A virtual power plant source and load scheduling method and system

Abstract

The application discloses a virtual power plant source-load scheduling method and system. By actively issuing standard test instructions to target resource points and analyzing response deviations, the application can effectively identify resource points with performance hidden attenuation, calculate quantitative performance attenuation coefficients, correct distorted reportable adjustable capacity data into reliable and credible adjustable capacity data, and enable resource points to effectively complete instructions, thereby ensuring the accuracy of scheduling decisions from the source.

Description

Technical Field

[0001] This invention relates to the technical field of power system operation control, specifically to a virtual power plant source-load dispatching method and system. Background Technology

[0002] The integration of distributed energy sources, represented by distributed photovoltaic and wind power, as well as user-side energy storage and flexible loads, is growing rapidly. The structure of the power system is shifting from the traditional centralized, unidirectional power supply model to a distributed, multi-element interactive model. Virtual power plants (VPPs), as a key technology for aggregating and coordinating the optimization of these massive distributed resources, can participate in the electricity market and grid dispatch as a whole, which is of great significance for improving the grid's ability to absorb new energy sources and enhancing system flexibility.

[0003] Currently, the performance of distributed resource points, especially energy storage systems, suffers from subtle degradation due to factors such as operating years, ambient temperature, and the number of charge-discharge cycles. This degradation manifests as increased internal resistance and a decrease in actual adjustable power. However, these changes are difficult to detect in a timely manner, as resource points often report only their rated or initial capacity data. When the dispatch center formulates dispatch plans based on this distorted reported adjustable capacity data, it often results in some resource points being unable to complete instructions, thus hindering the response performance of the entire virtual power plant cluster. Summary of the Invention

[0004] This application provides a virtual power plant source-load scheduling method and system to solve the problem mentioned in the background art that the implicit degradation of resource point performance in existing virtual power plant scheduling often leads to data distortion and causes some resource points to be unable to complete instructions.

[0005] To address the aforementioned technical problems, firstly, this application provides a virtual power plant source-load scheduling method, comprising:

[0006] Acquire the adjustable capacity data reported by each resource point in the virtual power plant; the resource points include power system resource points, energy storage system resource points, and controllable load system resource points.

[0007] A standard test command is issued to a pre-selected target resource point, and the actual response data of the target resource point is collected; based on the deviation between the actual response data and the expected data of the standard test command, the performance degradation coefficient of the target resource point is calculated, and the reported adjustable capability data of the target resource point is dynamically corrected accordingly to generate reliable adjustable capability data;

[0008] Receive power grid dispatch instructions and generate a resource dispatch scheme based on the reliable and adjustable capability data of each resource point;

[0009] The control commands in the resource scheduling scheme are sent to the corresponding resource points for execution.

[0010] Monitor and collect the actual operating data of each resource point after executing the control command, compare the actual operating data with the control command requirements, and generate a scheduling effect evaluation report.

[0011] In one embodiment, the method for issuing standard test commands to the target resource point is as follows:

[0012] Based on the type of the target resource point, match the corresponding instruction template from the standard test instruction library;

[0013] Based on the adjustable capability data reported by the target resource point, the amplitude and duration of the instruction template are scaled to generate a personalized standard test instruction.

[0014] In one embodiment, the method for calculating the performance degradation coefficient of the target resource point is as follows:

[0015] Calculate the root mean square error between the actual response power curve and the expected power curve;

[0016] Calculate the ratio of the actual response rate to the expected response rate;

[0017] The root mean square error and the ratio are weighted and fused to obtain the performance degradation coefficient.

[0018] In one embodiment, the method for dynamically correcting the reported adjustable capability data of the target resource point to generate reliable adjustable capability data is as follows: The reported adjustable capability data is multiplied by the performance degradation coefficient to obtain the reliable adjustable capability data.

[0019] In one embodiment, the method for receiving power grid dispatch instructions and generating a resource dispatch scheme based on the reliable and adjustable capability data of each resource point is as follows:

[0020] Construct a multi-objective optimization function model with the objectives of minimizing overall response delay and overall adjustment error;

[0021] The reliable and adjustable capability data of each resource point, the energy state and safe operation boundary of the energy storage system resource points, and the maximum power change rate of each resource point are used as constraints of the multi-objective optimization function model.

[0022] A multi-objective optimization algorithm is used to solve the multi-objective optimization function model to obtain a set of optimal power command allocation schemes for each resource point that achieves the best trade-off between the two objectives of response speed and regulation accuracy.

[0023] According to a preset strategy, one scheme is selected from the set of optimal power command allocation schemes for each resource point as the final optimal power command allocation scheme for each resource point.

[0024] In one embodiment, the method for monitoring and collecting actual operational data after each resource point executes control commands, comparing the actual operational data with the requirements of the control commands, and generating a scheduling effect evaluation report is as follows:

[0025] The actual operating data of each resource point is collected at a frequency no less than a preset frequency, and the collected actual operating data is filtered and outlier removed; the actual operating data includes actual power value, power direction and equipment status signal;

[0026] Based on the processed actual operating data and the issued control commands, at least one key performance indicator is calculated; the key performance indicator includes cluster-level indicators and resource point-level indicators.

[0027] The calculated key performance indicators are compared with preset performance thresholds. If the key performance indicators fail to meet the standards, the main reasons for the failure are analyzed and the specific resource points are located.

[0028] Automatically generate a scheduling performance evaluation report, which includes instruction completion rate, average response delay time, average steady-state error, and attribution analysis conclusions;

[0029] The scheduling effect evaluation report is sent to the monitoring platform, and the instruction completion rate, average response delay time, and average steady-state error are recorded in the historical performance archive of the corresponding resource point.

[0030] In one embodiment, after sending the evaluation report to the monitoring platform and recording the instruction completion rate, average response latency, and average steady-state error in the historical performance archive of the corresponding resource point, a credit rating is performed on the resource point based on the scheduling effect evaluation report. The credit rating method is as follows:

[0031] Based on the instruction completion rate, average response delay time, and average steady-state error in the resource point scheduling effect evaluation report, calculate the comprehensive performance score of the resource point.

[0032] Based on the comprehensive performance score, resource points are divided into different reputation levels;

[0033] When generating a resource scheduling scheme, priority is given to scheduling resource points with high reputation levels, and output power is limited for resource points with low reputation levels.

[0034] In one embodiment, the formula for calculating the overall performance score of the resource point is: , , , Here are the weighting coefficients for each indicator, and + + =1, Score the instruction completion rate. The average response delay time is scored. The average steady-state error score is given.

[0035] Secondly, this application also provides a virtual power plant source-load dispatching system, applied to a virtual power plant source-load dispatching method, the system comprising:

[0036] The resource point data acquisition module is used to acquire the reported adjustable capacity data of each resource point in the virtual power plant;

[0037] The target resource point data acquisition module is used to issue standard test commands to pre-selected target resource points and collect the actual response data of the target resource points;

[0038] The credibility assessment module is used to calculate the performance degradation coefficient of the target resource point based on the deviation between the actual response data and the expected data of the standard test command, and dynamically correct the reported adjustable capability data of the target resource point according to the performance degradation coefficient to generate credible adjustable capability data.

[0039] The scheduling decision module is used to receive power grid scheduling instructions and generate resource scheduling schemes based on the reliable and adjustable capability data of each resource point.

[0040] The instruction issuing module is used to issue control instructions in the resource scheduling scheme to the corresponding resource points for execution;

[0041] The effect evaluation module is used to monitor and collect the actual operation data of each resource point after executing the control command, compare the actual operation data with the control command requirements, and generate a scheduling effect evaluation report.

[0042] In one embodiment, the scheduling system further includes a reputation rating module, which is used to calculate the comprehensive performance score of each resource point based on the data in the scheduling effect evaluation report output by the effect evaluation module, and to perform dynamic reputation rating on the resource points based on the comprehensive performance score results.

[0043] Thirdly, this application also provides a computer device, including a processor and a memory, wherein the memory is used to store a computer program, and the computer program, when executed by the processor, implements the optical-grade PMMA raw material detection method.

[0044] Fourthly, this application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the aforementioned method for detecting optical-grade PMMA raw materials.

[0045] The beneficial effects of the above-mentioned virtual power plant source-load scheduling method and system are as follows:

[0046] This application can effectively identify resource points with implicit performance degradation by actively issuing standard test commands to target resource points and analyzing their response deviations. It can also calculate a quantified performance degradation coefficient, thereby correcting distorted reported adjustable capability data into reliable and trustworthy adjustable capability data, enabling resource points to effectively complete commands and ensuring the accuracy of scheduling decision-making basis from the source. Attached Figure Description

[0047] Figure 1 This is a flowchart illustrating a virtual power plant source-load scheduling method according to an embodiment of this application;

[0048] Figure 2 This is a schematic diagram illustrating the structure of a virtual power plant source-load dispatching system according to an embodiment of this application;

[0049] Figure 3 This is a schematic diagram of resource point information for a virtual power plant source-load scheduling system, as shown in an embodiment of this application.

[0050] Figure 4 for Figure 2 This application embodiment shows a schematic diagram of the working status of an energy storage device in a virtual power plant source-load dispatching system;

[0051] Figure 5 for Figure 2 This application embodiment shows a schematic diagram of a scheduling instruction for a virtual power plant source-load scheduling system. Detailed Implementation

[0052] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Many specific details are set forth in the following description to provide a thorough understanding of the present invention. However, the present invention can be practiced in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0053] It should be noted that when a component is considered to be "connected" to another component, it can be directly connected to the other component or there may be an intermediary component present. Conversely, when a component is said to be "directly" connected to another component, there is no intermediary component.

[0054] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0055] like Figure 1 As shown, in a first aspect, this application provides a virtual power plant source-load scheduling method, including:

[0056] S1. Obtain the reported adjustable capacity data of each resource point in the virtual power plant; the resource points include power system resource points, energy storage system resource points, and controllable load system resource points.

[0057] Power system resources include photovoltaic power plants, small wind turbines, natural gas combined heat and power units, and backup diesel generators. Their power generation capacity is adjustable. For photovoltaic power plants and small wind turbines, although the primary energy source is uncontrollable, their inverters can perform derating operation.

[0058] Energy storage system resources include centralized energy storage power stations on the grid side, large-scale energy storage cabinets on the user side, and charging pile clusters in the form of V2G for electric vehicles. Their charging and discharging power and capacity are adjustable. Bidirectional regulation of charging and discharging can be performed.

[0059] Controllable load system resource points include interruptible industrial loads, such as cement plant mills and electrolytic aluminum equipment; flexible loads in commercial buildings, such as smart central air conditioning and cold storage; and smart residential loads, such as smart water heaters and swimming pool pumps. Their power consumption is interruptible, can be reduced, or can be shifted.

[0060] like Figure 3 and Figure 4 As shown, the reported adjustable capacity data refers to the adjustable capacity parameters reported by each resource point itself, claiming that it possesses them. Adjustable capacity data for power system resource points includes, for example, maximum adjustable generating power (the maximum generating power allowed under the current environment); minimum technical output (the minimum generating power allowed for safe and stable operation of the equipment); and power regulation rate (the amount by which generating power can increase or decrease per unit time).

[0061] Adjustable capacity data of energy storage system resource points, such as maximum charging or discharging power, i.e., the maximum charging or discharging power allowed under the current charge state; current available capacity, i.e., the total energy that can be stored or released when fully charged or discharged under the current charge state; current state of charge, i.e. the remaining percentage of battery capacity;

[0062] Adjustable capacity data for controllable load system resource points, such as maximum power reduction, which is the maximum amount of power that can be reduced without affecting core functions; interruptible duration, which is the longest duration for which the load can be reduced or interrupted; and minimum response time, which is the time required from receiving an instruction to starting to reduce the load.

[0063] S2. Issue standard test commands to pre-selected target resource points and collect actual response data of the target resource points; based on the deviation between the actual response data and the expected data of the standard test commands, calculate the performance degradation coefficient of the target resource points, and dynamically correct the reported adjustable capability data of the target resource points accordingly to generate reliable adjustable capability data.

[0064] The method for issuing standard test commands to the target resource point is as follows:

[0065] S21. Match the corresponding instruction template from the standard test instruction library according to the type of the target resource point;

[0066] Assume the virtual power plant has three resource points: 1. An energy storage power station, belonging to the energy storage system resource point; 2. A rooftop photovoltaic power station, belonging to the power supply system resource point; 3. An industrial cooling load, belonging to the controllable load system resource point. The standard test instruction library pre-stores typical test templates for these resource types:

[0067] For example, for energy storage systems, the pre-stored instruction template is ESS_Power Step Test; for power supply systems, the pre-stored instruction template is DG_Power Reduction Test; and for controllable load systems, the pre-stored instruction template is CL_Load Interruption Test. For energy storage power stations, the system will match the ESS_Power Step Test template from the standard test instruction library; for photovoltaic power stations, the system will match the DG_Power Reduction Test template from the standard test instruction library; and for industrial cooling loads, the system will match the CL_Load Interruption Test template from the standard test instruction library.

[0068] S22. Based on the adjustable capability data reported by the target resource point, scale the amplitude and duration of the instruction template to generate a personalized standard test instruction.

[0069] For example, if the target resource is an energy storage power station, and the station reports adjustable capacity data of 500kW maximum discharge power and 500kW maximum charging power, then the matching template is ESS_Power Step Test. The original parameters of this template are a step signal with an amplitude of 100kW and a duration of 5 minutes. The next step is to scale the amplitude and duration of the command template. The system does not directly use the fixed value of 100kW. It reads the reported data from the energy storage power station, for example, 500kW, and then calculates the test amplitude according to a preset safety ratio. If the preset safety ratio is 20%, then the test amplitude = reported maximum power × safety ratio = 500kW × 20% = 100kW, generating a discharge step command with an amplitude of 100kW. Generally, when conducting power tests, the test duration is usually set to a fixed value, such as 5 minutes. This is because after scaling the power amplitude, 5 minutes is sufficient to observe whether the equipment can operate reliably in a stable state, without being too long and affecting the equipment's normal participation in market activities. A personalized standard test instruction can be generated as follows: Energy storage power station, please execute: Increase the discharge power from 0 to 100 kW within 1 second and maintain this power output for 5 minutes.

[0070] The method for calculating the performance degradation coefficient of the target resource point is as follows:

[0071] S23. Calculate the root mean square error between the actual response power curve and the expected power curve.

[0072] S24. Calculate the ratio of the actual response rate to the expected response rate;

[0073] S25. The root mean square error and the ratio are weighted and fused to obtain the performance degradation coefficient.

[0074] The expected power curve represents the ideal response data of an energy storage power station. The actual response power curve represents the actual response data of the energy storage power station as recorded by the monitoring system.

[0075] Suppose we are testing an energy storage power station that reports a maximum power of 500kW. We will execute the standard test command to increase the discharge power from 0 to 100kW within 1 second and maintain this power output for 5 minutes. The system collects data at a frequency of 10 times per second and calculates the error between the actual power and the expected power at each time point. As shown in Table 1:

[0076] Table 1

[0077] Calculate the average of the squares of all errors: (400+100+25+0+...+4) / N, where N is the total number of sampling points. Then take the square root of the mean to get the root mean square error value of 15. The larger this value is, the greater the overall deviation between the actual response power curve and the expected power curve, and the worse the performance.

[0078] Expected response rate: Calculated according to the standard test instruction requirements. The standard test instruction requires "from 0 to 100kW in 1 second", so the expected response rate is 100kW / 1s = 100kW / s.

[0079] Actual response rate: Calculated based on actual data. Assuming actual data shows that the device takes 1.2 seconds to reach 90% power (90kW) from 10% power (10kW), then the actual response rate = (90 - 10) kW / 1.2s ≈ 66.7kW / s. Therefore, the rate ratio = actual response rate / expected response rate = 66.7 / 100 = 0.667. The smaller this value is than 1, the slower the response speed and the worse the performance. The ideal rate ratio is 1.

[0080] The root mean square error (RMSE) value is normalized to the rate ratio. Let the RMSE value be K, then K=15. The RMSE value K is then converted into an accuracy score. Then S... 精度 =1 / (1+K)=1 / (1+15)≈0.0625. The speed score is S. 速度 =0.667. Therefore, the performance degradation coefficient = (W1*S) 精度 )+(W2*S 速度 The formula is: (0.7 * 0.0625) + (0.3 * 0.667) = 0.244. W1 and W2 are the weights of the accuracy score and speed score, respectively. Therefore, 0.244 is the calculated performance degradation coefficient. If the energy storage station has a perfect response, then K = 0, the rate ratio = 1, and the performance degradation coefficient will be (0.7 * 1) + (0.3 * 1) = 1.0. 0.244 is much less than 1.0, indicating that the energy storage station's performance has severely degraded.

[0081] The method for dynamically correcting the reported adjustable capability data of the target resource point to generate reliable adjustable capability data is as follows: S26. Multiply the reported adjustable capability data by the performance degradation coefficient to obtain the reliable adjustable capability data.

[0082] For example, if an energy storage power station reports a maximum power of 500kW, the corrected reliable adjustable capability will become: 500 kW * 0.244 = 122kW.

[0083] S3. Receive power grid dispatch instructions and generate a resource dispatch scheme based on the trusted and adjustable capability data of each resource point;

[0084] The method for receiving power grid dispatch instructions and generating a resource dispatch scheme based on the reliable and adjustable capability data of each resource point is as follows:

[0085] S31. Construct a multi-objective optimization function with the objectives of minimizing overall response delay and minimizing overall regulation error. The multi-objective optimization function model includes two objectives: first, minimizing overall response delay, for example, hoping that the total power of the cluster can reach the requirement of 200kW as quickly as possible; second, minimizing overall regulation error, for example, hoping that the total power of the cluster can be stably maintained at 200kW, without exceeding or falling below 200kW.

[0086] The horizontal axis of the function represents the overall response delay value f1, and the vertical axis represents the overall adjustment error value f2. The multi-objective optimization results form multiple optimal solutions in the function.

[0087] S32. The reliable and adjustable capability data of each resource point, the energy state and safe operation boundary of the energy storage system resource points, and the maximum power change rate of each resource point are used as constraints of the multi-objective optimization function model.

[0088] Constraints on reliable and adjustable capability data, such as 0 ≤ P a ≤100, meaning the commanded power of energy storage power station A cannot exceed 100kW; 0≤P b ≤80, meaning the commanded power of energy storage power station B cannot exceed 80kW; 0≤P c ≤50, meaning that the power reduction instruction for photovoltaic power station C cannot exceed 50kW;

[0089] Energy state and safety boundary constraints of energy storage systems. For example, assuming the energy state and safety boundary safety range of an energy storage system is 20% to 90%, this means that when issuing instructions to energy storage power station B, it must be ensured that its power level will not drop below 20% after the instructions are executed.

[0090] Maximum power change rate constraint, |dP a / dt|≤50kW / s, meaning the power change of energy storage power station A cannot be more drastic than 50kW / s; |dP b / dt|≤30kW / s, meaning the power change of energy storage power station B cannot be more drastic than 30kW / s; |dP c / dt|≤20kW / s, meaning that the power change of photovoltaic power station C cannot be more drastic than 20kW / s.

[0091] In addition to the above conditions, the sum of the power commands allocated to all resource points must be exactly equal to the power value required by the grid. For example, Pa +P b +P c =200 kW.

[0092] S33. The multi-objective optimization function model is solved by using a multi-objective optimization algorithm to obtain a set of optimal power command allocation schemes for each resource point that achieves the best trade-off between the two objectives of response speed and adjustment accuracy.

[0093] The multi-objective optimization function model will eventually output a solution set through the multi-objective optimization algorithm. This solution set contains multiple solutions, each of which is the best trade-off between response speed and adjustment accuracy.

[0094] For example, in Option 1, P a =100kW, P b =70kW, P c =30kW, allowing the best-performing energy storage power stations A and B to undertake most of the tasks, resulting in a very fast response, i.e., a very small overall response delay value f1. However, due to the drastic power changes, there is a slight overshoot, i.e., a slightly larger overall adjustment error value f2.

[0095] Option 2, P a =80kW, P b =60kW, P c =60kW, allowing for a more balanced distribution of power across the three stations, smoother power changes, and a slightly slower overall response delay (f1) but with very high stability and precision, and a very small overall adjustment error (f2). Neither of these two options is inherently better or worse; they represent the optimal balance between the two.

[0096] S34. Select a scheme from the set of optimal power command allocation schemes for each resource point according to the preset strategy, and use it as the final optimal power command allocation scheme for each resource point.

[0097] Based on the current needs of the power grid, select the most suitable option from the set of optimal trade-offs. For example, if extremely high speed is required, then option one is chosen; if higher accuracy is required, then option two is chosen.

[0098] S4. Send the control commands in the resource scheduling scheme to the corresponding resource points for execution;

[0099] The selected option, such as the power command (P) of option one. a =100kW, P b =70kW, P c =30kW) will be distributed to each resource point for execution.

[0100] S5. The method for monitoring and collecting the actual operating data of each resource point after executing the control command, comparing the actual operating data with the control command requirements, and generating a scheduling effect evaluation report is as follows:

[0101] S51. Collect the actual operating data of each resource point at a frequency not lower than a preset frequency, and perform filtering and outlier removal processing on the collected actual operating data; the actual operating data includes actual power value, power direction and equipment status signal;

[0102] Suppose a virtual power plant source-load dispatching system issues a control command to two resource points: Command content: "Starting at 10:00:00, please operate continuously for 10 minutes at a total power of 150kW." Allocation scheme: Energy storage power station A: Commanded power is P. a =100kW, Energy Storage Station B: Command Power P b =70kW. After the control command is executed, the source-load dispatching system collects the actual operating data of energy storage power station A and energy storage power station B 10 times per second. For example, the actual power values ​​collected for energy storage power station A are [98,101,99,105,102,...]kW; power direction: discharging; equipment status: normal. The actual power values ​​collected for energy storage power station B are [48,52,49,0,0,...]kW; power direction: discharging; equipment status: alarm after 10:02:15. The data is smoothed using a moving average method to remove small fluctuations. For example, the [98,101,99] of energy storage power station A is smoothed to 99.3. If the data of energy storage power station B is found to suddenly become 0 after a certain moment, the system will check whether this is accompanied by an equipment status alarm. If so, these outliers are retained; if it is a single outlier caused by communication interference, it is discarded.

[0103] S52. Based on the processed actual operating data and the issued control commands, calculate at least one key performance indicator; the key performance indicator includes cluster-level indicators and resource-point-level indicators; the cluster-level indicators include total command completion rate, cluster response latency, and cluster steady-state error; the resource-point-level indicators include single-point command completion rate, single-point average response latency, and single-point average steady-state error.

[0104] For example, the control command requires a total energy of 150kW*(10 / 60)h=25kWh. The actual total energy released is calculated, assuming that energy storage station A contributed 16.5kWh and energy storage station B contributed 4.5kWh before the fault. Then the total command completion rate = (16.5+4.5) / 25*100%=84%.

[0105] like Figure 5Looking at the resource point-level indicators, the instruction completion rate of energy storage power station A is 16.5kWh / (100kW * 10 / 60 h) = 99%, the average response delay time is 0.5 seconds, and the average steady-state error is (100-99.3) / 100*100%≈0.7%.

[0106] The command completion rate of energy storage power station B is 4.5kWh / (50kW*10 / 60h)=54%, the average response delay time is 2 seconds, and the average steady-state error is (50-49.6) / 50*100%=0.8%.

[0107] S53. Compare the calculated key performance indicators with the preset performance thresholds. If the key performance indicators do not meet the standards, analyze the main reasons for the non-compliance and locate the specific resource points.

[0108] For example, the preset performance thresholds are a 95% instruction completion rate and a 1.5-second response latency. In this case, the overall instruction completion rate (84%), the instruction completion rate of energy storage station B (54%), and the average response latency of energy storage station B (2 seconds) all fail to meet the targets. The main reason for the failure is that energy storage station B malfunctioned and exited shortly after the task began, resulting in a severe shortage of its energy contribution. Therefore, energy storage station B is directly identified as the failure point in this scheduling task.

[0109] S54. Automatically generate a scheduling effect evaluation report, which includes instruction completion rate, average response delay time, average steady-state error, and attribution analysis conclusions.

[0110] For example, the overall command completion rate is (16.5 + 4.5) / 25 * 100% = 84%, which is below the target. Energy storage station A has a command completion rate of 99%, an average response delay of 0.5 seconds, and an average steady-state error of 0.7%, which is excellent. Energy storage station B has a command completion rate of 54%, an average response delay of 2 seconds, and an average steady-state error of 0.8%, indicating a fault at energy storage station B. The main reason for the failure to meet the target is that energy storage station B experienced an unknown fault at 10:02:15, resulting in it going offline. It is recommended that this station be inspected and repaired immediately. All details are recorded in the dispatch performance evaluation report.

[0111] S55. Send the scheduling effect evaluation report to the monitoring platform and record the instruction completion rate, average response delay time, and average steady-state error to the historical performance file of the corresponding resource point.

[0112] The dispatch performance evaluation report is immediately pushed to the monitoring screen of the operation and maintenance personnel and an alarm is triggered, indicating that energy storage station B needs maintenance. The good performance of this task is recorded in the historical archive of energy storage station A, while the failure performance of this task is recorded in the historical archive of energy storage station B, and it is marked by the system as unavailable or pending maintenance. It will not be assigned tasks in the next dispatch until maintenance is completed and the test is passed.

[0113] In one embodiment, after sending the evaluation report to the monitoring platform and recording the instruction completion rate, average response latency, and average steady-state error in the historical performance archive of the corresponding resource point, a credit rating is performed on the resource point based on the scheduling effect evaluation report. The credit rating method is as follows:

[0114] Based on the instruction completion rate, average response delay time, and average steady-state error in the resource point scheduling effect evaluation report, calculate the comprehensive performance score of the resource point.

[0115] Based on the comprehensive performance score, resource points are divided into different reputation levels;

[0116] When generating a resource scheduling scheme, priority is given to scheduling resource points with high reputation levels, and output power is limited for resource points with low reputation levels.

[0117] The formula for calculating the comprehensive performance score of the resource point is as follows: , , , Here are the weighting coefficients for each indicator, and + + =1, Score the instruction completion rate. The average response delay time is scored. The average steady-state error score is given.

[0118] For example, the instruction completion rate is R C The average response delay time is T. d The average steady-state error is E s .So =R C *100; = ; = .

[0119] T1 is the optimal expected threshold for the average response delay time; E1 is the optimal expected threshold for the average steady-state error. These are the score decay coefficients, used to control the severity of score decline. The higher the value, the more severe the penalty for performance degradation.

[0120] For example, =0.5, =0.3, =0.2, the completion rate of instruction A of energy storage power station is 100%, then R C =1, =1*100=100; If T1=1S, the average response delay time is T d If it is 0.5s, then =100; if E1=1%, the average steady-state error is E s If it is 1%, then =100, then =0.5*100+0.3*100+0.2*100=100. If a score ≥ 90 is Grade A, 75 ≤ score < 90 is Grade B, 60 ≤ score < 75 is Grade C, and a score < 60 is Grade D, then energy storage power station A is Grade A. For Grade D resource points, their reliable and adjustable capabilities will be multiplied by a penalty coefficient, such as 0.5, severely limiting the amount of work they can undertake.

[0121] like Figure 2 As shown, in a second aspect, this application also provides a virtual power plant source-load scheduling system, applied to a virtual power plant source-load scheduling method, the system comprising:

[0122] Resource point data acquisition module 1 is used to acquire the reported adjustable capacity data of each resource point in the virtual power plant;

[0123] The target resource point data acquisition module 2 is used to issue standard test instructions to pre-selected target resource points and collect the actual response data of the target resource points;

[0124] The credibility assessment module 3 is used to calculate the performance degradation coefficient of the target resource point based on the deviation between the actual response data and the expected data of the standard test command, and dynamically correct the reported adjustable capability data of the target resource point according to the performance degradation coefficient to generate credible adjustable capability data.

[0125] The scheduling decision module 4 is used to receive power grid scheduling instructions and generate resource scheduling schemes based on the reliable and adjustable capability data of each resource point.

[0126] The instruction issuing module 5 is used to issue control instructions in the resource scheduling scheme to the corresponding resource points for execution.

[0127] The effect evaluation module 6 is used to monitor and collect the actual operation data of each resource point after executing the control command, compare the actual operation data with the control command requirements, and generate a scheduling effect evaluation report.

[0128] In one embodiment, the scheduling system further includes a reputation rating module, which is used to calculate the comprehensive performance score of each resource point based on the data in the scheduling effect evaluation report output by the effect evaluation module, and to perform dynamic reputation rating on the resource points based on the comprehensive performance score results.

[0129] The options described in the above method embodiments are also applicable to this embodiment, and will not be detailed here. The remaining content of this application's embodiments can be found in the above method embodiments, and will not be repeated in this embodiment.

[0130] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0131] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A virtual power plant source-load scheduling method, characterized in that: include: Acquire the adjustable capacity data reported by each resource point in the virtual power plant; the resource points include power system resource points, energy storage system resource points, and controllable load system resource points. Standard test commands are issued to pre-selected target resource points, and the actual response data of the target resource points is collected; Based on the deviation between the actual response data and the expected data of the standard test command, the performance degradation coefficient of the target resource point is calculated, and the reported adjustable capability data of the target resource point is dynamically corrected accordingly to generate reliable adjustable capability data. Multiply the reported adjustable capability data by the performance degradation coefficient to obtain the reliable adjustable capability data; Receive power grid dispatch instructions and generate a resource dispatch scheme based on the reliable and adjustable capability data of each resource point; The method for receiving power grid dispatch instructions and generating a resource dispatch scheme based on the reliable and adjustable capability data of each resource point is as follows: Construct a multi-objective optimization function model with the objectives of minimizing overall response delay and overall adjustment error; The reliable and adjustable capability data of each resource point, the energy state and safe operation boundary of the energy storage system resource points, and the maximum power change rate of each resource point are used as constraints of the multi-objective optimization function model. A multi-objective optimization algorithm is used to solve the multi-objective optimization function model to obtain a set of optimal power command allocation schemes for each resource point that achieves the best trade-off between the two objectives of response speed and regulation accuracy. According to a preset strategy, one scheme is selected from the set of optimal power command allocation schemes for each resource point as the final optimal power command allocation scheme for each resource point. The control commands in the resource scheduling scheme are sent to the corresponding resource points for execution. Monitor and collect the actual operating data of each resource point after executing control commands, compare the actual operating data with the requirements of the control commands, and generate a scheduling effect evaluation report; The method for monitoring and collecting the actual operational data of each resource point after executing control commands, comparing the actual operational data with the requirements of the control commands, and generating a scheduling effect evaluation report is as follows: The actual operating data of each resource point is collected at a frequency no less than a preset frequency, and the collected actual operating data is filtered and outlier removed. The actual operating data includes actual power values, power direction, and equipment status signals; Based on the processed actual operating data and the issued control commands, at least one key performance indicator is calculated; the key performance indicator includes cluster-level indicators and resource point-level indicators. The calculated key performance indicators are compared with preset performance thresholds. If the key performance indicators fail to meet the standards, the main reasons for the failure are analyzed and the specific resource points are located. Automatically generate a scheduling performance evaluation report, which includes instruction completion rate, average response delay time, average steady-state error, and attribution analysis conclusions; The scheduling effect evaluation report is sent to the monitoring platform, and the instruction completion rate, average response delay time, and average steady-state error are recorded in the historical performance archive of the corresponding resource point.

2. The virtual power plant source-load scheduling method according to claim 1, characterized in that: The method for issuing standard test commands to the target resource point is as follows: Based on the type of the target resource point, match the corresponding instruction template from the standard test instruction library; Based on the adjustable capability data reported by the target resource point, the amplitude and duration of the instruction template are scaled to generate a personalized standard test instruction.

3. The virtual power plant source-load scheduling method according to claim 1, characterized in that: The method for calculating the performance degradation coefficient of the target resource point is as follows: Calculate the root mean square error between the actual response power curve and the expected power curve; Calculate the ratio of the actual response rate to the expected response rate; The root mean square error and the ratio are weighted and fused to obtain the performance degradation coefficient.

4. The virtual power plant source-load scheduling method according to claim 3, characterized in that: The method for dynamically correcting the reported adjustable capability data of the target resource point to generate reliable adjustable capability data is as follows: multiply the reported adjustable capability data by the performance degradation coefficient to obtain the reliable adjustable capability data.

5. The virtual power plant source-load scheduling method according to claim 1, characterized in that, After sending the evaluation report to the monitoring platform and recording the instruction completion rate, average response latency, and average steady-state error in the historical performance archive of the corresponding resource point, a credit rating is performed on the resource point based on the scheduling effect evaluation report. The credit rating method is as follows: Based on the instruction completion rate, average response delay time, and average steady-state error in the resource point scheduling effect evaluation report, calculate the comprehensive performance score of the resource point. Based on the comprehensive performance score, resource points are divided into different reputation levels; When generating a resource scheduling scheme, priority is given to scheduling resource points with high reputation levels, and output power is limited for resource points with low reputation levels.

6. The virtual power plant source-load scheduling method according to claim 5, characterized in that, The formula for calculating the comprehensive performance score of the resource point is as follows: , , , Here are the weighting coefficients for each indicator, and + + =1, Score the instruction completion rate. The average response delay time is scored. The average steady-state error score is given.

7. A virtual power plant source-load dispatching system, characterized in that, A system applied to a virtual power plant source-load scheduling method as described in any one of claims 1 to 6, the system comprising: The resource point data acquisition module is used to acquire the reported adjustable capacity data of each resource point in the virtual power plant; The target resource point data acquisition module is used to issue standard test commands to pre-selected target resource points and collect the actual response data of the target resource points; The credibility assessment module is used to calculate the performance degradation coefficient of the target resource point based on the deviation between the actual response data and the expected data of the standard test command, and dynamically correct the reported adjustable capability data of the target resource point according to the performance degradation coefficient to generate credible adjustable capability data. The scheduling decision module is used to receive power grid scheduling instructions and generate resource scheduling schemes based on the reliable and adjustable capability data of each resource point. The instruction issuing module is used to issue control instructions in the resource scheduling scheme to the corresponding resource points for execution; The effect evaluation module is used to monitor and collect the actual operation data of each resource point after executing the control command, compare the actual operation data with the control command requirements, and generate a scheduling effect evaluation report.

8. The virtual power plant source-load dispatching system according to claim 7, characterized in that, The scheduling system also includes a reputation rating module, which is used to calculate the comprehensive performance score of each resource point based on the data in the scheduling effect evaluation report output by the effect evaluation module, and to perform dynamic reputation rating on the resource points based on the comprehensive performance score results.

Images