Electricity purchasing party electricity transaction method based on multi-dimensional data analysis

Abstract

The invention particularly relates to an electricity purchasing party electricity transaction method based on multi-dimensional data analysis, and relates to the technical field of electricity market transaction. Analyzing and predicting data; transaction strategy and execution; and risk management. According to the method, deep fusion and efficient management of multi-source data are realized, and the data-driven decision-making capability is improved; through high-frequency acquisition of multi-source data of power grid operation, market transaction, user side, external environment and the like, in combination with preprocessing means of abnormal value processing, missing value filling and the like, data quality is ensured; ordered storage and quick retrieval of the data are matched with metadata management auditing, so that the accuracy and traceability of the data are guaranteed; therefore, the electricity purchaser can insight the market dynamics based on comprehensive and accurate data and accurately pre-judge the risk, solid data support is provided for transaction decision making, and decision making errors caused by data isolation and insufficient analysis are effectively avoided.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power market transaction, and particularly relates to a power transaction method for a power purchasing party based on multi-dimensional data analysis. BACKGROUND

[0002] With the gradual opening of the power market and the large-scale access of new energy, the market environment faced by the power purchasing party is becoming increasingly complex: On the one hand, the uncertainty of new energy such as wind power and photovoltaic power leads to dynamic changes in power grid congestion risk, and traditional power flow calculation and risk assessment methods are difficult to adapt to large-scale scenarios; On the other hand, power transaction involves multi-source heterogeneous data (power grid operation, market transaction, user side, external environment, etc.), and the data management is scattered and the analysis dimension is single, which cannot provide comprehensive support for transaction decision-making.

[0003] The prior art has deficiencies in data integration efficiency, uncertainty analysis depth, risk quantification precision and transaction strategy adaptability, and it is urgent to build a complete method system covering data management, analysis and prediction, risk control and transaction execution, i.e. a power transaction method for a power purchasing party based on multi-dimensional data analysis, to cope with the above-mentioned problems. SUMMARY

[0004] The purpose of the present application is to solve the above-mentioned problems by providing a power transaction method for a power purchasing party based on multi-dimensional data analysis.

[0005] In order to achieve the above-mentioned purpose, the present application adopts the following technical solutions: The power transaction method for a power purchasing party based on multi-dimensional data analysis comprises: Data management and integration: through multi-source data acquisition, cleaning and standardization processing, using distributed storage and metadata management technology, efficient integration and orderly management of power transaction related data are realized; Data analysis and prediction: multi-dimensional deep analysis is performed on power transaction data, combined with market prediction and scenario simulation, to provide data support and trend prediction for transaction decision-making; Transaction strategy and execution: personalized transaction strategies are formulated according to market analysis and user demand, and order processing and transaction execution are automatically completed, and real-time monitoring and adjustment are performed; Risk management: identify various risks in power transaction, reduce the impact of risks through quantitative evaluation, hedging measures and early warning response mechanism.

[0006] Preferably, the data management and integration specifically comprises: Multi-source data acquisition, including power grid operation data acquisition, market transaction data acquisition, user side data acquisition and external environment data acquisition; and cleaning and preprocessing of data; In the Hadoop distributed file system, the storage directory is divided according to the data type and time period; When building a data warehouse, a star model is used to model user data, with user ID as the center, and industry, scale, and power consumption characteristics are associated; Use open source metadata management tools to record data definitions, sources, quality indicators, and update frequencies; Regularly audit metadata.

[0007] Preferably, the data analysis and prediction specifically includes: Combine LSTM with attention mechanism to capture long-term price trend and short-term fluctuation characteristics; Based on user historical power consumption data, a Markov chain model is constructed to predict user power consumption state transition probability under different price scenarios; Combining DC power flow type with Monte Carlo simulation, considering the uncertainty of wind power and photovoltaic output, the probability distribution of congestion risk is evaluated; Develop an index system for power grid vulnerability assessment, which integrates line load rate and short-circuit current level factors to identify weak links in the power grid and optimize transaction path planning; Combine Copula function to depict the dependence structure of multiple risk factors such as price and supply and demand; Use scenario analysis method to generate extreme market scenarios (such as extreme weather superimposed on policy changes) and calculate risk exposure; Build a credit scoring model based on machine learning, incorporating multi-dimensional features such as generator's historical default records, financial statement data, and public opinion information.

[0008] Preferably, the DC power flow type combined with Monte Carlo simulation considers the uncertainty of wind power and photovoltaic output to evaluate the probability distribution of congestion risk, specifically including: In each sampling of Monte Carlo simulation, the DC power flow model is used to calculate the line flow and determine whether it is congested, and finally the congestion probability is calculated; In the first According to the DC power flow model, the active power flow of the line is calculated as: ; : In the first The voltage phase angle of node i and node j; : The reactance of the line ; : The total number of scenarios in Monte Carlo simulation.

[0009] Preferably, the node voltage phase angle The node injection power equation is obtained by solving the node injection power equation for node i, and the injection power equation for node i is: ; : a set of nodes connected to node i.

[0010] 6. The electricity purchasing method based on multi-dimensional data analysis according to claim 5, wherein the congestion probability of the line is calculated according to the formula: ; ; The congestion probability of the line is combined with the active power flow to obtain ; : the rated transmission capacity of the line ; The node injection power equation is obtained by solving the node injection power equation, and in each scenario , the uncertainty of wind power and photovoltaic power output is embodied by generating different power generation scenarios through random sampling, thereby affecting the node injection power ; In each scenario , the line flow is calculated, and whether it is overloaded is recorded by an indicator function by comparing it with the rated capacity; the number of overloads and non-overloads is counted, and the overload depth is calculated based on the overload corresponding to the flow result; A preset overload depth threshold is subtracted from each overload depth, and the values less than 0 obtained are discarded, and the values greater than 0 are recorded as overload depth values; After counting the number of all overload depth values, the overload degree is obtained by dividing the overload number by the overload degree; The overload degree is obtained by dividing the number of all overloads by the number of non-overloads; After weighted calculation of the overload degree and the overload degree, the line congestion evaluation value is obtained; a plurality of threshold value ranges are preset, each threshold value range corresponds to a line congestion probability, and the line congestion evaluation value is matched with the plurality of threshold value ranges to obtain the corresponding line congestion probability.

[0011] Preferably, the diversified strategies are designed according to different user needs; The rule engine is used to preset the strategy template and the trigger condition; the optimization algorithm is combined to solve the optimal transaction combination according to the market prediction data; at the same time, the historical data backtest is used to evaluate the yield and risk exposure of the strategy, and the strategy parameters are continuously optimized; Based on the user portrait and the market prediction result, appropriate transaction strategies and contract periods are recommended for users; According to the transaction strategy, day-ahead market declaration quantity and price and real-time market bidding instructions are automatically generated, and the transaction center is connected through an API interface to obtain the transaction result in real time. The transaction process is monitored in real time, the transaction limit price deviation threshold is set, and when an exception occurs, the order is automatically blocked and a warning is issued. The full-link data of the order from declaration to transaction is recorded, including quotation information, transaction result, and fund flow, and a transaction account is generated.

[0012] Preferably, the risk management specifically includes: After identifying the risk, the risk is quantified, and a risk assessment index system is established to comprehensively score various risks; Financial derivatives are introduced to lock in price fluctuation risks through hedging operations; standby capacity purchase agreements are signed with peak shaving power sources to address power supply gaps caused by extreme weather or power grid failures; and a certain proportion of risk reserves is calculated based on historical risk event loss distribution; A risk warning index system is established, different levels of warning thresholds are set, and when the risk index exceeds the threshold, the corresponding level of warning is triggered, and the warning information is pushed to the relevant personnel.

[0013] Preferably, the user management and service: Users are divided into different roles, and data viewing and operation permissions are configured for each role through an RBAC model; A transaction data query function is provided, which supports user query by time, transaction type, and contract; transaction strategy reports, cost analysis reports, and risk assessment reports are automatically generated.

[0014] As described above, due to the adoption of the above technical solutions, the present application has the following advantages: 1、The present application improves data-driven decision-making capabilities through multi-source data deep integration and efficient management; through high-frequency collection of multi-source data such as power grid operation, market transactions, user side, and external environment, combined with preprocessing methods such as outlier processing and missing value filling, data quality is ensured; ordered storage and rapid retrieval of data, combined with metadata management audit, ensure the accuracy and traceability of data; enabling the power purchaser to accurately predict risks based on comprehensive and accurate data insights into market dynamics, providing solid data support for transaction decisions, and effectively avoiding decision-making errors caused by data isolation and insufficient analysis; 2、The application enhances the security and flexibility of electricity purchase transaction by constructing a full-process closed-loop management system; in terms of transaction strategy formulation, personalized strategies are designed for different user needs, and an optimization algorithm is used to solve the optimal transaction combination, and the parameters are continuously optimized through historical backtesting; in the execution process, automatic order processing, real-time monitoring and intelligent adjustment are realized, combined with electronic contract management and performance reminder to ensure efficient and accurate transaction; in terms of risk management, the system identifies various risks such as market and credit, uses VaR model for quantitative evaluation, introduces financial derivatives to hedge risks, and establishes a multi-level early warning and emergency response mechanism. BRIEF DESCRIPTION OF DRAWINGS

[0015] In the following description of exemplary embodiments in conjunction with the accompanying drawings, more details, features and advantages of the application are disclosed, in which: Figure 1 The system structure diagram of the application. DETAILED DESCRIPTION

[0016] Several embodiments of the application will be described in more detail below with reference to the accompanying drawings, so that those skilled in the art can implement the application. The application can be embodied in many different forms and purposes and should not be limited to the embodiments described herein. These embodiments are provided to make the application comprehensive and complete, and to fully convey the scope of the application to those skilled in the art. The embodiments do not limit the application.

[0017] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. It will be further understood that terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant art and / or the present specification, and should not be interpreted in an idealized or overly formal sense, unless expressly so defined herein.

[0018] Embodiment 1

[0019] The detailed description is combined with the accompanying drawings Figure 1 will be described in detail.

[0020] The accompanying drawings Figure 1 The electricity purchase method flowchart based on multi-dimensional data analysis provided by the embodiment of the application shows the complete steps from multi-source data collection to risk management.

[0021] In this embodiment, it includes: Data management and integration: through multi-source data collection, cleaning and standardization processing, using distributed storage and metadata management technology, realizing efficient integration and orderly management of power transaction related data, laying data foundation for system operation; Specifically includes: Multi-source data collection, including power grid operation data collection, market transaction data collection, user-side data collection, and external environment data collection; Power grid operation data collection: Through real-time communication interface with the power dispatch center, real-time load, voltage, frequency, and other data of the power grid are obtained every 15 minutes; for distributed power sources (such as photovoltaic power stations and wind farms), special data collectors are used to collect power generation power, equipment status, and other information at a minute level, providing a basis for real-time power supply and demand analysis; Market transaction data collection: Connect with the trading platform of the power trading center, download the transaction data of the day-ahead market and the real-time market daily, including transaction subjects, transaction volume, transaction price, transaction period, etc.; at the same time, through web crawler technology, real-time scraping of unstructured data such as power policy documents and market rule change announcements published on government department websites is carried out, and key information such as price adjustment range and subsidy policy changes is extracted using natural language processing (NLP) technology; User-side data collection: For users with smart meters, the power demand side management system collects user electricity load curves in hours; for large industrial users, access their internal energy management systems to obtain equipment operating status, production plans, and other data for analyzing the correlation between electricity consumption and production; in addition, historical electricity purchase contracts and credit rating data of users are collected to improve user portraits; External environment data collection: Cooperate with meteorological departments to obtain hourly weather data such as temperature, humidity, and wind speed, as weather factors have a significant impact on power demand (such as increased air conditioning load in hot weather); regularly download macroeconomic indicators such as GDP and industrial added value from the statistical bureau website for analyzing the correlation between economic situation and power demand; And clean and preprocess the data, including outlier treatment, missing value filling, and data standardization; In the Hadoop distributed file system, storage directories are divided according to data types and time periods, such as storing power grid operation data by day and market transaction data by month to improve data retrieval efficiency; use HBase as a columnar database to store real-time data and user portrait data that need to be quickly queried, meeting high concurrency access requirements; When building a data warehouse, use a star model to model user data, with user ID as the center, and associate industry, size, and electricity consumption characteristics; use ETL tools to extract, transform, and load data from source systems to the data warehouse, and establish a data lineage graph to clearly show the source and processing of data; Using open-source metadata management tools, record the definition, source, quality indicators, and update frequency of the data; regularly audit the metadata to ensure data accuracy and traceability, such as checking whether the update time of the data collection interface is consistent with the expected time; Data analysis and prediction: using various analysis models and techniques, multi-dimensional deep analysis of power trading data, combined with market prediction and scenario simulation, providing data support and trend prediction for trading decisions; Specifically including: Combining LSTM and attention mechanism (Attention), capturing long-term price trend and short-term fluctuation characteristics; introducing Transformer model to process unstructured policy text, quantifying the influence weight of policy on electricity price; Using social network analysis algorithm, draw the transaction relationship graph of power suppliers, identify the core participants and potential alliances in the market; simulate bidding strategies through game theory model, evaluate market power distribution; Build a policy impact model to analyze the transmission path of carbon emission quota price to the marginal cost of thermal power enterprises under the carbon trading policy; evaluate the impact degree and time lag effect of renewable energy subsidy reduction on new energy electricity price; Based on user historical electricity consumption data, build a Markov chain model to predict user electricity state transition probability (such as switching from high load to low load) under different electricity price scenarios; Use association rule mining (Apriori algorithm) to analyze the correlation between electricity consumption behavior and external factors (weather, holidays), such as the proportion of air conditioning load increase per 1℃ temperature rise; Introduce federated learning technology to train prediction models by combining multiple user electricity consumption data while protecting user privacy; combine enterprise production plan Gantt chart data to refine device-specific and time-specific electricity consumption forecasts; Combining DC power flow model with Monte Carlo simulation, considering the uncertainty of wind power and photovoltaic output, evaluating the probability distribution of congestion risk; Combining DC power flow model with Monte Carlo simulation, considering the uncertainty of wind power and photovoltaic output, evaluating the probability distribution of congestion risk, including: The DC power flow model calculates the line flow through linear equations, and the Monte Carlo simulation considers the uncertainty factors such as wind power and photovoltaic power through random sampling; In each sampling of Monte Carlo simulation, the DC power flow model is used to calculate the line flow and determine whether it is congested, and finally the congestion probability is calculated; In the first scenario, according to the DC power flow model, the active power flow of the line is calculated by the formula: ; : In the first scenario, the voltage phase angle of node i and node j; : The line reactance; : The total number of scenarios in Monte Carlo simulation.

[0022] The DC power flow model is based on a series of reasonable assumptions (ignoring line resistance, reactive power impact, assuming constant voltage amplitude), which converts the complex AC power flow nonlinear calculation into linear equation solving. Compared with AC power flow calculation, it greatly reduces the calculation complexity, significantly reduces the time of each power flow calculation in Monte Carlo simulation which needs a large number of scenarios (hundreds of times or even more sampling), improves the overall simulation efficiency, and completes the large-scale uncertainty analysis in an acceptable time. For example, when analyzing the power grid with a large number of new energy access, the power flow distribution under different output scenarios can be quickly traversed; Node voltage phase angle The node voltage phase angle needs to be obtained by solving the node injection power equation. For node i, the injection power equation is: ; : The set of nodes connected to node i.

[0023] The node injection power equation directly describes the balance relationship between node injection power and line power flow; Wind power, photovoltaic, load and other uncertain factors are ultimately reflected as random fluctuations in node injection power (such as when wind power is large, the injection power of the node where the wind farm is located increases significantly); Through the node injection power equation, these random fluctuations are "converted" into changes in voltage phase angle, which in turn affect the line power flow. This process fully restores the physical logic of "power / load changes to power flow adjustment", making the congestion risk assessment more realistic; Line Congestion probability The calculation formula is: ; The congestion probability of the line is combined with the active power flow to obtain ; : The rated transmission capacity of the line ; It is obtained by solving the node injection power equation, and in each scenario , the uncertainty of wind power, photovoltaic and other outputs is reflected by generating different power generation scenarios through random sampling, which in turn affects the node injection power ; ​​Calculate the line flow in each scenario , compare it with the line rated capacity, and record whether it is overloaded through an indicator function; count the number of overloads and non-overloads, and calculate the overload depth based on the flow results corresponding to the overload; ; represents the percentage of line flow exceeding the rated capacity in the scenario ; Pre-set the overload depth threshold, subtract the overload depth threshold from each overload depth, and eliminate the values less than 0 obtained from the values greater than 0, and record them as overload depth values; After counting all the overload depth values and dividing them by the number of overloads, the overload depth is obtained; Divide the number of overloads by the number of non-overloads to obtain the overload degree; After weighting the overload depth and the overload degree, the line congestion evaluation value is obtained; pre-set the value range of multiple sets of threshold values, and each set of threshold value range corresponds to a line congestion probability; match the line congestion evaluation value with the value range of multiple sets of threshold values to obtain the corresponding line congestion probability; Pre-set the weight factors of overload depth and overload degree, respectively multiply the overload depth and the overload degree by their corresponding weight factors, and then sum them up to obtain the line congestion evaluation value; The traditional method of only counting the congestion probability is relatively single. This method uses overload depth (reflecting the degree of flow exceeding rated capacity during single overload) to enable the buyer to clearly understand the "impact" of congestion on the line; for example, with a congestion probability of 10%, one scenario is occasional small overload (such as exceeding rated capacity by 5%), and the other is occasional large overload (such as exceeding rated capacity by 30%), which has a large difference in transaction risk impact, and overload depth can accurately distinguish between them; Overload depth (based on pre-set threshold, focusing on the proportion of overload depth exceeding threshold) further filters out extreme congestion scenarios that "seriously threaten transaction execution" to help the buyer identify the most vigilant risk points; overload degree (overload to non-overload ratio) reflects the frequency of congestion from a macro perspective. Multiple dimensions of indicators are combined to build a complete risk profile from "whether congestion" to "how serious is the congestion" and "how frequent is the serious congestion"; Develop a power grid vulnerability assessment index system to identify weak links in the power grid by considering factors such as line load rate and short-circuit current level, and optimize transaction path planning; Establish a power grid fault tree model to analyze the impact of fault propagation paths on the buyer; simulate power grid fault scenarios through digital twinning technology and develop differentiated backup power source activation strategies; Integrate Copula function to depict the dependence structure of multiple risk factors of electricity price and supply-demand; use scenario analysis to generate extreme market scenarios (e.g. extreme weather combined with policy shocks) and calculate risk exposure; Build a credit scoring model based on machine learning, incorporating multi-dimensional features such as the historical default records of power suppliers, financial statement data (asset-liability ratio, liquidity ratio), and public opinion information (negative media reports); Trading strategy and execution: develop personalized trading strategies based on market analysis and user needs, automatically complete order processing and transaction execution, and monitor and adjust in real time to ensure efficient, accurate, and compliant transaction processes; Specifically includes: Design diverse strategies for different user needs; For example, for cost-sensitive industrial users, develop peak-valley electricity price arbitrage strategies to increase electricity consumption during off-peak hours and reduce it during peak hours; for risk-averse users, use risk hedging strategies that combine long-term contracts and financial derivatives to lock in electricity costs; Use rule engines to pre-set strategy templates and trigger conditions, such as activating demand response when real-time electricity price > baseline price * 120%; use optimization algorithms such as linear programming and genetic algorithms to solve the optimal trading portfolio based on market forecast data, balancing cost, risk, and reliability; at the same time, evaluate the yield and risk exposure of the strategy through historical data backtesting, and continuously optimize the strategy parameters; Based on user profiling and market forecast results, recommend appropriate trading strategies and contract terms for users; for example, recommend long-term fixed electricity price contracts for large industrial users with stable loads, and flexible short-term contracts and spot trading portfolios for commercial users with volatile loads; Automatically generate day-ahead market bid prices and real-time market bidding instructions based on trading strategies, and interface with the power trading center through API to obtain real-time transaction results; for unexecuted orders, automatically adjust the price or bid amount and resubmit. At the same time, electronically manage long-term contracts and power purchase agreements, record contract terms, electricity delivery time, settlement methods, and other information, and set a compliance reminder function; Real-time monitoring of the trading process, setting transaction limits (such as maximum electricity volume per transaction and maximum transaction amount per day), price deviation thresholds (such as ±20%), and other risk control indicators, automatically blocking orders and issuing warnings when abnormalities occur; when market forecast deviation exceeds the threshold, automatically trigger strategy modification, such as increasing or decreasing spot purchases. At the same time, support manual intervention in the trading process and provide strategy adjustment suggestions; Record the full-link data of orders from bidding to transaction, including bid information, transaction results, and fund flow, generate a transaction ledger; record system operations, strategy changes, and exception handling events through a log management system for audit backtracking and problem positioning; Risk management: Identify various risks in power trading, reduce the impact of risks through quantitative assessment, hedging measures and early warning response mechanisms to ensure the safe and stable operation of trading; Specifically including: After identifying risks, risk quantification is performed, and a risk assessment index system is established to comprehensively score various risks; Risk type analysis: For market risk, historical data is used to simulate electricity price fluctuation scenarios to assess the impact of different fluctuation amplitudes on cost; For credit risk, data such as the financial status of trading partners and historical performance records are collected, and a credit scoring model is used to assess the probability of default; For operational risk, key nodes in the trading process are identified to identify potential human error or system failure; For policy risk, track policy changes and analyze the potential impact of policy adjustments on trading strategies.

[0024] Risk quantification assessment: Use VaR model, stress testing and other methods to quantify market risk and determine the maximum potential loss at a certain confidence level; For credit risk, set credit limits and dynamically adjust transaction size based on the credit rating of the trading partner; Introduce power futures and options financial derivatives to lock in electricity price fluctuation risk through hedging operations; Sign a standby capacity purchase agreement with peak shaving power sources to address power supply gaps caused by extreme weather or power grid failures; According to the loss distribution of historical risk events, a certain proportion of risk reserves is set aside to cover potential losses; Establish a risk early warning index system and set different levels of warning thresholds (such as yellow, orange, red warning); When the risk index exceeds the threshold, trigger the corresponding level of warning and push the warning information to relevant personnel; At the same time, develop an emergency strategy library to pre-set response plans for different risk events (such as electricity price surge, trading partner default) and automatically or manually initiate emergency measures.

[0025] The power trading method for power purchasers based on multi-dimensional data analysis according to claim 1, characterized in that user management and service: Users are divided into administrators, traders, analysts and other roles, and RBAC (Role-Based Access Control) model is used to configure data viewing and operation permissions for each role; For example, administrators have management permissions for all system functions, traders can only perform trading operations and view related data, and analysts can view analysis results but have no right to trade; Provide transaction data query function, support users to query by time, transaction type, contract; Automatically generate trading strategy report, cost analysis report, risk assessment report, present in the form of charts and text, convenient for users to understand the trading situation; Provide online manuals, video tutorials, and other training resources to help new users quickly familiarize themselves with system operations. Set up customer support channels to promptly answer users' questions during use; Support user-defined transaction preferences and parameters, such as setting price alert thresholds, selecting default transaction strategies, and adjusting data display methods, to meet individual user needs; Collect user opinions and suggestions through online questionnaires, feedback forms, and other means, regularly analyze user feedback data, and identify shortcomings in system functions and services. Based on user feedback, continuously optimize system functions, improve user interface design, and enhance user experience and satisfaction.

[0026] The above formulas are obtained by collecting a large amount of data for software simulation and selecting a formula close to the true value. The influence weight factor and specific coefficient value in the formula are set by a person skilled in the art according to the actual situation, and can be adjusted and modified later.

[0027] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, they produce the processes or functions described in the embodiments of the present application in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.

[0028] The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another, for example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center through wired or wireless (such as infrared, wireless, microwave, etc. ) way.

[0029] The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server, data center, etc. containing a set of one or more available media. The available media can be magnetic media, optical media, or semiconductor media. The semiconductor medium can be a solid state disk.

[0030] It should be understood that the term "and / or" in this document merely describes an associated relationship, which means that there can be three relationships, for example, A and / or B can mean that A exists alone, A and B exist together, and B exists alone, where A and B can be singular or plural. In addition, the character " / " in this document generally represents an "or" relationship between the front and rear associated objects, but it can also represent an "and / or" relationship. The specific meaning can be understood according to the context before and after.

[0031] In several embodiments provided in the present application, it should be understood that the disclosed system, device and method can be implemented in other ways.

[0032] For example, the device embodiments described above are only schematic, for example, the division of the units is only a logical function division, and actual implementation can have another division manner, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some interface, device or unit, which can be electrical, mechanical or other forms.

[0033] The units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, that is, they can be located in one place, or can be distributed on a plurality of network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment scheme.

[0034] In addition, the functional units in each embodiment of the present application can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit.

[0035] If the functions are realized in the form of software function units and sold or used as independent products, they can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application or the part of the present application which essentially contributes to the prior art or the part of the technical solutions can be embodied in the form of software products, which are stored in a storage medium and include a plurality of instructions for making a computer device (which can be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the method described in each embodiment of the present application.

[0036] And the aforementioned storage medium includes: a USB flash disk, a mobile hard disk, a read-only memory, a random access memory, a magnetic disk or an optical disk, and various media that can store program codes.

[0037] The foregoing description of the embodiments enables one of ordinary skill in the art to make or use the application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein can be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for power transaction of a power purchaser based on multi-dimensional data analysis, characterized in that, Comprise: Data management and integration: through multi-source data collection, cleaning, standardization processing, using distributed storage and metadata management technology, realize the efficient integration and orderly management of power transaction related data; Data analysis and prediction: multi-dimensional deep analysis of power transaction data, combined with market prediction and scenario simulation, provide data support and trend prediction for transaction decision; Transaction strategy and execution: according to market analysis and user demand, formulate personalized transaction strategy, automatically complete order processing and transaction execution, and real-time monitoring and adjustment; Risk management: identify various risks in power transaction, reduce the impact of risk through quantitative evaluation, hedging measures and early warning response mechanism. 2.The method of claim 1, wherein, Data management and integration, including: Multi-source data collection, including power grid operation data collection, market transaction data collection, user side data collection and external environment data collection; And data cleaning and preprocessing; In the Hadoop distributed file system, according to the data type and time period, the storage directory is divided; When building data warehouse, use star model to model user data, take user ID as center, associate industry, scale, power consumption characteristics; Use open source metadata management tools to record data definition, source, quality index, update frequency; Regularly audit metadata. 3.The power purchase method based on multi-dimensional data analysis according to claim 1, wherein, Data analysis and prediction, including: Combined with LSTM and attention mechanism, capture long cycle price trend and short-term fluctuation characteristics; Based on user historical power consumption data, build Markov chain model to predict user power consumption state transition probability under different price scenarios; Direct current flow type and Monte Carlo simulation are combined to consider the uncertainty of wind power and photovoltaic output, evaluate the probability distribution of congestion risk; Develop power grid vulnerability evaluation index system, integrate line load rate, short circuit current level factors, identify weak links of power grid, optimize transaction path planning; Combined with Copula function, depict the dependence structure of multiple risk factors of price and supply and demand; Use scenario analysis method to generate extreme market scenarios (such as extreme weather superimposed policy mutation), calculate risk exposure; Build credit scoring model based on machine learning, include multi-dimensional features of power generation company's historical default record, financial statement data and public opinion information. 4.The method of claim 3, wherein, Direct current flow type and Monte Carlo simulation are combined to consider the uncertainty of wind power and photovoltaic output, evaluate the probability distribution of congestion risk, including: In each sampling of Monte Carlo simulation, use direct current flow model to calculate line flow and judge whether it is congested, and finally count the congestion probability; In the first scenario, according to the direct current flow model, the active power flow of the line is calculated by the following formula: ; : In the first scenario, the phase angle of the voltage of node i and node j; : line of reactance; : Total number of scenarios for the Monte Carlo simulation.

5. The method for power transaction of electricity buyer based on multi-dimension data analysis according to claim 4, characterized in that, Node voltage phase angle The node injection power equation for node i is obtained by solving the node injection power equation: ; : Set of nodes connected to node i. 6.The method of power transaction of electricity buyer based on multi-dimension data analysis according to claim 5, characterized in that, Line The blocking probability The calculation formula is: ; The line blocking probability is combined with the active power flow to obtain ; : line of the rated transmission capacity; By solving the node injection power equation, and in each scenario The uncertainty of wind power and photovoltaic power is reflected by generating different power generation scenarios through random sampling, which further affects the node injection power ; In each scenario The line flow is calculated and compared with the line rating capacity, and the overload is recorded by the indicator function. Count all overload and non overload quantity, and calculate overload depth based on overload corresponding flow result; Predefine overload depth threshold, subtract overload depth threshold from each overload depth, remove values less than 0 from the obtained values, and record values greater than 0 as overload depth value; After counting all overload depth values, divide by overload quantity to get overload depth; Divide all overload quantity by non overload quantity to get overload degree; After weighted calculation of overload degree and overload degree, get line congestion evaluation value; The preset multiple sets of threshold value ranges, each set of threshold value range corresponding to a line congestion probability, matching the line congestion evaluation value with the multiple sets of threshold value ranges to obtain the corresponding line congestion probability. 7.The power trading method for power purchasers based on multi-dimensional data analysis according to claim 1, wherein, Transaction strategy and execution, specifically including: Designing diversified strategies for different user needs; Using rule engine to preset strategy template and trigger condition; combining optimization algorithm, solving the optimal transaction portfolio according to market prediction data; at the same time, through historical data backtest, evaluating the yield and risk exposure of the strategy, and continuously optimizing the strategy parameters; Based on user portrait and market prediction results, recommending appropriate transaction strategies and contract periods for users; According to the transaction strategy, automatically generating day-ahead market bidding price and real-time market bidding instruction, and through API interface, connecting with power trading center to obtain real-time transaction results; Real-time monitoring of the transaction process, setting transaction limit price deviation threshold, automatically blocking orders and issuing early warning when abnormality occurs; Recording the whole link data of the order from bidding to transaction, including quotation information, transaction results, fund flow, and generating transaction account. 8.The power trading method for power purchasers based on multi-dimensional data analysis according to claim 1, wherein, Risk management, specifically including: Risk quantification after risk identification, and at the same time, establishing a risk assessment index system to comprehensively score various risks; Introducing financial derivatives to lock in price fluctuation risk through hedging operations; signing standby capacity purchase agreement with peak shaving power to deal with power supply gap caused by extreme weather or power grid failure; according to the loss distribution of historical risk events, setting aside a certain proportion of risk reserve; Establishing a risk early warning index system and setting different levels of early warning thresholds; when the risk index exceeds the threshold, triggering the corresponding level of early warning and pushing the early warning information to relevant personnel. 9.The power trading method for power purchasers based on multi-dimensional data analysis according to claim 1, wherein, User management and service: Dividing users into different roles and configuring data viewing and operation permissions for each role through RBAC model; Providing transaction data query function, supporting users to query by time, transaction type and contract; automatically generating transaction strategy report, cost analysis report and risk assessment report.

Images