Retrospective Analysis Methods and Systems for Large Hydropower Projects Participating in Spot Market Operations

Abstract

This invention provides a post-mortem analysis method for large hydropower plants participating in the spot market, relating to the field of hydropower station management technology. Addressing the lack of an analytical system for large hydropower stations' operational effectiveness after participating in power trading, this invention achieves effective post-mortem analysis of production plans and trading strategies through three parts: data collection, data analysis, and results presentation. Data collection involves daily collection and standardized storage of transaction data, settlement data, forecast data, and hydropower station operation data. Data analysis uses the mean absolute error to measure the degree of deviation of indicators and combines the TOPSIS algorithm to identify key influencing factors, conducting analyses of revenue composition, forecast deviation, production efficiency, rule adaptability, and changing trends. Results presentation uses visual formats such as pie charts and line graphs to present revenue share, deviation indicators, assessment risks, and trend changes. This invention helps managers formulate cost-effective production plans, reduce power generation costs, and improve economic efficiency.

Description

Technical Field

[0001] This invention relates to the field of hydropower station management technology, specifically to a review method and system for large hydropower stations participating in spot market operations. Background Technology

[0002] Before the full rollout of the electricity spot market, large hydropower stations in southwestern my country primarily participated in medium- and long-term electricity market transactions. Under this model, the trading mechanism was relatively simple, and post-market analysis mainly revolved around the execution of medium- and long-term contracts. Trading centers typically did not release daily clearing statements for hydropower stations, and the methods for post-market analysis of hydropower station operations were relatively straightforward. However, with the deepening development of the electricity spot market, especially with the Southern Regional Market stipulating the participation of 27 large hydropower stations in the unified Southern Regional Market, the way large hydropower stations participate in the market has undergone a fundamental change. According to the rules and regulations of the Southern Regional Electricity Market, hydropower companies already in the market should participate in various transactions within the spot market system, just like other market participants. This includes "point-to-grid" hydropower companies that have signed medium- and long-term market-based contracts. This means that the operating environment for large hydropower has shifted from a single medium- and long-term market to a complex market system where medium- and long-term transactions, day-ahead spot transactions, and real-time spot transactions coexist. Furthermore, compared to the previous system where only medium- and long-term transaction revenue existed, the composition of settlement revenue for large hydropower has become more complex after entering the spot market. This fundamental shift presents significant challenges to traditional, long-term transaction-based retrospective methods in terms of both data dimensions and analytical logic. These methods are unable to effectively support the refined operation and decision-making of hydropower companies in the spot market environment, necessitating a completely new retrospective method adapted to the rules of the spot market.

[0003] The retrospective analysis method proposed in this patent is significantly innovative. Its innovation is first reflected in the targeted nature of the problem definition. It accurately addresses two core changes faced by hydropower after entering the spot market: first, the increased complexity of the transaction structure, requiring the simultaneous handling of multiple transaction and settlement results, including the breakdown of medium- and long-term contracts, day-ahead and real-time market bidding outputs, and the allocation of imbalanced funds; second, the increased sophistication of data requirements. The clearing and settlement of the spot market generates massive amounts of high-frequency detailed data (such as time-of-use electricity prices, unit time-of-use bidding outputs, real-time deviations, and various imbalanced fund costs). This data, which is irrelevant in non-spot market scenarios, has now become a key input for retrospective analysis. Faced with such multi-dimensional data, how should relevant business personnel collect, manage, and analyze it? These are all problems encountered in practical work. Secondly, the method's design concept is forward-looking. It is not a patching of old methods, but a native design based on the new market rule system, such as the Southern Regional Spot Market. These rules are entirely new, involving the connection between inter-provincial and intra-provincial markets, unified clearing, and optimization models that take into account the unique constraints of hydropower. Therefore, the accompanying review methodology must also be innovative.

[0004] Therefore, this invention proposes an analytical system for timely review of trading strategies and production plans when large hydropower stations actively participate in electricity spot market transactions. This system helps managers of large hydropower stations formulate more reasonable and efficient application schemes and develop production plans based on medium- and long-term holdings and spot prices, thereby ensuring a stable supply of electricity to the market while improving the economic benefits of hydropower enterprises. Summary of the Invention

[0005] The technical problem to be solved by this invention is to provide a review method and system for large hydropower stations participating in spot market operations. It solves the core problem that traditional review methods fail in the context of a surge in transaction data and complex analysis dimensions, and fills the technical gap in the field of review of large hydropower station operations in the spot market.

[0006] The technical solution adopted in this invention is a retrospective method for large hydropower projects participating in the spot market, comprising the following steps: S1. Data Acquisition: Daily collection of operational data from the power trading center and hydropower stations, and storage in the database; S2. Data analysis: Calculate total revenue by analyzing revenue composition, perform forecast deviation analysis, and identify core deviation indicators. S3. Results display: Visualized revenue breakdown pie chart, deviation heatmap, and risk warning panel; After large hydropower stations participate in power transactions in the southern region, data collection, analysis, and results presentation are conducted, and a retrospective analysis of the execution of trading strategies and production plans is carried out.

[0007] Preferably, in step S1, the data collected includes: The transaction-related data released by the power trading center includes day-ahead clearing prices, real-time clearing prices, day-ahead clearing volume, real-time clearing volume, actual power load of the province, and predicted power load of the province. Settlement data includes detailed data on the transaction volume and price of medium- and long-term contracts, the settlement volume and price of each period of day-ahead trading, the settlement volume and price of each period of real-time trading, and the return of imbalance funds. Forecast data includes future weather forecasts, reservoir inflow forecasts, and early warning information for extreme weather. Actual operating data of the hydropower station includes unit operating data, reservoir water level, reservoir water volume, and unit maintenance.

[0008] Preferably, in step S2, the calculation formulas for medium- and long-term income, daily income, and real-time income in the income composition analysis include: Medium and long-term income As shown in equation (1): (1); in, This represents the electricity price settled on day i of the k-th contract. This represents the electricity volume settled on day i of the k-th contract; Daily income As shown in equation (2): (2); in, This represents the day-ahead settlement price for the i-th day. This represents the day-ahead settlement electricity volume for day i. Real-time income As shown in equation (3): (3); in, This represents the real-time settlement price of electricity on day i. This represents the real-time electricity consumption on day i. In electricity trading, total revenue includes medium- and long-term revenue, day-ahead revenue, real-time revenue, electricity deviation assessment fees, and imbalance fund allocation and return. Among these, medium- and long-term revenue, day-ahead revenue, real-time revenue, and electricity deviation assessment fees are significantly affected by the active trading behavior of large hydropower operators, while most market imbalance funds are unrelated to the subjective will and trading behavior of market operators.

[0009] Preferably, in step S2, the prediction deviation analysis includes the day-ahead clearing price prediction deviation and the real-time clearing price prediction deviation, and the calculation formula is as follows: Recent clearing of price forecast deviations As shown in equation (4): (4); in, This represents the day-clearing electricity price predicted by the model for day i. This represents the actual day-ahead clearing price on day i; Real-time clearing price forecast bias As shown in equation (5): (5); in, This represents the real-time clearing electricity price predicted by the model for day i. This represents the actual real-time clearing price of electricity on day i. In the transaction declaration data, the power plant predicts the clearing price by combining its own characteristics. The main methods used include time series models, autoregressive models, wavelet decomposition and model fusion, etc. The model used for prediction is not involved. The prediction effect of the model is one of the important contents of the review.

[0010] Preferably, in step S2, the prediction deviation analysis also includes water level prediction deviation, operational transaction deviation, river inflow deviation, and reservoir inflow deviation, calculated using the following formulas: Water level prediction deviation As shown in equation (6): (6); in, This represents the water level height predicted by the model on day i. This represents the actual water level on day i. Running trading deviation As shown in equation (7): (7); in, This represents the actual online electricity consumption on day i. This represents the planned power generation on day i.

[0011] River inflow deviation As shown in equation (8): (8); in, This indicates that the model predicts the reservoir inflow on day i. This represents the actual inflow of water into the reservoir on day i. Reservoir inflow deviation As shown in equation (9): (9); in, This indicates that the model predicts the reservoir inflow on day i. This represents the actual reservoir inflow on day i. Based on all the above deviation values, we perform standardization and normalization to avoid deviations in the analysis results caused by different dimensions, and generate a set of deviation indicators including river inflow and reservoir storage.

[0012] Preferably, in step S2, the core deviation indicators include the TOPSIS algorithm using entropy weighting to calculate the weights of the deviation indicators, and vector standardization and normalization to eliminate the influence of dimensions, thus integrating all indicators. Convert to values ​​in the [0,1] interval. The standardized matrix contains each value. The formula is shown in equation (10) below: (10) Where i represents the number of days of the deviation sample, i=1,2,3,...,n, and j represents the deviation index, j=1,2,3,...,m. For the deviation of the j-th indicator on day i, construct a weighted decision matrix from the standardized data matrix, calculate the weight of each deviation indicator across all objects, and calculate the information entropy value of each indicator. As shown in equation (11): (11) in, It is the natural logarithm. The total number of evaluation subjects; Calculate the coefficient of variation. The formula reflecting the distinguishing ability of the indicator is shown in equation (12) below: (12) The weights are calculated based on the difference coefficients, and the calculation formula is shown in equation (13) below: (13) in, Let be the weight of the j-th indicator.

[0013] Preferably, the step of determining positive and negative ideal solutions involves the following steps: A positive ideal solution is composed of the optimal values ​​of all indicators, with benefit-type indicators taking their maximum values ​​and cost-type indicators taking their minimum values; a negative ideal solution is composed of the worst values ​​of all indicators, with benefit-type indicators taking their minimum values ​​and cost-type indicators taking their maximum values, in contrast to the positive ideal solution. The relative closeness is also considered. The calculation formula is shown in equation (14) below: (14) in, The distance to the positive ideal solution. Distance to the negative ideal solution, relative proximity A larger value indicates a smaller overall deviation for that day, while a smaller value indicates a larger overall deviation for that day. Identify the core deviation indicators and screen the 10% of days with the worst performance. Calculate the average deviation of each indicator on the low-scoring days, standardize and sort them in descending order of weight. The higher the weight of the indicator, the greater its impact on revenue. Select the indicators with a cumulative weight of 80% as the core deviation indicators, output the core deviation indicators and their impact weights, and finally generate an influencing factor report.

[0014] Preferably, by analyzing production efficiency, the risk of holding positions is detected, and an alarm is triggered when the deviation between the actual power generation and the total traded power exceeds the set threshold.

[0015] Preferably, in step S3, the visualization module is configured to use two BI tools, Tableau and FineReport, to visualize and display the analysis results in the form of dashboards, reports and charts, and output an influencing factor report containing core deviation indicators and their influence weights. Each project can also trace back to the past by selecting different date ranges.

[0016] This invention also provides a review system for large hydropower participation in spot market operations, used to execute the aforementioned review method for large hydropower participation in spot market operations, including: The data acquisition module is configured with an API interface to obtain data from the trading center, weather forecasts, and sensor data. The analysis engine module mainly includes a revenue calculation unit and a TOPSIS processing unit; The visualization module generates interactive BI dashboards and supports historical data backtracking.

[0017] The present invention provides a method and system for reviewing the participation of large hydropower projects in the spot market, which has the following beneficial effects: 1. The retrospective analysis method proposed in this patent is significantly innovative. Its innovation is first reflected in the targeted nature of the problem definition. It accurately addresses the two core changes faced by hydropower after entering the spot market: first, the increased complexity of the transaction structure, requiring the simultaneous handling of multiple transaction and settlement results, including the breakdown of medium- and long-term contracts, day-ahead and real-time market bidding outputs, and the allocation of imbalanced funds; second, the refined data requirements, involving massive amounts of detailed data such as time-of-use electricity prices, unit time-of-use bidding outputs, real-time deviations, and various imbalanced fund costs generated by spot market clearing and settlement. This data, which is unnecessary to consider in non-spot market scenarios, becomes a key input for retrospective analysis. Secondly, the method's design concept is forward-looking, based on a native design within the new market rule system of the Southern Regional Spot Market, involving the connection between inter-provincial and intra-provincial markets, unified clearing, and optimization models that consider the unique constraints of hydropower.

[0018] In terms of specific data, compared to markets that only offer medium- to long-term transactions, the spot market discloses more data—data that is larger in quantity, more diverse in dimensions, and more timely. This data places higher demands and challenges on the data collection, management, and analysis capabilities of power plant personnel. The table below lists some key data added after the commencement of spot trading. The data used in this invention was acquired through other automated means, and the acquisition methods are not within the scope of this invention's patent.

[0019]

[0020] 2. The data analysis module of this invention covers revenue composition analysis, forecast deviation analysis, core deviation indicator identification, and production efficiency and risk analysis of individual hydropower stations related to the spot market, comprehensively covering the key aspects of large hydropower participation in the spot market operation. Specifically, the revenue composition analysis clarifies the proportions of medium- and long-term, day-ahead, and real-time revenue, electricity deviation assessment fees, and imbalance fund allocation, eliminating non-management factors from profit and loss; the forecast deviation analysis covers multiple core deviation indicators such as price, water level, inflow, and operational transactions; and the core deviation indicator identification accurately identifies key influencing factors, forming a full-process, multi-dimensional retrospective analysis system to help managers fully grasp the operational status.

[0021] 3. By using the TOPSIS algorithm combined with the entropy weight method to screen core deviation indicators with a cumulative weight of 80%, key influencing factors in the operation process are identified, and targeted influencing factor reports are generated to help managers focus on core issues for improvement. At the same time, the risk warning function for the production efficiency and position assessment of individual hydropower stations related to the spot market can promptly detect abnormal deviations between actual power generation and the planned trading volume and trigger alarms, reduce assessment losses, and ultimately achieve the goal of reducing power generation costs and improving corporate economic benefits.

[0022] 4. By using BI tools such as Tableau and FineReport, the results of the review, such as revenue share, deviation distribution, and risk warning, can be visualized in the form of charts, heatmaps, and warning panels. At the same time, it supports backtracking historical data by date range, making complex analysis results more intuitive and easy to understand, reducing the difficulty for managers to interpret professional data, and helping to make quick and accurate decisions.

[0023] 5. This invention specifically collects large hydropower-specific data such as reservoir water level, inflow, and unit operation, and combines them with the medium- and long-term, day-ahead, and real-time trading scenarios of the electricity spot market to design and analyze dimensions. It is adapted to the operational characteristics of large hydropower participating in regional electricity trading, and can effectively review the rationality of trading strategies and production plan execution, providing technical support for the compliant operation and efficient participation of large hydropower in the spot market. Attached Figure Description

[0024] The present invention will be further described below with reference to the accompanying drawings and embodiments: Figure 1 This is a structural diagram of the retrospective method for large hydropower projects participating in the spot market operation according to the present invention; Figure 2 This is a flowchart of the TOPSIS evaluation method of the present invention. Detailed Implementation

[0025] To better understand the purpose, system architecture, and functional implementation of this embodiment, the embodiments and features described herein can be combined with each other without conflict. The exemplary embodiments disclosed herein will be described below with reference to the accompanying drawings, including specific technical details disclosed to aid understanding; however, these details should be considered exemplary rather than restrictive. Therefore, those skilled in the art should understand that various improvements and adjustments can be made to the embodiments described herein without departing from the scope and core ideas of the invention. Similarly, for clarity, detailed descriptions of well-known technologies, functions, and structures are omitted in the following description.

[0026] Example 1 Figure 1 This is a structural diagram of the retrospective method for large hydropower companies participating in spot market operations, as described in this invention.

[0027] like Figure 1 As shown, the review method 100 for large hydropower participating in spot market operations mainly includes S110~S130.

[0028] In operating S110, the basic data includes sunrise / real-time clearing electricity price, provincial electricity load and power generation, future weather forecast data, hydropower station transaction data, hydropower station actual operation data, and hydropower station actual operation data.

[0029] In operating S120, data processing and analysis include profit and loss analysis, forecast deviation prediction, production efficiency and rule assessment analysis, and revenue composition analysis.

[0030] When operating S130, the results are displayed including profit and loss amount, day-to-day / intraday clearing price deviation, position current, and pie charts and line charts of revenue composition.

[0031] According to an embodiment of the present invention, in operation S110, basic data is acquired before 12:00 noon every day. The data related to transactions includes the day-ahead cleared electricity volume and day-ahead cleared electricity price, the real-time cleared electricity volume and real-time cleared electricity price of the previous day, the electricity prices of all nodes in the province, the actual electricity load of the province, and the actual power generation of various types in the province. The relevant clearing node data includes the current day-ahead transaction volume and day-ahead transaction price, the real-time transaction volume and real-time transaction price of the previous day, the medium- and long-term transaction volume and medium- and long-term transaction price, the electricity price of the node where the hydropower station is located, and the market operation costs and ancillary service costs. The prediction data for future trends includes the day-ahead load forecast of the province, the day-ahead price forecast of the province, future meteorological forecast data, and future hydrological forecast data. The relevant operation data of the hydropower station includes the established production plan, the actual power generation data of the generating units, and the actual hydrological data of the hydropower station.

[0032] According to an embodiment of the present invention, in operation S120, the data collected in S110 is calculated, wherein medium- and long-term income... As shown in equation (15) below.

[0033] (15) in, This represents the electricity price settled on day i of the k-th contract. This represents the electricity volume settled on day i of the k-th contract.

[0034] Daily income As shown in equation (16) below.

[0035] (16) in, This represents the day-ahead settlement price for the i-th day. This represents the daily settlement electricity volume on day i.

[0036] Real-time income As shown in equation (17) below.

[0037] (17) in, This represents the real-time settlement price of electricity on day i. This represents the real-time electricity consumption on day i. Total revenue is the sum of medium- and long-term revenue, daily revenue, real-time revenue, and imbalance fund allocation returns. Based on the revenue data above, the percentage of each revenue item in the total revenue is calculated, and the revenue data for each day and each item is saved to a MySQL database, retaining at least 5 years of data for analysis and visualization of historical trends.

[0038] Recent clearing of price forecast deviations As shown in equation (18) below.

[0039] (18) in, This represents the day-clearing electricity price predicted by the model for day i. This represents the actual day-ahead clearing price on day i.

[0040] Real-time clearing price forecast bias As shown in equation (19) below.

[0041] (19) in, This represents the real-time clearing electricity price predicted by the model for day i. This represents the actual real-time clearing price of electricity on day i.

[0042] Water level prediction deviation As shown in equation (20) below.

[0043] (20) in, This represents the water level height predicted by the model on day i. This represents the actual water level on day i.

[0044] Running trading deviation As shown in equation (21) below.

[0045] (twenty one) in, This represents the actual online electricity consumption on day i. This represents the planned power generation on day i.

[0046] River inflow deviation As shown in equation (22) below.

[0047] (twenty two) in, This indicates that the model predicts the reservoir inflow on day i. This represents the actual inflow of water into the reservoir on day i.

[0048] Reservoir inflow deviation As shown in equation (23) below.

[0049] (twenty three) in, This indicates that the model predicts the reservoir inflow on day i. This represents the actual reservoir inflow on day i. Based on all the above deviation values, standardization and normalization are performed to avoid biases in the analysis results caused by different units of measurement. Then, the TOPSIS algorithm is used to evaluate which item has the greatest impact on the aforementioned revenue. Simultaneously, rule adaptability analysis is conducted to determine whether the assessment standards have been exceeded based on the specific rules of the trading center. Specific assessment items include position assessment and deviation assessment, analyzing the amounts of specific items such as market operation costs and ancillary service fees.

[0050] According to an embodiment of the present invention, in operation S130, BI tools such as Tableau and FineReport are used to visualize the revenue composition, prediction deviation result set, production efficiency and risk analysis results and trends through dashboards, reports and charts. Each item can be traced back to its history by selecting different date ranges.

[0051] Figure 2This is a flowchart of the TOPSIS evaluation method of the present invention.

[0052] like Figure 2 As shown, TOPSIS determines the weight of each deviation indicator's impact on revenue by calculating the relative distance between each deviation indicator and the ideal solution and the negative ideal solution. First, it collects data on various indicators such as day-ahead clearing price forecast deviation, real-time clearing price forecast deviation, water level forecast deviation, operational transaction deviation, river inflow deviation, and reservoir inflow deviation to form a decision matrix. As shown in equation (24).

[0053] (twenty four) Where i represents the deviation sample, i.e., the number of days, i=1,2,3,...,n; j represents the deviation index, j=1,2,3,...,m.

[0054] Secondly, vector standardization and normalization are used to eliminate the influence of dimensions, transforming all indicators into values ​​in the [0,1] interval. The standardized matrix is ​​shown in equation (25).

[0055] (25) in, For the deviation of the j-th indicator on day i, construct a weighted decision matrix from the standardized data matrix, calculate the weight of each deviation indicator across all objects, and calculate the information entropy value of each indicator. As shown in equation (26).

[0056] (26) in, For the natural logarithm, Let represent the entropy value of the j-th deviation index, where j ranges from [1, m].

[0057] Entropy The smaller the value, the greater the dispersion of the indicator and the more information it contains. If all objects of an indicator have the same value, when... If all values ​​are 0 or 1, then At this point, the indicator does not contribute to the evaluation.

[0058] The difference coefficient is calculated. The difference coefficient reflects the distinguishing ability of the indicator. The calculation formula is shown in the following formula (27).

[0059] (27) The weights are calculated based on the difference coefficients, and the calculation formula is shown in equation (28).

[0060] (28) in, Let j be the weight of the j-th indicator. This represents the total number of deviation indicators.

[0061] Determine the positive and negative ideal solutions. The positive ideal solution (PIS) consists of the optimal values ​​of all indicators, with benefit-type indicators taking the maximum value and cost-type indicators taking the minimum value. The negative ideal solution (NIS) consists of the worst values ​​of all indicators, with benefit-type indicators taking the minimum value and cost-type indicators taking the maximum value, which is the opposite of the positive ideal solution.

[0062] Next, calculate the Euclidean distance from each day's data to the positive ideal solution and the negative ideal value, and the distance from sample i to the positive ideal solution. As shown in equation (29) below.

[0063] (29) in, It is the weighted standardized value of the i-th sample for the j-th bias index. The value of the positive ideal solution at the j-th deviation index is the optimal value of the deviation index among all samples.

[0064] Distance from sample i to the negative ideal solution As shown in equation (30) below.

[0065] (30) in, The value of the negative ideal solution on the j-th deviation index is the worst value of the deviation index among all samples. By calculating the distance, the TOPSIS algorithm can objectively compare the degree of closeness of each sample to the ideal state. The smaller the distance, the better the sample, and the larger the distance, the worse the sample.

[0066] Relative proximity The calculation formula is shown in equation (31) below.

[0067] (31) in, The distance to the positive ideal solution. Distance to the negative ideal solution, relative proximity A larger value indicates a smaller overall deviation for that day, while a smaller value indicates a larger overall deviation for that day.

[0068] Identify core deviation indicators, screen the 10% of days with the worst performance, calculate the average deviation of each indicator on the low-scoring days, standardize and sort them in descending order of weight. The higher the weight of the indicator, the greater its impact on revenue. Typically, indicators with a cumulative weight of 80% are selected as core deviation indicators. Output the core deviation indicators and their impact weights, and finally generate an influencing factor report.

[0069] Visualize revenue breakdown charts, deviation heatmaps, and risk warning dashboards using BI tools. Utilize BI tools such as Tableau and FineReport to visualize revenue composition, forecast deviation results sets, and historical trends through dashboards, reports, and charts. Each item can be traced back to its historical origin by selecting different date ranges.

[0070] The production efficiency of a single hydropower station related to the electricity spot market is mainly reflected in the close coupling between its power generation plan and the output awarded in the market, as well as its ability to dynamically optimize production based on signals such as electricity prices and water inflow. Specifically, the production plan must strictly match the output awarded in the market transaction to avoid default or waste of resources. At the same time, when arranging actual production, the production and operation personnel need to use linkage mechanisms to generate more electricity during peak electricity price periods and schedule maintenance or troubleshooting during off-peak electricity price periods to maximize the benefits per kilowatt-hour. This deeply integrates the production plan with market prices and improves the overall production efficiency of a single power station in a market environment.

[0071] The specific logic of the analysis of actual power generation (production) deviation and rule assessment is that in many regions, the two detailed rules or provincial electricity spot market rules will restrict or assess the deviation of power plants. If it is an assessment, the planned and actual power generation will be compared. If the actual power generation exceeds the prescribed threshold, it will be assessed by the dispatch center (two detailed rules) or the trading center (electricity market) according to the specific rules. This invention will use a bar chart to display the planned power generation and the actual power generation to remind relevant business personnel to pay attention to the deviation of power generation and the risk of holding assessment. The planned power generation is shown in the following formula (32).

[0072] (32) The definition of the planned power generation on day i differs across different periods: for days i-2 and earlier, it represents the net contracted power generation for the medium to long term on day i, i.e. On day i-1, it represents the day-ahead clearing volume for day i, i.e. On day i, the real-time clearing of the electricity for day i is... For most spot markets, including the Southern Regional Unified Market, the current rules only take the real-time cleared electricity volume on the last i-th day as the final planned power generation. However, some provinces may impose restrictions or even assessments on the deviation between the medium- and long-term net contract volume and the actual power generation on i-2 days. Therefore, the formula here needs to be adjusted according to the rules of different provinces.

[0073] Example 2 A retrospective system for large hydropower participation in the spot market, used in the retrospective method for large hydropower participation in the spot market as described in Example 1, includes: Data acquisition module: Collects transaction-related data, settlement data, meteorological and hydrological forecast data, and hydropower station operation data published by the power trading center, and stores them in a MySQL database; Analysis engine module: mainly includes revenue calculation unit, prediction deviation analysis unit and TOPSIS processing unit. The revenue calculation unit is used for revenue composition analysis, the prediction deviation analysis unit is used for prediction deviation analysis, and the TOPSIS processing unit is used to identify core deviation indicators using the TOPSIS algorithm. Visualization module: Generates interactive BI dashboards, allowing users to visualize revenue breakdown charts, deviation heatmaps, and risk warning panels using BI tools.

[0074] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this is not limited herein.

[0075] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of this application.

Claims

1. A retrospective analysis method for large hydropower projects participating in the spot market, characterized by: Includes the following steps: S1. Data Acquisition: Daily collection of operational data from the power trading center and hydropower stations, and storage in the database; S2. Data analysis: Calculate total revenue by analyzing revenue composition, perform forecast deviation analysis, and identify core deviation indicators. S3. Results display: Visualized revenue breakdown pie chart, deviation heatmap, and risk warning panel; After large hydropower stations participate in power transactions in the southern region, data collection, analysis, and results presentation are conducted, and a retrospective analysis of the execution of trading strategies and production plans is carried out.

2. The post-analysis method for large hydropower participating in spot market operations according to claim 1, characterized in that, In step S1, the data collected includes: The transaction-related data released by the power trading center includes day-ahead clearing prices, real-time clearing prices, day-ahead clearing volume, real-time clearing volume, actual power load of the province, and predicted power load of the province. Settlement data includes detailed data on the transaction volume and price of medium- and long-term contracts, the settlement volume and price of each period of day-ahead trading, the settlement volume and price of each period of real-time trading, and the return of imbalance funds. Forecast data includes future weather forecasts, reservoir inflow forecasts, and early warning information for extreme weather. Actual operating data of the hydropower station includes unit operating data, reservoir water level, reservoir water volume, and unit maintenance.

3. The post-analysis method for large hydropower participating in spot market operations according to claim 1, characterized in that, In step S2, the calculation formulas for medium- and long-term income, daily income, and real-time income in the income composition analysis include: Medium and long-term income As shown in equation (1): (1)； in, This represents the electricity price settled on day i of the k-th contract. This represents the electricity volume settled on day i of the k-th contract; Daily income As shown in equation (2): (2)； in, This represents the day-ahead settlement price for the i-th day. This represents the day-ahead settlement electricity volume for day i. Real-time income As shown in equation (3): (3)； in, This represents the real-time settlement price of electricity on day i. This represents the real-time electricity consumption on day i. In electricity trading, total revenue includes medium- and long-term revenue, day-ahead revenue, real-time revenue, electricity deviation assessment fees, and imbalance fund allocation and return. Among these, medium- and long-term revenue, day-ahead revenue, real-time revenue, and electricity deviation assessment fees are significantly affected by the active trading behavior of large hydropower operators, while most market imbalance funds are unrelated to the subjective will and trading behavior of market operators.

4. The post-analysis method for large hydropower participating in spot market operations according to claim 1, characterized in that, In step S2, the prediction deviation analysis includes the day-ahead clearing price prediction deviation and the real-time clearing price prediction deviation, and the calculation formula is as follows: Recent clearing of price forecast deviations As shown in equation (4): (4)； in, This represents the day-clearing electricity price predicted by the model for day i. This represents the actual day-ahead clearing price on day i; Real-time clearing price forecast bias As shown in equation (5): (5)； in, This represents the real-time clearing electricity price predicted by the model for day i. This represents the actual real-time clearing price of electricity on day i. In the transaction declaration data, the power plant predicts the clearing price by combining its own characteristics. The main methods used include time series models, autoregressive models, wavelet decomposition and model fusion, etc. The model used for prediction is not involved. The prediction effect of the model is one of the important contents of the review.

5. The method for reviewing the participation of large hydropower projects in the spot market as described in claim 1, characterized in that, In step S2, the prediction deviation analysis also includes water level prediction deviation, operational transaction deviation, river inflow deviation, and reservoir inflow deviation, calculated using the following formulas: Water level prediction deviation As shown in equation (6): (6)； in, This represents the water level height predicted by the model on day i. This represents the actual water level on day i. Running trading deviation As shown in equation (7): (7)； in, This represents the actual online electricity consumption on day i. This represents the planned power generation on day i. River inflow deviation As shown in equation (8): (8)； in, This indicates that the model predicts the reservoir inflow on day i. This represents the actual inflow of water into the reservoir on day i. Reservoir inflow deviation As shown in equation (9): (9)； in, This indicates that the model predicts the reservoir inflow on day i. This represents the actual reservoir inflow on day i. Based on all the above deviation values, we perform standardization and normalization to avoid deviations in the analysis results caused by different dimensions, and generate a set of deviation indicators including river inflow and reservoir storage.

6. The post-analysis method for large hydropower participating in spot market operations according to claim 1, characterized in that, In step S2, the core deviation indicators include the TOPSIS algorithm which uses entropy weighting to calculate the weights of the deviation indicators, and vector standardization and normalization to eliminate the influence of dimensions, thus integrating all indicators. Convert to values ​​in the [0,1] interval. The standardized matrix contains each value. The formula is shown in equation (10) below: (10)； Where i represents the number of days in the deviation sample (i = 1, 2, 3, ..., n), and j represents the deviation index (j = 1, 2, 3, ..., m). A weighted decision matrix is ​​constructed from the standardized data matrix, and the weight of each deviation index across all objects is calculated. The information entropy value of each index is also calculated. As shown in equation (11): (11)； in, It is the natural logarithm. The total number of evaluation subjects; Calculate the coefficient of variation. The formula reflecting the distinguishing ability of the indicator is shown in equation (12) below: (12)； The weights are calculated based on the difference coefficients, and the calculation formula is shown in equation (13) below: (13)； in, Let be the weight of the j-th indicator.

7. The TOPSIS algorithm according to claim 6, characterized in that, The step of determining positive and negative ideal solutions involves the following steps: A positive ideal solution is composed of the optimal values ​​of all indicators, with benefit-type indicators taking their maximum values ​​and cost-type indicators taking their minimum values. A negative ideal solution is composed of the worst values ​​of all indicators, with benefit-type indicators taking their minimum values ​​and cost-type indicators taking their maximum values, in contrast to the positive ideal solution. The relative closeness is also considered. The calculation formula is shown in equation (14) below: (14)； in, The distance to the positive ideal solution. Distance to the negative ideal solution, relative proximity A larger value indicates a smaller overall deviation for that day, while a smaller value indicates a larger overall deviation for that day. Identify the core deviation indicators and screen the 10% of days with the worst performance. Calculate the average deviation of each indicator on the low-scoring days, standardize and sort them in descending order of weight. The higher the weight of the indicator, the greater its impact on revenue. Select the indicators with a cumulative weight of 80% as the core deviation indicators, output the core deviation indicators and their impact weights, and finally generate an influencing factor report.

8. The post-analysis method for large hydropower participating in spot market operations according to claim 1, characterized in that, It also includes production efficiency analysis, detection of position assessment risks, and triggering an alarm when the deviation between the actual power generation and the total traded power exceeds a set threshold.

9. The post-analysis method for large hydropower projects participating in spot market operations according to claim 1, characterized in that, In step S3, the results are displayed using two BI tools, Tableau and FineReport, which visualize the analysis results through dashboards, reports and charts and output a report containing core deviation indicators and their influence weights. Each item can also be traced back to the past by selecting different date ranges.

10. A retrospective system for large hydropower projects participating in the spot market, characterized in that: A review method for implementing any one of claims 1-9 regarding the participation of large hydropower projects in the spot market includes: The data acquisition module is configured with an API interface to obtain data from the trading center, weather forecasts, and sensor data. The analysis engine module mainly includes a revenue calculation unit and a TOPSIS processing unit; The visualization module generates interactive BI dashboards and supports historical data backtracking.

Images