Power transaction auxiliary decision-making system based on trusted data space

Abstract

The invention discloses a trusted data space-based power transaction auxiliary decision-making system, and relates to the technical field of power transaction auxiliary decision-making, and the system comprises a registration module which is used for distributing and registering block chain digital identities for power transaction participation nodes, and uploading an original power transaction data set to a trusted data space through a standardized interface; the calculation module is used for calling the original power transaction data set in the credible data space to carry out data preprocessing, executing short-term market clearing price prediction, network congestion risk assessment and green benefit index calculation in parallel, and outputting a composite decision factor set; and the solving module is used for integrating the composite decision factor set, inputting the integrated composite decision factor set into a dynamic rule engine, generating a preliminary transaction strategy set through constraint optimization solving, packaging the preliminary strategy set into a transaction strategy packet and issuing the transaction strategy packet to a corresponding user terminal. According to the method, the quantification and multi-target description capability of decision making can be improved, the transaction revenue and the scheduling feasibility are improved, the line risk is reduced, and the consumption of renewable energy sources is promoted.

Description

Technical Field

[0001] This invention relates to the field of electricity market technology, and in particular to an electricity trading auxiliary decision-making system based on a trusted data space. Background Technology

[0002] With the continued deepening of power market reform, the trading model is gradually shifting from planned to market-competitive, resulting in a significant increase in the number of trading entities and a continuous enrichment of trading products and settlement methods. Multiple business types, including ancillary services, green electricity, and spot trading, are operating in parallel. Against this backdrop, power trading data exhibits characteristics such as multi-source heterogeneity, cross-domain distribution, strong timeliness, and high sensitivity. Traditional data governance models relying on a single dispatch center or trading center are insufficient to support rapidly evolving business needs. In recent years, new digital infrastructures such as trusted data spaces, data element circulation technologies, and federated computing platforms have gradually become key means of building a trusted data sharing system. Meanwhile, technologies such as machine learning-based electricity price forecasting, network security verification, and green energy accounting have been used to assist in trading strategy formulation; however, most of these operate in isolated systems, lacking a unified data foundation and collaborative computing framework, making it difficult to form a reusable and reliable multi-factor decision-making chain.

[0003] However, existing power trading decision support technologies still face several limitations. On the one hand, cross-entity data sharing typically relies on centralized storage or anonymized exchange methods, making it difficult to ensure the separation of data ownership and usage rights. This results in insufficient data credibility, weak traceability, and limits the potential for in-depth data utilization. On the other hand, existing decision support tools often use fragmented model components, such as independent short-term electricity price forecasting models or regional power flow analysis tools. They lack the ability to perform parallel computation on key indicators such as price volatility, network congestion, and green benefits, and cannot construct a unified composite decision factor system. Furthermore, most optimization solutions rely on fixed rules or static parameter configurations, making the trading strategies generated in high-volatility scenarios lack dynamic adaptability and failing to meet the operational requirements of the spot market's "high-frequency clearing—rapid iteration" mechanism. These shortcomings prevent existing technologies from achieving effective synergy between a trusted environment, real-time data-driven approaches, and multi-objective optimization. Summary of the Invention

[0004] In view of the problems existing in a current power trading auxiliary decision-making system based on a trusted data space, this invention is proposed. Therefore, the problem this invention aims to solve is how to provide a power trading auxiliary decision-making system based on a trusted data space.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0006] In a first aspect, the present invention provides a power trading auxiliary decision-making system based on a trusted data space, which includes: a registration module, used to assign and register blockchain digital identities for power trading participating nodes, and to upload the original power trading dataset to the trusted data space through a standardized interface;

[0007] The calculation module is used to call the original power trading dataset in the trusted data space for data preprocessing, and to perform short-term market clearing price forecasting, network congestion risk assessment and green benefit index calculation in parallel. The output includes a composite decision factor set containing price forecast curves, network congestion risk coefficients and green benefit indicators.

[0008] The solution module integrates the set of composite decision factors and inputs them into the dynamic rule engine. Through constraint optimization, it generates a preliminary set of trading strategies, encapsulates the preliminary set of strategies into a trading strategy package, and publishes it to the corresponding user terminal.

[0009] As a preferred embodiment of the power trading auxiliary decision-making system based on trusted data space described in this invention, the original power trading dataset includes historical price curves, unit output plans, network topology constraint parameters, renewable energy predicted output, and historical market clearing price data.

[0010] As a preferred embodiment of the power trading auxiliary decision-making system based on trusted data space described in this invention, the short-term market clearing price forecast includes:

[0011] Supervised learning samples are constructed based on historical bids and historical clearing prices, and a prediction model based on long short-term memory networks is adopted.

[0012] The hidden state vector is generated based on the recursive structure of the input gate, forget gate, and output gate, resulting in the future predicted price sequence. The price prediction curve within the prediction time period is output, as follows:

[0013]

[0014] in: For the first The predicted market clearing price at time step; is the LSTM mapping function after training; for The market clearing price at any given moment. The number of moments.

[0015] As a preferred embodiment of the power trading auxiliary decision-making system based on trusted data space described in this invention, the network congestion risk assessment includes:

[0016] By invoking network topology constraints, node power information, line parameters, and real-time reported data, a set of AC power flow equations is constructed.

[0017] A series of pricing perturbation scenarios are generated using a heuristic strategy perturbation method, and a distributed optimal power flow solution is performed for each scenario.

[0018] Based on the line power flow distribution under various disturbance scenarios, the network congestion risk coefficient is calculated and expressed as:

[0019]

[0020] in: Network congestion risk coefficient; For the perturbation scenario, the first Trend of the line; For the first The maximum power flow allowed on this line; For the first The probability of this line exceeding the thermal stability current limit; and These are weighting coefficients; This refers to the number of lines.

[0021] As a preferred embodiment of the power trading auxiliary decision-making system based on trusted data space described in this invention, the step of performing distributed optimal power flow solution includes:

[0022] The power grid is divided into multiple computing nodes according to the region. Each node maintains its own local network model and is equipped with a local solver to perform local optimal power flow calculations.

[0023] When performing distributed solution, each regional node performs a local optimal power flow solution with local voltage phase angle, node power and line power flow as variables, and sends the node voltage and switching power of the boundary node to the coordinating node;

[0024] The coordinating node updates the global Lagrange multipliers and generates new boundary consistency quantities based on the constraint consistency conditions of the ADMM distributed optimization algorithm.

[0025] Each regional node updates the penalty term in its local objective function and performs local optimal power flow iteration again until the physical quantities of the boundary nodes satisfy the convergence criterion, generating power flow solutions and line power flow distributions.

[0026] As a preferred embodiment of the power trading auxiliary decision-making system based on trusted data space described in this invention, the calculation of the green benefit index includes:

[0027] Based on the renewable energy output forecast sequence, load forecast, and trading plan, the green electricity consumption ratio is calculated and expressed as:

[0028]

[0029] in: The proportion of green electricity consumption; Forecast available electricity for renewable energy sources; This refers to the reduction in electricity consumption due to forecasting errors and scheduling corrections. This represents the total electricity consumption during the same period;

[0030] Carbon emission intensity is calculated and expressed as:

[0031]

[0032] in, Carbon emission intensity; and For the first Emission factors of fossil fuel-like units; For the first Predicted power generation of fossil fuel-like units; The number of fossil fuel generating units;

[0033] The green energy consumption ratio and carbon emission intensity are combined according to predetermined weights to form a green benefit indicator, expressed as follows:

[0034]

[0035] in, As a green benefit indicator; and Predetermined weights.

[0036] As a preferred embodiment of the power trading auxiliary decision-making system based on trusted data space described in this invention, the generation of the preliminary trading strategy set includes:

[0037] The input composite decision factors are subjected to min-max normalization.

[0038] Pre-set market rules, corporate constraint strategies, and real-time policy documents are constructed as a mathematical constraint set. Under the constraints of this constraint set, an optimization problem with the goal of maximizing expected profit is solved.

[0039] The output includes a preliminary strategy set containing suggested price range, declared electricity volume, normalized expected profit, and risk assessment level.

[0040] As a preferred embodiment of the power trading auxiliary decision-making system based on trusted data space described in this invention, the step of encapsulating the preliminary strategy set into a trading strategy package and publishing it to the corresponding user terminal includes:

[0041] Based on the strategy generation logic, all intermediate results used to generate the initial strategy set are traced and extracted from the trusted data space, including price prediction curves, intermediate states of each trend solution, and key factors for carbon emission calculation.

[0042] The intermediate results are associated with the evidence stored on the blockchain and assembled into a data certificate chain according to the time sequence and logical dependencies.

[0043] The initial strategy set is encapsulated with the data credential chain to form a dynamic compliant transaction strategy package, which is then pushed to the user terminals of the corresponding nodes through a preset release channel, triggering a terminal interface update.

[0044] The beneficial effects of this invention are as follows: By assigning blockchain digital identities to transaction entities, this invention ensures that data sources are traceable, ownership is verifiable, and access is controllable; within a trusted data space, it completes standardized preprocessing and generates short-term clearing price predictions, distributed power flow congestion risk assessments, and green benefit indicators in parallel, converging them into composite decision factors with time-series and constraint information, thereby improving the quantitative and multi-objective characterization capabilities of decision-making, enhancing transaction profitability and scheduling feasibility, reducing line risks, and promoting the consumption of renewable energy. Attached Figure Description

[0045] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0046] Figure 1 This is a structural diagram of a power trading auxiliary decision-making system based on a trusted data space. Detailed Implementation

[0047] To make the above-mentioned objects, features, and advantages of the present invention more readily understood, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0048] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0049] Secondly, the term "an embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places throughout this specification does not necessarily refer to the same embodiment, nor is it a single embodiment or an embodiment selectively excluded from other embodiments.

[0050] Reference Figure 1 This is the first embodiment of the present invention, which provides a power trading auxiliary decision-making system based on a trusted data space, comprising:

[0051] S1: Assign and register blockchain digital identities for participating nodes in electricity trading, and upload the original electricity trading dataset to a trusted data space through a standardized interface.

[0052] S2: In the trusted data space, call the original power trading dataset for data preprocessing, and execute short-term market clearing price forecasting, network congestion risk assessment and green benefit index calculation in parallel. The output is a composite decision factor set containing price forecast curves, network congestion risk coefficients and green benefit indicators.

[0053] S3: Integrate the set of composite decision factors and input them into the dynamic rule engine. Generate a preliminary set of trading strategies through constraint optimization, encapsulate the preliminary set of strategies into a trading strategy package, and publish it to the corresponding user terminal.

[0054] Specifically, a unique and verifiable blockchain digital identity is generated for each participating node in the power trading process (including power generation companies, power sales companies, grid operators, and third-party institutions), and the identity is traceable and bound through on-chain registration.

[0055] Each node performs integrity verification on the original power trading dataset locally based on its digital identity, and then submits the data to the trusted data space through a standardized data upload interface.

[0056] The uploaded data includes historical price curves, unit output plans, network topology constraint parameters, renewable energy forecast output, and historical market clearing prices. Within a trusted data space, all newly uploaded data undergoes format mapping, field standardization, and data tagging to ensure that original transaction data from different sources are stored in a unified structure, forming the original electricity transaction dataset.

[0057] Retrieve the original power trading dataset from the trusted data space, including historical bid sequences, market clearing prices, unit output and declaration records, network topology constraint parameters, renewable energy forecast output and related trading plans, etc.

[0058] All data are aligned according to a unified set of reference times, so that time series from different sources and with different sampling frequencies are mapped to the same time axis.

[0059] After alignment, preprocessing is performed: missing values ​​are imputed using trend-preserving methods based on numerical attributes, and local curve fitting is used to ensure continuity for high-frequency fluctuating data;

[0060] For heterogeneous field differences caused by cross-device and cross-regional issues, a unified mapping is performed through a field standardization dictionary; outliers are corrected using interval constraints and dynamic upper and lower bound judgments, with the dynamic upper and lower bounds calculated from the autoregressive residual sequence within the historical window.

[0061] Subsequently, three types of multi-dimensional fusion analysis tasks were launched in parallel. Each type of task constructed an independent mathematical model and output composite decision factors during the fusion phase.

[0062] The short-term market clearing price prediction based on LSTM constructs historical quotes and historical clearing prices as supervised learning samples. For time series, the input sequence is constructed according to a predetermined sliding window length, and a multi-step prediction method is used to output the price prediction sequence for multiple future time periods.

[0063] The LSTM unit generates a hidden state vector based on a recursive structure of input gate, forget gate, and output gate, ultimately obtaining the future price prediction sequence. The final output is the price prediction curve for the prediction period, represented as:

[0064]

[0065] in: For the first The predicted market clearing price at time step; is the LSTM mapping function after training; for The market clearing price at any given moment. The number of moments.

[0066] Based on distributed optimal power flow, the network congestion risk assessment system utilizes network topology constraints, node power information, line parameters, and real-time reported data to construct a set of power flow equations. A heuristic strategy perturbation method is employed to generate a series of bidding perturbation scenarios, each corresponding to a set of power flow inputs.

[0067] For each disturbance scenario, the distributed optimal power flow module performs local solutions on the node side and completes consistency verification at the coordinating node.

[0068] The distributed optimal power flow solution employs the ADMM distributed optimization algorithm. The system divides the power grid into multiple computing nodes by region. Each node maintains its own local network model and is equipped with a local solver to perform local optimal power flow calculations. The local solver uses an AC power flow solver based on Newton-Raphson iteration to handle power balance, line constraints, and voltage constraints within its local region.

[0069] When performing distributed solution, each regional node first performs a local optimal power flow solution with local voltage phase angle, node power and line power flow as variables, and sends the node voltage and switching power of the boundary node to the coordinating node.

[0070] The coordinating node updates the global Lagrange multipliers and generates new boundary consistency quantities based on the constraint consistency conditions of the ADMM distributed optimization algorithm. Each regional node receives the consistency quantity, updates the penalty term in its local objective function, and performs local optimal power flow iteration again. This continues until the physical quantities of the boundary nodes satisfy the convergence criterion, ultimately generating the power flow solution and the corresponding line power flow distribution.

[0071] In all power flow solutions, the power flow distribution of each line under various disturbance scenarios is statistically analyzed to form a complete power flow set.

[0072] To measure the risk of line congestion, a comprehensive risk coefficient based on the probability of exceeding limits and the load rate is introduced, and the final output is a comprehensive network congestion risk factor, expressed as:

[0073]

[0074] in: Network congestion risk coefficient; For the perturbation scenario, the first Trend of the line; For the first The maximum power flow allowed on this line; For the first The probability of this line exceeding the thermal stability current limit; and These are weighting coefficients; This refers to the number of lines.

[0075] The calculation of green benefit indicators based on renewable energy output and trading plans involves calculating the available renewable energy consumption and actual planned consumption at future times based on the renewable energy output forecast sequence, load forecast and trading plan, and calculating carbon emission intensity using a weighted summation with unit emission factors.

[0076] The green benefit index is constructed by combining the green electricity consumption ratio and carbon emission intensity under fixed transformations, and is expressed as follows:

[0077]

[0078]

[0079]

[0080] in: The proportion of green electricity consumption; Forecast available electricity for renewable energy sources; This refers to the reduction in electricity consumption due to forecasting errors and scheduling corrections. This represents the total electricity consumption during the same period; Carbon emission intensity; and For the first Emission factors of fossil fuel-like units; For the first Predicted power generation of fossil fuel-like units; The number of fossil fuel generating units; As a green benefit indicator; and Predetermined weights.

[0081] After the analysis task is completed, the price prediction curve, network congestion risk coefficient, and green benefit indicators are obtained.

[0082] By integrating price forecast curves, network congestion risk coefficients, and green benefit indicators, a composite decision factor set is formed.

[0083] The system reads composite decision factors from the trusted data space, including price prediction curves, network congestion risk coefficients, and green benefit indicators, and simultaneously reads the corresponding on-chain credentials.

[0084] The rule engine normalizes the input factors using the min-max normalization method of historical windows.

[0085] The rules engine constructs a set of mathematical constraints based on pre-defined market rules, corporate constraint strategies, and real-time policy documents, and solves the constrained optimization problem.

[0086] The goal is to maximize expected profits while meeting constraints such as price ceilings and floor limits, priority consumption of clean energy, and corporate risk appetite.

[0087] The problem is formulated as a constrained linear / nonlinear optimization, and a solver is used to solve each policy case.

[0088] The dynamic rule engine first constructs a rule set in the condition-action-priority format, which expresses the running indicators, triggering conditions and adjustment actions in a structured way.

[0089] During system operation, forward chain reasoning is used to match state variables with rule conditions in real time, filter and resolve conflicting rules, and generate candidate control actions.

[0090] The actions are then converted into constraint parameters and initial optimization values, which are input into the distributed optimization solver. The solver then completes the final optimization solution under the premise of satisfying operating constraints such as voltage, load factor, and imbalance.

[0091] When the solver returns non-convergence or constraint conflict information, the rule engine dynamically adjusts the rule content.

[0092] The preliminary strategy set obtained from the solution includes: the suggested price range, the declared electricity volume, the corresponding normalized expected profit, and the risk assessment level for each forecast time.

[0093] The initial strategy set is packaged into a dynamic compliant trading strategy package and pushed to the corresponding user terminals through the publishing channel.

[0094] In summary, this invention assigns blockchain digital identities to transaction entities, ensuring traceability of data sources, verifiability of ownership, and controllable access. It completes standardized preprocessing within a trusted data space and generates short-term clearing price predictions, distributed power flow congestion risk assessments, and green benefit indicators in parallel. These are aggregated into composite decision factors with temporal and constraint information, enhancing the quantification and multi-objective characterization of decisions, improving transaction profitability and scheduling feasibility, reducing line risks, and promoting the consumption of renewable energy.

[0095] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A trusted data space based power transaction aided decision system, characterized in that: The application relates to a power transaction strategy optimization system based on blockchain, comprising the following modules: a registration module for assigning and registering a blockchain digital identity for a power transaction participant node, and uploading original power transaction data sets to a trusted data space through a standardized interface; a calculation module for calling the original power transaction data sets in the trusted data space for data preprocessing, and performing short-term market clearing price prediction, network congestion risk assessment and green benefit index calculation in parallel, and outputting a composite decision factor set containing a price prediction curve, a network congestion risk coefficient and a green benefit index; a solving module for integrating the composite decision factor set and inputting the same into a dynamic rule engine to generate a preliminary transaction strategy set through constraint optimization solving, and encapsulating the preliminary strategy set into a transaction strategy package and publishing the same to a corresponding user terminal.

2. The trusted data space based power transaction aided decision system as claimed in claim 1, wherein: The original power transaction data set comprises a historical bidding curve, a unit output plan, network topology constraint parameters, renewable energy prediction output and market clearing price historical data.

3. The trusted data space based power transaction aided decision system as claimed in claim 1, wherein: The short-term market clearing price prediction comprises the following steps: constructing a supervised learning sample based on historical bidding and historical clearing prices, and adopting a prediction model based on a long short-term memory network; generating a hidden state vector based on a recursive structure of an input gate, a forgetting gate and an output gate to obtain a future predicted price sequence, and outputting a price prediction curve in a prediction time, which is expressed as: ； wherein: is the predicted market clearing price at time is the trained LSTM mapping function is the market clearing price at time is the time at which the market clearing price is predicted, is the number of times.

4. The trusted data space based power transaction aided decision system as claimed in claim 1, wherein: The network congestion risk assessment comprises the following steps: calling network topology constraints, node power information, line parameters and real-time declaration data to construct an alternating current power flow equation set; adopting a heuristic strategy perturbation method to generate a series of bidding perturbation scenes, and performing distributed optimal power flow solving for each scene; calculating a network congestion risk coefficient based on line power flow distribution under each perturbation scene, which is expressed as: ； wherein: is the network congestion risk factor; is the first line power flow under the disturbance scenario; is the first line allowed maximum power flow; is the first line probability of exceeding the thermal steady load flow limit; and is the weighting factor; is the number of lines.

5. A trusted data space based power trading aided decision system as claimed in claim 4, wherein: The distributed optimal power flow solving comprises the following steps: dividing the power grid into multiple calculation nodes according to regions, and maintaining a local network model for each node and arranging a local solver to perform local optimal power flow calculation; when performing distributed solving, each regional node performs a local optimal power flow solving once with local voltage phase angle, node power and line power flow as variables, and sends node voltage and exchange power of a boundary node to a coordination node; the coordination node updates a global Lagrange multiplier and generates a new boundary consistency quantity according to a constraint consistency condition of an ADMM distributed optimization algorithm; each regional node updates a penalty term in a local objective function, and performs local optimal power flow iteration again until physical quantities of the boundary node meet a convergence criterion, and generates a power flow solution and line power flow distribution.

6. The trusted data space based power transaction aided decision system as claimed in claim 1, wherein: The green benefit index calculation comprises the following steps: calculating a green power consumption proportion according to a renewable energy prediction output sequence, load prediction and transaction plan, which is expressed as: ； Wherein: is the green power consumption proportion; is the renewable energy predicted available power; is the deducted power caused by prediction deviation and dispatch correction; is the total power consumption amount in the same period; calculating carbon emission intensity, which is expressed as: ； wherein, is the carbon intensity; and is the emission factor of the fossil fuel unit of class is the predicted electricity production of the fossil fuel unit of class is the number of fossil fuel units;​​ combining the green power consumption proportion and the carbon emission intensity according to a predetermined weight to form a green benefit index, which is expressed as: ； wherein, is a green benefit indicator; and is a predetermined weight.

7. The trusted data space based power transaction aided decision system as claimed in claim 1, wherein: The generation of the preliminary transaction strategy set comprises the following steps: performing minimum-maximum normalization processing on the input composite decision factor; constructing a mathematical constraint set from preset market rule clauses, enterprise constraint strategies and real-time policy files, and solving an optimization problem with the maximum expected profit as a target under the limitation of the constraint set. The preliminary strategy set includes a recommended bidding range, declared power, normalized expected profit, and risk assessment level.

8. The trusted data space based power transaction aided decision system as claimed in claim 1, wherein: The encapsulation of the preliminary strategy set into a transaction strategy package and the release to the corresponding user terminal include: According to the strategy generation logic, all intermediate results used to generate the preliminary strategy set are traced back and extracted from the trusted data space, including price prediction curves, intermediate states of each flow solution, and key factors of carbon emission calculation; The intermediate results are associated with the notarization credentials on the blockchain, and are assembled into a data credential chain in chronological order and logical dependency; The preliminary strategy set is encapsulated with the data credential chain to form a dynamic compliance transaction strategy package, which is pushed to the user terminal of the corresponding node through a preset release channel, and the terminal interface is updated at the same time.

Images