Electric power spot market annual security constraint unit combination rapid clearing method and system

Abstract

The invention is suitable for the field of electric power system operation and market clearing, and provides an electric power spot market annual security constraint unit combination rapid clearing method and system, and the method comprises the following steps: receiving input data; constructing an annual SCUC optimization problem based on the input data; the annual SCUC optimization problem is solved by adopting a two-stage calculation process, in the first stage, an initial solution is calculated by simplifying network topology and reducing time resolution, and in the second stage, the annual optimization problem is decomposed into a plurality of subtasks for parallel rolling calculation; according to the method, annual market and system data are received, an annual SCUC problem is constructed, a special solving method is applied, the optimal or near-optimal power generation plan and market clearing price in the annual period are determined in an actual time frame, more effective long-term planning is facilitated, and the method is suitable for popularization and application. The market efficiency is improved, and reliable operation of a power system is supported.

Description

Technical Field

[0001] The present invention relates to the field of power system operation and market clearing, and in particular to a method and system for rapid clearing of annual safety-constrained unit combinations in a power spot market. Background Art

[0002] Electricity spot markets rely on unit commitment (UC) to determine the optimal dispatch of generators to meet forecasted demand while satisfying various operational and security constraints. Security-constrained unit commitment (SCUC) extends UC by incorporating grid security constraints (e.g., transmission line power flow restrictions) to ensure system reliability. Performing SCUC on an annual timescale (8760 hours, or at finer temporal resolutions such as 15 minutes) creates a large and complex mixed-integer optimization problem.

[0003] Current methods for directly calculating such large-scale SCUC problems are often computationally intensive and time-consuming. This hinders analytical decision-making in the power market, reduces market research efficiency, and makes it difficult to simulate long-term spot electricity prices and unit commitment problems.

[0004] Existing methods may involve simplification, decomposition techniques, or heuristic algorithms, which may sacrifice optimality or accuracy for tractability. Even iterative calculations on a daily basis still require a long cycle to calculate the full year's results. There is an urgent need for a system that can effectively solve the annual SCUC problem without a significant loss of accuracy. Summary of the Invention

[0005] The purpose of the present invention is to provide a method and system for rapid clearing of annual safety-constrained unit combinations in the electricity spot market, aiming to solve the problems in the above-mentioned background technology.

[0006] The present invention is implemented as follows: a method and system for rapid clearing of annual safety-constrained unit combinations in the electricity spot market, the method comprising the following steps: Receive input data; Constructing an annual SCUC optimization problem based on the input data; A two-stage computational process is used to solve the annual SCUC optimization problem. In the first stage, the initial solution is calculated by simplifying the network topology and reducing the time resolution. In the second stage, the annual optimization problem is decomposed into multiple subtasks for parallel rolling calculation. Output power generation plan, power curve of each generator unit and node electricity price results.

[0007] As a further solution of the present invention: the input data includes generator set parameters, bus parameters, transmission line parameters, renewable energy parameters, load parameters, generator set declared price, renewable energy power generation forecast time series data, load forecast time series data and backup demand time series data.

[0008] As a further solution of the present invention, the first stage of simplifying the network topology and reducing the time resolution to calculate the initial solution specifically includes: The K-means algorithm is used to cluster and merge the generator sets, simplifying the complex network topology into a single bus model. The K-means algorithm formula is: Where: K is the number of clusters, C k is the set of k-th clusters, μ k is the center of the kth cluster and is calculated as the mean of the eigenvectors of all units in the cluster: Where: |C k ∣ is the number of units in cluster k, J is the objective function, which represents the sum of squared intra-cluster distances of all clusters; Superimpose the time series data of renewable energy output, load, external power and external power transmission into a single entity; Reduce the time resolution to a preset interval and perform rolling calculations to obtain an initial solution.

[0009] As a further solution of the present invention: the initial solution includes the power, state and continuous start and stop time of the generator set, and the initial solution of the aggregated generator set is distributed to a single unit.

[0010] As a further solution of the present invention: the second stage decomposes the annual optimization problem into multiple subtasks and performs parallel rolling calculations, specifically including: The full year data is broken down into multiple time periods based on the number of CPU cores; Each time period is assigned to an independent CPU core and a rolling calculation is performed based on the initial solution calculated in the first phase; Each subtask is solved on a daily basis using the complete model.

[0011] As a further solution of the present invention: the time window of the rolling calculation is 2 days, and the SCUC problem of 48 time points is optimized each time.

[0012] As a further solution of the present invention: the method also includes accelerating the solution process of the subtask through multi-core parallel computing.

[0013] The annual safety-constrained unit portfolio rapid clearing system for the power spot market operates the aforementioned annual safety-constrained unit portfolio rapid clearing method for the power spot market, and the system comprises: An input module, for receiving input data; Data preprocessing module, used to validate and clean input data and build the annual SCUC optimization model; A core optimization engine that executes a two-stage computational process, including a simplified computational module in the first stage and a parallel rolling computational module in the second stage; The post-processing and output module is used to output power generation plans, power curves and node electricity price results.

[0014] As a further solution of the present invention: the first-stage simplified calculation module is specifically used to: Cluster and merge the generators to simplify the network topology into a single bus model; reduce the time resolution and generate the initial solution through rolling calculation; The parallel rolling calculation module is specifically used for: The annual optimization problem is decomposed into multiple subtasks and assigned to multiple CPU cores for parallel computing. Each subtask is solved on a daily basis based on the initial solution.

[0015] As a further solution of the present invention: the system also includes a database module for storing input data, intermediate calculation results and final market clearing results.

[0016] Compared with the prior art, the present invention has the following beneficial effects: Significantly reduce calculation time: The calculation time for a full-year unit combination is reduced from tens of hours or days to just a few hours; Improved accuracy and optimality: Compared with the most simplified single busbar model calculation method, the second stage still uses the full model for calculation, but the initial conditions are generated by the simplified single busbar method; Enhanced long-term planning capabilities: enabling more reliable resource sufficiency assessments and investment signals; Improved market efficiency: more accurate price signals over a longer period; Scalability: Ability to handle large-scale power systems and increasing complexity (the approach can further incorporate energy storage); This invention aims to provide a more efficient and faster clearing mechanism that can perform annual SCUC calculations within a practical time frame, improve the accuracy of long-term generation scheduling and resource assessment, improve market efficiency by providing more accurate long-term price signals, promote better long-term planning of generation and transmission investments, and conduct more comprehensive scenario analysis for future power system development. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0018] Figure 1 This is the framework diagram of the annual SCUC clearing system in the present invention.

[0019] Figure 2 This is a flowchart of the rolling calculation in the present invention.

[0020] Figure 3 This is a mapping diagram between initial conditions and subtasks in the present invention.

[0021] Figure 4 Schematic diagram of the time division of rolling calculation in the present invention. DETAILED DESCRIPTION

[0022] The technical solution of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0023] The present invention will be further explained below with reference to specific embodiments.

[0024] See also Figures 1-4 The embodiment of the present invention provides a method for rapid clearing of annual safety-constrained unit combinations in the power spot market, the method comprising the following steps: Receive input data; the input data includes generator set parameters, bus parameters, transmission line parameters, renewable energy parameters, load parameters, generator set declared price, renewable energy power generation forecast time series data, load forecast time series data and standby demand time series data; Based on the input data, the annual safety-constrained unit commitment (SCUC) optimization problem is constructed as follows: Sets: G: The set of all generators (g=1,2,...,|G|).

[0025] T: the set of all time periods (t=1,2,...,|T|), for example, T={1,2,...,24}.

[0026] L: The set of all transmission lines (l=1,2,...,|L|).

[0027] Decision Variables: u_gt: Binary variable. The start and stop status of unit g in time period t.

[0028] u_gt=1 means unit g is online during time period t.

[0029] u_gt=0 means that unit g is shut down (offline) during period t.

[0030] p_gt: Continuous variable. Active power output (MW) of unit g during period t. When u_gt = 0, p_gt must be 0.

[0031] v_gt: binary variable (optional, but commonly used). The startup status of unit g in time period t.

[0032] v_gt=1 means that unit g starts at the beginning of period t (i.e., from shutdown at t-1 to startup at t).

[0033] v_gt=0 means that unit g was not started during period t.

[0034] w_gt: binary variable (optional, but commonly used); the shutdown state of unit g in time period t (shut-down).

[0035] w_gt=1 means that unit g is shut down at the beginning of period t (i.e., from startup at t-1 to shutdown at t).

[0036] w_gt=0 means that unit g does not shut down during period t.

[0037] Parameters: D_t: total active load demand of the system in time period t (MW).

[0038] R_t: The total spinning reserve capacity required by the system in period t (MW).

[0039] P_min_g, P_max_g: Minimum and maximum active power output (MW) of unit g (when powered on).

[0040] SUC_g: startup cost of unit g.

[0041] SDC_g: The shutdown cost of unit g (usually smaller than the startup cost or zero).

[0042] C_g(P): The production cost function of unit g, which is usually approximated as a quadratic function: , where a_g, b_g, c_g are cost coefficients (c_g is the no-load cost, which is incurred as long as the machine is turned on).

[0043] RU_g, RD_g: Ramp-Up and Ramp-Down limits of unit g (MW / h).

[0044] UT_g, DT_g: Minimum uptime (MinimumUpTime) and minimum downtime (MinimumDownTime) of unit g (hours).

[0045] PTDF_lg: Power Transfer Distribution Factor. This represents the change in power flow (in MW) on line l caused by a 1 MW increase in the output of unit g. (This is a key parameter in the linearized DC power flow model.)

[0046] F_max_l: Maximum allowable active power flow capacity of line l (MW).

[0047] F_base_lt: Initial power flow (MW) on line l at time period t, resulting from the base load distribution and other fixed injections. (This can be assumed to be zero for simplicity, or included in the load term.)

[0048] InitialStatus_g: The status (hours on or hours off) of unit g at time t=0 (before the problem started).

[0049] Objective Function: Minimize total system costs, which mainly include startup costs, shutdown costs and power generation production costs.

[0050]

[0051] Constraints: System power balance constraint (PowerBalance): , the sum of all unit outputs must equal the total system load; System spinning reserve constraints (SpinningReserve):

[0052] (or equivalent ); The total output of all units that can be increased during period t must meet the reserve requirements; Unit output upper and lower limit constraints (GenerationLimits): The output of the startup unit is between the minimum and maximum technical outputs; the output of the shutdown unit is 0; Unit ramp rate constraints (RampRateLimits): (upward climbing limit); (Downward climbing limit); the change in unit output between two consecutive periods cannot exceed the climbing capacity; Minimum start / stop time constraint of the unit (MinimumUp / DownTime): (It must run for at least UT_g hours after being turned on); (The machine must be stopped for at least DT_g hours after being shut down); there are many forms of logical implementation of these constraints, the above is one of the common representations; Unit start-up / shut-down logic constraints (Start-up / Shut-down Logic): (state change relationship); (Cannot start and stop at the same time); (Ensure that v_gt and w_gt accurately reflect state changes ); Security Constraints-Line Flow: Base Case-N: Ensures that the line is not overloaded under normal operating conditions.

[0053] ; Actual current = base current + current change caused by unit injection; Initial Conditions: It is necessary to set the states of u_g0, p_g0, etc. at time t = 0 according to InitialStatus_g, and ensure that the minimum start-up and shutdown time and ramp-up constraints are met at t = 1. This is usually achieved by fixing the variable values ​​at t = 0 or adding constraints specific to t = 1.

[0054] A two-stage computational process is used to solve the annual SCUC optimization problem. In the first stage, the initial solution is calculated by simplifying the network topology and reducing the temporal resolution. In the second stage, the annual optimization problem is decomposed into multiple subtasks for parallel rolling calculation. The first stage of simplifying the network topology and reducing the temporal resolution to calculate the initial solution specifically includes: The K-means algorithm is used to cluster and merge the generator sets, simplifying the complex network topology into a single busbar model. The details are as follows: To cluster unit data using the k-means algorithm, you need to represent each unit as a feature vector and define a distance metric (usually Euclidean distance) and the k-means objective function. The specific steps include: Step 1: Define the eigenvector of the unit Each unit i is represented by a d-dimensional feature vector Geographic location: longitude (loni) and latitude (lati), as two independent features, installed capacity (capi, in MW), minimum output (in MW), startup and shutdown time parameters: minimum operating time (in hours) and minimum downtime (in hours), marginal cost (in yuan / MWh); Therefore, the feature vector is x_ik, i=1,2,…,n, where n is the total number of units; d=7 (adjust d if features are added or deleted).

[0055] Feature standardization: Due to the different units and scales of features (e.g., a small range of longitude / latitude and a large range of installed capacity), k-means is sensitive to scale and each feature must be standardized before clustering. Commonly used z-score standardization: where x ik is the kth characteristic of the i-th unit, μ k is the mean of the kth feature of all units, σ k is the standard deviation. After standardization, all features have a mean of 0 and a variance of 1, ensuring fair contribution.

[0056] Step 2: Define the distance metric K-means uses Euclidean distance to measure the similarity between units. The distance between two units i and j is: Where: x i and x j is the normalized feature vector (the distance calculation is more reasonable after normalization).

[0057] The smaller the distance, the more similar the units are and they should be assigned to the same cluster.

[0058] Step 3: Mathematical expression of k-means algorithm Let K be the number of clusters (the number of categories, which must be specified in advance). The goal of k-means is to minimize the within-cluster sum of squares (WCSS): Where: C k is the set of the kth cluster (including the unit index assigned to the cluster), μ kis the center (centroid) of the kth cluster, calculated as the mean of the eigenvectors of all units in the cluster: ∣C k ∣ is the number of units in cluster k, and J is the objective function, which represents the sum of squared intra-cluster distances of all clusters.

[0059] Algorithm steps (iterative optimization): Initialization: Randomly select K units as the initial cluster centers μ1, μ2,…, μK.

[0060] Assignment step: Assign each unit to the nearest cluster center.

[0061] Update step: recalculate the center of each cluster.

[0062] Repeat steps 2 and 3 until the cluster center no longer changes or the change in J is less than a threshold.

[0063] Superimpose the time series data of renewable energy output, load, external power and external power transmission into a single entity; Reduce the time resolution to a preset interval, perform rolling calculations, and obtain an initial solution; the initial solution includes the power, status, and continuous start and stop time of the generator set, and distribute the initial solution of the aggregated generator set to a single unit; specifically, include: The initial state of each unit group calculated in the first stage is assumed to be: 1, 0, 1, 1, 1 at the boundary of the second stage time boundary (t), 1, 1, 0, 0, 0, 1; The initial output of the unit group at the demarcation moment is p(t), and the initial power of each unit is p(t) / N, where N is the number of units in the unit group. Therefore, the initial state of all units in this unit group at this moment is 0, and the initial prior continuous downtime is the number of 0s before the demarcation moment multiplied by the time interval calculated by the first stage of time resolution reduction (for example, 3 hours). So it is 0 zeros multiplied by 3 hours, and the initial downtime is 0 hours. The initial continuous power-on time is the number of 1s before the demarcation time multiplied by the time interval for calculating the time resolution reduction in the first stage (for example, 3 hours). This is 3 1s multiplied by 3 hours, and the initial power-on time is 9 hours. Then the initial state of the unit is set to initial downtime 0 hours, initial start time 9 hours, and initial power p(t) / N; The second phase decomposes the annual optimization problem into multiple subtasks for parallel rolling calculation, specifically including: The full year data is broken down into multiple time periods based on the number of CPU cores; Each time period is assigned to an independent CPU core and a rolling calculation is performed based on the initial solution calculated in the first phase; Each subtask uses the complete model to solve the problem on a daily rolling basis. The rolling calculation window is 2 days, and the SCUC problem is optimized at 48 time points each time. Output power generation plan, power curve of each generator set and node electricity price results; The method also includes accelerating the solution process of the subtasks through multi-core parallel computing.

[0064] The embodiment of the present invention provides a system for rapid clearing of annual safety-constrained unit combinations in the power spot market, which runs the aforementioned method for rapid clearing of annual safety-constrained unit combinations in the power spot market. The system includes: An input module, for receiving input data; Data preprocessing module, used to validate and clean input data and build the annual SCUC optimization model; The core optimization engine is used to execute a two-stage calculation process, including a first-stage simplified calculation module and a second-stage parallel rolling calculation module; the first-stage simplified calculation module is specifically used to: Cluster and merge the generators to simplify the network topology into a single bus model; reduce the time resolution and generate the initial solution through rolling calculation; The parallel rolling calculation module is specifically used for: The annual optimization problem is decomposed into multiple subtasks and assigned to multiple CPU cores for parallel computing. Each subtask is solved daily based on the initial solution. Post-processing and output module, used to output power generation plan, power curve and node electricity price results; The system also includes a database module for storing input data, intermediate calculation results and final market clearing results.

[0065] In embodiments of the present invention, the present invention is dedicated to providing a more efficient and faster clearing mechanism that can perform annual SCUC calculations within a practical time frame, improve the accuracy of long-term generation scheduling and resource assessment, improve market efficiency by providing more accurate long-term price signals, promote better long-term planning of generation and transmission investments, and conduct more comprehensive scenario analysis for future power system development.

[0066] In this example, the main goal is to significantly reduce the computation time required to obtain optimal or near-optimal generation dispatch plans on an annual time scale while complying with all relevant safety and operational constraints.

[0067] A two-stage computational process: the first stage simplifies the network topology and reduces the resolution to obtain the initial conditions for the second stage's computational subtasks. The second stage decomposes the annual optimization problem into subtasks based on time. Each subtask is combined with the initial conditions and then computed in parallel on the CPU cores. The system achieves this goal through a novel combination of: generator group aggregation, simplified optimization model, reduced time resolution calculation to obtain the initial solution, full-year time window decomposition, and multi-core parallel computing: Generator unit aggregation: Based on parameters such as the generator unit's spatial location, installed capacity, power generation cost, unit type, and maximum and minimum start and stop times, the K-means algorithm is used to cluster and merge them, and they are equivalent to a unit that can be combined. For example, there are four 300MW generator units in the same location, with a minimum output power of 120MW, a maximum and minimum start and stop time of 12 hours, and a marginal cost of 330-350 yuan / MWh. These can be aggregated into a single 1200MW generator unit with a minimum output power of 480MW, a maximum and minimum start and stop time of 12 hours, and a marginal cost of 330-350 yuan / MWh.

[0068] Simplified optimization model: The original complex network topology is simplified to a single bus model. All clustered generators are connected to this bus. All wind farms, solar power, loads, incoming power, and outgoing power are connected to this bus as a single entity. For example, if the original network model has 10 wind farms, each with a full-year 8760 curve, the values ​​of the curves for these 10 wind farms are added together and treated as a single wind farm connected to the bus. Similarly, all time series data for solar power, loads, incoming power, and outgoing power are directly superimposed and treated as a single element connected to the bus.

[0069] Calculate the initial solution by reducing the time resolution: Reduce the time resolution of the simplified model to 6 hours (for example, 4 hours is also possible) for unit combination calculation. At this time, the number of time stamps for the whole year is 1460 (8760 / 6=1460). Each unit combination calculation uses 96 time stamps (for example, 48 timestamps are also possible, and 96 timestamps are equivalent to 24 days). 15 rolling calculations are required (1440 / 96).

[0070] The initial solution of the simplified model with reduced time resolution is obtained. The initial solution mainly includes the power of the generator set, the status of each unit, etc. The initial solution is redistributed to the status and power of each unit. For example, an aggregated generator set 1, 1200MW, consists of four 300MW generator sets. The output power at a certain moment is 1000MW, and the continuous startup time beforehand is 12 hours. Then the output power allocated to a single unit is 250MW, and the previous startup time is also 12 hours.

[0071] Full-year time window decomposition: The original detailed SCUC annual optimization clearing problem is decomposed into monthly optimization problems in multiple time windows. Each monthly optimization is calculated on a daily rolling basis. The initial value of each simulation cycle is the initial value obtained from the first calculation, including the start and stop status of the unit, the initial continuous start and stop time, etc.

[0072] Multi-core parallel computing: Assign a core to each decomposed time window computing problem of the detailed model for parallel computing, and solve the problem of the day in each rolling The system will enable market operators to: Perform annual SCUC calculations within a realistic time frame; Improve the accuracy of long-term generation scheduling and resource assessment; Improve market efficiency by providing more accurate long-term price signals; Facilitate better long-term planning of generation and transmission investments; Conduct more comprehensive scenario analysis for future power system development.

[0073] The present invention is described in detail as follows: 1. System architecture: Input module: Receives input data, including: generator set parameters (name, bus location, installed capacity, maximum and minimum power, up and down ramp rates, minimum start and stop time, maintenance plan), bus (bus name, voltage level), transmission line parameters (line name, transmission capacity, start and end bus), renewable energy parameters (name, bus location, installed capacity), load parameters (name, bus location), generator set declared price, renewable energy power generation forecast (time series data), load forecast (time series data), and backup demand (time series data).

[0074] Data preprocessing module: validates and cleans input data; builds a safety-constrained unit combination model.

[0075] Core optimization engine: used to execute a two-stage calculation process, including a simplified calculation module in the first stage and a parallel rolling calculation module in the second stage.

[0076] The first stage simplifies the calculation: cluster the unit information, merge renewable energy generation and load, equate all grid topology nodes to one node, reduce the time resolution to solve the annual unit combination, and obtain the initial results at the monthly time boundary, mainly including the unit's pre-startup and shutdown time, initial power, etc.

[0077] Annual SCUC Problem Modeling: Objective function: Minimize the total annual system operating cost.

[0078] Constraints: Power balance (power generation meets all load demands + backup requirements); generator set operating constraints (output upper and lower limits, ramp rate, minimum start and stop time, etc.); transmission network constraints (DC power flow, transmission capacity limit).

[0079] Second stage calculation: Full year time decomposition algorithm: The full year data is divided into 12 time periods based on the number of CPU cores; Obtain corresponding static data, time series data, new energy output data and load forecast data for each time period; The initial value of each time period is the initial value calculated in the first step of the hierarchical solution; In each time period, 48 points of SCUC problems were optimized for 2 days each time; How it breaks down annual problems into manageable sub-problems.

[0080] How to achieve coordination between sub-problems.

[0081] Post-processing and output module: Calculate the node marginal price (LMP) or other relevant market clearing prices.

[0082] Database module: stores input data, intermediate calculation results, and market results.

[0083] 2. Process: 2.1 Describe step by step how the system operates from data input to final output: Data input for full-year simulation; 2.2 The first stage of calculation: simplify the model and reduce the time resolution to obtain the initial solution; Simplified unit clustering; The network topology is simplified to a single bus; Reduced temporal resolution; Rolling SCUC calculation; Get the initial solution.

[0084] 2.3 The complete model is segmented into time periods throughout the year based on the number of CPU cores.

[0085] 2.4 Each time segment uses the initial solution obtained from the first stage of calculation as part of the input data.

[0086] 2.5 Second stage calculation: Each CPU core is assigned a rolling calculation task.

[0087] 2.6 Results merging.

[0088] 3. Implementation (illustrative scenario): The traditional method of calculating a unit combination problem of 100 generator sets, 300 nodes, and 150 lines takes 10 minutes of cyclic calculation per day, so the total time for a year of 365 days is 60 hours.

[0089] In the first phase of this method, a simplified model was computed. Using unit clustering, the number of generators was reduced from 100 to 25, and the 300-node network topology was reduced to a single bus. The time resolution was reduced from 1 hour to 6 hours per interval, with 15 calculation cycles, each lasting approximately 4 minutes, for a total of 60 minutes. The second phase involved distributed computing, assuming 12 cores were used for parallel computing, with each core performing 30 calculations, taking approximately 5 hours. The total computation time was 6 hours.

[0090] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for rapid clearing of annual safety-constrained unit portfolios in the electricity spot market, characterized by: The method comprises the following steps: Receive input data; Constructing an annual SCUC optimization problem based on the input data; A two-stage computational process is used to solve the annual SCUC optimization problem. In the first stage, the initial solution is calculated by simplifying the network topology and reducing the time resolution. In the second stage, the annual optimization problem is decomposed into multiple subtasks for parallel rolling calculation. Output power generation plan, power curve of each generator unit and node electricity price results.

2. The method for rapid clearing of annual safety-constrained unit commitments in the electricity spot market according to claim 1, characterized in that: The input data includes generator set parameters, bus parameters, transmission line parameters, renewable energy parameters, load parameters, generator set declared price, renewable energy power generation forecast time series data, load forecast time series data and standby demand time series data.

3. The method for rapid clearing of annual safety-constrained unit commitments in the electricity spot market according to claim 1, characterized in that: The first stage of simplifying the network topology and reducing the time resolution to calculate the initial solution specifically includes: The K-means algorithm is used to cluster and merge the generator sets, simplifying the complex network topology into a single bus model. The K-means algorithm formula is: Where: K is the number of clusters, C k is the set of k-th clusters, μ k is the center of the kth cluster and is calculated as the mean of the eigenvectors of all units in the cluster: Where: |C k ∣ is the number of units in cluster k, J is the objective function, which represents the sum of squared intra-cluster distances of all clusters; Superimpose the time series data of renewable energy output, load, external power and external power transmission into a single entity; Reduce the time resolution to a preset interval and perform rolling calculations to obtain an initial solution.

4. The method for rapid clearing of annual safety-constrained unit commitments in the electricity spot market according to claim 3 is characterized in that: The initial solution includes the power, status and continuous start and stop time of the generator set, and the initial solution of the aggregated generator set is distributed to a single unit.

5. The method for rapid clearing of annual safety-constrained unit commitments in the electricity spot market according to claim 4 is characterized in that: The second phase decomposes the annual optimization problem into multiple subtasks for parallel rolling calculation, specifically including: The full year data is broken down into multiple time periods based on the number of CPU cores; Each time period is assigned to an independent CPU core and a rolling calculation is performed based on the initial solution calculated in the first phase; Each subtask is solved on a daily basis using the complete model.

6. The method for rapid clearing of annual safety-constrained unit commitments in the electricity spot market according to claim 5, characterized in that: The time window of the rolling calculation is 2 days, and the SCUC problem of 48 time points is optimized each time.

7. The method for rapid clearing of annual safety-constrained unit commitments in the electricity spot market according to claim 6, characterized in that: The method also includes accelerating the solution process of the subtasks through multi-core parallel computing.

8. The annual safety-constrained unit portfolio rapid clearing system for the electricity spot market is characterized by: The method for rapid clearing of annual safety-constrained unit commitments in the electricity spot market according to claim 7 is executed, wherein the system comprises: An input module, for receiving input data; Data preprocessing module, used to validate and clean input data and build the annual SCUC optimization model; A core optimization engine that executes a two-stage computational process, including a simplified computational module in the first stage and a parallel rolling computational module in the second stage; The post-processing and output module is used to output power generation plans, power curves and node electricity price results.

9. The annual safety-constrained unit combination rapid clearing system for the electricity spot market according to claim 8 is characterized in that: The first stage simplified calculation module is specifically used for: Cluster and merge the generators to simplify the network topology into a single bus model; reduce the time resolution and generate the initial solution through rolling calculation; The parallel rolling calculation module is specifically used for: The annual optimization problem is decomposed into multiple subtasks and assigned to multiple CPU cores for parallel computing. Each subtask is solved on a daily basis based on the initial solution.

10. The annual safety-constrained unit combination rapid clearing system for the electricity spot market according to claim 8, characterized in that: The system also includes a database module for storing input data, intermediate calculation results and final market clearing results.

Images