak electricity consumption periods (higher net load), each MG outputs higher power, and due to the higher electricity trading price during this period, BSDs sell electricity at high returns. During off-peak electricity consumption periods (lower net load), each MG outputs less power, and given the lower electricity price during this period, BSDs charge at low cost. The above process achieves "peak shaving and valley filling" and effectively reduces power supply pressure and producer-consumer costs. Comparing the SOC changes of BSDs alone, it can be seen that the method (Scheme 4) provided by this embodiment of the invention has a higher charging and discharging utilization rate throughout the entire optimization period T, which can, to a certain extent, ensure better returns for producer-consumers. The above results show that each BSD starts charging at 22:00 to ensure an initial SOC of 50% for subsequent use.

[0167] The optimized system node voltage distribution and reactive power optimization process are as follows: Figure 7 As shown. By Figure 7 (a) and Figure 7 As can be seen from (b), P2P transactions aimed solely at minimizing costs cannot guarantee the system's operational voltage safety. Both Scheme 1 and Scheme 2 experienced voltage exceedances to some extent, with exceedance rates around 7.50%. The maximum exceedance values ​​occurred at producer-consumer 10 at 19:00, with peak voltages dropping to 0.8986 pu and 0.8842 pu, respectively. In contrast, the ADN optimization operation models for coordinated reactive power / voltage in Schemes 3 and 4 additionally considered voltage deviation targets, effectively suppressing node voltage exceedances during the periods of minimum and maximum net load (11:00–14:00 and 18:00–24:00), thus ensuring the system's operational safety. Figure 7 (c) and Figure 7 As shown in (d). Individual comparison Figure 7 (c) and Figure 7 As can be seen from (d), the overall trends and numerical performance of the two are basically consistent, which indicates that the fully distributed ADN optimization operation strategy proposed in this embodiment has achieved performance comparable to global centralized reactive power / voltage optimization, proving the effectiveness of the proposed method. The reactive power flow results of Schemes 3 and 4 over the entire optimization period T are as follows: Figure 7 (e) and Figure 7 As shown in (f), the overall trends of the two are basically the same. However, in the period from 19:00 to 24:00, the reactive power absorbed by producers 2, 3 and 9 from downstream voltage lower limit producers 3, 9 and 10 in scheme 4 is significantly increased compared with scheme 3, thereby alleviating the voltage violation situation left over from simple P2P transactions.

[0168] (II) Analysis of the Effects of the Enhanced Price Update Mechanism; To further verify the effects of the enhanced price update mechanism proposed in this embodiment on the ADN decentralized optimization operation model based on P2P electricity trading, Figure 8 First, the P2P transaction electricity prices between producers and consumers under different comparative schemes within the optimization period T are presented. It should be noted that since the global centralized optimization uses the fixed ToU shown in Table 1 as the transaction electricity price, and no price updates are involved in the optimization process, the transaction price of scheme 3 will not be given in this section.

[0169] Depend on Figure 8 It can be seen that, influenced by the original ToU and net load, the P2P transaction prices between producers and consumers under different schemes generally maintain the same trend. That is, during peak electricity consumption periods, when the net load is large and the main grid retail electricity price is high, the producer-consumer transaction price is high; during off-peak electricity consumption periods, when the net load is small and the main grid retail electricity price is low, the producer-consumer transaction price is low. Comparing Schemes 2 and 4, which consider the enhanced price update mechanism of P2P transactions proposed in this paper, with Scheme 1, which only considers the supply and demand balance, it can be seen that the traditional P2P transaction price is significantly lower than the bulk market price during peak electricity consumption periods, and higher during off-peak electricity consumption periods, which is significantly deviating from the actual value, weakening the macro-control role of the electricity market mechanism, and failing to provide users with highly reliable price protection. The proposed enhanced price update mechanism can control the transaction electricity price between the main grid retail electricity price and the producer-consumer on-grid electricity price during different load levels, which can better reflect the impact of macro-market regulation on P2P transactions and effectively ensure transaction security. (Separate comparison) Figure 8 (b) and Figure 8 As can be seen from (c), under the method proposed in this embodiment of the invention, producers and consumers 3, 9, and 10, whose voltage exceeds the lower limit, adjust the active power trading limit and update the trading price according to formula (2) during the period from 18:00 to 21:00. In this way, relying on physical layer reactive power / voltage regulation, the operating voltage safety of the system is guaranteed. The update and iteration process of the trading price of producers and consumers 3, 9, and 10 at different times is as follows: Figure 9 As shown.

[0170] Within the optimization period T, the adaptive update process of the ADMM algorithm's iterative step size under the proposed enhanced price update mechanism is as follows: Figure 10 As shown. By Figure 10 It is evident that the iteration step size of the ADMM algorithm in the P2P transaction layer is... Able to adjust according to supply and demand ratio The adaptive adjustment, by optimizing the transaction power among prosumers, shows a continuous trend toward the supply-demand equilibrium point, thereby obtaining the optimal transaction volume and transaction price, as well as minimizing the costs of prosumers. This verifies the effectiveness of the proposed iterative step size adaptive adjustment method.

[0171] (III) Comparative Analysis of Prosumer-Consumer Operating Economics; Based on the P2P transaction model established according to this invention and the optimized producer-consumer transaction electricity price, the producer-consumer costs of each comparative scheme at different time periods can be obtained, such as... Figure 11 As shown in the figure, due to the increased output costs of MG and BSD during peak electricity consumption periods and the need to purchase electricity from the upper-level grid, the producer-consumer costs are highest during peak periods, followed by normal periods, and lowest during off-peak periods in all the comparative schemes. The total cost of each comparative scheme is shown by the red line in the figure. It can be seen that the cost of global centralized optimization is higher than that of decentralized optimization in schemes 1, 2, and 4. The method provided by the embodiment of the present invention has the best economic performance, reducing the total cost by RMB 0.14 million, RMB 0.08 million, and RMB 0.16 million respectively compared to the other three schemes, with reductions of 7.43%, 4.18%, and 8.43% respectively. In summary, it can be proven that the ADN decentralized optimization operation strategy based on enhanced price update mechanism P2P electricity trading proposed in the embodiment of the present invention can effectively reduce producer-consumer costs and has good operational economics.

[0172] (iv) Comparative Analysis of Model Solving Efficiency; In this embodiment of the invention, the ADMM algorithm is used to solve the models of the producer-consumer P2P transaction layer and the physical optimization operation layer, respectively. The two models reach their convergence thresholds of 0.1 and 0.001 after 133 and 287 iterations, respectively, achieving convergence. The convergence residual curves of the algorithm are shown below. Figure 12 As shown in Table 2, to verify the computational performance of the proposed model, a comparison was made with Scheme 3, which is a collaborative optimization model based on global centralized optimization of the economic and physical layers. The computational time of the two methods is statistically analyzed. As shown in Table 2, the total solution time of the proposed model is comparable to that of the global centralized optimization model, with a time saving of only 120 seconds, a reduction of approximately 3.51%, demonstrating certain engineering feasibility. It should be noted that the IEEE 33-node test case used in this embodiment is relatively small. For larger-scale systems, the proposed optimization model based on fully distributed computation may have a more significant advantage in solution efficiency.

[0173] Table 2 Online solution time for different schemes

[0174]

[0175] In summary, this invention's embodiments take into account the voltage regulation capabilities of P2P transactions, proposing a fully decentralized ADN optimization model for P2P electricity transactions based on an enhanced price update mechanism. The proposed decentralized reactive power / voltage control can effectively reduce the voltage exceedance rate and ensure system operational safety by automatically adjusting the active power trading limits and quantities among producers and consumers in voltage exceedance scenarios. Furthermore, compared to global centralized optimization, the decentralized ADN optimization model proposed in this invention has higher solution efficiency and convergence performance, demonstrating significant engineering application value. Additionally, the enhanced transaction price update mechanism based on the tanh function and anchoring term collaboration can adaptively update P2P transaction prices under varying degrees of supply-demand imbalance scenarios, keeping electricity transaction prices within a reasonable range, effectively ensuring the macro-control function of the electricity market, and providing highly reliable price protection for electricity users.

[0176] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A decentralized optimization method for ADN based on enhanced price update P2P transactions, characterized in that, Includes the following steps: Construct a two-layer decentralized optimization model that includes a producer-consumer P2P transaction layer and a physical optimization operation layer; In the producer-consumer P2P transaction layer, with the goal of minimizing the operating costs of producers and consumers, the ADMM algorithm is used to iteratively optimize the electricity transaction price and transaction volume between adjacent producers and consumers and / or between producers and consumers and the upper-level main grid based on the enhanced price update mechanism, and outputs the optimized active power transaction plan. The enhanced price update mechanism uses the tanh function to adaptively adjust the iteration step size of the ADMM algorithm and introduces a transaction price anchoring term with the upper-level main grid retail electricity price and the producer-consumer on-grid electricity price as the boundary. In the physical optimization operation layer, based on the optimized active power trading plan, the ADMM algorithm is used to solve the fully decentralized reactive power / voltage optimization model, calculate the node voltage, and optimize the reactive power output of producers and consumers. Determine whether the voltages of all nodes meet the safe operation constraints; If satisfied, output the final optimization result; If the conditions are not met, the active power trading limit for each producer-consumer will be adjusted according to the voltage limit violation, and the process will be returned to the producer-consumer P2P transaction layer for iterative optimization.

2. The ADN decentralized optimization method for P2P transactions based on enhanced price updates as described in claim 1, characterized in that, The prosumers include distributed photovoltaic (DPV), power generation equipment, and battery storage devices (BSD). In the prosumer P2P transaction layer, the method to minimize the operating cost of prosumers is as follows: For any prosumer i, construct its cost minimization objective function within the scheduling period T, as follows: ； ； ； In the formula, , These represent the total electricity sold by producer-consumer i at time t and the revenue per unit of electricity sold, respectively. , These represent the electricity purchased by producer-consumer i from producer-consumer j at time t, and the unit electricity purchase cost, respectively. , These represent the power purchased by producer-consumer i from the upper-level power grid at time t and the unit power purchase cost, respectively. , These represent the electricity sold by producer-consumer i to the upper-level power grid at time t and the unit electricity sales cost, respectively. , and , These represent the output cost and active power output of the power generation equipment and BSD in producer-consumer i at time t, respectively. The unit energy transaction cost charged by the distribution network operator (DSO) to the buyer at time t; , and This refers to the cost coefficient for the corresponding item; The total number of prosumers in the system; Let i be the set of neighboring nodes of node i; This indicates the cost of electricity transactions between producers and consumers; This represents the cost of electricity transactions between producers / consumers and the upper-level power grid.

3. The ADN decentralized optimization method for P2P transactions based on enhanced price updates as described in claim 2, characterized in that, The transaction price update formula for the enhanced price update mechanism is as follows: ； In the formula, This indicates the ratio of local power shortage or surplus among producers and consumers at time t; The hyperbolic tangent function is the adaptive gain factor. Anchor to the transaction price; is the iteration step size of the ADMM algorithm; k is the current iteration number.

4. The ADN decentralized optimization method for P2P transactions based on enhanced price updates according to claim 3, characterized in that, In the transaction price update formula: ； ； 。 5. The ADN decentralized optimization method for P2P transactions based on enhanced price updates according to claim 2, characterized in that, In the physical optimization operation layer, the reactive power / voltage optimization model aims to minimize the voltage deviation and active power loss at each producer-consumer node, specifically expressed as follows: ； In the formula, U is the square of the voltage at consumer node i at time t; max U min These are the upper and lower limits of the allowable node voltage, respectively; U N This is the system's rated voltage; , These are the active and reactive power injected by the upstream producer-consumer into producer-consumer i, respectively. r i The line resistance between producer-consumer i and the upstream producer-consumer; This is the penalty coefficient for voltage exceeding the limit.

6. The ADN decentralized optimization method for P2P transactions based on enhanced price updates according to claim 5, characterized in that, The constraints of the producer-consumer P2P transaction layer include power balance constraints, power generation equipment output constraints, BSD output constraints, and producer-consumer transaction constraints.

7. The ADN decentralized optimization method for P2P transactions based on enhanced price updates according to claim 6, characterized in that, The constraints of the physical optimization operation layer include linear Dist-Flow constraints, and capacity and limit constraints of power generation equipment and BSD.

8. The ADN decentralized optimization method for P2P transactions based on enhanced price updates according to claim 7, characterized in that, The formula for adjusting the active power trading limit is as follows: ； ； In the formula, The upper limit of the total electricity sold by producer-consumer i at time t; Let t be the upper limit of the electricity sold by producer-consumer j to producer-consumer i at time t; The trading power is adjusted for each iteration; U is the square of the voltage at consumer node i at time t; max U min These are the upper and lower limits of the allowable node voltage, respectively. This represents the allowed operating voltage range; k is the current iteration number.

9. The ADN decentralized optimization method for P2P transactions based on enhanced price updates according to any one of claims 1-8, characterized in that, The ADN distributed optimization method is applicable to improved IEEE 33-node distribution networks that include distributed photovoltaic (DPV), power generation equipment, and battery storage devices (BSD), with an allowable operating voltage range of 0.95-1.05 pu.

10. An ADN decentralized optimization system based on enhanced price update P2P transactions, used to implement the ADN decentralized optimization method according to any one of claims 1-9, characterized in that, include: The producer-consumer P2P transaction layer aims to minimize the operating costs of producers and consumers. Based on the enhanced price update mechanism, it uses the ADMM algorithm to iteratively optimize the electricity transaction price and transaction volume between adjacent producers and consumers and / or between producers and consumers and the upper-level main grid, and outputs the optimized active power transaction plan. The enhanced price update mechanism uses the tanh function to adaptively adjust the iteration step size of the ADMM algorithm and introduces a transaction price anchoring term with the upper-level main grid retail electricity price and the producer-consumer on-grid electricity price as the boundary. The physical optimization operation layer is used to solve the fully decentralized reactive power / voltage optimization model based on the optimized active power trading plan using the ADMM algorithm, calculate node voltage, and optimize the reactive power output of producers and consumers. The safety constraint module is used to determine whether the voltage of all nodes meets the safe operation constraints. If satisfied, output the final optimization result; If the conditions are not met, the active power trading limit for each producer-consumer will be adjusted according to the voltage limit violation, and the process will be returned to the producer-consumer P2P transaction layer for iterative optimization.

Images