Power transaction decision-making method and system

Abstract

The invention discloses a power transaction decision-making method and system, and the method comprises the steps: dividing market information issued by a power transaction center through employing a similarity calculation method, obtaining transaction attribute data, and carrying out the processing of the transaction attribute data through employing a classification and regression tree model, and obtaining a feature set; processing the feature set by using a Shapley value to obtain a plurality of target features; and inputting the plurality of target features into a constructed statistical arbitrage model to obtain a transaction decision, the construction process of the statistical arbitrage model comprising performing correlation analysis on a transaction target, a market transaction demand and a power grid technology by using a multi-factor model, and adjusting parameters of the statistical arbitrage model to realize optimization of the transaction decision. According to the power transaction decision-making method provided by the embodiment of the invention, the defect that transaction decision-making depends on artificial experience at present is overcome, the power transaction decision-making speed is improved, and the intelligent process of power transaction decision-making is promoted.

Description

Technical Field

[0001] This invention relates to the field of power technology, and in particular to a power trading decision-making method and system. Background Technology

[0002] With increasing emphasis on environmental protection and sustainable development, renewable energy sources such as solar and wind power are gradually accounting for a larger share of electricity supply. These energy sources are characterized by intermittency and instability, requiring more intelligent electricity trading decision-making methods to achieve their efficient utilization and grid connection.

[0003] In existing technologies, power trading decision-making methods rely on human experience, but human decision-making requires a lot of time and effort, especially when dealing with large amounts of trading data and complex market conditions. The decision-making speed cannot meet market demands, which can lead to missed trading opportunities.

[0004] Therefore, how to promote the intelligentization of power trading decision-making has become a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] This invention provides a power trading decision-making method and system to solve the current technical problem of relying mainly on human experience, so as to achieve the effect of intelligent trading decision-making.

[0006] To address the aforementioned technical problems, embodiments of the present invention provide a power trading decision-making method, comprising:

[0007] The market information released by the power trading center is divided into transaction attribute data by using a similarity calculation method. The transaction attribute data is then processed using a classification and regression tree model to obtain a feature set.

[0008] The feature set is processed using the Shapley value to obtain several target features;

[0009] Several target features are input into a constructed statistical arbitrage model to obtain trading decisions. The construction process of the statistical arbitrage model includes using a multi-factor model to perform correlation analysis on trading objectives, market trading demand and power grid technology, and adjusting the parameters of the statistical arbitrage model to optimize the trading decisions.

[0010] As one preferred embodiment, the step of processing the feature set using the Shapley value to obtain several target features includes:

[0011] Traverse all feature subsets in the feature set, calculate the marginal contribution value of each feature subset, and process the marginal contribution value corresponding to each feature subset to obtain the Shapley value of each feature subset;

[0012] Sort all feature subsets in the feature set according to the Shapley value, and obtain several features with Shapley values ​​greater than a first threshold as the target features.

[0013] As one preferred embodiment, the method of using a multi-factor model to perform correlation analysis on trading objectives, market trading demand, and power grid technology, and adjusting the parameters of the statistical arbitrage model to optimize the trading decision, includes:

[0014] Principal component analysis was used to process the transaction objective, the market transaction demand, and the power grid technology to obtain several factors;

[0015] The correlation coefficient between the factors is calculated using the Pearson correlation coefficient, and the parameters of the statistical arbitrage model are adjusted based on the correlation coefficient to optimize the trading decision.

[0016] As a preferred embodiment, after inputting several of the target features into the constructed statistical arbitrage model to obtain a trading decision, the electricity trading decision method further includes:

[0017] The trading decision is validated using cross-validation to assess its accuracy.

[0018] As a preferred embodiment, before using a similarity calculation method to segment the market information released by the power trading center to obtain trading attribute data, the power trading decision-making method further includes:

[0019] The market information is preprocessed, and the preprocessing steps include data cleaning, data integration, data transformation, and data reduction.

[0020] Another embodiment of the present invention provides a power trading decision system, comprising:

[0021] The processing module is used to divide the market information released by the power trading center using a similarity calculation method to obtain transaction attribute data, and to process the transaction attribute data using a classification and regression tree model to obtain a feature set.

[0022] The calculation module is used to process the feature set using the Shapley value to obtain several target features;

[0023] The acquisition module is used to input several target features into a pre-constructed statistical arbitrage model to obtain trading decisions. The construction process of the statistical arbitrage model includes using a multi-factor model to perform correlation analysis on trading objectives, market trading demand and power grid technology, and adjusting the parameters of the statistical arbitrage model to optimize the trading decisions.

[0024] As one preferred embodiment, the step of processing the feature set using the Shapley value to obtain several target features includes:

[0025] Traverse all feature subsets in the feature set, calculate the marginal contribution value of each feature subset, and process the marginal contribution value corresponding to each feature subset to obtain the Shapley value of each feature subset;

[0026] Sort all feature subsets in the feature set according to the Shapley value, and obtain several features with Shapley values ​​greater than a first threshold as the target features.

[0027] As one preferred embodiment, the method of using a multi-factor model to perform correlation analysis on trading objectives, market trading demand, and power grid technology, and adjusting the parameters of the statistical arbitrage model to optimize the trading decision, includes:

[0028] Principal component analysis was used to process the transaction objective, the market transaction demand, and the power grid technology to obtain several factors;

[0029] The correlation coefficient between the factors is calculated using the Pearson correlation coefficient, and the parameters of the statistical arbitrage model are adjusted based on the correlation coefficient to optimize the trading decision.

[0030] As a preferred embodiment, after inputting several of the target features into the constructed statistical arbitrage model to obtain a trading decision, the power trading decision system further includes:

[0031] The verification module is used to verify the trading decision using cross-validation to assess the accuracy of the trading decision.

[0032] As one preferred embodiment, before using a similarity calculation method to segment the market information released by the power trading center to obtain trading attribute data, the power trading decision-making system further includes:

[0033] The preprocessing module is used to preprocess the market information. The preprocessing steps include data cleaning, data integration, data transformation, and data reduction.

[0034] Compared with the prior art, the beneficial effects of the embodiments of the present invention are at least one of the following:

[0035] (1) This embodiment takes into account the market information, trading objectives, market trading demands and current power grid technology released by the power trading center. By constructing a series of methods and processes to process the data, it can overcome the shortcomings of relying on human experience to make trading decisions, greatly improve the speed of power trading decisions, and promote the intelligent process of power trading decisions.

[0036] (2) This embodiment integrates multiple methods such as similarity calculation, classification and regression tree model, Shapley value analysis and multi-factor model, which can realize comprehensive analysis and prediction of the power trading market and provide intelligent decision support for power trading. Attached Figure Description

[0037] Figure 1 This is a schematic diagram of the power trading decision-making method in one embodiment of the present invention;

[0038] Figure 2 This is a schematic diagram of the power trading decision system structure in one embodiment of the present invention.

[0039] Figure label:

[0040] Among them, 11 is the processing module; 12 is the calculation module; and 13 is the acquisition module. Detailed Implementation

[0041] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0042] In the description of this application, the terms "first," "second," "third," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined with "first," "second," "third," etc., may explicitly or implicitly include one or more of that feature. In the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0043] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "installation," "connection," and "joint" should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections or electrical connections. The terms "vertical," "horizontal," "left," "right," "upper," "lower," and similar expressions used herein are for illustrative purposes only and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as limiting the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0044] In the description of this application, it should be noted that, unless otherwise defined, all technical and scientific terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in this specification is for the purpose of describing specific embodiments only and is not intended to limit the invention. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0045] One embodiment of the present invention provides a power trading decision-making method; for details, please refer to [link to relevant documentation]. Figure 1 , Figure 1 The diagram shown is a schematic flowchart of a power trading decision-making method according to one embodiment of the present invention. The method includes:

[0046] S1: The market information released by the power trading center is divided into transaction attribute data by using a similarity calculation method, and the transaction attribute data is processed by a classification and regression tree model to obtain a feature set;

[0047] S2: Process the feature set using the Shapley value to obtain several target features;

[0048] S3: Input several target features into the constructed statistical arbitrage model to obtain a trading decision. The construction process of the statistical arbitrage model includes using a multi-factor model to perform correlation analysis on the trading target, market trading demand and power grid technology, and adjusting the parameters of the statistical arbitrage model to optimize the trading decision.

[0049] Before performing step S1, the market information released by the power trading center is preprocessed before the similarity calculation method is used to classify it. The preprocessing steps include data cleaning, data integration, data transformation and data reduction.

[0050] Data cleaning is the first step in data preprocessing, aiming to detect and correct errors, incompleteness, or inaccuracies in the dataset. For market information in power trading centers, data cleaning includes, depending on the business context or statistical analysis, selecting options such as deleting samples containing missing values, filling in missing values ​​using the mean, median, or other statistics, or estimating missing values ​​using interpolation methods (such as linear interpolation or model-based interpolation).

[0051] Data integration involves merging data from different data sources into a consistent data store to support subsequent analysis. This is achieved using database join operations (such as SQL's JOIN operation) to merge multiple tables into one, combining data from different data sources based on common unique identifiers (such as transaction IDs, timestamps, etc.).

[0052] Data transformation is to ensure that the dataset meets the requirements of subsequent analysis, while data reduction reduces the complexity of the dataset while maintaining its integrity and representativeness. This helps to reduce the storage and computational costs of the dataset, while preserving the important information in the dataset.

[0053] In step S1, the main purpose of the similarity calculation method is to identify transaction data with similar characteristics from the large amount of market information released by the power trading center. The market information released by the power trading center is segmented using a similarity calculation method to obtain transaction attribute data. The similarity calculation method can be cosine similarity, Euclidean distance, or Manhattan distance; the appropriate similarity calculation method is determined based on actual needs.

[0054] Based on the similarity calculation results, transaction attribute data is obtained.

[0055] The transaction attribute data, after being segmented by similarity, is input into the classification and regression tree model. The classification and regression tree model recursively selects the optimal feature for splitting to construct a decision tree.

[0056] In each split, the model evaluates the impact of different features on the classification or regression task and selects the feature that best distinguishes different categories or minimizes the regression error for splitting.

[0057] As the decision tree is built, the model gradually determines which features have a significant impact on trading decisions, and these features constitute the feature set.

[0058] The specific process of processing the feature set using the Shapley value to obtain several target features in step S2 includes:

[0059] Step 1: Traverse all feature subsets in the feature set, calculate the marginal contribution value of each feature subset, and process the marginal contribution value of each feature subset to obtain the Shapley value of each feature subset;

[0060] Step 2: Sort all feature subsets in the feature set according to the Shapley value, and obtain several features with Shapley values ​​greater than the first threshold as target features.

[0061] The first threshold and the number of features are determined by the objectives of this electricity transaction.

[0062] Shapley value is a concept in cooperative game theory. It can be used to measure the contribution or weight of each participant in the cooperation. When processing feature sets, Shapley value can be used to obtain several target features that have an important impact on the outcome.

[0063] After obtaining several target features, these features are input into the constructed statistical arbitrage model to obtain trading decisions. The construction process of the statistical arbitrage model includes using a multi-factor model to conduct correlation analysis on trading objectives, market trading demand, and power grid technology, and adjusting the parameters of the statistical arbitrage model to optimize trading decisions.

[0064] Among them, the multi-factor model is used to identify and quantify various factors affecting trading objectives. After conducting correlation analysis on trading objectives, market trading demand, and power grid technology using the multi-factor model, the parameters of the statistical arbitrage model are adjusted to optimize trading decisions, including:

[0065] Principal component analysis was used to process the trading objectives, market trading demand, and power grid technology to obtain several factors;

[0066] The correlation coefficient between factors is calculated using the Pearson correlation coefficient, and the parameters of the statistical arbitrage model are adjusted based on the correlation coefficient to optimize trading decisions.

[0067] The factors obtained by processing transaction objectives, market transaction demands, and power grid technology through principal component analysis are actually projections of the original data into a low-dimensional space. These factors represent the main directions of change in the original data.

[0068] Factors represent both the correlation or trend between trading objectives and market trading demand, and also reflect certain characteristics of power grid technology or its impact on market trading demand.

[0069] The Pearson correlation coefficient is used to quantify the degree of linear correlation between two variables. In statistical arbitrage models, it can help us identify which factors have significant correlations that may affect the accuracy of trading decisions.

[0070] Based on the Pearson correlation coefficient, identify which factors are highly correlated. Select one or more representative factors from these highly correlated factors to construct a statistical arbitrage model. The selection of representative factors can be based on factors such as importance, stability, and interpretability. Adjust the parameters of the statistical arbitrage model, such as factor weights, trading thresholds, and stop-loss points, according to the selected representative factors. These parameter adjustments can be based on backtesting results using historical data and model performance evaluation.

[0071] Based on the results of the correlation analysis, the parameters of the statistical arbitrage model can be adjusted. For example, factors that are highly correlated with the trading objective can be given higher weights, while factors that are less correlated with market trading demand and power grid technology can be appropriately reduced in weight.

[0072] By adjusting parameters, the performance of the statistical arbitrage model can be optimized to more accurately reflect the actual situation of trading objectives, market trading demands, and power grid technology. Historical data can also be used for backtesting to evaluate the model's performance under different parameters in order to find the optimal parameter combination.

[0073] After inputting several target features into the constructed statistical arbitrage model to obtain trading decisions, cross-validation is used to verify the trading decisions in order to evaluate their accuracy.

[0074] One embodiment of the present invention provides a power trading decision system; for details, please refer to [link to documentation]. Figure 2 , Figure 2 The diagram shown is a schematic representation of the power trading decision system structure in one embodiment of the present invention, including:

[0075] Processing module 11 is used to divide the market information released by the power trading center using a similarity calculation method to obtain transaction attribute data, and to process the transaction attribute data using a classification and regression tree model to obtain a feature set;

[0076] Calculation module 12 is used to process the feature set using Shapley values ​​to obtain several target features;

[0077] The acquisition module 13 is used to input several target features into the constructed statistical arbitrage model to obtain trading decisions. The construction process of the statistical arbitrage model includes using a multi-factor model to perform correlation analysis on trading targets, market trading demand and power grid technology, and adjusting the parameters of the statistical arbitrage model to optimize the trading decisions.

[0078] Compared with the prior art, the beneficial effects of the embodiments of the present invention are at least one of the following:

[0079] (1) This embodiment takes into account the market information, trading objectives, market trading demands and current power grid technology released by the power trading center. By constructing a series of methods and processes to process the data, it can overcome the shortcomings of relying on human experience to make trading decisions, greatly improve the speed of power trading decisions, and promote the intelligent process of power trading decisions.

[0080] (2) This embodiment integrates multiple methods such as similarity calculation, classification and regression tree model, Shapley value analysis and multi-factor model, which can realize comprehensive analysis and prediction of the power trading market and provide intelligent decision support for power trading.

[0081] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A power trading decision-making method, characterized in that, include: The market information released by the power trading center is divided into transaction attribute data by using a similarity calculation method. The transaction attribute data is then processed using a classification and regression tree model to obtain a feature set. The feature set is processed using the Shapley value to obtain several target features; Several target features are input into a constructed statistical arbitrage model to obtain trading decisions. The construction process of the statistical arbitrage model includes using a multi-factor model to perform correlation analysis on trading objectives, market trading demand and power grid technology, and adjusting the parameters of the statistical arbitrage model to optimize the trading decisions.

2. The power trading decision-making method as described in claim 1, characterized in that, The process of processing the feature set using the Shapley value to obtain several target features includes: Traverse all feature subsets in the feature set, calculate the marginal contribution value of each feature subset, and process the marginal contribution value corresponding to each feature subset to obtain the Shapley value of each feature subset; Sort all feature subsets in the feature set according to the Shapley value, and obtain several features with Shapley values ​​greater than a first threshold as the target features.

3. The power trading decision-making method as described in claim 1, characterized in that, The method of using a multi-factor model to perform correlation analysis on trading objectives, market trading demand, and power grid technology, and adjusting the parameters of the statistical arbitrage model to optimize the trading decision, includes: Principal component analysis was used to process the transaction objective, the market transaction demand, and the power grid technology to obtain several factors; The correlation coefficient between the factors is calculated using the Pearson correlation coefficient, and the parameters of the statistical arbitrage model are adjusted based on the correlation coefficient to optimize the trading decision.

4. The power trading decision-making method as described in claim 1, characterized in that, After inputting several of the target features into the constructed statistical arbitrage model to obtain a trading decision, the electricity trading decision method further includes: The trading decision is validated using cross-validation to assess its accuracy.

5. The power trading decision-making method as described in claim 1, characterized in that, Before using a similarity calculation method to segment the market information released by the power trading center to obtain trading attribute data, the power trading decision-making method further includes: The market information is preprocessed, and the preprocessing steps include data cleaning, data integration, data transformation, and data reduction.

6. A power trading decision-making system, characterized in that, include: The processing module is used to divide the market information released by the power trading center using a similarity calculation method to obtain transaction attribute data, and to process the transaction attribute data using a classification and regression tree model to obtain a feature set. The calculation module is used to process the feature set using the Shapley value to obtain several target features; The acquisition module is used to input several target features into a pre-constructed statistical arbitrage model to obtain trading decisions. The construction process of the statistical arbitrage model includes using a multi-factor model to perform correlation analysis on trading objectives, market trading demand and power grid technology, and adjusting the parameters of the statistical arbitrage model to optimize the trading decisions.

7. The power trading decision system as described in claim 6, characterized in that, The process of processing the feature set using the Shapley value to obtain several target features includes: Traverse all feature subsets in the feature set, calculate the marginal contribution value of each feature subset, and process the marginal contribution value corresponding to each feature subset to obtain the Shapley value of each feature subset; Sort all feature subsets in the feature set according to the Shapley value, and obtain several features with Shapley values ​​greater than a first threshold as the target features.

8. The power trading decision system as described in claim 6, characterized in that, The method of using a multi-factor model to perform correlation analysis on trading objectives, market trading demand, and power grid technology, and adjusting the parameters of the statistical arbitrage model to optimize the trading decision, includes: Principal component analysis was used to process the transaction objective, the market transaction demand, and the power grid technology to obtain several factors; The correlation coefficient between the factors is calculated using the Pearson correlation coefficient, and the parameters of the statistical arbitrage model are adjusted based on the correlation coefficient to optimize the trading decision.

9. The power trading decision system as described in claim 6, characterized in that, After inputting several of the target features into the constructed statistical arbitrage model to obtain a trading decision, the power trading decision system further includes: The verification module is used to verify the trading decision using cross-validation to assess the accuracy of the trading decision.

10. The power trading decision system as described in claim 6, characterized in that, Before using a similarity calculation method to segment the market information released by the power trading center to obtain trading attribute data, the power trading decision system also includes: The preprocessing module is used to preprocess the market information. The preprocessing steps include data cleaning, data integration, data transformation, and data reduction.

Images