Adaptive curtailment method, system, device and medium for regional mutual assistance power transaction

Abstract

The application discloses a kind of regional mutual aid electric power transaction self-adapting reduction method, system, equipment and medium, method includes, according to time sequence, power market clearing, obtain each market subject in each transaction period Initial transaction power data;Build the power reduction optimization model considering all uncertainty scenarios;Solve model, obtain each market subject in each transaction period After reduction transaction power data;The reduction transaction power data is as each market subject in each transaction period Maximum tradable power upper limit, re-performs power market clearing for each transaction period, obtains new transaction power data;Repeat iteration until the difference of the clearing result of before and after two times is less than set threshold or reaches maximum iteration number, with the transaction power data at this time as final transaction result.The application maximizes the social welfare of regional mutual aid under the premise of guaranteeing power grid safety, effectively solves the transaction efficiency and fairness problem under supply and demand fluctuation.

Description

Technical Field

[0001] This invention belongs to the field of power grid dispatching technology, and specifically relates to an adaptive reduction method, system, equipment and medium for regional mutual assistance power trading. Background Technology

[0002] Regional power exchange is a core means for the national unified power market system to achieve inter-provincial surplus and shortage adjustment, clean energy consumption and resource optimization. According to relevant policy requirements, cross-regional power exchange transactions must follow the principles of safety and reliability, green and low-carbon, fairness and efficiency, and adopt a technical approach that combines market clearing and grid safety verification.

[0003] Currently, regional mutual aid transaction clearing mainly consists of two mainstream technical solutions: sorting clearing and optimization modeling clearing, which provide stable support for the allocation of regional power resources.

[0004] Sequential clearing is a mature and widely used clearing method in regional mutual aid transactions. It uses time-segmented sequential matching as its core logic, prioritizing buyers and sellers based on their price spreads from highest to lowest, and completing transaction matching in each time slot. For entities with the same price spread, transaction capacity is allocated through an equal distribution of electricity. This solution boasts significant advantages such as intuitive calculation logic, high execution efficiency, ease of understanding by market participants, and strong engineering feasibility. It aligns with the large-scale and routine operation needs of medium- and long-term inter-provincial mutual aid transactions and is currently the mainstream and practical solution in the market.

[0005] The optimized modeling clearing approach, centered on a centralized optimization model, solves for the clearing result that maximizes either the price difference or social welfare by constructing conditions such as power balance, line flow, and ramp constraints. Theoretically, this can achieve global optimal resource allocation. Existing optimization models, with minimum adjustment as their core objective and built on deterministic data, offer high computational accuracy. However, their robustness can be further improved when dealing with scenarios such as fluctuations in renewable energy sources and sudden load changes.

[0006] While the two mainstream technologies for regional mutual power trading—ranking clearing and optimization modeling clearing—can meet basic trading needs, they still have specific shortcomings in new power system scenarios.

[0007] Regarding the ranking and clearing process, this method uses price difference as the core ranking criterion and simple equal distribution of electricity volume for the same price difference. While generally practical, it has two shortcomings. First, the allocation mechanism for the same price difference is simplistic, allocating only proportionally based on declared electricity volume without considering the comprehensive contributions of the entity's electricity volume stability, performance credit, clean energy attributes, and tolerance to reduction. This results in insufficient fairness and policy adaptability, failing to reflect the trading value of green electricity and high-quality entities. Second, the reduction constraints after clearing are rigid, using fixed values ​​to constrain electricity ramp-up fluctuations in adjacent time periods without dynamically adjusting thresholds based on load, renewable energy output, and power flow congestion. In complex scenarios, this can easily lead to either overly strict constraints with no feasible solution or overly lenient constraints causing grid fluctuation risks.

[0008] Optimization modeling clearing, centered on mathematical programming models, solves for the globally optimal transaction solution by constructing an objective function and various constraints. Theoretically, it can achieve optimal allocation of regional mutual aid resources. However, it has several core shortcomings in practical engineering applications and market implementation. Among them, the lack of interpretability and the inability to effectively choose from multiple solutions are the biggest problems restricting its large-scale application. The specific shortcomings are as follows:

[0009] First, the models suffer from significant limitations in determinism and insufficient robustness. Existing optimization reduction models are built solely based on deterministic data such as predicted load and fixed unit output, without incorporating uncertainties such as fluctuations in renewable energy output, sudden load changes, and tie-line faults. As a result, the reduction results deviate significantly from actual operating conditions, making it difficult to withstand grid security risks caused by random disturbances and unable to adapt to the operational needs of new power systems with a high proportion of renewable energy integration.

[0010] Secondly, the model has extremely poor interpretability and low market acceptance. The optimization modeling clearing process is a "black box" solution process. The clearing logic, reduction basis, and variable weights are completely opaque to market participants. Regulatory agencies and trading entities cannot trace the path of the solution generation or understand the reasons for the decision, which violates the basic principles of "openness, fairness, and impartiality" in the electricity market, thus limiting its market promotion and acceptance by market participants.

[0011] Third, the problem of multiple solutions cannot be effectively handled, resulting in poor stability of the results. Optimization models often have multiple sets of optimal or near-optimal solutions within the feasible region. Existing technologies lack a clear ranking and selection mechanism and cannot select the optimal feasible solution based on dimensions such as market rules, social welfare, and principal contributions. This leads to random, discontinuous, and inconsistent clearing results, which seriously affects the transaction order and execution.

[0012] The aforementioned shortcomings of existing technologies, in the context of high-proportion new energy access and the normalization of cross-regional transactions, have constrained the synergistic improvement of the fairness, security and efficiency of regional mutual assistance transactions, and urgently need to be improved through mechanism optimization. Summary of the Invention

[0013] The purpose of this invention is to provide an adaptive reduction method, system, device, and medium for regional mutual assistance power trading. Addressing the ramp-up constraints and uncertainties in inter-regional trading, this invention constructs a dual-objective collaborative model of price difference and social welfare, combining an adaptive reduction threshold and a robust optimization mechanism to achieve a dynamic closed loop from time-based clearing to power reduction. This invention maximizes the social welfare of regional mutual assistance while ensuring grid security, effectively solving the challenges of trading efficiency and fairness under supply and demand fluctuations.

[0014] To achieve the above objectives, the solution of the present invention is:

[0015] An adaptive reduction method for regional mutual aid power trading includes,

[0016] Step A: According to the time sequence, the electricity market is cleared for each trading period to obtain the initial trading electricity data of each market participant in each trading period;

[0017] Step B: Construct an optimization model for power reduction that considers all uncertain scenarios, wherein the constraints of the model include chance constraints.

[0018] Step C: Solve the power reduction optimization model to obtain the power reduction transaction data of each market participant in each trading period;

[0019] Step D: The reduced traded electricity data is used as the maximum tradable electricity limit for each market participant in each trading period. The electricity market is cleared again for each trading period to obtain new traded electricity data.

[0020] Step E: Repeat steps A-D until the difference between the two clearing results is less than the set threshold or the maximum number of iterations is reached. Use the transaction power data at this point as the final transaction result.

[0021] In step A, electricity market clearing is performed for each trading period, including:

[0022] The clearing rule prioritizes the largest bid-ask spread, meaning that transactions are executed sequentially from largest to smallest bid-ask spread until all possible combinations of transactions have been traversed or a safety limit has been reached.

[0023] In step B, a power reduction optimization model considering all uncertainties is constructed, including:

[0024] The objective function of the model is:

[0025]

[0026] in, As the normalization factor, , This provides the initial trading electricity matrix for each market participant in each trading session. Indicates market entities, Indicates the trading session; These represent positive and negative reduction amounts, respectively. ; This represents the total number of uncertain scenarios; For the scene Social welfare under the following circumstances; For scene probabilities; These are the weighting coefficients;

[0027] The constraints of the model include adaptive ramping / reverse superposition constraints, opportunity constraints for power flow security and CVaR control constraints, opportunity constraints for regional power balance, and upper and lower limits for trading power.

[0028] Among them, the scene social welfare Defined as the sum of "buyer's quote - seller's quote" for all completed transactions. ,in, The decreasing weights are assigned based on the price difference from largest to smallest, and they are positively correlated with the initial transaction price difference. For market entities During the period The reduction in traded electricity.

[0029] The opportunity constraints and CVaR control constraints for power flow safety mentioned above include the following:

[0030] For each critical section and time period Define a random variable.

[0031]

[0032] in, The power transfer distribution factor, For market entities During the period Reduced trading of electricity; For the scene The contribution of new energy sources and load fluctuations to cross-sectional power;

[0033] The opportunity constraint requires that,

[0034]

[0035] in, Indicates the probability of taking the value. For the allowed probability of violation, This represents the cross-sectional transmission limit;

[0036] Define the scenario based on CVaR constraints. Off-limit variables for,

[0037]

[0038] in, Indicated at the critical section Time period Scene The contribution of new energy sources and load fluctuations to cross-sectional power;

[0039] Then the CVaR constraint is,

[0040]

[0041] in, For key sections Time period The variable that exceeds the limit, The acceptable average threshold for exceeding limits;

[0042] After linearization, we get...

[0043]

[0044]

[0045]

[0046] in, As an auxiliary variable for Value at Risk (VaR), it represents the confidence level. The threshold for exceeding the limit quantile; For scene probabilities; For the scene The excess and over-limit variables, It is the set of real numbers.

[0047] In step C, if no feasible solution is found when solving the power reduction optimization model, slack variables are introduced. Relax dynamic threshold constraints

[0048]

[0049] in, For market entities During the period Reduced trading of electricity For market entities During the period Reduced trading of electricity and For any adjacent time period; For dynamic thresholds; Indicates the buyer,

[0050] Add a penalty item to the target. ,in, The penalty coefficient is a large positive number. It is a small coefficient.

[0051] Among them, the dynamic threshold According to the following formula,

[0052]

[0053] in, Based on the threshold, This is an adjustable coefficient; For time period The weighted average bid-ask spread, This represents the historical / period minimum of the bid-ask spread. This represents the historical / period maximum value of the bid-ask spread; For regional power balance margin, This represents the historical / period maximum value of the power balance margin; The congestion index is used to indicate the flow of traffic.

[0054] In step B, Monte Carlo sampling combined with historical error distribution is used to generate... Each scene It includes the renewable energy output multiplier, load fluctuation term, and scenario probability for each time period.

[0055] In steps A and D, when clearing the electricity market, a weighted allocation scheme based on the comprehensive contribution of entities with the same price difference is adopted.

[0056] For market entities Its overall contribution It consists of three weighted parts.

[0057]

[0058] in, Contribution to green development To adjust flexibility, Historical contract fulfillment rate; These are the weighting coefficients.

[0059] An adaptive reduction system for regional mutual aid power trading includes,

[0060] The initial electricity market clearing module is configured to clear the electricity market for each trading period in chronological order, thereby obtaining the initial trading electricity data of each market participant for each trading period.

[0061] The power reduction optimization model building module is configured to build a power reduction optimization model that considers all uncertain scenarios, wherein the constraints of the model include chance constraints.

[0062] The model solving module is configured to solve the power reduction optimization model to obtain the power reduction transaction data of each market participant in each trading period;

[0063] The repeated electricity market clearing module is configured to use the reduced traded electricity data as the maximum tradable electricity limit for each market participant in each trading period, and to perform a new electricity market clearing for each trading period to obtain new traded electricity data; and...

[0064] The iteration module is configured to repeat the actions of the aforementioned modules until the difference between two clearing results is less than a set threshold or the maximum number of iterations is reached, and the transaction power data at this point is used as the final transaction result.

[0065] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor; the processor executes the computer program to implement the steps of the adaptive reduction method for regional mutual aid power trading as described above.

[0066] A computer-readable storage medium storing a computer program; when executed by a processor, the computer program implements the steps of the adaptive reduction method for regional mutual aid power trading as described above.

[0067] Compared with the prior art, the present invention has the following advantages after adopting the above solution:

[0068] (1) Dynamic adaptive threshold reduction to overcome the rigidity of fixed thresholds: In existing technologies, fixed values ​​(such as 600MW) are usually used as limits for power changes or reverse power superposition in adjacent time periods. This fixed threshold cannot adapt to the differences in power supply and demand in different time periods and regions. The power fluctuation ranges during peak and off-peak periods are very different, and the power flow characteristics of new energy rich areas and load centers are also very different. Fixed thresholds either lead to excessive risk of exceeding limits during peak periods or are too strict during off-peak periods, resulting in the cancellation of a large number of effective transactions. This invention constructs an adaptive threshold reduction dynamic adjustment model based on real-time power data. By calculating the price difference between buyers and sellers, regional power balance margin, and power flow congestion in real time, the threshold range is dynamically adjusted: when the price difference is larger and the balance margin is more sufficient, the threshold is appropriately relaxed to improve the economic efficiency of transactions; when the power flow is congested, the threshold is automatically tightened to ensure grid security. At the same time, the threshold relaxation gradient is set in combination with the relaxation treatment when there is no solution to avoid optimization deviation caused by single relaxation. Compared with existing fixed threshold methods, this invention can improve the transaction completion rate, significantly reduce the risk of power flow exceeding the limit, and achieve "dynamic adjustment on demand" of the threshold reduction, adapting to the characteristics of large supply and demand fluctuations in new power systems.

[0069] (2) Robust optimization combining opportunity constraints and CVaR overcomes the defect of deterministic methods ignoring uncertainty: Existing power reduction models are usually based on deterministic forecast data (such as load forecasts and new energy output forecasts), completely ignoring uncertainties such as photovoltaic / wind power output fluctuations and load mutations. This deterministic method is feasible when the forecast is accurate, but in actual operation, the output of new energy sources deviates significantly, leading to frequent power flow overruns and section congestion in the actual implementation of the reduced trading scheme. The grid dispatching agency is forced to temporarily cut loads or curtail wind and solar power, which in severe cases threatens the safe operation of the system. If the safety boundary is artificially tightened to avoid risks, it will result in excessive reduction and sacrifice of a large amount of social welfare. This invention abandons the deterministic assumption and introduces a robust optimization framework combining opportunity constraints and conditional value at risk (CVaR): it guarantees the satisfaction of power flow safety constraints with a high confidence level probability, allows slight overruns in a very small number of scenarios, and strictly limits the expected magnitude of overruns through CVaR. Compared to deterministic methods, which face the dilemma of high risk of exceeding limits and difficulty in balancing safety and economy in actual operation, this invention controls the probability of power flow exceeding limits at a low level and has a clear upper bound on the amount of exceeding limits. At the same time, by allowing a small probability of controllable exceeding limits, it avoids the social welfare loss caused by excessive reduction, thus achieving an effective balance between safety and economy. In addition, all constraints of this invention can be expressed linearly, and the solution efficiency is comparable to that of deterministic models. The parameters (confidence level, exceeding limit threshold) can be flexibly adjusted according to the actual safety requirements of the power grid, overcoming the dilemma of deterministic methods that are "either not safe enough or overly conservative".

[0070] (3) A multi-objective weighted collaborative optimization model to achieve a balance between maximizing social welfare and minimizing adjustment: Existing power reduction models take "minimizing the overall adjustment" as the single optimization objective, without considering the differences in social welfare contributions of different transactions. This results in high-value transactions with high price differences and low-value transactions with low price differences being treated and reduced equally, which seriously deviates from the core orientation of the power market of "maximizing the price difference between buyers and sellers". This invention constructs a dual-objective collaborative optimization framework in the reduction model: on the one hand, minimizing the overall power adjustment to control market disturbances, and on the other hand, maximizing the expected value of social welfare under various uncertainty scenarios. Social welfare is assigned a decreasing weight according to the initial transaction price difference to ensure that high-price-difference transactions are preferentially retained during reduction. The dual objectives are transformed into a single-objective linear programming through a weighted sum method, and the weight coefficients can be adjusted according to actual operating preferences (such as emphasizing economy or stability). Under the same adjustment, compared with the single-objective model, this invention truly realizes the design intention of "optimal social welfare ranking" while ensuring the feasibility of the transaction results, making the market clearing results more in line with the requirements of resource allocation efficiency.

[0071] (4) Weighted allocation based on comprehensive contribution to solve the problem of fair ranking of entities with the same price difference: When multiple sellers or buyers have the same marginal price (i.e., the same price difference), price signals alone cannot distinguish priorities. Existing technologies handle this by equal distribution according to the bottleneck side ratio, random allocation, simple round-robin, or sorting by application time, which lacks transparent and quantifiable basis for fairness, easily leads to disputes among market participants, and cannot form any positive incentives. This invention proposes a comprehensive contribution index, which quantifies and normalizes each entity to the 0~1 range from three dimensions: green contribution (new energy ratio, carbon emission reduction), adjustment flexibility (adjustable capacity, ramp-up rate), and historical performance rate (transaction execution deviation). The weight coefficients of the three dimensions can be adjusted according to actual policy guidance (for example, the weight of green contribution can be increased in areas with high new energy penetration). In the initial clearing stage, entities with the same price difference are allocated the traded electricity according to the proportion of comprehensive contribution; in the reduction stage, entities with the same price difference are allocated the reduction amount inversely proportional to the comprehensive contribution, that is, entities with higher contribution receive less reduction. This method provides a transparent and traceable ranking basis for scenarios with similar price differences, while incentivizing market players to improve green performance, enhance regulatory capabilities, and improve compliance records, thus having a good economic guiding effect.

[0072] (5) Closed-loop iterative mechanism to ensure consistency between reduction results and clearing logic: Existing technologies directly output results after power reduction without re-substituting the reduced power ceiling into the clearing process for verification and correction. This can lead to conflicts between the reduced trading results and the original clearing logic (price difference priority principle), requiring manual intervention for adjustment during actual execution. In this invention, after solving the reduction model to obtain the reduced trading power for each subject in each time period, the results are used as the maximum tradable power ceiling for each subject in each time period, and the process is returned to step 1 to perform time-segmented clearing again. This iterative process of "clearing → reduction → re-clearing" is repeated until the difference between the two clearing results is less than a set threshold (e.g., 1MW) or the maximum number of iterations (e.g., 5 times) is reached. This closed-loop iterative mechanism ensures that the final trading results simultaneously satisfy reduction constraints (ramp-up, power flow, uncertainty robustness) and clearing rules (price difference priority, social welfare maximization), eliminating the need for manual intervention and improving the automation level and reliability of the trading system. Detailed Implementation

[0073] The technical solution and beneficial effects of the present invention will be described in detail below.

[0074] This invention provides an adaptive reduction method for regional mutual assistance power trading, employing a technical approach of "time-segmented clearing - power reduction model - repeated clearing," specifically including the following steps:

[0075] Step 1: Clear electricity in time slots to obtain initial trading power.

[0076] Electricity market clearing is performed sequentially for each trading session (e.g., 15 minutes per session). The clearing rule follows the principle of "maximum bid-ask spread priority," meaning transactions are executed in descending order of bid-ask spread until all possible combinations of transactions have been exhausted or a safety limit has been reached. The initial electricity trading matrix for each market participant (power generation companies, electricity retailers, large users, etc.) is output for each session. ,in Indicates market entities, Indicates a time period.

[0077] Step 2: Construct a robust power reduction optimization model based on opportunity constraints and CVaR risk metric.

[0078] This model considers all uncertain scenarios simultaneously, replaces the traditional "all scenarios are satisfied" constraint with opportunity constraints, and controls the severity of exceeding the limit through conditional value at risk (CVaR), thus significantly improving economic efficiency while ensuring safety and controllability.

[0079] 2.1 Decision Variables

[0080] Market entities During the period Reduced traded electricity (continuous variable, in MW);

[0081] Positive and negative adjustment amounts (reduction amounts) to meet the requirements. ,and Usually only reductions are allowed. However, a small increase is permissible during relaxation treatment.

[0082] 2.2 Objective Function (Dual Objective Synergy)

[0083] The weighted sum method transforms "minimizing the overall adjustment" and "maximizing the expected social welfare" into a single objective:

[0084]

[0085] in:

[0086] The first item is the overall adjustment (normalization factor). );

[0087] For the scene The social welfare under this condition is defined as the sum of "buyer's offer - seller's offer" for all completed transactions. In actual calculations, a linear sorting method is introduced: decreasing weights are assigned based on the price difference from largest to smallest. ,but ,in Positively correlated with the initial trade spread;

[0088] Weighting coefficients (typical values) ).

[0089] 2.3 Constraints

[0090] Constraint 1: Adaptive Climbing / Reverse Stacking Constraint (Deterministic)

[0091] For any adjacent time period and :

[0092] If the total power of all sellers and the total power of all buyers have the same sign (both net selling or net buying) in two time periods, then:

[0093]

[0094]

[0095] If one period is characterized by net selling and the other by net buying, then:

[0096]

[0097]

[0098] in, For the buyer, For the seller, This is a dynamic threshold.

[0099] Constraint 2: Opportunity Constraints for Power Flow Safety and CVAR Control

[0100] For each critical section (such as lines, transformers) and time periods Define a random variable:

[0101]

[0102] in The power transfer distribution factor, For the scene The contribution of new energy sources and load fluctuations to cross-sectional power.

[0103] Opportunity constraints require:

[0104]

[0105] in, Indicates the probability of taking the value. The allowed probability of violation (typical value 0.05). This represents the cross-sectional transmission limit.

[0106] To control the severity of boundary violations, CVaR constraints are introduced. First, the scenario is defined. Off-limit variables :

[0107]

[0108] in, Indicated at the critical section Time period Scene The contribution of new energy sources and load fluctuations to the cross-sectional power, i.e., the uncertain power component;

[0109] The CVaR constraint is then:

[0110]

[0111] in, For key sections Time period The over-limit variable, whose value is determined by all uncertain scenarios. Specific out-of-bounds variables Decide; The acceptable average exceedance threshold is (e.g., 2% of cross-sectional capacity). The above constraint can be linearized as:

[0112]

[0113]

[0114]

[0115] in, As an auxiliary variable for Value at Risk (VaR), it represents the confidence level. The threshold for exceeding the limit, i.e., "the maximum amount that is not included in the tail risk calculation", is a continuous variable introduced when linearizing CVaR, corresponding to the risk quantile under the scenario method; For the scene The excess limit variable is defined as follows: , indicating a scene The China-Vietnam limit exceeds the VaR threshold. The part if and only if It takes a positive value when it is positive and 0 otherwise, and is used to calculate the average over-limit of the tail scene; It is the set of real numbers.

[0116] Constraint 3: Opportunity Constraints for Regional Power Balance

[0117] Within each region, power generation and load must be approximately balanced for each time period. This can be achieved using a CVaR form similar to constraint 2, or simplified to a deterministic balance with relaxation variables to reduce complexity. This invention provides a simplified version:

[0118]

[0119] in, A collection of conventional generator sets. For time period Within the region, the combined output of all conventional generating units; Contribute to new energy forecasting; For the set of load nodes, This indicates traversing all load nodes within the region. Indicates time period Within the region, the total power demand of all loads; For load nodes During the period The load power, As a slack variable, it is used as a penalty term in the objective. Included, This is the penalty coefficient.

[0120] Constraint 4: Upper and lower limits for traded electricity

[0121]

[0122] Constraint 5: Relaxation measures (for cases with no solution)

[0123] If the above constraints result in no feasible solution, slack variables can be further introduced. Relaxing dynamic threshold constraints:

[0124]

[0125] The same applies to reverse superposition constraints, and a penalty term is added to the objective. ,in For small coefficients (e.g., 0.01), it is encouraged to be as close as possible to the original threshold; The penalty coefficient is a large positive number.

[0126] 2.4 Solution Method

[0127] The above model is a linear programming problem (the absolute value can be linearized by introducing positive and negative deviation variables, and the CVaR constraint is linear). It can be solved using commercial solvers (Gurobi, CPLEX) or open-source solvers (HiGHS) to obtain the reduced traded electricity for each entity in each time period. .

[0128] In constraint 1, the dynamic threshold can be calculated by constructing an adaptive threshold reduction dynamic adjustment model:

[0129] Traditional fixed threshold (assuming) The function is replaced with a dynamic threshold function.

[0130]

[0131] in:

[0132] Time period The weighted average bid-ask spread (RMB / MWh) indicates that the larger the spread, the higher the economic efficiency of the transaction, and the threshold should be appropriately relaxed.

[0133] Regional power balance margin is defined as follows: The larger the margin, the more relaxed the threshold can be; among them, For time period The maximum available power supply capacity in the area, For time period Total load power within the area;

[0134] The congestion index is calculated based on the load rate of key sections; the threshold tightens as congestion worsens.

[0135] Specific threshold calculation formula:

[0136]

[0137] in, The base threshold (can be set to 600MW). Adjustable coefficient (value range 0~1, typical value) ); This represents the historical / period minimum of the bid-ask spread. This represents the historical / period maximum value of the bid-ask spread; This represents the historical / periodic maximum value of the power balance margin. The threshold value range is set to... For example, 200MW~1000MW.

[0138] This includes generating a set of uncertainty scenarios for the model to use:

[0139] To account for fluctuations in new energy output and sudden load changes, Monte Carlo sampling combined with historical error distribution is used to generate... Scenario (typical value) Each scene Include:

[0140] New energy output multiplier at different times (Follows a Beta distribution, with parameters obtained by fitting historical prediction errors);

[0141] Load fluctuation term , This represents the standard deviation of the load forecasting error.

[0142] Scenario Probability Determined by kernel density estimation or equal weighting ( ).

[0143] Step 3, repeat clearing

[0144] The result obtained in step 2 As the maximum tradable electricity limit for each entity in each time period, return to step 1 to re-clear by time period. Repeat the iteration until the difference between two clearing results is less than a set threshold (e.g., 1MW), or the maximum number of iterations is reached (e.g., 5 times). Finally, output the trading results.

[0145] In a preferred embodiment of the present invention, in the clearing stage of step 1 and the repeated clearing stage of step 3, a weighted allocation scheme based on the comprehensive contribution of entities with the same price difference is adopted for electricity trading, as follows:

[0146] When multiple sellers (or buyers) have the same marginal offer (i.e., the same price spread) at a certain time, price signals alone cannot distinguish priorities. This invention proposes a comprehensive contribution index to allocate reductions or traded electricity when price spreads are the same.

[0147] For market entities Its overall contribution It consists of three weighted parts:

[0148]

[0149] in:

[0150] Green contribution, based on the normalized value (0~1) of the entity's proportion of new energy or carbon emission reduction.

[0151] : Adjustment flexibility, based on its adjustable capacity, ramp rate, etc. (0~1);

[0152] Historical fulfillment rate, based on the reciprocal (0~1) of past transaction execution deviations.

[0153] Weighting coefficients (typical values) ).

[0154] In the initial clearing phase of step 1, when multiple buyers or sellers with the same price difference appear, the transaction power is allocated according to the proportion of their overall contribution. In the reduction phase of step 2, if it is necessary to reduce the transaction volume of the same price difference group, the reduction is carried out inversely according to the proportion of the overall contribution (that is, the higher the contribution, the less the reduction).

[0155] Specific allocation formula (taking reduction as an example):

[0156] Let the set of principals within the same price difference group be... The total amount to be reduced is Then the main body The reduction amount is:

[0157]

[0158] If the reduction exceeds the initial traded electricity for that entity, the remainder is redistributed to other entities.

[0159] This invention also provides another computer device, including a processor and a memory configured to store a computer program capable of running on the processor; wherein, when the processor is configured to run the computer program, it performs the method steps described in the foregoing embodiments.

[0160] In practical applications, the aforementioned processor includes a Field-Programmable Gate Array (FPGA), and the processor can be a Central Processing Unit (CPU) or a Digital Signal Processor (DSP). It is understood that for different devices, the electronic devices used to implement the above-mentioned processor functions can also be other types, and this embodiment of the invention does not impose specific limitations.

[0161] The aforementioned memory can be volatile memory, such as random-access memory (RAM); or non-volatile memory, such as read-only memory (ROM), flash memory, hard disk drive (HDD), or solid-state drive (SSD); or a combination of the above types of memory, and provides instructions and data to the processor.

[0162] In an exemplary embodiment, the present invention also provides a computer-readable storage medium for storing a computer program.

[0163] Optionally, the computer-readable storage medium can be applied to any of the methods in the embodiments of the present invention, and the computer program causes the computer to execute the corresponding processes implemented by the processor in the various methods of the embodiments of the present invention. For the sake of brevity, these will not be described in detail here.

[0164] In the several embodiments provided by this invention, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.

Claims

1. An adaptive reduction method for regional mutual assistance power trading, characterized in that: include, Step A: According to the time sequence, the electricity market is cleared for each trading period to obtain the initial trading electricity data of each market participant in each trading period; Step B: Construct an optimization model for power reduction that considers all uncertain scenarios, wherein the constraints of the model include chance constraints. Step C: Solve the power reduction optimization model to obtain the power reduction transaction data of each market participant in each trading period; Step D: The reduced traded electricity data is used as the maximum tradable electricity limit for each market participant in each trading period. The electricity market is cleared again for each trading period to obtain new traded electricity data. Step E: Repeat steps A-D until the difference between the two clearing results is less than the set threshold or the maximum number of iterations is reached, and use the transaction power data at this time as the final transaction result. In step B, a power reduction optimization model considering all uncertainties is constructed, including: The objective function of the model is: ， in, As the normalization factor, , This provides the initial trading electricity matrix for each market participant in each trading session. Indicates market entities, Indicates the trading session; These represent positive and negative reduction amounts, respectively. ; This represents the total number of uncertain scenarios; For the scene Social welfare under the following circumstances; For scene probabilities; These are the weighting coefficients; The constraints of the model include adaptive ramping / reverse superposition constraints, opportunity constraints for power flow security and CVaR control constraints, opportunity constraints for regional power balance, and upper and lower limits for trading power. In steps A and D, when clearing the electricity market, a weighted allocation scheme based on the comprehensive contribution of entities with the same price difference is adopted. For market entities Its overall contribution It consists of three weighted parts. ， in, Contribution to green development To adjust flexibility, Historical contract fulfillment rate; These are the weighting coefficients.

2. The method as described in claim 1, characterized in that: In step A, electricity market clearing is performed for each trading period, including: The clearing rule prioritizes the largest bid-ask spread, meaning that transactions are executed sequentially from largest to smallest bid-ask spread until all possible combinations of transactions have been traversed or a safety limit has been reached.

3. The method as described in claim 1, characterized in that: Scene Social welfare Defined as the sum of "buyer's quote - seller's quote" for all completed transactions. ,in, The decreasing weights are assigned based on the price difference from largest to smallest, and they are positively correlated with the initial transaction price difference. For market entities During the period The reduction in traded electricity.

4. The method as described in claim 1, characterized in that: The opportunity constraints and CVaR control constraints for power flow safety include the following: For each critical section and time period Define a random variable. ， in, The power transfer distribution factor, For market entities During the period Reduced trading of electricity; For the scene The contribution of new energy sources and load fluctuations to cross-sectional power; The opportunity constraint requires that, ， in, Indicates the probability of taking the value. For the allowed probability of violation, This represents the cross-sectional transmission limit; Define the scenario based on CVaR constraints. Off-limit variables for, ， in, Indicated at the critical section Time period Scene The contribution of new energy sources and load fluctuations to cross-sectional power; Then the CVaR constraint is, ， in, For key sections Time period The variable that exceeds the limit, The acceptable average threshold for exceeding limits; After linearization, we get... ， ， ， in, As an auxiliary variable for Value at Risk (VaR), it represents the confidence level. The threshold for exceeding the limit quantile; For scene probabilities; For the scene The excess or over-limit variable, It is the set of real numbers.

5. The method as described in claim 1, characterized in that: In step C, if no feasible solution is found when solving the power reduction optimization model, slack variables are introduced. Relax dynamic threshold constraints ， in, For market entities During the period Reduced trading of electricity For market entities During the period Reduced trading of electricity and For any adjacent time period; For dynamic thresholds; Indicates the buyer, Add a penalty item to the target. ,in, The penalty coefficient is a large positive number. It is a small coefficient.

6. The method as described in claim 5, characterized in that: The dynamic threshold According to the following formula, ， in, Based on the threshold, This is an adjustable coefficient; For time period The weighted average bid-ask spread, This represents the historical / period minimum of the bid-ask spread. This represents the historical / period maximum value of the bid-ask spread; For regional power balance margin, This represents the historical / period maximum value of the power balance margin; The congestion index is used to indicate the flow of traffic.

7. The method as described in claim 1, characterized in that: In step B, Monte Carlo sampling combined with historical error distribution is used to generate... Each scene It includes the renewable energy output multiplier, load fluctuation term, and scenario probability for each time period.

8. An adaptive reduction system for regional mutual aid power trading, used to implement the adaptive reduction method for regional mutual aid power trading as described in any one of claims 1 to 7; characterized in that: include, The initial electricity market clearing module is configured to clear the electricity market for each trading period in chronological order, thereby obtaining the initial trading electricity data of each market participant for each trading period. The power reduction optimization model building module is configured to build a power reduction optimization model that considers all uncertain scenarios, wherein the constraints of the model include chance constraints. The model solving module is configured to solve the power reduction optimization model to obtain the power reduction transaction data of each market participant in each trading period; The repeated electricity market clearing module is configured to use the reduced traded electricity data as the maximum tradable electricity limit for each market participant in each trading period, and to perform electricity market clearing again for each trading period to obtain new traded electricity data. as well as, The iteration module is configured to repeat the actions of the aforementioned modules until the difference between two clearing results is less than a set threshold or the maximum number of iterations is reached, and the transaction power data at this time is used as the final transaction result.

9. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor; characterized in that: When the processor executes the computer program, it implements the steps of the adaptive reduction method for regional mutual aid power trading as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program; characterized in that: When the computer program is executed by a processor, it implements the steps of the adaptive reduction method for regional mutual aid power trading as described in any one of claims 1 to 7.

Images