New energy market operation auxiliary decision system based on data interaction and coordinated operation

Abstract

The present application relates to the technical field of power market operation management, and discloses a new energy market operation auxiliary decision system based on data interaction and coordinated operation, comprising: a data management module for collecting data of a remote monitoring system, a power prediction system, a competitive bidding system and a maintenance management system, determining a data missing tolerance level according to a current decision type, and outputting unified standard operation data; a market analysis module for identifying a high electricity price period according to a market price prediction result, judging the overlap condition of an approved maintenance window and the high electricity price period, generating a maintenance adjustment suggestion when the overlap condition is satisfied, and updating market declared power in coordination; and a price prediction module for predicting medium and long term electricity prices and spot electricity prices based on the unified standard operation data, and outputting market price prediction results with confidence intervals. The present application effectively improves the fault tolerance and declaration timeliness of new energy market operation management.

Description

Technical Field

[0001] This invention relates to the field of power market operation and management technology, and more specifically, to a new energy market operation auxiliary decision-making system based on data interaction and coordinated operation. Background Technology

[0002] With the continuous development and improvement of the electricity market, participation of new energy power generation companies in the medium- and long-term market and the spot market has become commonplace. Regional new energy control centers undertake the unified monitoring and operation management functions of multiple wind farms and photovoltaic power stations within their jurisdiction. They need to simultaneously process data from multiple business systems, including remote monitoring systems, power forecasting systems, competitive bidding systems, and the group management platform, and based on this data, complete key decision-making tasks such as market application and maintenance scheduling. Against this backdrop, how to efficiently integrate multi-source data and support regional companies' participation in electricity market transactions has become an important issue for improving the level of new energy operation and management.

[0003] Taking Southwest my country as an example, this region is rich in new energy resources, with wind power and photovoltaic installed capacity continuously expanding. The regional new energy control center needs to manage a large number of scattered wind farms and photovoltaic power stations within its jurisdiction, with grid connection capacity already reaching the megawatt level and still growing rapidly. At the same time, the region's electricity market features both medium- and long-term trading and spot trading operating in parallel. In its daily operations, the control center needs to frequently connect with provincial trading centers, grid dispatching agencies, and group management platforms. The parallel operation of multiple business systems and frequent data interaction place high demands on the multi-source data coordination and processing capabilities of the auxiliary decision-making system.

[0004] To address the aforementioned needs, there is already a certain research foundation for auxiliary decision-making systems for new energy participation in the electricity market. Existing systems typically employ a centralized data aggregation approach, unifying and storing data from various business systems, and building upon this foundation to establish functional modules such as power generation forecasting, price forecasting, and trading strategy generation. Furthermore, some systems are also attempting to integrate maintenance management with market trading functions, allowing users to view maintenance plans and trading applications on a single platform.

[0005] However, existing systems still have significant shortcomings in data coordination. On the one hand, data sources such as remote monitoring, power prediction, competitive bidding, and group platforms vary significantly in terms of collection frequency, update cycle, and transmission latency. Existing systems generally use a fixed time window alignment method to process multi-source data. Once a certain data source is delayed or missing, the system cannot identify the extent to which the missing data affects the current decision-making process. It can only passively wait or directly use expired data. In scenarios where the market reporting window is extremely limited, this can easily lead to decision failure or missing the reporting opportunity.

[0006] On the other hand, the maintenance management system and the market quotation system operate independently. The formulation of maintenance plans fails to be linked with market price forecasts. When the approved maintenance window happens to overlap with the period of high electricity price, the system cannot automatically identify whether the maintenance plan is eligible for adjustment, nor can it trigger the coordinated update of the maintenance plan and the market-declared electricity volume. As a result, the regional company lacks effective coordination and decision support between market revenue optimization and maintenance safety management.

[0007] No effective solutions have yet been proposed to address the problems in the relevant technologies. Summary of the Invention

[0008] To address the problems in related technologies, this invention proposes a new energy market operation auxiliary decision-making system based on data interaction and coordinated operation, in order to overcome the aforementioned technical problems existing in the existing related technologies.

[0009] Therefore, the specific technical solution adopted by the present invention is as follows:

[0010] This new energy market operation auxiliary decision-making system, based on data interaction and coordinated operation, includes: a data management module for collecting data from various data sources, including remote monitoring systems, power prediction systems, bidding systems, and maintenance management systems; real-time monitoring of the update cycle and transmission status of each data source; identification of data missing or delayed states; determination of data missing tolerance levels based on the current decision type; and output of standardized operational data. A market analysis module is used to acquire standardized operational data and approved maintenance plans, identify high-electricity-price periods based on market price forecasts, and determine... The system analyzes the overlap between approved maintenance windows and high-electricity-price periods, generates maintenance adjustment suggestions when the overlap condition is met, and updates market-declared electricity volume to generate a trading strategy package. The price prediction module forecasts medium- and long-term electricity prices and spot prices based on unified operational data, outputs market price prediction results with confidence intervals, and provides these results to the market analysis module. The data management module handles the missing data status of each data source according to the data missing tolerance level, performing waiting, downgrading, or early warning processing. The market analysis module determines that the overlap condition is met when the maintenance window and high-electricity-price period overlap and the overlap duration exceeds a preset threshold.

[0011] Furthermore, this new energy market operation auxiliary decision-making system based on data interaction and coordinated operation also includes: a cockpit module, used to centrally display unified-caliber operation data, execution status of trading strategy packages, and key market indicators in the form of visual charts; a green trading module, used to statistically calculate the green certificates and CCER emission reductions of each station, track the application, listing, and transaction status of green certificates, and provide green certificate revenue data to the review and evaluation module; an economic operation module, used to statistically analyze the power generation, utilization hours, and market revenue of each station based on unified-caliber operation data and executed trading strategy packages, and output economic operation reports; and a review and evaluation module, used to retrieve data status record tables, correction logs, and collaborative update operation records, and generate review and analysis reports.

[0012] The beneficial effects of this invention are as follows:

[0013] (1) This invention introduces a data missing tolerance mechanism centered on decision type in the data management module, and explicitly models the dependency relationship between decision type and each data source as a dependency table. For the missing data sources under different decision scenarios, waiting, downgrading replacement or skipping processing is performed respectively. This breaks the one-size-fits-all logic of "not proceeding if data is not complete" in the existing data platform or SCADA system. In the actual scenario of short application window and uneven data arrival in the new energy market, the system can adaptively coordinate the quality problems of multi-source data according to the current decision intention, thereby avoiding the overall decision process from being blocked due to the lack of non-critical data, and effectively improving the fault tolerance capability and application timeliness of new energy market operation and management.

[0014] (2) This invention establishes a maintenance transferable rule base in the market analysis module and uses the price prediction results with confidence intervals output by the price prediction module as the triggering basis for maintenance adjustment, thereby realizing automatic collaboration between the two traditionally separate business domains of maintenance management and market quotation. The system will actively compare the approved maintenance window with the price prediction results. Once it is found that the maintenance time coincides with the high electricity price period, it will automatically determine whether the maintenance task can be moved back based on the operation type and equipment type. If it can be moved, the optimal transfer period and potential revenue increment will be directly calculated. At the same time, the available power generation during this period will be recalculated into the market declaration power to complete the constraint verification. The maintenance adjustment and the report update will be completed synchronously. There is no need for manual back-and-forth between the two systems. Thus, the power generation revenue missed due to information silos in the past will be transformed into executable market decisions, thereby improving the market operation revenue level of new energy power plants. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a principle block diagram of a new energy market operation auxiliary decision-making system based on data interaction and coordinated operation according to an embodiment of the present invention;

[0017] Figure 2 This is a schematic diagram of the algorithm flow of the core module in the new energy market operation auxiliary decision-making system based on data interaction and coordinated operation according to an embodiment of the present invention;

[0018] Figure 3 This is a schematic diagram of a data missing hierarchical processing mechanism in a new energy market operation auxiliary decision-making system based on data interaction and coordinated operation according to an embodiment of the present invention.

[0019] Figure 4 This is a schematic diagram illustrating the collaborative optimization of maintenance windows and high-electricity-price periods in a new energy market operation auxiliary decision-making system based on data interaction and coordinated operation, according to an embodiment of the present invention.

[0020] Figure 5 This is a schematic diagram of the structure of a computer device according to an embodiment of the present invention. Detailed Implementation

[0021] To further illustrate the various embodiments, the present invention provides accompanying drawings, which are part of the disclosure of the present invention. These drawings are mainly used to illustrate the embodiments and can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these drawings, those skilled in the art should be able to understand other possible implementation methods and the advantages of the present invention. The components in the drawings are not drawn to scale, and similar component symbols are generally used to represent similar components.

[0022] According to an embodiment of the present invention, a new energy market operation auxiliary decision-making system based on data interaction and coordinated operation is provided.

[0023] The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments, such as... Figure 1 and Figure 2As shown, according to an embodiment of the present invention, a new energy market operation auxiliary decision-making system based on data interaction and coordinated operation is provided. This system includes: a data management module, used to collect data from various data sources including a remote monitoring system, a power prediction system, a bidding system, and a maintenance management system; to monitor the update cycle and transmission status of each data source in real time; to identify data missing or delayed states of each data source; to determine the data missing tolerance level based on the current decision type; and to output unified-caliber operation data; and a market analysis module, used to obtain unified-caliber operation data and approved maintenance plans, and to analyze market price prediction results. The system identifies high-electricity-price periods, determines the overlap between approved maintenance windows and high-electricity-price periods, generates maintenance adjustment suggestions when the overlap condition is met, and updates market-declared electricity volume to generate a trading strategy package. The price prediction module is used to predict medium- and long-term electricity prices and spot electricity prices based on unified operational data, outputs market price prediction results with confidence intervals, and provides the market price results to the market analysis module. The data management module performs waiting, downgrading, or warning processing on the missing status of each data source according to the data missing tolerance level. The market analysis module determines that the overlap condition is met when there is an overlap between the maintenance window and the high-electricity-price period and the overlap duration exceeds a preset threshold.

[0024] In one embodiment, when the data management module monitors the update cycle and transmission status of each data source in real time and identifies the data missing or delayed status of each data source, the process includes: configuring a heartbeat detection cycle and a timeout threshold for each data source in the remote monitoring system, power prediction system, bidding system, and maintenance management system; comparing the actual data arrival time with the rated update cycle of each data source; if the actual data arrival time exceeds the sum of the rated update cycle and the timeout threshold, then marking the data source as delayed; if the number of consecutive timeouts exceeds a preset threshold, then marking the data source as missing; writing the delayed or missing status of each data source and the current missing duration into a data status record table, generating a corresponding data status tag, and transmitting the data status tag to the subsequent processing flow of the data management module.

[0025] It should be noted that the remote monitoring system, power prediction system, bidding system, and maintenance management system are all existing systems widely deployed in the operation and management of new energy power plants. The remote monitoring system continuously collects real-time power generation, equipment operating status, and fault alarm data from each power plant through sensors and communication networks, typically refreshing at second or minute intervals. The power prediction system, based on weather forecasts and historical output patterns, generates time-segmented power generation predictions for the next few hours to days, outputting them at hourly or 15-minute granularity. The bidding system connects to the power trading platform, storing the bidding data and transaction results for each power plant within each trading cycle, with an update frequency based on the trading day. The maintenance management system records maintenance applications, approval status, and planned maintenance time windows for each power plant's equipment, with data updates triggered by the maintenance plan approval node. These four systems have significant differences in update frequency and transmission methods, and are the objects that the data management module of this application needs to coordinate and process uniformly.

[0026] In one embodiment, when the data management module determines the data missing tolerance level based on the current decision type, it includes: querying the impact level of each data source on the current decision from a preset decision type and data source dependency table; the impact level includes intolerable, degradeable, and negligible levels; for data sources marked as delayed or missing, executing corresponding processing strategies based on the impact level, specifically including: if the impact level is intolerable, starting a waiting timer and triggering a decision termination warning after the timeout; if the impact level is degradeable, replacing the current missing data with the most recent valid historical data of the data source and adding a degraded mark to the unified caliber running data; if the impact level is negligible, skipping the data source and directly proceeding to subsequent processing; and recording the execution results of the processing strategies and degraded marks of each data source in a data status record table for retrieval and analysis by the review and evaluation module.

[0027] Specifically, the decision type and data source dependency table is a mapping table pre-configured and stored in the system by the system operator based on the actual data needs of various business decisions. Its content includes day-ahead market declarations, monthly medium- and long-term contract declarations, and maintenance plan collaborative updates. Each row in the table records the importance of each data source to the decision outcome under a specific decision type. In this invention, the "unacceptable impact level" means that the information provided by the data source is indispensable input for the current decision. In the day-ahead market declaration scenario of this embodiment, the power prediction system data directly determines the calculation basis of the declared electricity volume. Once missing, the declaration result cannot be generated. At this time, the system must wait for data recovery. If the data is not recovered after a timeout, the current decision is terminated and an early warning is issued to avoid submitting erroneous declarations based on incomplete information. The "degradable use level" means that if the data source is missing, the most recently successfully acquired valid historical data from that data source is used as a temporary substitute. The decision can continue, but a degradation mark needs to be added to the unified caliber operating data to indicate that the current data of downstream modules has quality risks. For example, when the remote monitoring system is temporarily interrupted, the operating status data from the previous collection cycle can be used as a degradation substitute, without affecting the main processes of price prediction and maintenance collaboration. The "ignorable" level means that the data from this data source has no substantial impact on the current decision type, and the system directly skips this data source and proceeds to subsequent processing. For example, when performing a collaborative update of the maintenance plan, the historical bidding data of the bidding system has no direct effect on the current maintenance window judgment and can be ignored.

[0028] In one embodiment, when the data management module outputs unified-caliber operational data, it includes: after processing the missing tolerance level of each data source, normalizing the time granularity, metering boundary, and unit of each data source according to the preset data caliber specification; wherein, the normalized data includes power prediction data of each station in different time periods and historical electricity price data of each time period; merging and encapsulating the normalization result with the data status mark and degradation mark of each data source to generate unified-caliber operational data with complete quality labeling; writing the unified-caliber operational data into the database, and synchronously writing the correction content of the current normalization process into the correction log for subsequent settlement deviation traceability.

[0029] Specifically, the pre-defined data caliber specifications are a set of data standards that are uniformly formulated and solidified in the system configuration by the operator during the system deployment phase, taking into account the metering boundaries of each station and the grid settlement rules. The specifications mainly stipulate requirements in three dimensions: the time granularity is uniformly set to a time-segmented format with hours as the basic unit, and cross-source data can only be combined after each data source is aligned with this granularity; the metering boundaries are clearly defined, and the power data of each station is based on the metering point at the grid connection port, eliminating differences in the caliber of station power consumption and line loss; the unit specifications require that the power data be uniformly set in megawatts (MW), and the electricity price data be uniformly set in yuan per megawatt-hour (yuan / MWh). In this invention, after normalization, the unified operational data actually includes the following: First, power prediction data for each power station in different time periods, i.e., the predicted power generation values ​​for each future time period output by the power prediction system, in MW, used to support the calculation of spot electricity price prediction and market-declared electricity volume; Second, historical electricity price data for each time period, including the medium- and long-term contract electricity prices for each historical month and the spot electricity prices for each historical day and time period, all in yuan / MWh, used to support the construction of similar daily reference samples in medium- and long-term electricity price prediction and spot electricity price prediction; Third, data status markers and degradation markers for each data source, which are encapsulated along with the data as quality labels, for downstream modules to determine whether the current data is complete and reliable when using the data.

[0030] The data status marker is a structured annotation field added to each data record by the data management module after monitoring and identifying missing data sources. It indicates the source quality of the data. The data status marker has three values: normal, delayed, and missing, corresponding to on-time arrival, exceeding the rated update cycle but not yet exceeding the consecutive timeout threshold, and exceeding the consecutive timeout threshold, respectively. The degradation marker is an additional annotation field added by the data management module after degrading missing or delayed data. It indicates that the current data is not from the real-time collection results of the current collection cycle, but is obtained by substituting the most recent valid historical data, and is therefore of questionable quality. In this invention, both types of markers are written into the unified standard operating data along with the data and ultimately stored in the database. The market analysis module can read these markers when using the unified standard operating data for electricity price forecasting and market declaration calculations to determine the completeness of the current input data. Similarly, the review and evaluation module, when conducting historical decision analysis afterward, retrieves these two types of markers to assess the overall quality level of the data relied upon for the current decision, providing a reference for subsequent decision optimization.

[0031] In one embodiment, when the market analysis module identifies high-electricity-price periods based on market price forecast results and determines the overlap between approved maintenance windows and high-electricity-price periods, the process includes: obtaining price forecast results with confidence intervals output by the price forecast module; marking periods where the lower bound of the predicted electricity price confidence interval exceeds the preset electricity price threshold as high-electricity-price periods, based on a preset electricity price threshold; obtaining approved maintenance plans from the maintenance management system; extracting the planned start and end times of each maintenance task as approved maintenance windows; comparing the planned start and end times with high-electricity-price periods one by one; if the planned time period of a maintenance task overlaps with any high-electricity-price period, recording the overlap duration of the maintenance task and the corresponding average predicted electricity price, and generating a maintenance electricity price overlap record; wherein, the maintenance electricity price overlap record includes the maintenance task identifier, overlap duration, the average predicted electricity price of the overlap period, and the corresponding lower bound of the confidence interval.

[0032] Specifically, the preset electricity price threshold is a reference price level, measured in yuan / MWh, pre-set and stored in the system parameter table by the operator during the system configuration phase based on the historical price distribution of the regional electricity market and the company's revenue management requirements. In this embodiment, its value is based on the statistical average of the spot market clearing price in the region over the past year or two, and can also be adjusted by the operator based on the marginal cost per kilowatt-hour and revenue targets of each power station. This invention uses the lower bound of the confidence interval to compare with the electricity price threshold, rather than using the estimated value of the prediction point. The purpose is to place the judgment standard on a more conservative basis. Only when the lower bound of the prediction result still exceeds the threshold is the period confirmed as a high electricity price period, thereby reducing the risk of maintenance plans being erroneously adjusted due to prediction errors.

[0033] In one embodiment, when generating maintenance adjustment suggestions, the market analysis module includes: based on the maintenance task identifier in the maintenance electricity price overlap record, reading the corresponding maintenance task's operation type, involved equipment type, and current approval status from the maintenance management system; matching the maintenance task according to the maintenance shiftable rule library; the maintenance shiftable rule library stores the maximum allowable delay time under different combinations of operation type and equipment type; if the maintenance task meets the matching conditions in the maintenance shiftable rule library, marking the maintenance task as adjustable; calculating the earliest feasible period after the high electricity price period as the suggested shift period; multiplying the difference between the predicted average electricity price of the suggested shift period and the predicted average electricity price of the overlapping period in the maintenance electricity price overlap record by the predicted power generation to obtain the potential revenue increment; encapsulating the adjustable status mark, the suggested shift period, and the potential revenue increment into a maintenance adjustment suggestion, and outputting it to the market analysis module for collaborative update.

[0034] Specifically, the maintenance relocation rule base is a set of rules pre-compiled and stored in the system by the operator during the deployment phase, based on power grid dispatching procedures, equipment manufacturer requirements, and historical maintenance experience. Each rule uses a combination of job type and equipment type as the key and the maximum allowed delay time as the value. In this embodiment, for wind turbine periodic inspections, the maximum allowed delay time recorded in the rule base can be 72 hours; for main transformer protection scheduled maintenance, due to strict safety control requirements, the allowed delay time can be set to 0, indicating that this type of task cannot be relocated. When the system performs condition matching on maintenance tasks, it directly uses the job type and equipment type fields read from the maintenance management system as the query key. If a matching record exists in the rule base and the allowed delay time is greater than 0, the task is marked as adjustable.

[0035] It should also be noted that the calculation method for the proposed shift period in this invention is as follows: Taking the end time of the original maintenance window as the starting point, the earliest time period that meets the following conditions is continuously searched as the proposed shift start time, namely, the average predicted electricity price corresponding to this time period is lower than the preset electricity price threshold, and the time periods from this start time that are not less than the length of the original maintenance window are not within the range of high electricity price periods, and the interval from the end time of the original maintenance window to the proposed shift start time does not exceed the maximum allowable delay time corresponding to the task in the rule base. After the proposed shift period is determined, the potential revenue increment is calculated as follows: Let the average predicted electricity price of the high electricity price periods overlapping with the original maintenance period be P1, in yuan / MWh; the average predicted electricity price of the proposed shift period be P2, in yuan / MWh; the predicted power generation estimated by the station based on the power prediction data within the original maintenance period is Q, in MWh, which is calculated by summing the cumulative power prediction values ​​of each time period within the original maintenance window and multiplying by the length of each time period. The formula for calculating the potential revenue increment ΔR is: ΔR = (P1 - P2) × Q, in yuan. In this formula, the difference between P1 and P2 reflects the price difference between the high-price period and the shifted period, Q is the predicted power generation affected by the maintenance, and the product of the two is ΔR, which is the expected additional market revenue that can be obtained after shifting the maintenance.

[0036] In one embodiment, when the market analysis module collaboratively updates the market-declared electricity volume and generates a trading strategy package, it includes: taking the suggested shift period in the maintenance adjustment proposal as the updated maintenance occupancy period; extracting the power forecast data of the corresponding power station within the suggested shift period from the unified caliber operation data; recalculating the available power generation of the power station within the original maintenance period; superimposing the recalculated available power generation with the existing market-declared electricity volume to generate the updated market-declared electricity volume; and verifying whether the updated market-declared electricity volume exceeds the current medium- and long-term contract's electricity volume limit and grid section constraints; encapsulating the updated market-declared electricity volume, the corresponding bidding strategy, and the maintenance adjustment proposal into a trading strategy package, and simultaneously generating a collaborative update operation record; the operation record includes the electricity volume change and constraint verification results, and is written to the database for retrieval by the review and evaluation module.

[0037] Specifically, after extracting the power prediction data of the corresponding power station within the suggested shift period from the unified operational data, the method for recalculating the available power generation of the power station within the original maintenance period is as follows: Since the maintenance adjustment suggestion has shifted the maintenance-occupied period from the original maintenance window to the suggested shift period, and the power station has returned to a normal power generation state within the original maintenance period, the power prediction values ​​of the power station for each period stored in the unified operational data are used as the basis. The product of the power prediction value of each period and the corresponding period length is summed period by period to obtain the available power generation E of the power station within the original maintenance period, in MWh, using the formula E=Σ(P i ×Δt), where the upper limit of the summation is the total number of time periods N included in the original maintenance period, and the lower limit is i=1, that is, the power prediction values ​​from the first time period to the Nth time period within the original maintenance window are accumulated segment by segment; P i Δt represents the predicted power value for the i-th time period in MW, and Δt represents the duration of the time period in hours.

[0038] It should be added that the existing market-declared electricity volume refers to the market declaration volumes for each time period that were generated and temporarily stored by the market analysis module based on the original maintenance plan before this collaborative update was implemented. These volumes are stored in the system database and are measured in MWh. The grid section constraint is the maximum allowable power transmission limit for each time period, determined and issued by the power dispatching agency based on the thermal stability limits of the regional power grid transmission channels. It is a mandatory technical boundary condition in the current power market operation, and the system obtains this constraint parameter from the maintenance management system or dispatch data interface. When verifying the updated market-declared electricity volume, the system compares the updated declared electricity volume with the minimum value of the power limit and the section constraint for each time period. If any time period exceeds the constraint, the system marks the over-limit status in the constraint verification result field of the operation record for operators to review during post-mortem evaluation.

[0039] In one embodiment, the medium- and long-term electricity price is the contract electricity price formed through centralized bidding or bilateral negotiation within the monthly and annual medium- and long-term contract trading cycle. When forecasting the medium- and long-term electricity price, the price forecasting module includes: extracting the historical monthly medium- and long-term contract electricity prices from unified operational data at a monthly time granularity; using the forecast start month as the cutoff point, extracting a continuous historical monthly electricity price sequence for a preset number of months; using the monthly electricity price in the historical monthly electricity price sequence as the dependent variable and the corresponding month's sequence number as the independent variable, and analyzing the historical monthly electricity price sequence... Linear regression fitting is performed to obtain regression coefficients and intercept terms. The month number of each forecast month is substituted into the regression coefficients and intercept terms to calculate the medium- and long-term electricity price point estimate for each forecast month. A residual sequence is constructed using the difference between the actual electricity price of each month in the historical monthly electricity price series and the corresponding fitted value. The standard deviation of the residual sequence is calculated. The standard deviation is multiplied by the normal distribution quantile corresponding to the preset confidence level to obtain the upper and lower bounds of the confidence interval for each forecast month. The point estimate and the upper and lower bounds of the confidence interval are combined and output as the medium- and long-term market price forecast result.

[0040] Specifically, in this invention, the starting point for extracting the historical monthly electricity price sequence is set to the earliest complete monthly data available to the system, extending up to the month preceding the prediction start month. The number of consecutive preset months can be configured according to the available data range; in this embodiment, the nearest 24 months are used. During linear regression fitting, the month number is used as the independent variable, i.e., the first month in the historical sequence is denoted as number 1, the second month as number 2, and so on. The actual medium- and long-term contract electricity price for each month is used as the dependent variable. The least squares method is used to calculate the regression coefficient k and the intercept term b, ensuring that the fitted value is equal to the sum of the squares of the difference between k multiplied by the month number plus b and the actual electricity price. Substituting the month numbers of the months to be predicted into the above regression equation sequentially yields the estimated medium- and long-term electricity price points for each prediction month, in yuan / MWh. The confidence interval is calculated as follows: The actual electricity price for each month in the historical monthly electricity price series is subtracted from the fitted value for that month to construct a residual series. The standard deviation σ of this residual series is then calculated. In this embodiment, the pre-set confidence level is 95%, corresponding to a two-sided quantile of 1.96 for the standard normal distribution. Therefore, the upper bound of the confidence interval for each predicted month is the point estimate plus 1.96 multiplied by σ, and the lower bound is the point estimate minus 1.96 multiplied by σ. Combining the point estimate, upper bound, and lower bound yields the medium- to long-term market price forecast with confidence intervals, which is then output to the market analysis module for identifying periods of high electricity prices. The linear regression method described above is a mature engineering practice in monthly electricity price forecasting and will not be elaborated upon further here.

[0041] In one embodiment, the spot electricity price is the electricity price for each time period formed by the time-of-use electricity supply and demand clearing mechanism in the day-ahead and real-time spot markets; when the price forecasting module forecasts the spot electricity price, it includes: extracting power forecast data for each time period of the forecast day from unified caliber operating data at an hourly time granularity; extracting power forecast data and spot electricity prices for the corresponding time periods of each historical day within a preset number of days before and after the forecast day from historical data, and constructing a reference sample set of the same type of day using the power forecast data and spot electricity prices for the corresponding time periods of each historical day; and for each time period of the forecast day... The system calculates the mean of the spot electricity price for the corresponding period in the reference sample set of the same type of day, and uses it as the point estimate of the spot electricity price for that period. It constructs the residual series for each period by the difference between the spot electricity price for the corresponding period of each historical day in the reference sample set of the same type of day and the mean of the spot electricity price for the corresponding period. It calculates the standard deviation of the residual series for each period, and multiplies it by the standard deviation according to the normal distribution quantile corresponding to the preset confidence level to obtain the upper and lower bounds of the confidence interval for each period. It combines the point estimate with the upper and lower bounds of the confidence interval and outputs it as the spot market price prediction result, which is then provided to the market analysis module.

[0042] Specifically, when extracting power prediction data for each time period of the prediction day from the unified caliber operational data, the predicted power generation value for all 24 time periods of the prediction day is extracted at the hourly level, in MW. The construction method for the similar day reference sample set is as follows: Centered on the prediction day, extract data within a preset number of days before and after the prediction day; in this embodiment, 30 days before and after are selected, totaling approximately 60 historical days. The power prediction data for each time period of these historical days and the corresponding actual cleared spot electricity price for each time period are read from the unified caliber operational data. These time-segmented data from these historical days are then aggregated into a similar day reference sample set. For each time period of the prediction day, taking the t-th hour as an example, the spot electricity price for all historical days in the same time period t is extracted from the similar day reference sample set, and its arithmetic mean is calculated as the estimated spot electricity price point value for that time period, in yuan / MWh. The construction method for the residual sequence for each time period is as follows: Subtract the mean of the time period from the spot electricity price for each historical day in the similar day reference sample set in time period t to obtain a set of residual values. The standard deviation σ of this set of residual values ​​is then calculated. t Multiply by σ according to the normal distribution quantile corresponding to the preset confidence level. t The upper and lower bounds of the confidence interval for that period are obtained. Finally, the point estimates for all periods of the forecast day and the corresponding upper and lower bounds of the confidence interval are merged to form the spot market price forecast result with confidence intervals, which is provided to the market analysis module in the form of time-segmented electricity price series for the identification and judgment of high electricity price periods.

[0043] In one embodiment, such as Figure 1As shown, this new energy market operation auxiliary decision-making system based on data interaction and coordinated operation also includes: a cockpit module, used to centrally display unified-caliber operation data, execution status of trading strategy packages, and key market indicators in the form of visual charts, to support users in panoramic monitoring of the operation status and market application progress of each station; a green trading module, used to statistically calculate the green certificates and CCER emission reductions of each station, track the application, listing, and transaction status of green certificates, and provide green certificate revenue data to the review and evaluation module; an economic operation module, used to statistically analyze the power generation, utilization hours, and market revenue of each station based on unified-caliber operation data and executed trading strategy packages, and output economic operation reports; and a review and evaluation module, used to retrieve data status record tables, correction logs, and collaborative update operation records, to comprehensively evaluate the accuracy of historical trading decisions, data quality, and maintenance and adjustment benefits, and generate a review and analysis report.

[0044] In this invention, the cockpit module provides operators with a unified visual monitoring interface in the form of a web page. The interface includes a line graph of real-time power generation at each station for the current period, a summary table of data status markers for each data source in the unified operational data, the execution progress and application status of the trading strategy packages for the current day and the next day, and a dashboard of key market indicators, including expected daily revenue and a comparison of the deviation between real-time spot prices and predicted prices. Through the cockpit module, operators can directly grasp the overall operational status of each station and the progress of market applications without consulting a database. Abnormal conditions are highlighted with alarm colors to prompt timely handling. The green trading module interfaces with the national green certificate trading platform and the CCER registration system, periodically capturing the green certificate application status, listing prices, and transaction records of each station. It calculates the revenue from green certificate sales and CCER emission reduction transfers monthly, providing the green revenue data to the review and evaluation module for comprehensive revenue analysis. The economic operation module, based on the executed portions of the time-segmented power forecast data for each power station and the executed trading strategy packages in the unified caliber operation data, summarizes the actual power generation, utilization hours, spot market winning bid volume, and average winning bid price by power station and trading cycle, generating monthly and annual economic operation reports to assist management in assessing the market-based profitability of each power station. The review and evaluation module, after the end of each trading cycle, retrieves the data quality records for that cycle from the data status record table, reads the correction log to confirm the data corrections made by the unified caliber processing, and reads the collaborative update operation records to verify the actual adoption of maintenance adjustment suggestions and the corresponding changes in power volume. It compares this information with the actual settlement data, calculates the absolute and relative deviations between the predicted and actual clearing prices, and the degree of agreement between the actual and potential revenue increases brought about by the maintenance adjustment suggestions, generating a review and analysis report for operators to optimize subsequent decision-making strategies.

[0045] To facilitate understanding of the above-mentioned technical solution of the present invention, the following is a specific explanation using a new energy power generation company in a southwestern region of China as an example:

[0046] This new energy power generation company manages six wind farms and two photovoltaic power stations in northwestern Yunnan Province, with a total installed capacity of approximately 900MW. All of these are located within the coverage area of ​​the region's power transmission channels and are traded in the region's spot electricity market, holding several monthly and annual medium- and long-term contracts. Because the region is located in the southwest monsoon zone, there is a significant difference in hydropower output between the dry and flood seasons. During the flood season, a large influx of hydropower into the market leads to frequent declines in spot prices, while reduced hydropower generation during the dry season significantly increases the value of new energy power. Simultaneously, constrained by the capacity of the west-to-east power transmission channels, the grid's absorption capacity is limited during peak market periods, and the choice of maintenance windows has a significant impact on the actual amount of electricity fed into the grid. These characteristics make the coordinated optimization of maintenance plans and electricity price forecasts highly valuable in this region.

[0047] The company deployed the system described in this invention in its operation management platform to unify and coordinate the operational data of various stations with market declaration decisions. After the system was put into operation, the data management module continuously monitored the data arrival status of four data sources—the remote monitoring system, the power prediction system, the bidding system, and the maintenance management system—with a heartbeat detection cycle of 15 minutes. During the preparation stage of a day-ahead declaration in a dry season, the power prediction system experienced a data interruption of approximately 25 minutes due to a temporary failure of the meteorological data interface. After the data management module identified that the data source had entered a missing state, it determined, based on the decision type and data source dependency table, that the power prediction data had an intolerable level of dependency on the day-ahead declaration and immediately started a waiting timer.

[0048] like Figure 3 As shown in the figure, the horizontal axis represents time in minutes, and the vertical axis represents data status. The upper horizontal line indicates that the data has arrived normally, and the lower horizontal line indicates that the data is missing. It can be seen that this invention does not use a fixed time window alignment method in data processing. Instead, it generates data status markers by using heartbeat detection and timeout threshold judgment, and determines that the power prediction data belongs to an intolerable level dependency in the day-ahead reporting scenario by combining the decision type and data source dependency table, thereby triggering a graded tolerance processing mechanism. After the interface is restored, the latest power prediction data is used to complete the unified standard processing, ensuring the complete execution of the day-ahead reporting process.

[0049] In the price forecasting phase, the price forecasting module extracts historical medium- and long-term contract electricity prices from nearly 24 months of standardized operational data. It then uses linear regression to predict the contract electricity prices for the next three months, outputting forecast results with 95% confidence intervals (both upper and lower bounds are in yuan / MWh). Simultaneously, using the next day as the forecast date, the module constructs a reference sample set for similar days from data from the 30 days before and after the same historical period, predicting the spot clearing electricity price for all 24 time periods on the next day and outputting the 95% confidence interval for each time period. After obtaining the above forecast results, the market analysis module, considering the historical pattern that Yunnan's dry season electricity prices are typically higher than the annual average, identifies five time periods from 16:00 to 21:00 on the next day where the lower bound of the spot price confidence interval is consistently higher than the preset electricity price threshold, marking these as high-price periods.

[0050] like Figure 4 As shown, this invention identifies high-price periods based on the lower bound of the confidence interval of the spot electricity price forecast results, and performs time-axis overlay analysis on the high-price interval and the maintenance window. When the overlap between the maintenance window and the high-price period exceeds a preset threshold, the system triggers the maintenance shift rule matching mechanism and calculates the earliest feasible shift period and the corresponding change in electricity volume. The figure shows the time sequence relationship between the original maintenance window, the suggested shift window, and the high-price period. The horizontal axis represents the number of hours in a day, and the vertical axis represents the electricity price, in yuan / MWh. The curves in the figure represent the predicted electricity price change trend; the horizontal line represents the preset electricity price threshold; the shaded area represents the identified high-price period; one shaded section represents the original maintenance window, and the other shaded section represents the maintenance window after the suggested shift. At the same time, the quarterly regular maintenance plan of a wind farm was read from the maintenance management system. The planned maintenance time is from 17:00 to 20:00 the next day, which overlaps with the identified high-price period by 3 hours, exceeding the preset 2-hour overlap threshold.

[0051] The market analysis module then calls the maintenance relocation rule base to match the maintenance task, confirming that the task belongs to the wind turbine periodic inspection category, with a maximum allowable delay of 72 hours, meeting the relocation conditions. Considering that the cross-sectional constraints of the Yunnan West-to-East Power Transmission Channel are relatively relaxed during the nighttime off-peak hours, the system calculates that shifting the maintenance window to 22:00 to 25:00 the following day (i.e., 1:00 AM the following day) is the earliest feasible time period, as the average predicted electricity price during this period is significantly lower than the average predicted electricity price during the original overlapping high-electricity-price period. Multiplying the price difference between the two time periods by the predicted power generation Q (in MWh, obtained by summing the predicted power value MW of each time period by the time period length h) within the original maintenance period, the potential revenue increment ΔR is estimated to be approximately 150,000 yuan. The system encapsulates the maintenance adjustment suggestion and the collaboratively updated market declaration electricity volume into a trading strategy package, and verifies that the updated declaration volume does not exceed the upper limit of the grid cross-sectional constraints. After confirmation by the operator in the cockpit module, the declaration is submitted, and the relevant maintenance postponement application is simultaneously pushed to the Yunnan Power Dispatch Platform.

[0052] After the transaction settlement, the review and evaluation module retrieves the collaborative update operation record and compares the actual settlement revenue with the predicted potential revenue increment, as shown in Table 1. Through the maintenance window shift and market-reported electricity volume collaborative update, the predicted potential revenue increment in this embodiment is about 150,000 yuan. The actual settlement revenue deviates from the predicted value by about 10%. The main reason is that the actual clearing price of the period after the shift is slightly lower than the predicted average, reflecting that there is still some uncertainty in the supply and demand margin during the nighttime period of Yunnan's dry season.

[0053] Table 1. Comparison of Benefits Before and After Maintenance Collaboration Optimization

[0054] The results demonstrate that, in the market environment of Southwest China with high hydropower penetration and significant price fluctuations, the present invention can effectively improve the market profitability of new energy power plants while maintaining good predictive stability. The aforementioned review data, along with the green certificate revenue calculation results provided by the green trading module, are included in this period's review analysis report, providing quantitative basis for operators to subsequently adjust parameters such as preset electricity price thresholds and the range of reference days for similar days. This embodiment shows that, in the market environment of high hydropower penetration in the monsoon region of Southwest my country, the system described in this invention can effectively coordinate data management, price forecasting, and maintenance plan decision-making, achieving refined operation and management of new energy power plants in the spot market.

[0055] In one embodiment, an electronic device is provided. For example... Figure 5 As shown, the electronic device includes: a processor; a system bus; a main memory; an auxiliary memory; a system human-machine interface for displaying the unified standard operating data and the execution status of the trading strategy package to the user; a database for storing the unified standard operating data, the trading strategy package, and the execution receipt data; and a data exchange interface for enabling data interaction with the remote monitoring system, the power prediction system, the bidding system, and the maintenance management system.

[0056] The processor provides computing and control capabilities and communicates with both main memory and auxiliary memory via a system bus. Main memory is internal memory that provides an environment for the execution of computer programs; auxiliary memory is a non-volatile storage medium that stores the operating system and computer programs. The computer programs are executed by the processor to implement the system described in the above embodiment.

[0057] Those skilled in the art will understand that Figure 5 The structure shown is only a block diagram of a part of the structure related to the present invention and does not constitute a limitation. The specific electronic device may include more or fewer components, or have different component arrangements.

[0058] In addition, the present invention also provides a computer-readable storage medium having computer instructions stored thereon, which, when executed by a processor, implement the steps in the above embodiments.

[0059] The computer-readable storage medium includes main memory or secondary memory. The main memory may be random access memory (RAM), and the secondary memory may be a non-volatile storage medium such as read-only memory (ROM), flash memory, hard disk, or optical storage.

[0060] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware.

[0061] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A new energy market operation auxiliary decision-making system based on data interaction and coordinated operation, characterized in that: include: The data management module is used to collect data from various data sources of the remote monitoring system, power prediction system, bidding system and maintenance management system, monitor the update cycle and transmission status of each data source in real time, identify the data missing or delayed status of each data source, determine the data missing tolerance level according to the current decision type, and output unified standard operating data. The market analysis module is used to acquire the unified standard operation data and approved maintenance plans, identify high electricity price periods based on market price forecast results, determine the overlap between approved maintenance windows and high electricity price periods, generate maintenance adjustment suggestions when the overlap conditions are met, and update market-declared electricity volume to generate a trading strategy package. The price forecasting module is used to forecast medium- and long-term electricity prices and spot electricity prices based on the unified operating data, output market price forecast results with confidence intervals, and provide the market price results to the market analysis module. The data management module performs waiting, downgrading, or early warning processing on the missing status of each data source according to the data missing tolerance level. The market analysis module determines that the overlap condition is met when the maintenance window and the high electricity price period overlap and the overlap duration exceeds a preset threshold.

2. The new energy market operation auxiliary decision-making system based on data interaction and coordinated operation according to claim 1, characterized in that, The data management module, when monitoring the update cycle and transmission status of each data source in real time and identifying data loss or delays in each data source, includes: Configure heartbeat detection cycles and timeout thresholds for each data source of the remote monitoring system, power prediction system, bidding system, and maintenance management system, and compare the actual data arrival time according to the rated update cycle of each data source; If the actual data arrival time exceeds the sum of the rated update cycle and the timeout threshold, the data source is marked as delayed; if the number of consecutive timeouts exceeds the preset threshold, the data source is marked as missing. Write the delay or missing status of each data source and the current missing duration into the data status record table, generate the corresponding data status flag, and pass the data status flag to the subsequent processing flow of the data management module.

3. The new energy market operation auxiliary decision-making system based on data interaction and coordinated operation according to claim 2, characterized in that, When determining the data missing tolerance level based on the current decision type, the data management module includes: Based on the currently triggered decision type, query the impact level of each data source on the current decision from the preset decision type and data source dependency table; the impact level includes intolerable level, degradeable level, and negligible level; For data sources marked as delayed or missing, corresponding processing strategies are executed based on the impact level. Specifically, if the impact level is intolerable, a waiting timer is started, and a decision termination warning is triggered after the timeout. If the impact level is degradeable, the current missing data is replaced with the most recent valid historical data of the data source, and a degraded flag is added to the unified caliber operating data. If the impact level is negligible, the data source is skipped and the subsequent processing is initiated directly. Record the execution results of the processing strategies for each data source and the degradation flags to the data status record table.

4. The new energy market operation auxiliary decision-making system based on data interaction and coordinated operation according to claim 3, characterized in that, When the data management module outputs standardized operational data, it includes: After processing the missing tolerance level of each data source, the time granularity, metering boundary and unit of each data source are normalized according to the preset data standard; wherein, the normalized data includes the power prediction data of each station in different time periods and the historical electricity price data of each time period. The normalization results are merged and encapsulated with the data status tags and degradation tags of each data source to generate unified caliber operational data with complete quality labels; Write the unified standard operating data into the database, and synchronously write the correction content of the current normalization process into the correction log.

5. The new energy market operation auxiliary decision-making system based on data interaction and coordinated operation according to claim 1, characterized in that, When the market analysis module identifies high-electricity-price periods based on market price forecasts and determines the overlap between approved maintenance windows and high-electricity-price periods, it includes: Obtain the price prediction result with confidence interval output by the price prediction module, and mark the period when the lower bound of the predicted electricity price confidence interval exceeds the preset electricity price threshold as a high electricity price period, based on the preset electricity price threshold. The approved maintenance plans are obtained from the maintenance management system. The start and end times of each maintenance task are extracted as the approved maintenance window. The start and end times of the plans are compared with the high electricity price periods one by one. If the planned time period of a certain maintenance task overlaps with any high electricity price period, the overlapping duration of the maintenance task and the corresponding average predicted electricity price are recorded to generate a maintenance electricity price overlap record. The maintenance electricity price overlap record includes maintenance task identifier, overlap duration, average predicted electricity price for the overlap period, and the corresponding lower bound of the confidence interval.

6. The new energy market operation auxiliary decision-making system based on data interaction and coordinated operation according to claim 5, characterized in that, When generating maintenance and adjustment suggestions, the market analysis module includes: Based on the maintenance task identifier in the maintenance electricity price overlap record, the operation type, equipment type and current approval status of the corresponding maintenance task are read from the maintenance management system, and the maintenance task is matched according to the maintenance transferable rule base; the maintenance transferable rule base stores the maximum allowable delay time under different combinations of operation type and equipment type. If a maintenance task meets the matching conditions in the maintenance shift rule base, the maintenance task is marked as adjustable. The earliest feasible period after the high electricity price period is calculated as the suggested shift period. The difference between the predicted average electricity price of the suggested shift period and the predicted average electricity price of the overlapping period in the maintenance electricity price overlap record is multiplied by the predicted power generation to obtain the potential revenue increment. The adjustable status flags, suggested shift periods, and potential revenue increments are encapsulated as maintenance and adjustment suggestions and output to the market analysis module for collaborative updates.

7. The new energy market operation auxiliary decision-making system based on data interaction and coordinated operation according to claim 6, characterized in that, When the market analysis module collaboratively updates market-reported electricity volume and generates trading strategy packages, it includes: The suggested shift period in the maintenance adjustment proposal is taken as the updated maintenance occupancy period. The power prediction data of the corresponding power station within the suggested shift period is extracted from the unified caliber operation data, and the available power generation of the power station within the original maintenance period is recalculated. The recalculated available power generation is superimposed on the existing market-declared power to generate an updated market-declared power, and it is verified whether the updated market-declared power exceeds the current medium- and long-term contract power limit and grid section constraints. The updated market-reported electricity volume, corresponding pricing strategies, and maintenance adjustment suggestions are packaged into a trading strategy package, and a collaborative update operation record is generated. The operation record includes the electricity volume change and constraint verification results, and is written to the database for the review and evaluation module to retrieve.

8. The new energy market operation auxiliary decision-making system based on data interaction and coordinated operation according to claim 1, characterized in that, The medium- and long-term electricity price mentioned above refers to the contract electricity price formed through centralized bidding or bilateral negotiation within the monthly and annual medium- and long-term contract trading cycles. The price forecasting module, when forecasting medium- and long-term electricity prices, includes: The historical monthly electricity prices of medium- and long-term contracts are extracted from the unified operational data at a monthly time granularity. The historical monthly electricity price sequence for a consecutive preset number of months is extracted with the prediction start month as the cutoff time. Using the electricity price of each month in the historical monthly electricity price series as the dependent variable and the corresponding month number as the independent variable, a linear regression is performed on the historical monthly electricity price series to obtain the regression coefficients and intercept term. Substitute the month number of each forecast month into the regression coefficient and intercept term to calculate the estimated medium- and long-term electricity price point for each forecast month. A residual sequence is constructed by the difference between the actual electricity price of each month and the corresponding fitted value in the historical monthly electricity price sequence. The standard deviation of the residual sequence is calculated. The standard deviation is multiplied by the normal distribution quantile corresponding to the preset confidence level to obtain the upper and lower bounds of the confidence interval for each prediction month. The point estimate and the upper and lower bounds of the confidence interval are combined and output as the medium- and long-term market price prediction result.

9. The new energy market operation auxiliary decision-making system based on data interaction and coordinated operation according to claim 8, characterized in that, The spot electricity price is the electricity price for each time period formed in the day-ahead and real-time spot markets according to the time-of-use electricity supply and demand clearing mechanism. The price forecasting module, when forecasting spot electricity prices, includes: Power prediction data for each time period of the prediction day is extracted from the unified caliber operation data at the hourly time granularity. Power forecast data and spot electricity prices for each historical day within a preset number of days before and after the forecast date are extracted from historical data. A reference sample set for the same day is constructed using the power forecast data and spot electricity prices for each historical day. For each time period of the forecast day, the average spot electricity price of the corresponding time period in the reference sample set of the same type of day is calculated as the estimated spot electricity price point for that time period; The residual series for each time period is constructed by the difference between the spot electricity price for each historical day and the mean spot electricity price for each time period in the reference sample set of the same type of day. The standard deviation of the residual series for each time period is calculated. The standard deviation is multiplied by the normal distribution quantile corresponding to the preset confidence level to obtain the upper and lower bounds of the confidence interval for each time period. The point estimate and the upper and lower bounds of the confidence interval are combined and output as the spot market price prediction result, and provided to the market analysis module.

10. The new energy market operation auxiliary decision-making system based on data interaction and coordinated operation according to claim 1, characterized in that, Also includes: The cockpit module is used to centrally display the unified operational data, execution status of the trading strategy package, and key market indicators in the form of visual charts; The green trading module is used to statistically calculate the green certificates and CCER emission reductions of each site, track the application, listing and transaction status of green certificates, and provide the green certificate revenue data to the review and evaluation module. The economic operation module is used to perform statistical analysis on the power generation, utilization hours and market revenue of each power station based on the unified standard operation data and the executed trading strategy package, and output economic operation reports. The debriefing and evaluation module is used to retrieve data status records, correction logs, and collaborative update operation records, and generate a debriefing analysis report.

Images