Power spot market bidding decision method, system and device considering deviation assessment, electronic equipment and storage medium

Abstract

The application discloses a power spot market bidding decision-making method considering bias assessment, and belongs to the technical field of power market transaction decision-making. The method comprises the following steps: acquiring benchmark transaction parameters of arbitrage bidding and market bias income recovery rules; constructing a settlement net income model of arbitrage bidding, wherein the net income model models the bias income recovery rules as a function affecting net income; constructing a distribution robust optimization model considering market price difference prediction uncertainty, wherein the distribution robust optimization model is based on the net income model and takes arbitrage bidding quantity as a decision variable; and solving the distribution robust optimization model to output the optimal arbitrage bidding quantity. The application embeds market segmented bias income recovery rules in the decision-making model, and adopts a distribution robust optimization framework to cope with uncertainty, thereby achieving the technical effect of significantly improving decision-making robustness and income calculation accuracy under the premise of strict compliance.

Description

Technical Field

[0001] This application relates to the field of electricity spot market transaction decision-making technology, and in particular to an electricity spot market declaration decision-making method, system, device, electronic equipment and storage medium that takes deviation assessment into account. Background Technology

[0002] With the deepening of power market reforms, the spot market has become a crucial link in the optimal allocation of power resources. When power retailers participate in day-ahead-real-time arbitrage transactions, their bidding behavior directly impacts system balance and market stability. To regulate market behavior, power market regulatory agencies in some regions (such as Shanxi) have formulated clear rules for deviation assessment and profit recovery. For example, if a power retailer engages in arbitrage by amplifying its bidding volume, leading to a significant increase in system balancing costs, its corresponding excess profits will be recovered according to regulations. This segmented deviation profit recovery rule directly determines the final settlement profit of arbitrage bids and is a core regulatory constraint that arbitrage decisions must adapt to.

[0003] Currently, the existing technical solutions for arbitrage reporting by electricity sales companies mainly have the following defects:

[0004] Insufficient rule adaptability: Traditional deterministic optimization schemes for arbitrage reporting based on mixed integer linear programming (MILP) only use the deviation exemption threshold as a hard constraint, assuming that arbitrage profits can be fully retained. This scheme only performs compliance verification in the post-settlement stage, failing to fully and explicitly embed the profit recovery rules into the pre-decision optimization process. This results in a serious disconnect between the expected returns calculated by the model and the actual settlement results, leading to distorted decision-making.

[0005] Weak risk resistance: Traditional deterministic optimization methods or general robust optimization schemes that do not consider local rules are unable to effectively cope with the significant forecasting uncertainty caused by the "fat-tailed" distribution of electricity spot market prices. In extreme market scenarios where price spread forecasts deviate significantly or even reverse, such methods have a high risk of decision failure and may lead to serious losses.

[0006] Therefore, there is an urgent need for a technical solution that can deeply integrate localized regulatory rules with cutting-edge optimization theories, while simultaneously addressing the two core pain points of compliance adaptation and sound decision-making. Summary of the Invention

[0007] This application provides a method, system, device, electronic device, and storage medium for electricity spot market declaration decision-making that considers deviation assessment. It accurately models the deviation revenue recovery rules of the electricity market and fully embeds them into the decision optimization process, ensuring the compliance of the decision and the accuracy of the revenue calculation from the root. Furthermore, it introduces an advanced sub-Bruker optimization framework to effectively cope with the uncertainty of price difference prediction and significantly improve the risk resistance and revenue robustness of the decision scheme in extreme market environments.

[0008] On the one hand, embodiments of this application provide a method for electricity spot market declaration decision-making that considers deviation assessment, including:

[0009] Obtain the benchmark trading parameters for arbitrage applications and the rules for recovering market deviation profits;

[0010] A settlement net profit model for arbitrage declarations is constructed, wherein the deviation profit recovery rule is used as a function affecting the net profit in the model.

[0011] Construct a sub-Brussels bar optimization model that considers the uncertainty of market price spread prediction. The sub-Brussels bar optimization model is based on the net profit model and uses the arbitrage application volume as the decision variable.

[0012] Solve the aforementioned arbitrage bar optimization model and output the optimal arbitrage declaration quantity.

[0013] In one possible embodiment, obtaining the benchmark trading parameters and market deviation profit recovery rules for arbitrage applications includes:

[0014] Obtain the benchmark trading parameters for arbitrage applications, which include the benchmark application electricity volume and the predicted market price difference;

[0015] as well as,

[0016] Obtain the market deviation profit recovery rule, which is used to define the piecewise function rule of the profit recovery logic, including the exemption threshold and the profit recovery ratio coefficient.

[0017] In one possible embodiment, the construction of the settlement net profit model for arbitrage declarations includes:

[0018] Based on the aforementioned benchmark transaction parameters, calculate the expected declared profit without triggering profit recovery.

[0019] Based on the market deviation profit recovery rule, the profit recovery calculation logic corresponding to the arbitrage declaration volume is determined, and the profit recovery amount is calculated.

[0020] Based on the expected declared revenue and the revenue recovery amount, the settlement net revenue model is constructed.

[0021] In one possible embodiment,

[0022] The method for calculating the expected declared revenue without triggering revenue recovery includes: multiplying the benchmark declared electricity volume, the predicted market price difference, and the arbitrage declaration amplification factor as the expected declared revenue;

[0023] The logic for determining the profit recovery calculation corresponding to the arbitrage declaration volume includes the following method for calculating the profit recovery amount: when the deviation rate of the arbitrage declaration amplification coefficient relative to the benchmark value exceeds the exemption threshold, the profit recovery amount is calculated based on the profit recovery ratio, the predicted market price difference, and the benchmark declared electricity volume.

[0024] In one possible embodiment, constructing the sub-Bruker optimization model that considers the uncertainty of market spread prediction includes:

[0025] Based on historical price spread data, a fuzzy set containing multiple possible price spread probability distributions is constructed. The fuzzy set is used to characterize the uncertainty of the market price spread prediction.

[0026] Using the arbitrage declaration volume as the decision variable and the net profit calculated by the settlement net profit model as the optimization objective, an optimization model is constructed to find the optimal net profit under the worst probability distribution among all possible probability distributions defined by the fuzzy set.

[0027] In one possible embodiment, constructing a fuzzy set containing multiple possible price spread probability distributions includes:

[0028] A baseline probability distribution is determined based on historical price spread data;

[0029] A set of probability distributions is constructed with the baseline probability distribution as the center and a preset distance tolerance as the radius.

[0030] Wherein, the probability distance between the probability distribution within the set and the baseline probability distribution is not greater than the distance tolerance.

[0031] In one possible embodiment, solving the arbitrage bar optimization model and outputting the optimal arbitrage declaration quantity includes:

[0032] By performing a dual transformation on the aforementioned sub-Bruker optimization model, it is transformed from a minimax problem containing fuzzy sets of probability distributions into a deterministic problem that can be directly solved by a standard mathematical programming solver.

[0033] The mathematical programming solver is invoked to solve the deterministic problem, thereby obtaining the optimal arbitrage application quantity.

[0034] On the one hand, embodiments of this application provide a power spot market bidding decision system that considers deviation assessment, including:

[0035] The parameter acquisition module is used to obtain the benchmark trading parameters for arbitrage applications and the rules for recovering market deviation profits;

[0036] The model building module is used to build a settlement net profit model for arbitrage declarations. The net profit model uses the deviation profit recovery rule as a function affecting the net profit.

[0037] The model optimization module is used to construct a sub-Brussels bar optimization model that considers the uncertainty of market price spread prediction. The sub-Brussels bar optimization model is based on the net profit model and uses the arbitrage application volume as the decision variable.

[0038] The solution output module is used to solve the arbitrage bar optimization model and output the optimal arbitrage declaration quantity.

[0039] On the one hand, embodiments of this application provide an electronic device, which includes a processor and a memory, wherein the memory stores program code, and when the program code is executed by the processor, the processor executes any of the above-mentioned methods for making decisions on electricity spot market declarations that take into account deviation assessment.

[0040] On the one hand, this application provides a computer-readable storage medium including program code, which, when the storage medium is run on an electronic device, causes the electronic device to execute any of the above-mentioned methods for making decisions on electricity spot market declarations that take into account deviation assessments.

[0041] On one hand, an embodiment of this application provides a computer program product, which includes computer instructions stored in a computer-readable storage medium; when the processor of an electronic device reads the computer instructions from the computer-readable storage medium, the processor executes the computer instructions, causing the electronic device to execute any of the above-mentioned methods for making decisions on electricity spot market declarations that take into account deviation assessment.

[0042] The beneficial effects of this application are as follows:

[0043] 1. This application is the first to explicitly and solvably model the segmented deviation revenue recovery rules of the electricity market, and fully embeds them as rigid constraints into the entire decision optimization process to ensure that arbitrage reporting behavior automatically meets regulatory requirements, while accurately calculating the actual retainable revenue, fundamentally solving the problem of revenue calculation distortion.

[0044] 2. By introducing the sub-Bruker optimization framework, this application can effectively characterize and resist the uncertainty of electricity price forecasting. Even in extreme market environments where price difference forecasts deviate significantly, the decision-making scheme can still guarantee positive returns or significantly reduce the loss margin, demonstrating excellent risk resistance capabilities.

[0045] 3. This application transforms the complex bibliometric optimization model into a standard deterministic optimization problem through dual transformation. It can be solved efficiently using mature commercial solvers (such as GUROBI and COPT), with short computation time. It can meet the strict timeliness requirements of electricity spot market declarations and can be directly integrated into the automated trading decision system of electricity sales companies, making it highly practical.

[0046] Other features and advantages of this application will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the application. The objectives and other advantages of this application may be realized and obtained by means of the structures particularly pointed out in the written description, claims, and drawings. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the drawings described below are only embodiments of this application. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0048] Figure 1 This is a flowchart illustrating the implementation of a power spot market declaration decision-making method that considers deviation assessment in an embodiment of this application.

[0049] Figure 2 This is a schematic diagram of the structure of a power spot market declaration decision system that considers deviation assessment in an embodiment of this application;

[0050] Figure 3 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of this application. Detailed Implementation

[0051] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application. Unless otherwise specified, the embodiments and features in the embodiments of this application can be arbitrarily combined with each other. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than that shown here.

[0052] The preferred embodiments of this application are described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit this application. Furthermore, the embodiments and features in the embodiments of this application can be combined with each other without conflict.

[0053] In the electricity spot market, electricity retailers typically employ an "arbitrage reporting" strategy to generate profits. This involves reporting electricity volumes that deviate from their real-time physical delivery demand in the day-ahead market, profiting from the price difference between the day-ahead and real-time markets. To regulate market behavior, some regional electricity regulatory authorities, such as those in Shanxi, have introduced localized deviation assessment and profit recovery rules. These rules stipulate that if an electricity retailer's arbitrage reporting behavior significantly increases system balancing costs, its excess profits will be recovered according to specific segmented rules. Currently, the mainstream solutions for the "arbitrage reporting" strategy mainly fall into two representative technical categories: The first is a deterministic optimization method based on mixed integer linear programming (MILP). This method aims to maximize arbitrage profits but only uses the exemption threshold in the rules as a simple constraint, assuming that profits within the threshold are fully retained. Its technical flaw lies in the disconnect between the internal profit calculation logic of the model and the actual market settlement rules, failing to internalize the crucial settlement link of profit recovery through modeling, leading to inaccurate decision-making results. The second category is electricity purchase decision models based on stochastic programming or robust optimization. Although these models can handle the uncertainty of electricity prices or loads, their framework does not incorporate a segmented revenue recovery cost function for arbitrage applications, and therefore cannot solve the problem of local rule adaptability.

[0054] To address the aforementioned deficiencies in existing technologies, this application proposes a decision-making method for electricity spot market declarations that considers deviation assessment. For the first time, it creatively expresses the electricity market's unique "segmented deviation revenue recovery rule" in a calculable formula and embeds it as an endogenous constraint into an advanced sub-Bruker optimization decision-making framework, forming a closed-loop technical solution from rule adaptation to robust solution.

[0055] The core idea of ​​this application is to transform market regulatory rules (taking Shanxi rules as an example, with core parameters being the exemption threshold and the profit recovery ratio) into an explicit, piecewise linear mathematical function, and directly introduce it as a cost item into the settlement net profit model. Endogenous modeling of the rules ensures that the objective function of the optimization model is completely consistent with the final settlement caliber, eliminating measurement distortion at its source. Simultaneously, addressing the "leggy" distribution characteristics of price spread prediction, this application introduces a biblical optimization framework based on Wasserstein distance. Instead of relying on a single predicted value, it constructs a "fuzzy set of probability distribution" centered on historical experience distributions and with Wasserstein distance as its radius. This set describes all possible price spread scenarios. The decision objective is to seek the optimal arbitrage application quantity under the worst-case distribution covered by this fuzzy set, thus providing performance assurance for the worst-case scenario. For this complex sub-Bruker optimization model with piecewise constraints, the dual transformation is used as a key technical means to transform the "maximum minima" problem in the original problem, which is difficult to solve directly, into a standard, deterministic convex optimization problem (such as linear programming). This allows the solution to be completed in seconds with the help of mature high-performance commercial mathematical solvers (such as GUROBI and COPT), which fully meets the timeliness requirements of power market application.

[0056] like Figure 1 The diagram shown is an implementation flowchart of a method for making decisions in the electricity spot market that considers deviation assessment, provided in an embodiment of this application. The specific implementation process of this method is as follows:

[0057] S101, obtain the benchmark trading parameters for arbitrage applications and the rules for recovering market deviation profits.

[0058] This step provides accurate and reliable input data for subsequent modeling and optimization processes. Its implementation is not dependent on specific hardware and can be integrated into the electricity sales company's transaction decision system (TMS), and can be completed automatically or semi-automatically through a data interface.

[0059] In a preferred embodiment, step S101, which involves obtaining the benchmark trading parameters for arbitrage applications and the market deviation profit recovery rules, includes:

[0060] Obtain the benchmark trading parameters for arbitrage applications, which include the benchmark application electricity volume and the predicted market price difference;

[0061] as well as,

[0062] Obtain the market deviation profit recovery rule, which is used to define the piecewise function rule of the profit recovery logic, including the exemption threshold and the profit recovery ratio coefficient.

[0063] The benchmark trading parameters are the basis for preliminary calculation of arbitrage profits and determination of the application scale. In one specific embodiment of this application, the benchmark trading parameters include at least: benchmark application electricity volume. And the predicted market spread ( Among them, the benchmark declared electricity volume. This refers to the amount of electricity a power retailer plans to declare in the day-ahead market based on its retail contracts with electricity users, load forecasts, and physical delivery needs, without any arbitrage motive. This parameter typically originates from the power retailer's internal load forecasting system or retail contract management system. For example, for a specific trading period, the system can obtain the predicted total load for that period from the load forecasting module, subtract the amount of electricity already locked in through medium- to long-term contracts, and consider necessary reserves to calculate the baseline declared electricity amount. In one example, =100 MWh; Forecast market spread ( This refers to the difference between the predicted price in the current-day market and the real-time market, representing a potential source of profit for arbitrage activities. This parameter originates from the market analysis and forecasting module. Based on historical market price data, supply and demand forecasts, network congestion information, and renewable energy output forecasts, this module uses time series analysis and machine learning algorithms to generate predicted values ​​for the current-day and real-time prices for a specific future period and calculates their difference. For example, the predicted market price is... The real-time market price is Then predict the price difference = 150 yuan / MWh.

[0064] The market deviation return recovery rules are external policy constraints formulated by regulatory agencies to regulate the reporting behavior of market participants, and their acquisition must ensure authority and accuracy. This application is the first to transform such rules from textual clauses into quantitative parameters that can be embedded in mathematical models.

[0065] In a preferred embodiment for the Shanxi electricity spot market, the market deviation revenue recovery rule is a piecewise function rule. Its core logic is as follows: when the deviation rate between the actual declared electricity volume of a power sales company and its "benchmark value" (usually the declared day-ahead electricity volume) exceeds a certain exemption threshold, the revenue generated from the excess portion (calculated based on the price difference) will be recovered in proportion. This rule is specifically manifested in obtaining at least two core parameters: the exemption threshold... and the rate of return ( Specifically,

[0066] Exemption threshold This refers to the upper limit of the allowed reporting deviation rate, within which arbitrage profits are exempt from being recovered. This parameter is either a fixed value explicitly published by the regulatory authority or can be calculated based on publicly available rules. For example, according to the "Implementation Rules for Shanxi Electricity Spot Market Transactions," this threshold is explicitly set at 6% (i.e., = 0.06). This parameter can be directly read from the official rule document database published by the trading center or from the verified rule knowledge base; Profit recovery ratio coefficient ( This refers to the percentage of profits recovered for any deviation exceeding the exemption threshold. This parameter is also explicitly defined by regulatory rules. For example, in the aforementioned Shanxi market rules, the excess profit recovery percentage is explicitly set at 100% (i.e., η = 1.0). This parameter, along with the exemption threshold, is obtained as part of the rule quantification package.

[0067] It should be noted that a rule adapter module can be set up during the software system implementation. This module is either built into or connected to a market rule base, which stores quantitative templates of deviation assessment rules for different electricity markets (such as Shanxi, Guangdong, Zhejiang, etc.). The system automatically retrieves the corresponding parameter package from the rule base based on the selected trading market (e.g., "Shanxi"). Simultaneously, the baseline declared electricity volume is obtained from the load forecasting system via the data bus. Obtain the predicted price spread from the price spread prediction engine. At this point, all the necessary input data has been collected. The data is accurately collected and transmitted to the next processing module, laying the data foundation for building a net income model that truly reflects the settlement logic.

[0068] S102, Construct a settlement net profit model for arbitrage declarations, wherein the net profit model models the deviation profit recovery rule as a function affecting the net profit.

[0069] The main flaw in traditional arbitrage decision-making models lies in the logical disconnect between their objective function (usually maximizing expected returns) and the actual market settlement rules. To fundamentally address this issue, this embodiment creatively transforms the "deviation return recovery rule" from a textual clause into a computable mathematical function and embeds it directly into the net return calculation as an endogenous variable. The aim is to internalize external regulatory rules as mathematical constraints for the decision-making model, thereby constructing a return calculation model that accurately reflects the final settlement result and achieving a unified "decision-settlement" logic.

[0070] This embodiment constructs a settlement net revenue model, functionalizing abstract regulatory rules and embedding them into the optimization objective. This ensures that any subsequent optimization process aimed at maximizing this net revenue automatically satisfies the actual economic considerations under the rule constraints. The net revenue output by the model is the settlement revenue that the electricity sales company is expected to actually retain under the corresponding reporting decision, fundamentally eliminating the defects in revenue calculation of traditional models.

[0071] S103, construct a split-bar optimization model that considers the uncertainty of market price difference prediction. The split-bar optimization model is based on the net profit model and uses the arbitrage application volume as the decision variable.

[0072] In practical applications, perfect point prediction is difficult to achieve in complex markets. Traditional deterministic optimization (such as using only point prediction) or simple scenario analysis methods are insufficient to effectively mitigate prediction errors, especially in extreme cases where electricity market price spreads exhibit "leggy" statistical characteristics. Even a single probability distribution estimated based on limited historical data may be biased. Therefore, this embodiment constructs a set (called the fuzzy set or uncertainty set) containing all reasonably possible price spread probability distributions based on historical data. The optimization objective is set as follows: in the worst-case scenario among all possible distributions covered by this set, find the decision that maximizes net profit. The core is to seek performance guarantees for the worst-case scenario.

[0073] This embodiment elevates the robustness of decision-making to the distribution level by constructing a distributed bar optimization model. Instead of relying on a single prediction, it considers a family of possible distributions and optimizes for the worst-case scenario. This enables the final decision-making scheme to effectively address prediction biases and significantly reduces the risk of huge losses in extreme price difference scenarios (such as prediction errors, market price spikes or reversals), thereby achieving a bottom-line approach to returns.

[0074] S104, Solve the arbitrage bar optimization model and output the optimal arbitrage declaration quantity.

[0075] This embodiment, through the equivalent transformation of complex models and the invocation of standardized solvers, enables the entire decision-making process to be completed within seconds, fully meeting the stringent timeliness requirements of day-ahead declarations in the electricity spot market. This method can be modularized and integrated into the automated trading system of electricity sales companies, achieving unattended intelligent decision-making and possessing significant engineering practical value.

[0076] In some implementations, step S102, which involves constructing the settlement net profit model for arbitrage declarations, includes:

[0077] Based on the aforementioned benchmark transaction parameters, calculate the expected declared profit without triggering profit recovery.

[0078] Based on the market deviation profit recovery rule, the profit recovery calculation logic corresponding to the arbitrage declaration volume is determined, and the profit recovery amount is calculated.

[0079] Based on the expected declared revenue and the revenue recovery amount, the settlement net revenue model is constructed.

[0080] This embodiment aims to provide a specific and implementable method for constructing a net settlement return model, addressing the problem that traditional methods treat regulatory rules as "post-event verification" rather than "pre-event endogenous variables" when calculating returns. Through this method, the "theoretical return" and "regulatory cost" of arbitrage activities can be clearly and calculably separated, thereby constructing a return calculation kernel that can accurately simulate the actual market settlement process from a mathematical perspective, providing a target function that is completely consistent with real settlement for subsequent optimization decisions.

[0081] Based on the aforementioned benchmark trading parameters, the total theoretical economic returns that a power sales company could generate under ideal market conditions (i.e., completely ignoring the impact of regulatory penalties) due to executing arbitrage reporting decisions are pre-quantified. Specifically, within the "safe zone" where market deviation returns are not triggered, the economic return of arbitrage reporting is linearly determined by the objective market trading conditions (price difference) and the reporting behavior itself (reporting volume). Therefore, the core of calculating this expected return lies in establishing a deterministic mathematical mapping relationship from "reporting decision" to "theoretical return." Preferably, a calculation function can be constructed based on the aforementioned benchmark trading parameters. This function uses the arbitrage reporting volume as the core input variable and the benchmark trading parameters as key coefficients to perform the calculation. The calculated expected reporting return is a continuously changing value whose magnitude directly and uniquely depends on the input reporting decision variable. This value represents an idealized "gross profit" before deducting regulatory costs. In subsequent model construction, this expected reporting return will be used as a positive "revenue item" and combined with the revenue recovery amount as a negative "cost item" to form a complete settlement net profit model.

[0082] Then, the market deviation profit recovery rules are parsed, transforming the textual rules (such as "profits exceeding the exemption threshold will be recovered according to the regulatory recovery ratio") into an executable mathematical judgment and calculation logic regarding decision variables. Specifically, the market deviation profit recovery rules are deconstructed to identify the core conditional parameters used to trigger judgments (e.g., threshold parameters used to define whether the behavior is "violation" or "requires assessment") and the core calculation parameters used to quantify penalties (e.g., parameters specifying the severity of penalties or recovery ratios). Extracting and parameterizing these parameters from the rule text forms the basis for constructing the computable logic. Based on the parsed parameters, a mathematical function is constructed with the arbitrage declaration volume (or its representative variable, such as the amplification factor) as the core input variable. This function inherently embodies the logical structure of conditional judgment and quantitative calculation: its function value (i.e., the recovery amount) has different expressions or value ranges depending on whether the declaration volume meets the triggering conditions defined in the rules. In other words, the function is a piecewise function or a composite function containing conditional judgments. Its segmentation points or judgment conditions are determined by the triggering condition parameters in the rules, and the calculation formula of each segment is determined by the quantization calculation parameters in the rules.

[0083] Finally, through a specific mathematical operation, the expected declared revenue and the revenue recovery amount of the two input functions are combined into a single, new objective function. In a preferred embodiment, based on the actual market settlement process (calculating total revenue, deducting the illegal recovery portion, and obtaining net revenue), the settlement net revenue model is constructed as: expected declared revenue - revenue recovery amount. Through this model, the reduction effect of regulatory rules on revenue is endogenously and structurally embedded into the objective of the decision-making model.

[0084] In some implementations, the method for calculating the expected declared revenue without triggering revenue recovery includes: multiplying the benchmark declared electricity volume, the predicted market price difference, and the arbitrage declaration amplification factor as the expected declared revenue;

[0085] The logic for determining the profit recovery calculation corresponding to the arbitrage declaration volume includes the following method for calculating the profit recovery amount: when the deviation rate of the arbitrage declaration amplification coefficient relative to the benchmark value exceeds the exemption threshold, the profit recovery amount is calculated based on the profit recovery ratio, the predicted market price difference, and the benchmark declared electricity volume.

[0086] The purpose of calculating the expected declared profit is to precisely quantify the total theoretical economic value that arbitrage reporting can generate under ideal conditions where profit recovery is not required. In a practical application scenario, the calculation process may include:

[0087] The baseline trading parameters for arbitrage applications are obtained, and after parsing, the baseline declared electricity volume from the load forecasting system is obtained. And the predicted market spread from the spread prediction engine Simultaneously, the arbitrage reporting amplification factor (K) is determined as the decision variable, where K = actual day-ahead market reported electricity volume / The calculation expression for the expected declared revenue is as follows: Expected Declared Revenue (K-1) represents the deviation of the electricity volume or the amount of arbitrage declaration relative to the benchmark declaration volume caused by the arbitrage declaration behavior; when K=1, it means there is no arbitrage declaration; when K>1, it means the declaration volume is greater than the benchmark electricity volume (forward arbitrage); when K<1, it means the declaration volume is less than the benchmark electricity volume (reverse arbitrage). (K-1) is the change of the amplification factor relative to the benchmark state (1). The economic meaning of this product is the unit predicted market price difference ( ) and total arbitrage power The product of these factors, expressed as an expression, precisely reflects the total theoretical arbitrage profit achievable based on the price spread prediction and reporting decision, without considering any regulatory penalties. The value calculated by this expression is the expected reporting profit, a positive component (profit item) in subsequent calculations of net profit, and it varies linearly with the amplification factor K.

[0088] The purpose of calculating the aforementioned revenue recovery amount is to transform the segmented deviation revenue recovery rules of the electricity market (such as the Shanxi electricity market) into a precise and executable mathematical judgment and calculation process. In a practical application scenario, the calculation process may include: obtaining the arbitrage declaration amplification coefficient (K) of the aforementioned determined decision variables, and the market rule parameters (exemption threshold) from the rule base. and the rate of return ( )) and the aforementioned benchmark trading parameters (predicted market spread ( ), benchmark declared electricity Calculate the absolute deviation rate of the arbitrage reporting amplification factor K relative to the benchmark value 1: |K-1|. Then, compare this deviation rate with the exemption threshold. The comparison involves determining whether the conditions are met. :like If the declaration behavior does not trigger the profit recovery rule, the profit recovery amount is directly set to 0, and the calculation ends; if If the declaration behavior triggers the profit recovery rule, the next step of calculation will proceed. When the triggering condition is met, the calculation expression for the profit recovery amount is as follows, according to the Shanxi market rules:

[0089] Revenue Recovery Amount ,

[0090] in, Calculate the excess deviation rate exceeding the exemption threshold;

[0091] Convert the excess deviation rate into the excess electricity declared;

[0092] To calculate the theoretical benefit that excess electricity can generate under the predicted price difference, i.e., the benchmark value of the "excess profit" that should be recovered, the absolute value of the price difference is used. This is because it recovers the absolute value of the returns from arbitrage.

[0093] Based on the profit recovery ratio coefficient η stipulated in the rules, the above-mentioned excess theoretical profit is proportionally deducted to obtain the final amount to be recovered.

[0094] Further integrating the above judgment and calculation process into a piecewise function, the revenue recovery function is constructed as follows:

[0095] Revenue recovery function ,

[0096] in, The function encapsulates the judgment logic triggered by the above declaration behavior.

[0097] In some implementations, step S103, which involves constructing a sub-Bruker optimization model that considers the uncertainty of market price spread prediction, includes:

[0098] Based on historical price spread data, a fuzzy set containing multiple possible price spread probability distributions is constructed. The fuzzy set is used to characterize the uncertainty of the market price spread prediction.

[0099] Using the arbitrage declaration volume as the decision variable and the net profit calculated by the settlement net profit model as the optimization objective, an optimization model is constructed to find the optimal net profit under the worst probability distribution among all possible probability distributions defined by the fuzzy set.

[0100] This embodiment constructs a set (fuzzy set) containing all reasonably possible future price spread probability distributions, and optimizes under the worst-case distribution defined by this set. This goes beyond single or random prediction of price spreads and establishes a high-level decision-making framework that can resist prediction risks at the probability distribution level. This ensures that the final optimal bid volume decision can still achieve relatively optimal net return performance even in the worst but "reasonable" market scenario, thus achieving decision robustness.

[0101] This embodiment constructs a fuzzy set of probability distributions based on historical price spread data. This fuzzy set is not a single distribution, but rather a collection containing multiple (theoretically infinite) possible price spread probability distributions. The core technical function of this set is to comprehensively characterize the uncertainty inherent in market price spread prediction. It acknowledges that future price spread distributions may appear in multiple forms, rather than simply being a repetition of historical experience distributions. Various specific techniques can be used to construct this fuzzy set, but their common purpose is to define a reasonable range of distributions; any distribution belonging to this set is considered a possible future scenario.

[0102] After defining the uncertainty range (fuzzy set), the decision optimization criteria under this uncertainty are further established, and an optimization model is constructed. First, the optimization model explicitly uses the arbitrage application quantity (or its functional representation, such as the amplification coefficient K) as the decision variable. The optimization objective of the model directly adopts the net profit calculated by the settlement net profit model, ensuring that the optimization process is always guided by real and compliant settlement profits. Then, among all possible probability distributions defined by the fuzzy set, the "worst-case probability distribution" that minimizes the expected net profit is sought. Furthermore, among all feasible arbitrage application quantity decisions, an optimal decision is sought that maximizes the expected net profit under this "worst-case probability distribution."

[0103] The sub-Bruker optimization model constructed in this embodiment does not aim to maximize the average return, but rather to maximize the return under the worst-case scenario. This reflects a prudent risk management approach, making it particularly suitable for financial market decisions with low tolerance for extreme losses. Its robustness assurance does not rely on simple estimations of the range of uncertain parameters, but rather elevates to the level of combating changes in the entire family of probability distributions, resulting in stronger robustness. It perfectly encapsulates the net return model with embedded rules, enabling compliance considerations and uncertainty management to be solved within a unified mathematical framework.

[0104] In some implementations, constructing a fuzzy set containing multiple possible price spread probability distributions includes:

[0105] A baseline probability distribution is determined based on historical price spread data;

[0106] A set of probability distributions is constructed with the baseline probability distribution as the center and a preset distance tolerance as the radius.

[0107] Wherein, the probability distance between the probability distribution within the set and the baseline probability distribution is not greater than the distance tolerance.

[0108] In practical applications, the historical price spread data is collected from the electricity sales company's market data center or historical database, extracting the day-to-day / real-time market price spread sequence within a past time window for the target trading period (e.g., the same day and time). For example, price spread data from the most recent 30, 60, or 90 trading days can be extracted. As a historical sample, where n is the number of samples, This represents the actual price difference on the i-th trading day, and these data form the original empirical basis for constructing the fuzzy set. Based on the above historical price difference samples, a benchmark probability distribution is constructed as an optimal guess or reference point for the unknown true distribution. Specifically, this is the historical sample... Each sample point in the dataset is assigned the same probability mass. The resulting discrete probability distribution This is the desired baseline probability distribution (also known as the empirical distribution, i.e., the empirical distribution obtained based on historical price difference samples). Here... It is a direct depiction of the statistical regularity of price differences reflected in historical data. It does not presuppose any parameterized distribution form (such as normal distribution), thus avoiding model assumption bias and better capturing atypical characteristics such as peaks and heavy tails that may exist in electricity market price differences.

[0109] In this embodiment, the baseline probability distribution is used. The reason for centering on this is that the actual distribution in the future may not be exactly equal to the historical distribution, therefore we do not center it. Instead of considering it as the only possible distribution, we define a neighborhood around it and construct a set of probability distributions. We assume that all distributions falling within this neighborhood are reasonable possible distributions; this set constitutes a fuzzy set. The radius of this neighborhood is constructed using a preset distance tolerance, and specific construction methods can include:

[0110] To effectively measure the difference or proximity between two probability distributions (especially complex distributions like empirical distributions), this embodiment preferably uses the Wasserstein distance (also known as bulldozer distance) as the metric. This distance measures the minimum amount of work required to transform one distribution into another, is relatively insensitive to local perturbations in probability quality, and is well-suited for building robust models based on finite samples. Definition The Wasserstein distance metric is used to measure the distance between two probability distributions (or their corresponding random variables ξ and ξ). The difference between ) is defined by a non-negative parameter ε, called the distance tolerance or robust radius. This parameter is a key adjustable risk parameter in this model, used to control the conservatism of the uncertainty set, and can be flexibly adjusted according to the risk preference of the electricity sales company. ε=0: indicates that the fuzzy set only contains the baseline distribution. The model itself degenerates into a traditional stochastic programming model based on empirical distributions, lacking robustness. ε > 0: The larger ε is, the greater the difference between the future distribution and the historical empirical distribution is allowed, the wider the range of the constructed fuzzy set, and the stronger the model's robustness (corresponding to more conservative decision-making). However, it may lose revenue due to excessive conservatism. The parameter ε can be calibrated according to the electricity sales company's risk preference through methods such as cross-validation and backtesting. For example, in the embodiment, ε can be set to 20 to cover approximately ±15% of the price difference prediction error.

[0111] Based on the aforementioned baseline probability distribution Centered on a given point, with a predetermined distance tolerance ε as the radius, and using the Wasserstein distance as the metric, construct a set of probability distributions. The mathematical definition of this set can be expressed by the following formula: ,

[0112] Wherein, P on the left side of the equation represents the fuzzy set to be constructed, that is, the set of all possible future price difference probability distributions; P inside the curly braces on the right side of the equation represents any candidate probability distribution within the fuzzy set.

[0113] Represents the Wasserstein distance metric function;

[0114] This represents the empirical distribution (baseline probability distribution) obtained based on historical samples.

[0115] ε represents the robust radius (distance tolerance).

[0116] This formula defines a fuzzy set P as consisting of all probability distributions P that satisfy the following condition: the distribution P is the same as the empirical distribution. The Wasserstein distance between them does not exceed ε.

[0117] In some implementations, step S104, which involves solving the arbitrage bar optimization model and outputting the optimal arbitrage declaration quantity, includes:

[0118] By performing a dual transformation on the aforementioned sub-Bruker optimization model, it is transformed from a minimax problem containing fuzzy sets of probability distributions into a deterministic problem that can be directly solved by a standard mathematical programming solver.

[0119] The mathematical programming solver is invoked to solve the deterministic problem, thereby obtaining the optimal arbitrage application quantity.

[0120] The aforementioned arbitrage bar optimization model is mathematically expressed as a minimax problem, the objective of which is to maximize the minimum net profit under the worst possible price spread distribution. Specifically, this problem can be abstractly represented as: solving for the decision variable (arbitrage reporting amplification coefficient K) to maximize the expected net profit under the worst possible price spread distribution (belonging to the fuzzy set P). This is a two-level optimization problem, with the inner level minimizing the probability distribution P and the outer level maximizing the decision variable K.

[0121] Because the aforementioned minima problem involves fuzzy sets of probability distributions and has a complex form, it cannot be directly handled by commercial solvers. Therefore, the original Bruker minima problem is transformed into a deterministic optimization problem that is easier to solve through dual transformation. Specifically, duality theory from mathematical optimization theory is applied to dualize the inner minimization problem (finding the worst distribution on the fuzzy set P). This process introduces dual variables (e.g., λ) and auxiliary variables related to historical samples (e.g., ...). Through rigorous mathematical derivation, the constraints and minimization operations regarding the probability distribution P in the original problem are transformed into a series of linear or convex constraints on these newly introduced variables, as well as additional terms in the objective function. After dual transformation, the original problem is equivalently transformed into a deterministic optimization problem that does not explicitly include probability distribution variables or fuzzy set optimization operations. Its objective function and all constraints are deterministic mathematical expressions about the decision variable K, the dual variable, and the auxiliary variables. The transformed core objective function is:

[0122] ,

[0123] Where, λ and The variables introduced for dual transformation are ε, where ε is the robust radius and n is the number of historical samples. The objective function needs to be optimized under a series of deterministic constraints (such as dual constraints corresponding to the original net profit model and recovery rule, probabilistic dual constraints, and nonnegativity constraints of decision variables).

[0124] Once the deterministic optimization problem is determined, it can be solved efficiently using mature computational tools. In practical applications, commercial solvers such as GUROBI and COPT (a domestically developed solver) can be used. These are standard mathematical programming solvers, adept at efficiently solving deterministic optimization problems such as linear programming and quadratic programming. The complete mathematical model of the transformed deterministic optimization problem (including the objective function, all constraints, and variable boundaries) is input into the selected commercial solver. The solver's built-in advanced algorithms (such as the simplex method and interior point method) will automatically perform the calculations, outputting the global optimum in a very short time.

[0125] The optimal solution calculated by the solver corresponds to the optimal arbitrage reporting amplification factor K. The algorithm can automatically converge the optimal K factor to the edge of the profit recovery red line, achieving the optimal balance between risk and return within the rules-permitted range, and outputting compliant and robust arbitrage reporting decisions. This result can be directly transmitted to the electricity sales company's transaction reporting system to generate and submit the day-ahead market arbitrage reporting curve to the electricity trading platform, completing the automated decision-making closed loop.

[0126] The electricity spot market declaration decision method based on deviation assessment proposed in this application can be applied to the robust arbitrage declaration decision in the Shanxi electricity spot market. It aims to solve the problem of how Shanxi electricity sales companies can accurately adapt to the deviation profit recovery rules of the Shanxi electricity market when making price difference arbitrage declarations in the day-ahead market, and effectively resist the uncertainty of price difference prediction, so as to output a compliant and stable optimal declaration volume.

[0127] Assuming a specific application scenario is the "Implementation Rules for Shanxi Electricity Spot Market Transactions," obtain the market rule parameters: exemption threshold. = 6% (i.e., 0.06) and the profit recovery ratio factor = 100% (i.e., 1.0), based on the benchmark trading parameters for arbitrage declarations obtained from the electricity sales company's internal forecasting system: benchmark declared electricity volume for a single time period. The spread between the current day and the real-time market forecast (Assuming the current day price is higher than the real-time price.)

[0128] Based on historical data and risk preferences, the following settings were used: historical spread sample size n = 30 (groups), probabilistic distance metric: 1st order Wasserstein distance, robust radius ε = 20 (calibrated, this value can cover approximately ±15% of spread prediction error), and the COPT commercial mathematical programming solver was employed.

[0129] Based on the above application scenario, the parameter acquisition task of the electricity spot market declaration decision system considering deviation assessment is as follows: read the benchmark declaration electricity for the target period of the next day from the "load forecasting system". Obtain the predicted price spread for this period from the "Price Spread Prediction Engine". ; Retrieve rule parameters by calling the "Shanxi Market" template from the "Market Rules Knowledge Base": = 0.06, = 1.0.

[0130] A net profit model is constructed using the arbitrage reporting amplification factor K as the decision variable. This model accurately reflects the segmented settlement rules of the Shanxi market: first, the expected reporting profit (theoretical gross profit) is calculated, and the calculation method is as follows: expected reporting profit... Then calculate the revenue recovery amount. According to Shanxi rules, the revenue recovery function Recov(K) is:

[0131]

[0132] when hour, .

[0133] when hour, .

[0134] The expected declared profit and the profit recovery function are combined to form the net profit model NetProfit(K), whose expression is:

[0135] ,

[0136] This model clearly shows that: Within the range, net profit grows linearly with K; once K exceeds this range, the excess profit is fully recovered, and net profit reaches its upper limit at the boundary.

[0137] In response To address uncertainty, a robust decision-making framework is constructed.

[0138] First, construct a fuzzy set: based on price difference samples from the most recent 30 trading days. Constructing an experience distribution ,by With the center as the metric, the Wasserstein distance as the metric, and ε=20 as the radius, construct a fuzzy set. This set contains all possible future price spread distributions that are not far from the historical experience distribution.

[0139] Then, a sub-Bruker optimization model is established to find the optimal K, such that the expected net profit is maximized under the worst possible price difference distribution defined by the fuzzy set P. Since the original problem cannot be solved directly, duality theory is applied to transform the above minima problem into an equivalent deterministic linear programming problem. The core objective function after transformation is as follows: ,in and These are the introduced dual and auxiliary variables. Simultaneously, the transformation process generates a series of deterministic linear constraints, including constraints on the original net income model, piecewise linear constraints on the recovery rule, and probabilistic dual constraints.

[0140] Finally, the complete mathematical description of the linear programming problem (objective function, constraint matrix, variable boundaries) is input into the COPT solver. The solver runs in approximately 0.12 seconds and outputs the global optimal solution with an optimal arbitrage reporting amplification factor K=1.058. This corresponds to a reporting deviation rate of 5.8%, strictly controlled within the exemption threshold, and aligns with the day-ahead market reporting requirements.

[0141] Analysis of the rationality and accuracy of the results obtained by applying the above methods to decision-making:

[0142] Compliance verification: The optimal K value corresponds to a deviation rate of 5.8%, which is lower than the 6% exemption threshold and does not trigger profit recovery, fully complying with Shanxi market supervision rules.

[0143] Verification of revenue accuracy: The net revenue in the benchmark scenario is 870 yuan, which is only 3.3% lower than the maximum revenue threshold, reaching the upper limit of revenue allowed by the rules; backtesting of the actual settlement electricity price shows that the deviation between the pre-calculated net revenue and the actual net revenue is only 2.8%, and the calculation accuracy is significantly better than that of the traditional model.

[0144] Risk resistance verification: In the extreme scenario of a 15% error in price spread prediction, this plan still guarantees positive returns; in the extreme scenario of price spread reversal, the loss is only 38.6% of that of the traditional over-threshold reporting plan, demonstrating a significant advantage in risk resistance.

[0145] Based on the same inventive concept, embodiments of this application also provide a power spot market bidding decision system that considers deviation assessment. For example... Figure 2 As shown, this is a schematic diagram of the structure of a power spot market declaration and decision-making system 200 that considers deviation assessment, which may include:

[0146] The parameter acquisition module 201 is used to acquire the benchmark trading parameters for arbitrage applications and the market deviation profit recovery rules.

[0147] The model building module 202 is used to build a settlement net income model for arbitrage declarations. The net income model uses the deviation income recovery rule as a function affecting the net income.

[0148] Model optimization module 203 is used to construct a sub-Bruker optimization model that considers the uncertainty of market price spread prediction. The sub-Bruker optimization model is based on the net profit model and uses the arbitrage application volume as the decision variable.

[0149] The solution output module 204 is used to solve the arbitrage bar optimization model and output the optimal arbitrage declaration quantity.

[0150] Based on the same inventive concept, this application also provides an electronic device that can realize the functions of the aforementioned electricity spot market declaration decision-making method system that considers deviation assessment. (Refer to...) Figure 3 The electronic device includes:

[0151] At least one processor 301 and a memory 302 connected to at least one processor 301. In this embodiment, the specific connection medium between the processor 301 and the memory 302 is not limited. Figure 3The example shown is the connection between processor 301 and memory 302 via bus 300. Bus 300 is... Figure 3 The connections between other components are indicated by thick lines and are for illustrative purposes only, not as limiting information. Bus 300 can be divided into address bus, data bus, control bus, etc., for ease of representation. Figure 3 The term 301 is represented by a single thick line, but this does not imply that there is only one bus or one type of bus. Alternatively, the processor 301 can also be called a controller; there is no restriction on the name.

[0152] In this embodiment, memory 302 stores instructions executable by at least one processor 301. By executing the instructions stored in memory 302, at least one processor 301 can execute the electricity spot market declaration decision method considering deviation assessment discussed above. Processor 301 can implement... Figure 2 The system shown illustrates the functions of each module.

[0153] The processor 301 is the control center of the system. It can connect to various parts of the control device through various interfaces and lines. By running or executing instructions stored in memory 302 and calling data stored in memory 302, the system can perform various functions and process data, thereby monitoring the system as a whole.

[0154] In one possible design, processor 301 may include one or more processing units. Processor 301 may integrate an application processor and a modem processor, wherein the application processor mainly handles the operating system, user interface, and applications, and the modem processor mainly handles wireless communication. It is understood that the modem processor may also not be integrated into processor 301. In some embodiments, processor 301 and memory 302 may be implemented on the same chip; in some embodiments, they may also be implemented on separate chips.

[0155] Processor 301 can be a general-purpose processor, such as a central processing unit (CPU), digital signal processor, application-specific integrated circuit, field-programmable gate array or other programmable logic device, discrete gate or transistor logic device, or discrete hardware component, capable of implementing or executing the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the electricity spot market declaration decision method considering deviation assessment disclosed in the embodiments of this application can be directly reflected in the execution by a hardware processor, or executed by a combination of hardware and software modules within the processor.

[0156] Memory 302, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer-executable programs, and modules. Memory 302 may include at least one type of storage medium, such as flash memory, hard disk, multimedia card, card-type memory, random access memory (RAM), static random access memory (SRAM), programmable read-only memory (PROM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), magnetic storage, magnetic disk, optical disk, etc. Memory 302 can be any other medium capable of carrying or storing desired program code in the form of instructions or data structures that can be accessed by a computer, but is not limited thereto. In the embodiments of this application, memory 302 can also be a circuit or any other system capable of implementing storage functions for storing program instructions and / or data.

[0157] By designing and programming the processor 301, the code corresponding to the electricity spot market declaration decision method considering deviation assessment described in the foregoing embodiments can be embedded into the chip, thereby enabling the chip to execute the code during operation. Figure 1 The illustrated embodiment describes the steps of the electricity spot market declaration decision method considering deviation assessment. How to design and program the processor 301 is a technique well-known to those skilled in the art and will not be described further here.

[0158] Based on the same inventive concept, embodiments of this application also provide a storage medium storing computer instructions that, when executed on a computer, cause the computer to perform the electricity spot market declaration decision method that considers deviation assessment as described above.

[0159] In some possible implementations, various aspects of the electricity spot market declaration decision method considering deviation assessment provided in this application can also be implemented in the form of a program product, which includes program code that, when the program product is run on a system, causes the control device to perform the steps in the electricity spot market declaration decision method considering deviation assessment according to the various exemplary embodiments of this application described above.

[0160] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0161] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A system that specifies functions in one or more boxes.

[0162] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including an instruction set implemented in a process. Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0163] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0164] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for electricity spot market bidding decision-making that considers deviation assessment, characterized in that, include: Obtain the benchmark trading parameters for arbitrage applications and the rules for recovering market deviation profits; A settlement net profit model for arbitrage declarations is constructed, wherein the deviation profit recovery rule is used as a function affecting the net profit in the model. Construct a sub-Brussels bar optimization model that considers the uncertainty of market price spread prediction. The sub-Brussels bar optimization model is based on the net profit model and uses the arbitrage application volume as the decision variable. Solve the aforementioned arbitrage bar optimization model and output the optimal arbitrage declaration quantity.

2. The method as described in claim 1, characterized in that, The benchmark trading parameters for obtaining arbitrage applications and the rules for recovering market deviation profits include: Obtain the benchmark trading parameters for arbitrage applications, which include the benchmark application electricity volume and the predicted market price difference; as well as, Obtain the market deviation profit recovery rule, which is used to define the piecewise function rule of the profit recovery logic, including the exemption threshold and the profit recovery ratio coefficient.

3. The method as described in claim 1, characterized in that, The method for constructing the settlement net profit model for arbitrage declarations includes: Based on the aforementioned benchmark transaction parameters, calculate the expected declared profit without triggering profit recovery. Based on the market deviation profit recovery rule, the profit recovery calculation logic corresponding to the arbitrage declaration volume is determined, and the profit recovery amount is calculated. Based on the expected declared revenue and the revenue recovery amount, the settlement net revenue model is constructed.

4. The method as described in claim 3, characterized in that, The method for calculating the expected declared revenue without triggering revenue recovery includes: multiplying the benchmark declared electricity volume, the predicted market price difference, and the arbitrage declaration amplification factor as the expected declared revenue; The logic for determining the profit recovery calculation corresponding to the arbitrage declaration volume includes the following method for calculating the profit recovery amount: when the deviation rate of the arbitrage declaration amplification coefficient relative to the benchmark value exceeds the exemption threshold, the profit recovery amount is calculated based on the profit recovery ratio, the predicted market price difference, and the benchmark declared electricity volume.

5. The method as described in claim 1, characterized in that, The construction of the sub-Bruker optimization model considering the uncertainty of market price spread prediction includes: Based on historical price spread data, a fuzzy set containing multiple possible price spread probability distributions is constructed. The fuzzy set is used to characterize the uncertainty of the market price spread prediction. Using the arbitrage declaration volume as the decision variable and the net profit calculated by the settlement net profit model as the optimization objective, an optimization model is constructed to find the optimal net profit under the worst probability distribution among all possible probability distributions defined by the fuzzy set.

6. The method as described in claim 5, characterized in that, The construction of a fuzzy set containing multiple possible price spread probability distributions includes: A baseline probability distribution is determined based on historical price spread data; A set of probability distributions is constructed with the baseline probability distribution as the center and a preset distance tolerance as the radius. Wherein, the probability distance between the probability distribution within the set and the baseline probability distribution is not greater than the distance tolerance.

7. The method according to claim 1, characterized in that, Solving the arbitrage bar optimization model and outputting the optimal arbitrage declaration quantity includes: By performing a dual transformation on the aforementioned sub-Bruker optimization model, it is transformed from a minimax problem containing fuzzy sets of probability distributions into a deterministic problem that can be directly solved by a standard mathematical programming solver. The mathematical programming solver is invoked to solve the deterministic problem, thereby obtaining the optimal arbitrage application quantity.

8. A power spot market bidding decision system that considers deviation assessment, characterized in that, include: The parameter acquisition module is used to obtain the benchmark trading parameters for arbitrage applications and the rules for recovering market deviation profits; The model building module is used to build a settlement net profit model for arbitrage declarations. The net profit model uses the deviation profit recovery rule as a function affecting the net profit. The model optimization module is used to construct a sub-Brussels bar optimization model that considers the uncertainty of market price spread prediction. The sub-Brussels bar optimization model is based on the net profit model and uses the arbitrage application volume as the decision variable. The solution output module is used to solve the arbitrage bar optimization model and output the optimal arbitrage declaration quantity.

9. An electronic device, characterized in that, It includes a processor and a memory, wherein the memory stores program code that, when executed by the processor, causes the processor to perform any of the methods described in claims 1 to 7.

10. A computer-readable storage medium, characterized in that, Includes program code that, when the storage medium is run on an electronic device, causes the electronic device to perform any of the methods described in claims 1 to 7.

Images