Source-network collaborative optimization system of electricity market

Abstract

The invention discloses a source-network collaborative optimization system of a power market, relates to the technical field of power markets, and aims to generate a prediction interval set, a flexibility curve and a security constraint set, form a source-network joint feasible region and complete realizability check, complete multi-product joint clearing in the source-network joint feasible region and generate a time sequence power track and a price signal. According to the method, through source-network joint feasible region construction, security constraint preposition is integrated into market clearing, the operation safety and economy are improved, multi-product joint clearing and refined instruction decomposition are adopted, and the operation efficiency is improved. The market efficiency and the control precision are enhanced, the deviation immunity and the online optimization are realized by depending on closed-loop evaluation and data driving, and the fair and ordered market is guaranteed through a compliance mechanism.

Description

Technical Field

[0001] This invention relates to the field of power market technology, and more particularly to a power market source-grid coordinated optimization system. Background Technology

[0002] With the advancement of dual-carbon goals and the construction of new power systems, the power system is undergoing profound changes. The high proportion of renewable energy connected to the grid has brought about strong uncertainty and volatility. At the same time, the deepening of power market reforms and the increasing complexity of trading instruments pose significant challenges to the traditional source-follow-load dispatching model and the sequential, independently optimized operation framework.

[0003] Currently, Chinese invention patent application number CN202311366991.0 discloses a method and system for multi-source collaborative optimization of distribution networks. The method includes: constructing a multi-source collaborative strategy for the distribution network; determining the total resource consumption of distribution network source-load scheduling based on the resource consumption model of distribution network source-load scheduling; establishing a multi-source collaborative optimization model for the distribution network; the model includes: a day-ahead planned scheduling model and an intraday rolling correction scheduling model; the day-ahead planned scheduling model aims to minimize the total operating resource consumption of the distribution network; the intraday rolling correction scheduling model aims to minimize the adjustment of distribution network source-load output and maximize the wind power and photovoltaic absorption rate; and performing multi-source collaborative optimization of the distribution network based on the multi-source collaborative optimization model of the distribution network.

[0004] The aforementioned technologies are insufficient to deeply integrate power grid physical security constraints with multi-timescale economic dispatch during the power market clearing process, thus hindering the achievement of closed-loop optimization and risk control through source-grid coordination. Summary of the Invention

[0005] To address the aforementioned problems, this invention provides a source-grid coordinated optimization system for the power market.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0007] A source-grid coordinated optimization system for the electricity market, comprising:

[0008] The data acquisition module is used to collect source-side operational data, network-side monitoring data, market quotation data, and execution status data, and to perform spatiotemporal alignment and quality verification, and store them in a unified data pool;

[0009] The state prediction module is used to make predictions based on data in the data pool and generate a set of prediction intervals, flexibility curves, and a set of safety constraints.

[0010] The source-network joint construction module is used to generate the source-network joint feasible region based on the prediction interval set, flexibility curve and security constraint set and to complete the realizability verification.

[0011] The market scheduling coupling module is used to complete the joint clearing of multiple products and generate time-series power trajectories and price signals in the source-network joint feasible domain.

[0012] The instruction decomposition and monitoring module is used to translate the time-series power trajectory and price signal into executable instructions for the device and monitor the execution deviation in real time.

[0013] The closed-loop evaluation and correction module is used to diagnose and continuously correct execution deviations.

[0014] Preferably, the data acquisition module includes:

[0015] The spatiotemporal alignment unit is used to perform cross-source comparison and unify source-side operational data, network-side monitoring data, market quotation data, and execution status data to the same time base, and adds alignment method identifiers and missing data markers;

[0016] The quality verification unit is used to verify the source-side operation data, network-side monitoring data, market quotation data, and execution status data according to preset boundary rules.

[0017] The unified data pool unit is used to classify and store verified source-side operational data, network-side monitoring data, market quotation data, and execution status data, and to add maintenance version numbers, data time window indexes, and data integrity verification codes.

[0018] Preferably, the verification content includes identifying abnormal rise and fall points according to mutation rules, identifying delays or packet loss according to communication delay rules, generating trusted data labels for qualified data, and attaching abnormal type labels to unqualified data.

[0019] Preferably, the state prediction module includes:

[0020] The load and renewable energy forecasting unit is used to generate a set of time-segmented forecast intervals with upper and lower boundaries and confidence labels based on data in the data pool, combined with historical data and unit operation and maintenance plans.

[0021] The flexibility profiling unit is used to convert unit parameters into flexibility curves.

[0022] The safety status identification unit is used to generate a safety constraint set by hourly detection in the network-side monitoring dataset, combining the power flow calculation results and the real-time load level of the cross section.

[0023] Preferably, the flexibility curve includes a climbing ability curve, a sustained ability curve, and a non-functional ability curve in a continuous time dimension.

[0024] Preferably, the source-network joint construction module includes:

[0025] The source-side feasible region generation unit is used to generate the source-side feasible region based on the flexibility curve on a unified time series, and in the generation process, the ramp rate constraint, minimum start-up and shutdown time constraint and energy storage operating condition constraint are mapped into time series boundary conditions to form a set of source-side feasible regions.

[0026] The grid-side feasible region generation unit is used to combine the power flow balance equation, the cross-sectional transmission capacity limit, the upper and lower limits of DC transmission and receiving ends, and the voltage and reactive power operating range in the safety constraint set to establish a set of grid-side feasible regions that meet the boundary conditions of the distribution network and the transmission network on an hourly basis, and output constraint indicators.

[0027] The domain coupling scaling unit is used to perform set intersection on the source-side feasible domain set and the network-side feasible domain set on a unified time axis to obtain the source-network joint feasible domain, and to tighten the boundary of the source-network joint feasible domain hourly according to the constraint activity index and risk coefficient.

[0028] The feasibility verification unit is used to match and compare the salable power segments corresponding to the market quotation dataset with the source-network joint feasible domain hourly, and output the source-network joint feasible domain that meets the threshold.

[0029] Preferably, after performing hourly matching and comparison between the reportable power segments corresponding to the market quotation dataset and the source-network joint feasible domain, the method further includes:

[0030] If the reportable power segment does not meet the threshold, the reportable power segment is marked as unrealizable and removed, or adjusted to the effective range within the joint boundary.

[0031] Preferably, the market scheduling coupling module includes:

[0032] The multi-product joint clearing unit is used to simultaneously allocate resources through a unified optimization model based on actual needs within the source-network joint feasible domain, and generate time-sharing and product-sharing clearing results.

[0033] The network-endogenous pricing unit is used to calculate shadow prices based on the activity constraints of cross sections and nodes during the process of generating clearing results, and output node electricity prices and cross section congestion shadow values.

[0034] The trajectory generation unit is used to convert the cleared power of time-sharing products into a time-series power trajectory covering the settlement period based on the clearing results, and to generate a backup reserved trajectory and a reactive power reserved trajectory, and to add a tracking error threshold and tolerance rules to each time interval in the trajectory.

[0035] Preferably, the instruction decomposition and monitoring module includes:

[0036] The instruction decomposition unit, after receiving the timing power trajectory, combines the corresponding tolerance rules and tracking error threshold to convert the timing power trajectory into setting instructions that the device can directly execute:

[0037] The boundary dispatch unit is used to generate transmission boundary commands and uplink capacity curves based on the cross-section congestion shadow and the distribution network transmission boundary;

[0038] The execution monitoring unit is used to read real-time data from the execution status dataset according to a preset sampling period, compare it with the corresponding trajectory and boundary instructions, generate execution deviation values ​​and durations, and mark them according to tolerance rules to form execution deviation records.

[0039] Preferably, the closed-loop evaluation correction module includes:

[0040] The deviation diagnosis unit is used to analyze the execution deviation records item by item, classify the deviations into prediction errors, equipment limitations and network mutations, and locate the corresponding links while classifying them.

[0041] The rolling reconfiguration unit is used to trigger a local reconfiguration process when the execution deviation continuously exceeds the tracking error threshold or an out-of-bounds trend occurs. It performs local recalculation of the affected time period and constraints, and outputs a new time-series power trajectory and reserve and reactive power configuration when the emergency re-clearing function is called.

[0042] The Behavioral Compliance Unit is used to continuously assess the consistency between market participants' pricing and performance.

[0043] The beneficial effects of this invention are as follows:

[0044] This invention constructs a feasible domain by combining source and network, integrates security constraints into market clearing in advance, improves operational security and economy, adopts multi-product joint clearing and refined instruction decomposition to enhance market efficiency and control precision, relies on closed-loop evaluation and data-driven approach to achieve deviation immunity and online optimization, and ensures a fair and orderly market through compliance mechanisms. Attached Figure Description

[0045] Figure 1 This is a block diagram of a power market source-grid coordinated optimization system in a specific embodiment of the present invention. Detailed Implementation

[0046] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0047] Please see Figure 1 As shown, the present invention relates to a source-grid coordinated optimization system for the power market, comprising:

[0048] The module comprises: data acquisition and calibration module, predictive state identification module, source-network joint construction module, market scheduling coupling module, instruction decomposition and monitoring module, and closed-loop evaluation and handling module.

[0049] The data acquisition and calibration module is used to perform spatiotemporal alignment and quality verification on source-side operational data, network-side monitoring data, market quotation data, and execution status data, and store them in a unified data pool;

[0050] Specifically, it includes:

[0051] Granularity of a unified time base:

[0052] The system uses the sampling period of the power grid SCADA system as a unified time reference. Typically, this reference granularity is set to one data point every 15 or 30 seconds. All data sources must be aligned to this timestamp sequence.

[0053] Cross-source matching algorithms:

[0054] A primary key-based association matching method is employed. Specifically, a globally unique asset code is assigned to each physical device (such as a generator, line, or transformer). Data from different data sources (such as monitoring systems, asset management systems, and market quotation systems) are associated and concatenated using this asset code to form a complete spatiotemporal data record for that device.

[0055] Threshold for determining missing test markers:

[0056] The system scans in time windows (e.g., 5 minutes). If the number of valid values ​​for a data point within the current window is less than 60% of the window length, the time window is marked as "missing data". Simultaneously, for channels that have not reported any data for more than 30 consecutive reference time points, a "communication interruption" alarm is triggered.

[0057] The prediction state identification module is used to generate prediction interval sets, flexibility curves, and safety constraint sets.

[0058] The source-network joint construction module is used to form the source-network joint feasible domain and complete the feasibility verification.

[0059] The market scheduling coupling module is used to complete the joint clearing of multiple products and generate time-series power trajectories and price signals within the source-network joint feasible domain.

[0060] The instruction decomposition and monitoring module is used to translate the timing power trajectory into executable instructions for the device and monitor the execution deviation in real time.

[0061] The closed-loop assessment and handling module is used to diagnose and continuously correct execution deviations.

[0062] This invention constructs a feasible domain by combining source and network, integrates security constraints into market clearing in advance, improves operational security and economy, adopts multi-product joint clearing and refined instruction decomposition to enhance market efficiency and control precision, relies on closed-loop evaluation and data-driven approach to achieve deviation immunity and online optimization, and ensures a fair and orderly market through compliance mechanisms.

[0063] The data acquisition and calibration module includes a spatiotemporal alignment unit, a quality verification unit, and a unified data pool unit.

[0064] The spatiotemporal alignment unit is used to perform cross-source comparison of source-side operational data, network-side monitoring data, market quotation data, and execution status data and unify them to the same time base. Each data record is marked with an alignment method identifier and a missing data mark.

[0065] The spatiotemporal alignment unit achieves time-series synchronization and benchmark unification of multi-source data. Through cross-source comparison and time alignment processing, this unit eliminates time-series discrepancies caused by differences in acquisition frequency, communication delays, and timestamps, providing a unified time-series framework for the fusion analysis of multimodal data. The added alignment markers and missing data markers enhance the transparency and traceability of the data process.

[0066] The quality verification unit is used to detect whether the source-side operation data, network-side monitoring data, market quotation data and execution status data exceed the operating range according to the preset boundary rules, identify abnormal rise and fall points according to the mutation rules, identify delay or packet loss according to the communication delay rules, generate reliable data tags for qualified data, and attach anomaly type tags to unqualified data.

[0067] Specifically, the magnitude of abnormal spikes / drops is defined using a dynamic threshold method. A data point is marked as abnormal if it meets one of the following conditions:

[0068] |x t -x t-1 |>3·σ;

[0069] |x t -μ historical |>5·σ historical ;

[0070] Where σ is the standard deviation of recent data, μ historical σ historical These represent the mean and standard deviation of long-term historical data under the same operating conditions.

[0071] The quality verification unit has established an automated data quality assessment and labeling system. Based on predefined business rules, this unit checks the validity of data and identifies and classifies abnormal data. By attaching trust labels to qualified data and labeling the type of abnormal data, it provides clear data quality identifiers for downstream modules, supporting the application of differentiated data processing strategies.

[0072] The unified data pool unit is used to classify and store the verified source-side operation data, network-side monitoring data, market quotation data, and execution status data. These data are written into the source-side operation dataset, network-side monitoring dataset, market quotation dataset, and execution status dataset, respectively, and the unit maintains the version number, data time window index, and data integrity verification code.

[0073] The unified data pool unit constructs a systematic and versioned data center, classifying, storing, and centrally managing verified data to form a structured data set. By maintaining data versions, time indexes, and integrity check codes, it ensures data reusability, retrieval, and consistency, providing efficient and reliable data access services for all modules of the system.

[0074] As the foundational data layer of the system, the data acquisition and calibration module effectively addresses the standardization and reliability issues of multi-source heterogeneous data. Its core effect lies in constructing a high-quality, traceable, unified data source through standardized processing and quality governance of raw data. This provides an accurate and reliable data foundation for optimization calculations and decision analysis in upper-layer applications, ensuring the overall reliability of the system from the source.

[0075] The predictive status identification module includes a load and renewable energy prediction unit, a flexibility profiling unit, and a safety status identification unit.

[0076] The load and renewable energy forecasting unit is used to generate a set of time-segmented forecast intervals with upper and lower boundaries and confidence level labels based on historical load curves, weather data, meteorological forecasts and renewable energy output records, combined with unit operation and maintenance plans. The set of time-segmented forecast intervals labels the upper boundary value, lower boundary value and confidence level of each time interval.

[0077] Specifically, the generation of prediction intervals and confidence labels:

[0078] The specific criteria for dividing the day into time periods are as follows: Based on the characteristics of the power grid load, the day is divided into three periods: peak, flat, and valley. For example: peak period (08:00-12:00, 17:00-21:00), flat period (12:00-17:00, 21:00-24:00), and valley period (00:00-08:00). A prediction model is built independently for each time period.

[0079] Confidence labels are calculated by using quantile regression to generate prediction intervals.

[0080] The load and renewable energy forecasting unit provides quantitatively described uncertainty forecasting information. This unit not only generates point forecasts, but more importantly, it outputs a set of time-segmented forecast intervals with confidence level labels, quantifying the uncertainty of load and renewable energy output in boundary form. This output provides crucial input for subsequent risk-averse optimization decisions, enabling the system to perform robust optimization within the foreseeable fluctuation range and improving the ability of the decision-making scheme to cope with uncertainty.

[0081] The flexibility profiling unit is used to convert the unit's minimum start-up time, minimum downtime, ramp-up and ramp-down rates, energy storage system's charge and discharge power boundaries and SOC constraints, and interruptible load's allowable interruption duration and recovery interval parameters into continuous time-dimensional ramp-up capability curves, continuous capability curves, and non-functional capability curves.

[0082] Specifically, the generation of the climbing ability curve:

[0083] Based on the unit's technical ramp rate R tech and current output P current Generate upper and lower bound curves for climbing ability that vary over time:

[0084] Uphill boundary: P up (t)=min(P max ,P current +R tech ·t);

[0085] Downhill boundary: P down (t)=max(P min ,P current -R tech ·t)

[0086] Generation of the sustainability curve:

[0087] The maximum possible duration of the computer group at the current output is determined based on the unit's minimum start / stop time and fuel inventory / storage capacity.

[0088] The flexibility profiling unit enables unified modeling and quantitative evaluation of system flexibility resources. This unit transforms the physical operational constraints of various heterogeneous resources into standardized capability curves over continuous time, thereby integrating dispersed and abstract flexibility parameters into system-level callable and continuously adjustable capabilities, providing a unified flexibility resource model for collaborative optimization across multiple time scales.

[0089] The safety status identification unit is used to combine the power flow calculation results and the real-time load level of the cross section in the grid-side monitoring data set to detect the safety constraints of the N-1 cross section, the capacity boundaries of critical lines and transformers, the upper and lower limits of DC power transmission and reception, and the voltage and reactive power operating range hourly, forming a safety constraint set. The safety constraint set includes constraint type, constraint boundary value and constraint activity index, and is output to the source-grid joint construction module.

[0090] The safety status identification unit enables accurate perception and dynamic boundary extraction of the real-time safety status of the power grid. Based on real-time grid-side data and power flow calculations, this unit dynamically identifies and quantifies the set of key constraints affecting system safety. By outputting a set of safety constraints that includes constraint boundaries and their activity indicators, this unit provides clear, dynamic, and prioritized safety boundary conditions for the source-grid joint construction module, ensuring that all subsequent optimization schemes are built on a reliable safety foundation.

[0091] The predictive state identification module enables quantitative perception of the future system operating state and dynamic identification of safety boundaries. Its core effect lies in transforming uncertain predictive results, dispersed flexibility characteristics, and complex power grid safety constraints into standardized, structured datasets that can be directly utilized by optimization models. This provides accurate input and clear boundaries for subsequent collaborative optimization and safety constraints, laying the foundation for the forward-looking and safe nature of system decision-making.

[0092] The source-network joint construction module includes a source-side feasible domain generation unit, a network-side feasible domain generation unit, a domain coupling scaling unit, and a feasibility verification unit.

[0093] The source-side feasible region generation unit is used to generate the source-side feasible region based on the flexibility curve, unit start-up and shutdown sequence and energy storage SOC boundary on a unified time sequence. In the generation process, the ramp rate constraint, minimum start-up and shutdown time constraint and energy storage operating condition constraint are mapped to time series boundary conditions to form a set of source-side feasible regions.

[0094] The processing logic of the source-side feasible region generation unit is as follows:

[0095] When constructing the feasible domain on the source side, the minimum start-up time, minimum downtime, and corresponding start-up and shutdown states of the unit are written into the time-coupled constraint sequence.

[0096] The unit's ramp-up and ramp-down rates are written into the power variation boundaries of adjacent intervals to limit the power variation range within consecutive time intervals.

[0097] The upper and lower bounds of the SOC of the energy storage unit and the boundary of the charge and discharge power are written into the SOC evolution boundary set to keep the SOC within a safe range.

[0098] Write the operating condition transition constraints of the pumped storage unit into the operating condition transition table, define the switching conditions and minimum hold time for the three types of operating conditions: pumping, power generation and shutdown, and generate the source-side feasible region set.

[0099] The source-side feasible domain generation unit enables refined and time-series modeling of generation-side flexibility resources. This unit transforms the complex physical constraints of various generating units and energy storage systems into a set of power feasible ranges across the entire time series. Its output is no longer an isolated parameter, but rather represents the set of all possible, safe, and executable power points for all power sources at every future moment, providing a precise decision space for market optimization.

[0100] The grid-side feasible region generation unit is used to combine the power flow balance equation, the cross-sectional transmission capacity limit, the upper and lower limits of DC transmission and receiving ends, and the voltage and reactive power operating range in the safety constraint set to establish the grid-side feasible region that satisfies the boundary conditions of the distribution network and the transmission network on an hourly basis, and output information including the node voltage constraint status and the cross-sectional load margin.

[0101] The processing logic of the network-side feasible region generation unit is as follows:

[0102] When generating the feasible region on the network side, a node power balance boundary is established for each time interval based on the set of security constraints.

[0103] Set a line thermal stability boundary for each transmission line.

[0104] Set the cross-sectional transport capacity boundary at the critical cross-section.

[0105] Establish the sending and receiving end boundaries for the DC converter station.

[0106] Write the allowable voltage range and reactive power support range hourly to generate a set of feasible regions on the grid side, and output constraint indicators, which include node voltage constraint status, cross-sectional load margin and DC power boundary information.

[0107] The grid-side feasible region generation unit provides a dynamic and quantitative description of the power grid's safe operating space. This unit combines static grid topology and parameters with dynamic real-time operating conditions, transforming complex power flow equations and safety rules into a full-time series set of transmission capacity boundaries. Its output clearly defines the limit of power that the grid can safely transmit at each moment, along with margin information, providing explicit network boundary conditions for market optimization.

[0108] The domain coupling scaling unit is used to perform set intersection on the source-side feasible domain and the network-side feasible domain on a unified time axis to obtain the source-network joint feasible domain, and tighten the boundary of the joint feasible domain hourly according to the constraint activity index and risk coefficient.

[0109] The feasibility verification unit is used to match and compare the salable power segments corresponding to the market quotation dataset with the joint feasible domain of the source network hourly. If the quotation segment is found to exceed the joint boundary, the salable power segment is marked as unrealizable and removed.

[0110] The processing logic for the domain coupling scaling unit and the realizability verification unit is as follows:

[0111] On a unified timeline, perform set intersection on the source-side feasible region and the network-side feasible region to obtain the initial source-network joint feasible region.

[0112] Risk coefficients are assigned to each time interval based on the activity of each constraint in the security constraint set, and these risk coefficients are mapped to the boundary tightening magnitude, thereby tightening the joint feasible region of source network hourly.

[0113] Specifically, the range of the risk coefficient ρ is obtained through simulation of historical accident scenarios. The system simulates N typical failure scenarios, and the probability of exceeding the limit P and the severity S of each constraint are statistically analyzed. Then, ρ... i =p i ×s i ;

[0114] The formula for calculating the boundary tightening range is:

[0115]

[0116] Where α is the global risk aversion coefficient (e.g., 0.1), which can be set by the scheduler.

[0117] In the tightened source-network joint feasible domain, the salable power segments corresponding to the market quotation dataset are compared with the joint feasible domain hourly. If the salable power segment exceeds the joint boundary, it is marked as unrealizable and removed, or adjusted to the effective range within the joint boundary.

[0118] Output source network joint feasible domain.

[0119] The domain coupling scaling unit achieves precise coupling and risk control of the security spaces on both the source and network sides. This unit uses set intersection operations to identify the common portion of the feasible regions on both the source and network sides, i.e., the globally feasible decision space of the system. Furthermore, it dynamically tightens the boundaries based on risk preferences, generating a compressed feasible region that incorporates uncertainty and defensive requirements. This process is a core step in ensuring the system's safe and stable operation under anticipated incidents, significantly improving the robustness of decision-making.

[0120] The feasibility verification unit performs pre-emptive verification and filtering of market bids and their physical feasibility. Before market clearing, this unit matches the bidding behavior of market participants with the physical feasible domain, identifying and eliminating bids that cannot be executed due to technical constraints or network congestion. This ensures that all bids entering the market model are executable, greatly improving the actual feasibility of the market clearing result and reducing the workload of post-event scheduling and adjustments, as well as market disputes.

[0121] The source-grid joint construction module achieves deep integration and unified modeling of the physical operating constraints of the power system and the scope of market transactions. Its core effect lies in aggregating the dispersed source-side regulation capabilities and grid-side safety boundaries into a unified, precise, and executable source-grid joint feasible domain through mathematical set operations. This feasible domain rigorously defines the safe operating space for market transactions and dispatch operations, mathematically ensuring that all subsequent market clearing results are both economical and secure. It fundamentally avoids the production of unrealistic trading plans and serves as a crucial bridge connecting predictive sensing, market decision-making, and physical execution.

[0122] The market scheduling coupling module includes a multi-product joint clearing unit, a network-endogenous pricing unit, and a trajectory generation unit.

[0123] The multi-product joint clearing unit is used to allocate resources simultaneously according to a unified optimization model within the source-network joint feasible domain that has undergone risk tightening and consistency verification, based on energy demand, reserve demand, ramp-up demand and non-functional capacity demand, and generate time-sharing and product-sharing clearing results.

[0124] The multi-product joint clearing unit achieves coordinated optimization and unified allocation of ancillary services and energy commodities across multiple time scales. Within a unified mathematical optimization framework, this unit comprehensively considers multiple objectives such as energy balance, reserve capacity, ramp-up capability, and reactive power demand, avoiding resource conflicts and optimization benefit losses that might result from sequential clearing. Its output, while ensuring the safe and reliable operation of the system, minimizes the total electricity purchase cost for society or maximizes social welfare, significantly improving the overall economic efficiency of the market.

[0125] The network-endogenous pricing unit is used to calculate shadow prices based on the activity constraints of cross sections and nodes during the clearing process, and outputs node electricity prices and cross section congestion shadow values.

[0126] Specifically, the logic for solving shadow prices includes:

[0127] After solving the source-network collaborative optimization model, the Lagrange dual variables of the optimization problem are obtained. Among them, the dual variable corresponding to the node power balance constraint is the node marginal electricity price, and the dual variable corresponding to the line power flow constraint is the section congestion shadow value.

[0128] Quantification standard for cross-sectional congestion shadow value:

[0129] When the power flow of a certain line reaches its limit, and this constraint is an "active constraint" at the optimal solution, its corresponding dual variable is the non-zero congestion shadow value. The physical meaning of this value is the marginal cost that the entire system needs to increase in order to satisfy the safety constraint of this line.

[0130] The network-endogenous pricing unit enables spatial electricity price discovery that accurately reflects the degree of grid congestion and marginal costs. Based on the Lagrange multipliers in the optimization model, this unit automatically calculates the marginal electricity price and congestion costs for different nodes and sections. These price signals accurately reveal the degree of resource scarcity and congestion at different locations in the grid, providing transparent and fair economic guidance for congestion management, financial transmission rights, and the investment and consumption behavior of market participants.

[0131] The trajectory generation unit is used to convert the cleared power of time-sharing products into a time-series power trajectory covering the settlement period based on the clearing results, and to generate a backup reserved trajectory and a reactive power reserved trajectory. In the trajectory, a tracking error threshold and tolerance rules are added to each time interval.

[0132] The trajectory generation unit transforms discrete market clearing results into a continuous, executable sequence of scheduling and control commands. This unit converts abstract market products such as time-segmented cleared electricity and reserve capacity into continuous time-series power commands and reserved trajectories covering the entire settlement cycle, and attaches error control thresholds and tolerance rules required for execution. This transformation process provides directly deployable control objectives with clear quality requirements for the next level of command execution and real-time monitoring, ensuring a smooth transition from market transaction results to physical execution.

[0133] The market dispatch coupling module achieves coordinated optimization and precise pricing of multiple types of power resources under strict physical security constraints. Its core effect lies in the fact that, based on the unified security decision space of the source-grid joint feasible domain, it synchronously clears various products such as energy, reserves, and ramp-up resources through a unified optimization model, and endogenously generates precise spatial price signals. This module ensures that the market allocation results theoretically possess global optimality, physical feasibility, and financial settlement basis, serving as the command center connecting forward-looking planning and real-time operation.

[0134] The instruction decomposition and monitoring module includes an instruction decomposition unit, a boundary issuance unit, and an execution monitoring unit.

[0135] The instruction decomposition unit, upon receiving the timing power trajectory, the reserved backup trajectory, and the reactive power reserved trajectory, combines the corresponding tolerance rules and tracking error thresholds to convert the above trajectories hourly into setting instructions that the equipment can directly execute:

[0136] Generate AGC settings for the time-series power trajectory. In the AGC settings, specify the target power value, ramp time period, tolerance range, and effective sampling period.

[0137] Convert the reactive power reserved trajectory into reactive power voltage setting. The reactive power voltage setting includes the target voltage value or reactive power value and its tolerance range, and marks the corresponding control object.

[0138] The energy storage-related trajectory is converted into a SOC target sequence, which specifies the SOC target level, charging and discharging direction, and power limitation conditions.

[0139] The relevant trajectories of pumped storage are converted into operating condition sequences, which clearly define the switching conditions and minimum holding time for pumping, power generation, and shutdown.

[0140] The AGC settings, reactive voltage settings, and SOC target sequences include timestamps and effective window information.

[0141] The instruction decomposition unit achieves the standardization and refinement of the transformation from optimization results to executable control instructions. This unit decomposes the unified time-series power trajectory, based on the control characteristics of different resource types, and generates personalized instruction sets with clearly defined control objectives, tolerance ranges, and timeliness information. This process ensures that the high-level decision-making intent can be unambiguously understood and executed by the low-level control equipment, improving the accuracy and reliability of control.

[0142] The boundary dispatch unit is used to generate transmission boundary instructions and transmission capacity curves based on the cross-section congestion shadow and the distribution network transmission boundary. The transmission boundary instructions include cross-section or line identification, allowable power flow direction and upper limit, temporary derating factor and effective window conditions. The transmission capacity curves include the upper limit of transmission and the lower limit of reactive power support for each time period of the distribution network feeder.

[0143] Specifically, the formula for calculating the temporary reduction factor is as follows:

[0144]

[0145] Where P available P is the current actual available output. forecast_error To account for positive reserve requirements after taking into account prediction errors.

[0146] Specifically, the rules for determining the effective window conditions are as follows:

[0147] Time condition: Within the warning period issued by the system (e.g., issued 2 hours in advance).

[0148] Space conditions: The unit is located within a predefined affected electrical area.

[0149] Status conditions: The unit is currently in grid-connected operation and there are no trip alarms.

[0150] The boundary issuing unit issues instructions to the dispatch master station, distribution network automation or field control terminal, and records the instruction receipt status. Instructions that are not received or rejected are marked as abnormal and the original boundary settings are retained.

[0151] The boundary dispatch unit enables proactive management and coordinated control of the power grid's safe operation boundaries. This unit transforms the grid congestion information revealed during market clearing into specific transmission boundary commands and distribution network transmission capacity curves, and proactively dispatches them to relevant control systems. This is equivalent to defining real-time, dynamic safe operation boundaries for each component of the power grid, achieving coordinated safety constraints between the main grid and distribution networks, and between dispatching and power stations, thus moving preventative control forward and effectively reducing safety risks.

[0152] The execution monitoring unit is used to read real-time data from the execution status dataset according to a preset sampling period. The real-time data includes active power, reactive power, energy storage SOC, pumped storage operating conditions, node voltage and cross-sectional power. It compares each item with the corresponding trajectory and boundary instructions to generate execution deviation values ​​and durations. According to the tolerance rules, it is marked as normal, alarm or out of bounds to form an execution deviation record. The record includes timestamp, equipment identification, target value, measured value, deviation amount and cause label.

[0153] When the execution deviation continuously exceeds the tracking error threshold or an out-of-bounds trend occurs, an out-of-bounds trend flag is output and the record is written to the unified data pool.

[0154] The execution monitoring unit enables real-time measurement, evaluation, and deviation early warning of the instruction execution process. This unit collects actual operational data at high frequency and compares it in real-time with expected instructions and trajectories to quantitatively assess execution deviations and mark status according to rules. Its output forms a complete execution deviation record, providing accurate and objective data for subsequent closed-loop evaluation, responsibility determination, and determining whether rolling optimization or emergency control is needed. This is a crucial guarantee for achieving knowable, controllable, and adjustable system status.

[0155] The instruction decomposition and monitoring module accurately translates macro-level market clearing results into micro-level equipment control instructions and monitors their execution throughout the entire process. Its core effect lies in decomposing the abstract power curve generated by optimization calculations into specific control instructions that can be directly executed by various resources, with timeliness and tolerance ranges. Through real-time monitoring and comparison of the execution status, it forms a precise perception of the system's actual operating status and provides early warnings of deviations, ensuring that market transaction results are accurately and reliably implemented, thus forming a key link in the closed-loop process.

[0156] The closed-loop assessment and handling module includes a deviation diagnosis unit, a rolling reconstruction unit, and a behavior compliance unit.

[0157] The deviation diagnosis unit is used to analyze the execution deviation records item by item, classify the deviations into prediction errors, equipment limitations and network mutations, and locate the corresponding links at the same time as the classification.

[0158] The deviation diagnosis unit enables intelligent attribution and precise location of system execution deviations. By establishing multi-dimensional diagnostic rules, this unit can effectively distinguish whether the root cause of the deviation stems from inaccurate predictions, physical limitations of equipment, or sudden power grid conditions, and accurately pinpoint the specific link. This refined diagnosis provides a basis for subsequent targeted handling decisions, avoids blind adjustments, and improves the system's intelligence level in responding to abnormal operating conditions.

[0159] The rolling reconfiguration unit is used to trigger a local reconfiguration process when the execution deviation continuously exceeds the tracking error threshold or an out-of-bounds trend occurs. It performs local recalculation of the affected time period and constraints, and calls the emergency re-clearing function when necessary to output a new time-series power trajectory and reserve and reactive power configuration.

[0160] The rolling reconfiguration unit enables dynamic adjustment and forward-looking correction of system optimization decisions. When it detects persistent deviations or out-of-bounds trends, this unit proactively triggers a re-optimization process, rollingly revising future plans based on the latest actual conditions. This online reconfiguration capability allows the system to quickly adapt to changes in actual conditions, promptly eliminate the risk of accumulated deviations, and significantly improve the real-time adaptability of the scheduling plan and the overall robustness of the system.

[0161] The behavioral compliance unit is used to continuously assess the consistency between the market entity's quotations and performance. The assessment methods include comparing the reported power segments with the actual executed time-series power trajectories one by one, statistically analyzing the frequency and magnitude of deviations, and adjusting the market entity's reporting scope or credit score based on the frequency and magnitude of deviations, and recording inconsistent behaviors in the compliance record database.

[0162] The behavioral compliance unit enables quantitative supervision and credit management of market participants' transaction behavior and performance capabilities. This unit establishes an objective and impartial credit evaluation system by continuously comparing bids and performance, and dynamically adjusts the scope of eligible market participants accordingly. This mechanism effectively constrains speculative behavior, ensures market fairness, and incentivizes market participants to fulfill their obligations in good faith through a credit score mechanism, thus promoting the healthy and orderly development of the market.

[0163] The closed-loop assessment and handling module drives the system to perform rolling optimization and strategy adjustment through intelligent diagnosis and root cause analysis of execution deviations. At the same time, it conducts compliance supervision of market entities' behavior, ultimately achieving continuous improvement in system operation efficiency and standardized and orderly market order, ensuring that the entire collaborative optimization system has dynamic adaptability and continuous evolution capabilities.

[0164] The power market source-grid coordinated optimization system proposed in this invention brings significant benefits through its unique modular design and coordinated processing logic:

[0165] The predictive state identification module generates prediction intervals and flexibility profiles, quantifying source-load uncertainty and providing risk-aware boundary conditions for optimization decisions. The source-grid joint construction module integrates grid security constraints into the market optimization model in the form of feasible regions, achieving market clearing under security constraints and fundamentally preventing the generation of unsafe solutions. The domain coupling scaling unit dynamically tightens the feasible region based on constraint activity and risk coefficients, enhancing the system's ability to defend against potential risks. The market scheduling coupling module performs joint clearing of multiple products within the joint feasible region, realizing value discovery and optimal allocation of various resources and improving overall market efficiency. The network-endogenous pricing unit calculates node electricity prices and congestion costs based on activity constraints, providing more accurate price signals, effectively guiding resource flow, alleviating network congestion, and reducing overall social electricity costs. The instruction decomposition and monitoring module tracks the macroscopic clearing power trajectory. The traces are transformed into precise, device-level executable instructions, along with tolerance rules, greatly improving the executability and control accuracy of the instructions. The boundary-issuing unit actively issues transmission boundaries and uplink capacity curves, providing clear operational guidelines for distribution networks and field control, achieving coordinated control of source, grid, load, and storage. From the source of data acquisition and calibration, it ensures the spatiotemporal consistency and high quality of multi-source heterogeneous data, providing a reliable data foundation for optimized decision-making. The execution monitoring unit tracks instruction execution deviations in real time, and the closed-loop evaluation and handling module intelligently diagnoses and continuously reconstructs deviations, enabling the system to learn online and self-correct, exhibiting strong adaptability to prediction errors and emergencies. The behavior compliance unit continuously evaluates the consistency between market participants' quotations and execution, constraining speculative behavior through mechanisms such as credit scores and reporting ranges, ensuring market fairness, incentivizing honest participation from market participants, and maintaining a healthy market order.

[0166] Specifically, this includes the rules for distinguishing between prediction errors, equipment limitations, and network mutations.

[0167] Prediction error dominates: Most resources within the system simultaneously show the same direction of deviation, and the power flow of the main power grid line does not exceed the limit.

[0168] Equipment-constrained dominance: Deviations are concentrated in a specific unit or load, and the telemetry data of that equipment (such as valve opening degree, circuit breaker status) indicates abnormalities.

[0169] Network mutation-driven: A line or transformer in the power grid trips (switch position signal), causing drastic and structural changes in power flow distribution.

[0170] The algorithm logic for local recalculation:

[0171] Once the disturbance source is identified, the system initiates a partial recalculation:

[0172] Constraint adjustment: Remove power flow constraints associated with tripped devices and add security constraints under the new network topology.

[0173] Fixed decision variables: The planned output of units that are unaffected and cannot be adjusted is fixed at the current value.

[0174] Optimization scope definition: Narrow the optimization scope to a local area that is directly connected electrically, and only re-optimize the resource output that can still be adjusted within this area.

[0175] Fast solution: Using a simplified model, this local optimization problem can be solved quickly, generating a corrective plan.

[0176] The above embodiments are merely descriptions of preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made by those skilled in the art to the technical solutions of the present invention without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A source-grid coordinated optimization system for the electricity market, characterized in that, include: The data acquisition module is used to collect source-side operational data, network-side monitoring data, market quotation data, and execution status data, and to perform spatiotemporal alignment and quality verification, and store them in a unified data pool; The state prediction module is used to make predictions based on data in the data pool and generate a set of prediction intervals, flexibility curves, and a set of safety constraints. The source-network joint construction module is used to generate the source-network joint feasible region based on the prediction interval set, flexibility curve and security constraint set and to complete the realizability verification. The market scheduling coupling module is used to complete the joint clearing of multiple products and generate time-series power trajectories and price signals in the source-network joint feasible domain. The instruction decomposition and monitoring module is used to translate the time-series power trajectory and price signal into executable instructions for the device and monitor the execution deviation in real time. The closed-loop evaluation and correction module is used to diagnose and continuously correct execution deviations.

2. The power market source-grid coordinated optimization system according to claim 1, characterized in that, The data acquisition module includes: The spatiotemporal alignment unit is used to perform cross-source comparison and unify source-side operational data, network-side monitoring data, market quotation data, and execution status data to the same time base, and adds alignment method identifiers and missing data markers; The quality verification unit is used to verify the source-side operation data, network-side monitoring data, market quotation data, and execution status data according to preset boundary rules. The unified data pool unit is used to classify and store verified source-side operational data, network-side monitoring data, market quotation data, and execution status data, and to add maintenance version numbers, data time window indexes, and data integrity verification codes.

3. The power market source-grid coordinated optimization system according to claim 2, characterized in that, The verification process includes identifying abrupt increases and decreases according to mutation rules, identifying delays or packet loss according to communication delay rules, generating trusted data labels for qualified data, and attaching anomaly type labels to unqualified data.

4. The power market source-grid coordinated optimization system according to claim 1, characterized in that, The state prediction module includes: The load and renewable energy forecasting unit is used to generate a set of time-segmented forecast intervals with upper and lower boundaries and confidence labels based on data in the data pool, combined with historical data and unit operation and maintenance plans. The flexibility profiling unit is used to convert unit parameters into flexibility curves. The safety status identification unit is used to generate a safety constraint set by hourly detection in the network-side monitoring dataset, combining the power flow calculation results and the real-time load level of the cross section.

5. The power market source-grid coordinated optimization system according to claim 4, characterized in that, The flexibility curves include the climbing ability curve, the endurance curve, and the non-functionality curve in the continuous time dimension.

6. The power market source-grid coordinated optimization system according to claim 1, characterized in that, The source-network joint construction module includes: The source-side feasible region generation unit is used to generate the source-side feasible region based on the flexibility curve on a unified time series, and in the generation process, the ramp rate constraint, minimum start-up and shutdown time constraint and energy storage operating condition constraint are mapped into time series boundary conditions to form a set of source-side feasible regions. The grid-side feasible region generation unit is used to combine the power flow balance equation, the cross-sectional transmission capacity limit, the upper and lower limits of DC transmission and receiving ends, and the voltage and reactive power operating range in the safety constraint set to establish a set of grid-side feasible regions that meet the boundary conditions of the distribution network and the transmission network on an hourly basis, and output constraint indicators. The domain coupling scaling unit is used to perform set intersection on the source-side feasible domain set and the network-side feasible domain set on a unified time axis to obtain the source-network joint feasible domain, and to tighten the boundary of the source-network joint feasible domain hourly according to the constraint activity index and risk coefficient. The feasibility verification unit is used to match and compare the salable power segments corresponding to the market quotation dataset with the source-network joint feasible domain hourly, and output the source-network joint feasible domain that meets the threshold.

7. The power market source-grid coordinated optimization system according to claim 6, characterized in that, After performing hourly matching and comparison of the reportable power segments corresponding to the market quotation dataset with the source-network joint feasible domain, the method further includes: If the reportable power segment does not meet the threshold, the reportable power segment is marked as unrealizable and removed, or adjusted to the effective range within the joint boundary.

8. The power market source-grid coordinated optimization system according to claim 1, characterized in that, The market scheduling coupling module includes: The multi-product joint clearing unit is used to simultaneously allocate resources through a unified optimization model based on actual needs within the source-network joint feasible domain, and generate time-sharing and product-sharing clearing results. The network-endogenous pricing unit is used to calculate shadow prices based on the activity constraints of cross sections and nodes during the process of generating clearing results, and output node electricity prices and cross section congestion shadow values. The trajectory generation unit is used to convert the cleared power of time-sharing products into a time-series power trajectory covering the settlement period based on the clearing results, and to generate a backup reserved trajectory and a reactive power reserved trajectory, and to add a tracking error threshold and tolerance rules to each time interval in the trajectory.

9. The power market source-grid coordinated optimization system according to claim 1, characterized in that, The instruction decomposition and monitoring module includes: The instruction decomposition unit, after receiving the timing power trajectory, combines the corresponding tolerance rules and tracking error threshold to convert the timing power trajectory into setting instructions that the device can directly execute: The boundary dispatch unit is used to generate transmission boundary commands and uplink capacity curves based on the cross-section congestion shadow and the distribution network transmission boundary; The execution monitoring unit is used to read real-time data from the execution status dataset according to a preset sampling period, compare it with the corresponding trajectory and boundary instructions, generate execution deviation values ​​and durations, and mark them according to tolerance rules to form execution deviation records.

10. The power market source-grid coordinated optimization system according to claim 1, characterized in that, The closed-loop evaluation and correction module includes: The deviation diagnosis unit is used to analyze the execution deviation records item by item, classify the deviations into prediction errors, equipment limitations and network mutations, and locate the corresponding links while classifying them. The rolling reconfiguration unit is used to trigger a local reconfiguration process when the execution deviation continuously exceeds the tracking error threshold or an out-of-bounds trend occurs. It performs local recalculation of the affected time period and constraints, calls the emergency re-clearing function, and outputs a new time-series power trajectory and reserve and reactive power configuration. The Behavioral Compliance Unit is used to continuously assess the consistency between market participants' pricing and performance.

Images