A method, system, equipment, and medium for calculating the reserve capacity demand of a high-proportion renewable energy power system considering hierarchical sharing.
By using a hierarchical and shared method for calculating reserve capacity demand, and leveraging dynamic clustering algorithms and weather information, the problem of distorted reserve capacity calculation in high-proportion renewable energy power systems has been solved, achieving accurate calculation of reserve capacity and cost control.
Patent Information
- Application Number
- CN202611065106.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-07-17
- Publication Date
- 2026-11-17
AI Technical Summary
The strong fluctuations and large prediction errors of high-proportion renewable energy sources in traditional power systems lead to distorted reserve capacity calculations. The simple superposition of multiple types of disturbance demand results in reserve capacity redundancy, increasing electricity costs.
A hierarchical and shared reserve capacity demand calculation method is adopted. By analyzing the historical predicted power deviation of new energy sources and loads through dynamic clustering algorithm, and combining the fault status of power sources and weather information, the reserve capacity demand is calculated in a hierarchical manner.
It enables accurate calculation of the reserve capacity of high-proportion new energy power systems, avoids reserve capacity redundancy, and reduces reserve service costs.
Smart Images

Figure CN123121479A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of power load calculation technology, and in particular to a method, system, equipment and medium for calculating the reserve capacity demand of a high-proportion new energy power system that considers hierarchical sharing. Background Technology
[0002] In power grid dispatching, reserving generation reserves is a fundamental framework for the power system to cope with source-load uncertainties and ensure power balance. Currently, power grid dispatching typically uses a simple estimation and reserve based on a fixed percentage of the maximum daily load, usually between 8% and 10%. This setting mainly considers two factors: firstly, the deviation between actual load demand and planned values (load forecasting error), known as load reserve; and secondly, the power deficit caused by power source failures such as unit / DC feed-in faults within the system, typically calculated based on single unit tripping, single-pole DC faults, or bipolar DC faults, known as contingency reserve.
[0003] However, the proportion of new energy sources in the current power system has rapidly increased to over 50%. Wind and solar power generation exhibits greater volatility and prediction errors, and faces new disturbances such as widespread wind power outages caused by extreme weather conditions. The strong power disturbances brought by new energy sources are far greater than those of the load, strongly correlated with weather, and their temporal variation is far less predictable than that of the load. Therefore, the required operational reserve capacity for high-proportion new energy power systems is significantly increased, and it is not suitable to calculate based on a fixed proportion of the load. Furthermore, in the electricity market environment, reserve services require users to pay for both capacity and electricity. Simply adding the reserve demand generated by new energy generation to the existing load reserve and conventional power failure reserve needs to calculate the grid reserve capacity can easily lead to excessive reserve provisions, resulting in an increase in the overall electricity cost for society. Summary of the Invention
[0004] In view of the aforementioned existing problems, the present invention is proposed.
[0005] Therefore, this invention provides a method and system for calculating the reserve capacity demand of a high-proportion renewable energy power system that considers hierarchical sharing. This solves the problems of traditional calculation methods being unable to adapt to the strong fluctuations and large prediction errors of high-proportion renewable energy, which leads to calculation distortion, and the problems of simply superimposing multiple types of disturbance reserve demands, which leads to reserved capacity redundancy and increases electricity costs.
[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides a method for calculating the reserve capacity demand of a high-proportion renewable energy power system considering hierarchical sharing, including: Acquire power grid operation load data, divide the characteristic time periods for reserve capacity demand calculation, and establish a sample library of wind and solar load prediction and actual measurement for the characteristic time periods; Based on the historical predicted power and historical measured power of new energy and load in each characteristic period, a dynamic clustering algorithm is used to perform nearest neighbor search to obtain the deviation of new energy predicted power and load predicted power in each characteristic period of the evaluation day. Based on the predicted power deviation of new energy sources and the predicted power deviation of load corresponding to each characteristic time period, the target time with the largest actual net load within the assessment day is determined, and the negative net load prediction deviation corresponding to the target time is obtained. Based on the negative deviation of the net load prediction and the preset maximum power deficit of unit components of different power types under unit component failure conditions, the first-level standby capacity requirement is calculated. Based on the preset maximum power deficit of multi-components of different power types under multi-component fault conditions, and combined with the sharing rules of the maximum power deficit of unit components in the first-level reserve capacity requirement, the second-level reserve capacity requirement is calculated. Based on weather information, the expected power reduction of new energy sources at the target time under weather changes is estimated, and the expected power reduction is combined with the first-level reserve capacity demand and the second-level reserve capacity demand to obtain the third-level reserve capacity demand.
[0007] As a preferred embodiment of the method for calculating the reserve capacity demand of a high-proportion renewable energy power system considering hierarchical sharing as described in this invention, the method includes: using a dynamic clustering algorithm to perform nearest neighbor search to obtain the renewable energy predicted power deviation and load predicted power deviation for each characteristic time period within the evaluation day, including: For any characteristic time period, calculate the weighted distance between the predicted power sequence of the evaluation day within the corresponding characteristic time period and the predicted power sequences of each historical sample in the corresponding historical sample library; Sort the samples according to the weighted distance and select a set number of nearby historical samples to form a nearest neighbor set; By using the deviation between the measured power sequence and the predicted power sequence corresponding to each historical sample in the nearest neighbor set, a weighted average is calculated to obtain the positive deviation sequence of the predicted power of new energy sources and the negative deviation sequence of the predicted power of load within the corresponding characteristic time period.
[0008] As a preferred embodiment of the method for calculating the reserve capacity demand of a high-proportion renewable energy power system considering hierarchical sharing as described in this invention, the method includes: determining the target time with the largest actual net load within the assessment day based on the renewable energy predicted power deviation and the load predicted power deviation corresponding to each characteristic time period, and obtaining the negative net load predicted deviation corresponding to the target time, including: For any characteristic period, the positive deviation sequence of the new energy predicted power and the negative deviation sequence of the load predicted power within the corresponding characteristic period are superimposed on the predicted net load sequence of the assessment date to obtain the actual net load estimation sequence for the corresponding characteristic period. By comparing the values in the actual net load estimation sequence within all characteristic time periods, the time corresponding to the maximum value in each actual net load estimation sequence is determined as the target time, and the difference between the maximum value and the predicted net load at the target time is taken as the negative deviation of the net load prediction corresponding to the target time.
[0009] As a preferred embodiment of the method for calculating the reserve capacity demand of a high-proportion renewable energy power system considering hierarchical sharing as described in this invention, the first-level reserve capacity demand is calculated based on the negative deviation of the net load forecast and the preset maximum power deficit of unit components of different power types under unit component failure conditions, including: The target time is obtained as follows: the single-unit planned power of conventional generator sets in operation, the single-pole planned power of inter-regional DC transmission, the single-circuit predicted power of the on-site collection line of a single wind farm, and the single-circuit predicted power of the on-site collection line of a single photovoltaic power station. The maximum value is selected from the single-unit planned power, single-pole planned power, and single-circuit predicted power of each field collection line, and is taken as the maximum power deficit of the unit. The negative deviation of the net load forecast corresponding to the target time is added to the maximum power deficit of the unit to obtain the first-level reserve capacity requirement.
[0010] As a preferred embodiment of the method for calculating the reserve capacity demand of a high-proportion new energy power system considering hierarchical sharing as described in this invention, the second-level reserve capacity demand is calculated based on the preset maximum power deficit of multiple power sources of different power types under multi-component fault conditions, and in conjunction with the sharing rules of the maximum power deficit of unit components in the first-level reserve capacity demand, including: The target time is to obtain the planned power of dual-machine failure of conventional generator units in operation, the planned power of bipolar blocking failure of inter-regional DC transmission, and the predicted power of a single new energy power station completely disconnected from the grid. The maximum value is selected from the planned power of dual-machine failure, the planned power of bipolar blocking failure, and the predicted power of a single new energy power station being completely disconnected from the grid, and is used as the maximum power deficit of multiple components. Calculate the difference between the required reserve for each type of severe fault and the reserve reserved for a single fault of the same type in the first level; take the maximum value of the difference as the reserve capacity requirement for the second level.
[0011] As a preferred embodiment of the method for calculating the reserve capacity demand of a high-proportion renewable energy power system considering hierarchical sharing as described in this invention, the method includes: estimating the expected power reduction of renewable energy at a target time under weather changes based on weather information, and combining the expected power reduction with the first-level reserve capacity demand and the second-level reserve capacity demand to obtain the third-level reserve capacity demand, including: Based on rolling weather forecast information, weather change events are identified, and combined with the range of new energy power plants covered by the weather change events and historical similar samples, the expected value of the power reduction of new energy caused by weather change at the target time is estimated. Calculate the difference between the expected power reduction and the sum of the Tier 1 and Tier 2 reserve capacity requirements; If the difference is greater than zero, the difference will be used as the third-level reserve capacity requirement; if the difference is less than or equal to zero, the third-level reserve capacity requirement will be set to zero.
[0012] As a preferred embodiment of the method for calculating the reserve capacity demand of a high-proportion renewable energy power system considering hierarchical sharing as described in this invention, the method further includes: before calculating the weighted distance, a rolling update of the historical sample database is performed, specifically: Collect the day-ahead forecast power series and the day-ahead measured power series of new energy sources and loads over several historical days; According to the rules for dividing characteristic time periods, the predicted power sequence of the day and the measured power sequence of the day are cropped into time periods to construct a historical sample library corresponding to each characteristic time period; The historical sample databases are updated on a first-in, first-out (FIFO) basis to ensure that each database always retains historical samples from a fixed number of days prior to the evaluation date.
[0013] Secondly, the present invention provides a system for calculating the reserve capacity demand of a high-proportion renewable energy power system that considers hierarchical sharing, comprising: The sample library construction module is used to acquire power grid operation load data, divide the characteristic time periods for reserve capacity demand calculation, and establish a sample library of wind and solar load prediction and actual measurement for the characteristic time periods. The clustering module is used to perform nearest neighbor search using a dynamic clustering algorithm based on the historical predicted power and historical measured power of new energy and load in each characteristic time period, to obtain the deviation of new energy predicted power and load predicted power in each characteristic time period of the evaluation day. The determination module is used to determine the target time with the largest actual net load within the assessment day based on the new energy predicted power deviation and load predicted power deviation corresponding to each characteristic time period, and to obtain the net load predicted negative deviation corresponding to the target time. The first calculation module is used to calculate the first-level reserve capacity requirement based on the net load prediction negative deviation and the preset maximum power deficit of unit components of different power types under unit component failure conditions. The second calculation module is used to calculate the second-level reserve capacity requirement based on the maximum power deficit of multi-element components of different power types under the preset multi-element fault conditions, and in combination with the sharing rules of the maximum power deficit of unit components in the first-level reserve capacity requirement. The third calculation module is used to estimate the expected power reduction of new energy sources at the target time under weather changes based on weather information, and to combine the expected power reduction with the first-level reserve capacity demand and the second-level reserve capacity demand to obtain the third-level reserve capacity demand.
[0014] Thirdly, the present invention provides a computer device, comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, they implement the steps of a method for calculating the reserve capacity demand of a high-proportion new energy power system that considers hierarchical sharing.
[0015] Fourthly, the present invention provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the steps of the method for calculating the reserve capacity demand of a high-proportion new energy power system considering hierarchical sharing.
[0016] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention sets characteristic time periods for reserve capacity demand calculation based on the peak-valley shape of the load curve, and proposes a method for estimating the prediction deviation of wind power and photovoltaic power generation during these characteristic time periods based on dynamic clustering. Secondly, considering the differences in the response speed of different power source disturbances to reserve demand, and the correlation between extreme weather causing multiple renewable energy power plants to disconnect from the grid successively, a method for calculating reserve capacity demand in high-proportion renewable energy power systems with tiered reserve capacity and demand sharing is proposed. The proposed method comprehensively assesses renewable energy disturbances and power balance assurance while also taking into account the need for total reserve capacity control, thus avoiding excessively high reserve service costs. Attached Figure Description
[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0018] Figure 1 This is a schematic diagram of the overall process of a method for calculating the reserve capacity demand of a high-proportion new energy power system considering hierarchical sharing, as described in one embodiment of the present invention.
[0019] Figure 2 This is a schematic diagram of the overall framework of a method for calculating the reserve capacity demand of a high-proportion new energy power system considering hierarchical sharing, as described in one embodiment of the present invention. Detailed Implementation
[0020] 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. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.
[0021] Example 1, referring to Figures 1-2 As an embodiment of the present invention, a method for calculating the reserve capacity demand of a high-proportion renewable energy power system considering hierarchical sharing is provided, comprising: S100: Acquire power grid operation load data, divide the characteristic time periods for reserve capacity demand calculation, and establish a sample library of wind and solar load prediction and actual measurement for the characteristic time periods; S200: Based on the historical predicted power and historical measured power of new energy and load in each characteristic period, a dynamic clustering algorithm is used to perform nearest neighbor search to obtain the deviation of new energy predicted power and load predicted power in each characteristic period of the evaluation day. S300: Based on the predicted power deviation of new energy sources and the predicted power deviation of load corresponding to each characteristic time period, determine the target time when the actual net load is the largest within the assessment day, and obtain the negative net load prediction deviation corresponding to the target time. S400: Based on the negative deviation of the net load prediction and the preset maximum power deficit of unit components of different power types under unit component failure conditions, the first-level standby capacity requirement is calculated. S500: Based on the preset maximum power deficit of multiple components of different power types under multi-component fault conditions, and combined with the sharing rules of the maximum power deficit of unit components in the first-level reserve capacity requirement, the second-level reserve capacity requirement is calculated. S600: Based on weather information, estimate the expected power reduction of new energy sources at the target time under weather changes, and combine the expected power reduction with the first-level reserve capacity demand and the second-level reserve capacity demand to obtain the third-level reserve capacity demand.
[0022] Before elaborating on the specific plan, it should be noted that the reserve requirement comes from all unplanned power disturbances from load, wind power, and photovoltaics.
[0023] Hereinafter referred to as any time The load power minus the wind power and photovoltaic power generation at the corresponding moment is: The net load power at any given time. The theoretical value of the grid's reserve power demand at any given time is equal to the difference between the actual net load power at that time and the net load power predicted when the generation plan was formulated. Furthermore, considering that the net load is provided by controllable resources such as conventional power sources, adjustable energy storage, and virtual power plants, therefore... The theoretical formula for calculating the reserve capacity of the power grid at any given time is shown below: (1) In the formula, They are respectively The actual net load power at any given time and the net load power predicted when the power generation plan was formulated; for Planned power generation capacity of controllable resources in the power grid at any given time; This represents the length of the daily load and generation schedule time series. For example, typically one point every 15 minutes. .
[0024] In actual operation, the maximum generating capacity of controllable power generation resources in the power grid is fixed at any given time. Therefore, the time when controllable resources can provide the lowest reserve capacity within a day is... This must be the moment when the actual net load is at its maximum; therefore, the daily reserve capacity requirement needs to be calculated in power grid dispatching. Corresponding to The system reserve capacity requirement at any given time is expressed mathematically as follows: ; (2) In the formula, the expression Output The maximum value corresponding to value.
[0025] Based on the above analysis, this invention will determine the daily reserve capacity requirement value. The estimation process is divided into two stages.
[0026] The first stage is the identification moment. And estimate under normal operating conditions Actual net load at time The challenge lies in how to account for the temporal misalignment of wind, solar, and load peaks and troughs, and more accurately estimate the deviation between actual wind and solar power and predicted values at different times. To address this, refer to... Figure 1 The key solution proposed in steps S100-S300 is a negative bias estimation method for net load prediction during characteristic periods based on dynamic clustering.
[0027] The second phase considers the impact of unforeseen events such as conventional power source failures and renewable energy shutdowns caused by extreme weather, and adjusts the planned power generation capacity of controllable grid resources accordingly. This involves completing a comprehensive estimate of the daily reserve capacity requirement value R. One of the technical challenges at this stage is how to account for the differences in predictability of different events to form a tiered assessment of reserve capacity. A second challenge is how to account for the independence and correlation of different power supply failures, and how to consider the sharing of contingency reserve requirements during reserve estimation. To address this, refer to... Figure 1 The key solution proposed in the S400-S600 steps is a method for calculating reserve capacity requirements that takes into account meteorological relevance and hierarchical sharing.
[0028] In summary, such as Figure 2 As shown, the dynamic calculation method for reserve capacity demand of high-proportion new energy power systems proposed in this invention includes two main frameworks.
[0029] Example 2, based on the above implementation method, referring to Figures 1-2 This invention provides a method for calculating the reserve capacity demand of a high-proportion new energy power system that considers hierarchical sharing, as one embodiment of the present invention.
[0030] S100: Acquire power grid operation load data, divide the characteristic time periods for reserve capacity demand calculation, and establish a sample library of wind and solar load prediction and actual measurement for the characteristic time periods; It should be noted that the moment of maximum actual net load during the day is always when the load is relatively high and the power generation of new energy sources such as wind power and photovoltaics is relatively low. Therefore, it must fall within one of the morning peak, noon peak, and evening peak periods of the load curve.
[0031] Specifically, considering the morning / afternoon ramp-up characteristics of photovoltaic power and the midnight peak characteristics of charging load, the time period division can be appropriately relaxed. For example, the following three characteristic time periods are proposed:
T1
T2
T3
[0032] Taking the wind power prediction / measured sample library WT1 for time period T1 as an example, its first Sample The structure is as follows: (3) In the formula, , The first Measured wind power time series and predicted wind power time series for the T1 period; For the first The date on which each sample occurred.
[0033] In this application's implementation, the historical sample database is also updated on a rolling basis, specifically as follows: Collect the day-ahead forecast power series and the day-ahead measured power series of new energy sources and loads over several historical days; According to the rules for dividing characteristic time periods, the predicted power sequence of the day and the measured power sequence of the day are cropped into time periods to construct a historical sample library corresponding to each characteristic time period; The historical sample databases are updated on a first-in, first-out (FIFO) basis to ensure that each database always retains historical samples from a fixed number of days prior to the evaluation date.
[0034] For example, each sample library maintains a fixed sample size N, typically N=60. It is updated on a first-in, first-out (FIFO) basis during scheduling, meaning it always retains historical samples from [evaluation day - 1, evaluation day - N] days. This sample library will be used for dynamic estimation of prediction errors on the evaluation day.
[0035] S200: Based on the historical predicted power and historical measured power of new energy and load within each characteristic time period, a dynamic clustering algorithm is used to perform nearest neighbor search to obtain the deviation of new energy predicted power and load predicted power within each characteristic time period of the evaluation day.
[0036] Specifically, for the S100-based technical solution, the evaluation date is recorded as... The wind power forecast time series for that day is known to be... The photovoltaic power prediction time series is The load forecast time series is Similarly, respectively for Extract the predicted power feature segments belonging to the characteristic time periods T1, T2, and T3, and denote them as follows: Estimate the prediction bias of each of the six power characteristic sequences in a specific direction using the following steps.
[0037] In this embodiment of the application, S200 uses a dynamic clustering algorithm to perform nearest neighbor search to obtain the predicted power deviation of new energy sources and the predicted power deviation of load for each characteristic period of the evaluation day, including the following steps A1-A3: A1: For any characteristic time period, calculate the weighted distance between the predicted power sequence of the evaluation day within the corresponding characteristic time period and the predicted power sequences of each historical sample in the corresponding historical sample library; A2: Sort according to the weighted distance and select a set number of nearest historical samples to form a nearest neighbor set; A3: The weighted average of the deviations between the measured power sequence and the predicted power sequence corresponding to each historical sample in the nearest neighbor set is used to calculate the positive deviation sequence of the predicted power of new energy sources and the negative deviation sequence of the predicted power of load within the corresponding characteristic time period.
[0038] For example, steps A1-A3, taking the characteristic time period T1 as an example: ① Positive deviation of wind power forecast: Each sample was retrieved one by one from the historical sample database WT1. Calculate its relationship with Wind power forecast segment for the T1 time period. Weighted distance : (4) In the formula, The weighting coefficients are those derived from sample k. The day before Heaven, then take ,otherwise Typical values can be taken as follows: .
[0039] All in ascending order Sort the samples and take the first p samples to form a whole. nearest neighbor set Typical values can be taken as follows: .
[0040] The aforementioned nearest neighbor search process and results are based on a dynamically changing sample library and power prediction fragments, hence the term dynamic clustering.
[0041] but, Positive deviation segment of wind power forecast during the T1 period of the day. Estimate using the following formula: (5) ② Estimate the positive deviation of photovoltaic power forecast: The method for estimating the positive deviation of photovoltaic power forecasting is the same as the previous step for estimating the positive deviation of wind power forecasting. The only difference is the replacement of the corresponding historical sample database. Positive deviation segment of photovoltaic power forecast during the T1 period of the day. Estimate using the following formula: (6) ③ Estimated negative deviation of load power forecast: The method for estimating the negative deviation of load power forecasting remains the same as the process for estimating the negative deviation of wind power forecasting. However, Negative deviation segment of load power forecast during the T1 period of the day The estimation formula is slightly different, as follows: (7) Repeat the above steps to complete the wind / solar / load prediction bias estimation for the T2 and T3 characteristic periods.
[0042] S300: Based on the predicted power deviation of new energy sources and the predicted power deviation of load corresponding to each characteristic time period, determine the target time when the actual net load is the largest within the assessment day, and obtain the negative deviation of net load prediction corresponding to the target time.
[0043] In this embodiment of the application, step S300 determines the target time with the largest actual net load within the assessment day based on the new energy predicted power deviation and load predicted power deviation corresponding to each characteristic time period, and obtains the negative net load predicted deviation corresponding to the target time, including the following steps B1-B2: B1: For any characteristic period, the positive deviation sequence of the new energy predicted power and the negative deviation sequence of the load predicted power within the corresponding characteristic period are superimposed on the predicted net load sequence of the assessment date to obtain the actual net load estimation sequence for the corresponding characteristic period. B2: Compare each value in the actual net load estimation sequence within all characteristic time periods, determine the time corresponding to the maximum value in each actual net load estimation sequence as the target time, and take the difference between the maximum value and the predicted net load at the target time as the negative deviation of the net load prediction corresponding to the target time.
[0044] Specifically, for B1 and B2, positive deviations (actual power lower than predicted value) are used for wind and solar power generation forecasts and negative deviations (actual load higher than predicted value) are used for load forecasts in all three characteristic time periods. Maximum actual net load of the day Negative deviation of net load power forecast The estimate can be expressed by the formula: (8) It should be noted that steps S100-S300 mainly combine the peak and valley load patterns to divide characteristic time periods, and use dynamic clustering algorithms to accurately estimate the prediction deviation of wind power and photovoltaic power. This effectively addresses the strong fluctuations and large errors of new energy sources, solves the problem of calculation distortion in traditional methods, and provides accurate data support for reserve capacity configuration.
[0045] Furthermore, it should be noted that traditionally, dispatching operations divide grid reserve requirements into two levels: spinning reserve (hot reserve) and cold reserve. Hot reserve response time is within minutes, and cold reserve also needs to be able to start generating power within 10 minutes to cope with emergency reserve needs. Therefore, only fast-response resources such as peak-shaving gas turbines, hydropower units, and pumped storage units can be considered. However, in high-proportion renewable energy power systems, renewable energy outages or sharp power drops caused by extreme weather involve many power plants with large capacities. If all of them are reserved according to traditional fast-response reserve, it will result in significant operating reserve costs and construction investment requirements. In fact, combined with rolling weather forecasts, it is often possible to accurately predict and warn of large-scale disturbances in renewable energy power caused by extreme weather several hours in advance. Therefore, some reserve resources that require longer preparation time, such as large industrial users, can also be included in the corresponding emergency reserve resource queue.
[0046] Accordingly, this invention proposes the following alternative three-level partitioning method and provides a hierarchical demand evaluation algorithm.
[0047] Level I standby (Level 1 standby): Standby capacity that can be adjusted to the required level within minutes after receiving a control command, without prior notice, corresponding to the existing concepts of hot standby and spinning standby.
[0048] Level II standby (second-level standby): Requires prior notification, and can accept control commands 10-15 minutes after receiving the prior notification. The standby capacity can be adjusted to the required level within minutes after receiving the control commands, corresponding to the existing concept of emergency cold standby.
[0049] Level III standby (third-level standby): Requires prior notification, and can accept control commands one hour after receiving the prior notification, and can adjust the standby capacity to the required level within minutes after receiving the control commands.
[0050] S400: Based on the negative deviation of the net load prediction and the preset maximum power deficit of unit components of different power types under unit component failure conditions, the first-level standby capacity requirement is calculated.
[0051] In this embodiment of the application, S400 specifically includes the following steps C1-C3: C1: Obtain the single-unit planned power of the conventional generator set in operation at the target time, the single-pole planned power of the inter-regional DC transmission, the single-circuit predicted power of the on-site collection line of a single wind farm, and the single-circuit predicted power of the on-site collection line of a single photovoltaic power station. C2: Select the maximum value from the single-unit planned power, single-pole planned power, and single-circuit predicted power of each field collection line, and use it as the maximum power deficit of the unit; C3: Add the negative deviation of the net load forecast corresponding to the target time to the maximum power deficit of the unit to obtain the first-level reserve capacity requirement.
[0052] Specifically, for C1-C3, the Level I reserve is estimated based on the net load forecast deviation calculated by S200-S300 and the impact of the N-1 fault, which has a high probability of occurrence. The calculation formula is as follows: (9) In the formula, , These represent any one of the conventional generator sets currently in operation within the system. Any DC transmission converter station in operation exist Planned power at any given time; , They represent any wind farm in the system. And any photovoltaic power station exist Predicted power at time step. , They represent wind farms Photovoltaic power station The number of collection lines within the station.
[0053] S500: Based on the preset maximum power deficit of multiple components of different power types under multi-component fault conditions, and combined with the sharing rules of the maximum power deficit of unit components in the first-level reserve capacity requirement, the second-level reserve capacity requirement is calculated.
[0054] In this embodiment of the application, S500 specifically includes the following steps D1-D3: D1: Obtain the planned power of dual-machine failure of conventional generator sets in operation at the target time, the planned power of bipolar blocking failure of inter-regional DC transmission, and the predicted power of a single new energy power station completely disconnected from the grid. D2: Select the maximum value from the planned power of dual-machine failure, the planned power of bipolar blocking failure, and the predicted power of a single new energy power station being completely disconnected from the grid, as the maximum power deficit of multiple components; D3: Calculate the difference between the required reserve for each type of severe fault condition and the reserve reserved for a single fault of the same type in the first level; take the maximum value of the difference as the reserve capacity requirement for the second level.
[0055] Specifically, in D1-D3, Level II reserve is estimated based on conventional power supply N-2 faults, DC bipolar blocking, and single renewable energy plant disconnection. Considering that the above types of power supply accidents fully share resources, and deducting the emergency reserve reserved in Level I reserve, we have: (10) It should be noted that since Level I has already reserved reserves for the less severe N-1 faults and fully shares resources, these reserves can offset part of the capacity required for more severe faults. The deduction is not reflected in the difference between each item and the corresponding item in Level I.
[0056] S600: Based on weather information, estimate the expected power reduction of new energy sources at the target time under weather changes, and combine the expected power reduction with the first-level reserve capacity demand and the second-level reserve capacity demand to obtain the third-level reserve capacity demand.
[0057] In this embodiment of the application, S600 estimates the expected power reduction of new energy sources at the target time based on weather information, and combines the expected power reduction with the first-level reserve capacity demand and the second-level reserve capacity demand to calculate the third-level reserve capacity demand, including the following steps E1-E3: E1: Based on rolling weather forecast information, identify weather change events, and combine the range of new energy power plants covered by the weather change events with historical similar samples to estimate the expected power reduction of new energy at the target time due to weather change. E2: Calculate the difference between the expected power reduction and the sum of the Tier 1 and Tier 2 reserve capacity requirements; E3: If the difference is greater than zero, the difference will be used as the third-level reserve capacity requirement; if the difference is less than or equal to zero, the third-level reserve capacity requirement will be set to zero.
[0058] Specifically, in E1-E3, the dispatcher identifies sudden weather changes based on rolling weather forecasts and, in conjunction with the new energy power stations covered by the weather changes, estimates the impact using methods such as historical sample analogy. The expected value of the decrease in renewable energy power compared to the planned value at any given time is denoted as . ; It should be noted that in this embodiment, sudden weather changes refer to drastic changes in meteorological elements on a minute- or hour-scale timescale, which may cause a significant drop in wind and solar power output in a short period of time, and the degree of this drop is significantly beyond the range of conventional prediction errors. For example, the average wind speed decreases by Δv ≥ 8 m / s within 10 minutes, accompanied by a sudden change in wind direction angle > 60°.
[0059] It should be noted that the prediction of expected values involves other scheduling technology fields and may employ methods such as distance-weighted averaging; however, this embodiment will not discuss its prediction methods. Under the known conditions, the calculation method for Level III reserve is given as follows: (11) That is, if the difference is greater than zero, the difference will be used as the third-level reserve capacity requirement; if the difference is less than or equal to zero, the third-level reserve capacity requirement will be set to zero.
[0060] It should be noted that the S400-S600 steps are calculation methods that consider meteorological factors and tiered sharing of reserve capacity requirements, constructing a tiered sharing calculation mechanism that balances grid security and operational economy. This approach comprehensively considers multiple risks, including renewable energy power disturbances, load fluctuations, conventional power source failures, and cascading grid disconnections of renewable energy sources under extreme weather conditions. Based on differences in disturbance levels and response speeds, tiered calculations and resource sharing are performed, ensuring grid power balance while avoiding reserve capacity redundancy and rationally controlling the total reserve scale. This enables accurate dynamic calculation of reserve capacity in high-proportion renewable energy power systems.
[0061] In summary, S100-S600 divides the day into several characteristic time periods based on the peak-valley characteristics of the load. A sample library is constructed based on historical measured and predicted samples of wind power, photovoltaic power, and load in each time period. A dynamic clustering nearest neighbor search algorithm is introduced to perform weighted estimation of power deviations of various types, locating the moment with the largest actual net load during the day, and achieving dynamic prediction of the negative deviation of the net load forecast at the corresponding moment. Simultaneously, considering the disturbance characteristics of conventional power source failures, DC blocking, and the cascading disconnection of renewable energy sources under extreme weather conditions, a new reserve resource classification is introduced. Based on the sharing rules of differentiated disturbances, a method for calculating the reserve demand capacity at each level is proposed, balancing grid security and reserve cost control.
[0062] Example 3, referring to Figure 1 In conjunction with Tables 1-4, based on the above embodiments, this embodiment provides a calculation example of a method for calculating the reserve capacity demand of a high-proportion new energy power system considering hierarchical sharing, in order to verify the feasibility of our solution.
[0063] Taking the operation data of a power grid with a high proportion of new energy as an example, with June 5th of a certain year as the evaluation date. Based on the parameter N=60 recommended by S100, the sample database is dynamically updated. The samples in the sample database are from the time series of the predicted values of wind, solar and load days and the actual values of the day from April 6th to June 4th of the current year, and are composed of samples according to formula (3).
[0064] Taking wind power data from the [T1] time period (9:00-12:00) as an example, this paper illustrates the process of estimating the positive bias of wind power prediction based on dynamic clustering: Sample data for the corresponding time period were selected from the sample database according to the time period division rules. The weighted distance between the wind power forecast time series curve for time period T1 on June 5th and the forecast power time series curves for the same time period in the sample database was calculated. Table 1 shows the forecast power time series curve for June 5th and sample instances for the corresponding time period on April 19th in the sample database.
[0065] Table 1: Forecast power time series curves for the T1 period on April 19 and June 5 (unit: MW)
[0066] According to formula (4), the weighted distance of the wind power forecast time series curve between April 19 and June 5 is 0.002136628.
[0067] After calculating the weighted distance between all samples in the wind power sample library for the T1 period and the wind power prediction sequence for the T1 period on June 5, the three closest dates were April 19, May 29, and April 15.
[0068] Using the same method, weighted distances were calculated for the time series curves of wind, solar, and load for each time period, and the three dates corresponding to each time period were obtained. The results are shown in Table 2.
[0069] Table 2: Weighted nearest distance to wind / solar / load for various time periods and corresponding dates
[0070] As can be seen from the table above, similar samples from different time periods come from different historical days and are dynamically changing.
[0071] The prediction deviation was estimated using formulas (5)-(7). Taking the wind power data at 11:55 during the [T1] period (9:00-12:00) as an example, the positive deviation between the measured power curve and the predicted power curve of wind power on April 19, May 29 and April 15 was calculated according to formula (5), and the statistical average result of 72.66MW was taken as the positive deviation of prediction on the evaluation date of June 5. Using the same method, the positive deviation of prediction for wind power and photovoltaic power at each time point and the negative deviation of prediction for load at each time point were calculated, and the results are shown in Table 3.
[0072] Table 3: Positive deviations in wind / solar forecasts and negative deviations in load forecasts for the T1 period on June 5 (unit: MW)
[0073] Using the same method, the positive deviations of wind and solar forecasts and the negative deviations of load forecasts for the entire time period (T1, T2, T3) on the assessment date of June 5 were calculated.
[0074] According to formula (6), the analysis shows that the maximum net load deviation on that day occurred at 11:15, with a corresponding value of 131.67MW.
[0075] Based on the above analysis and calculations, the maximum net load deviation for the province on June 5th, the assessment date, occurred at 11:15, corresponding to a value of 131.67 MW. Key information parameters provided by relevant departments at that time, including the start-up plans of conventional generating units, DC transmission power, and renewable energy power plants, are shown in Table 4.
[0076] Table 4: Key Information Parameters at 11:15 AM on June 5th
[0077] Based on the hierarchical demand assessment algorithm proposed by steps S400-S600, reserve requirements at each level can be calculated progressively: First-level standby: According to formula (9), we can obtain: Second-level backup According to formula (10), we can obtain: Level 3 standby The weather was good that day, and there was no typhoon warning, so TR=0. The calculation for the third-level standby is as follows: Based on the above examples, it can be seen that the solution of the present invention is applicable to regional power grids with a high proportion of new energy grid connection, covering the daily dispatch and operation of the power grid, power generation reserve planning, real-time power balance management and control, and power market reserve auxiliary service accounting scenarios. It focuses on the reserve demand calculation of new disturbances such as new energy power fluctuations, prediction deviations, and grid disconnection due to extreme weather, while also taking into account the reserve demand of traditional disturbances such as load fluctuations, conventional unit failures, and DC faults.
[0078] Example 4 illustrates a schematic scheme for calculating the reserve capacity demand of a high-proportion renewable energy power system considering tiered sharing. It should be noted that the technical solution of this system for calculating the reserve capacity demand of a high-proportion renewable energy power system considering tiered sharing is based on the same concept as the aforementioned method for calculating the reserve capacity demand of a high-proportion renewable energy power system considering tiered sharing. Details not described in detail in this embodiment can be found in the description of the aforementioned method for calculating the reserve capacity demand of a high-proportion renewable energy power system considering tiered sharing.
[0079] This embodiment provides a system for calculating the reserve capacity demand of a high-proportion renewable energy power system that considers hierarchical sharing, including: The sample library construction module is used to acquire power grid operation load data, divide the characteristic time periods for reserve capacity demand calculation, and establish a sample library of wind and solar load prediction and actual measurement for the characteristic time periods. The clustering module is used to perform nearest neighbor search using a dynamic clustering algorithm based on the historical predicted power and historical measured power of new energy and load in each characteristic time period, to obtain the deviation of new energy predicted power and load predicted power in each characteristic time period of the evaluation day. The determination module is used to determine the target time with the largest actual net load within the assessment day based on the new energy predicted power deviation and load predicted power deviation corresponding to each characteristic time period, and to obtain the net load predicted negative deviation corresponding to the target time. The first calculation module is used to calculate the first-level reserve capacity requirement based on the net load prediction negative deviation and the preset maximum power deficit of unit components of different power types under unit component failure conditions. The second calculation module is used to calculate the second-level reserve capacity requirement based on the maximum power deficit of multi-element components of different power types under the preset multi-element fault conditions, and in combination with the sharing rules of the maximum power deficit of unit components in the first-level reserve capacity requirement. The third calculation module is used to estimate the expected power reduction of new energy sources at the target time under weather changes based on weather information, and to combine the expected power reduction with the first-level reserve capacity demand and the second-level reserve capacity demand to obtain the third-level reserve capacity demand.
[0080] This embodiment also provides a computer device applicable to the calculation of reserve capacity demand of a high-proportion renewable energy power system considering hierarchical sharing, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the method for calculating reserve capacity demand of a high-proportion renewable energy power system considering hierarchical sharing as proposed in the above embodiment.
[0081] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements a method for calculating the reserve capacity demand of a high-proportion new energy power system considering hierarchical sharing, as proposed in the above embodiments.
[0082] The storage medium proposed in this embodiment and the method for calculating the reserve capacity demand of a high-proportion new energy power system considering hierarchical sharing proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.
[0083] From the above description of the implementation methods, those skilled in the art will clearly understand that the present invention can be implemented using software and necessary general-purpose hardware. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.
[0084] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.
Claims
1. A method for calculating the reserve capacity demand of a high-proportion renewable energy power system considering hierarchical sharing, characterized in that, include: Acquire power grid operation load data, divide the characteristic time periods for reserve capacity demand calculation, and establish a sample library of wind and solar load prediction and actual measurement for the characteristic time periods; Based on the historical predicted power and historical measured power of new energy and load in each characteristic period, a dynamic clustering algorithm is used to perform nearest neighbor search to obtain the deviation of new energy predicted power and load predicted power in each characteristic period of the evaluation day. Based on the predicted power deviation of new energy sources and the predicted power deviation of load corresponding to each characteristic time period, the target time with the largest actual net load within the assessment day is determined, and the negative net load prediction deviation corresponding to the target time is obtained. Based on the negative deviation of the net load prediction and the preset maximum power deficit of unit components of different power types under unit component failure conditions, the first-level standby capacity requirement is calculated. Based on the preset maximum power deficit of multi-components of different power types under multi-component fault conditions, and combined with the sharing rules of the maximum power deficit of unit components in the first-level reserve capacity requirement, the second-level reserve capacity requirement is calculated. Based on weather information, the expected power reduction of new energy sources at the target time under weather changes is estimated, and the expected power reduction is combined with the first-level reserve capacity demand and the second-level reserve capacity demand to obtain the third-level reserve capacity demand.
2. The method for calculating the reserve capacity demand of a high-proportion renewable energy power system considering hierarchical sharing as described in claim 1, characterized in that, The nearest neighbor search is performed using a dynamic clustering algorithm to obtain the predicted power deviation of new energy sources and the predicted power deviation of load for each characteristic period of the evaluation day, including: For any characteristic time period, calculate the weighted distance between the predicted power sequence of the evaluation day within the corresponding characteristic time period and the predicted power sequences of each historical sample in the corresponding historical sample library; Sort the samples according to the weighted distance and select a set number of nearby historical samples to form a nearest neighbor set; By using the deviation between the measured power sequence and the predicted power sequence corresponding to each historical sample in the nearest neighbor set, a weighted average is calculated to obtain the positive deviation sequence of the predicted power of new energy sources and the negative deviation sequence of the predicted power of load within the corresponding characteristic time period.
3. The method for calculating the reserve capacity demand of a high-proportion new energy power system considering hierarchical sharing as described in claim 2, characterized in that, Based on the predicted power deviation of new energy sources and the predicted power deviation of load corresponding to each characteristic time period, the target time with the largest actual net load within the assessment day is determined, and the negative net load prediction deviation corresponding to the target time is obtained, including: For any characteristic period, the positive deviation sequence of the new energy predicted power and the negative deviation sequence of the load predicted power within the corresponding characteristic period are superimposed on the predicted net load sequence of the assessment date to obtain the actual net load estimation sequence for the corresponding characteristic period. By comparing the values in the actual net load estimation sequence within all characteristic time periods, the time corresponding to the maximum value in each actual net load estimation sequence is determined as the target time, and the difference between the maximum value and the predicted net load at the target time is taken as the negative deviation of the net load prediction corresponding to the target time.
4. The method for calculating the reserve capacity demand of a high-proportion new energy power system considering hierarchical sharing as described in claim 3, characterized in that, Based on the negative deviation of the net load forecast and the preset maximum power deficit of unit types under different power supply conditions, the first-level reserve capacity requirement is calculated, including: The target time is obtained as follows: the single-unit planned power of conventional generator sets in operation, the single-pole planned power of inter-regional DC transmission, the single-circuit predicted power of the on-site collection line of a single wind farm, and the single-circuit predicted power of the on-site collection line of a single photovoltaic power station. The maximum value is selected from the single-unit planned power, single-pole planned power, and single-circuit predicted power of each field collection line, and is taken as the maximum power deficit of the unit. The negative deviation of the net load forecast corresponding to the target time is added to the maximum power deficit of the unit to obtain the first-level reserve capacity requirement.
5. The method for calculating the reserve capacity demand of a high-proportion new energy power system considering hierarchical sharing as described in claim 4, characterized in that, Based on the preset maximum power deficit of multiple components of different power types under multi-component fault conditions, and combined with the sharing rules of the maximum power deficit of unit components in the first-level reserve capacity requirement, the second-level reserve capacity requirement is calculated, including: The target time is to obtain the planned power of dual-machine failure of conventional generator units in operation, the planned power of bipolar blocking failure of inter-regional DC transmission, and the predicted power of a single new energy power station completely disconnected from the grid. The maximum value is selected from the planned power of dual-machine failure, the planned power of bipolar blocking failure, and the predicted power of a single new energy power station being completely disconnected from the grid, and is used as the maximum power deficit of multiple components. Calculate the difference between the required reserve for each type of severe fault and the reserve reserved for a single fault of the same type in the first level; take the maximum value of the difference as the reserve capacity requirement for the second level.
6. The method for calculating the reserve capacity demand of a high-proportion new energy power system considering hierarchical sharing as described in claim 5, characterized in that, Based on weather information, the expected power reduction of new energy sources at the target time under weather changes is estimated. This expected power reduction is then combined with the first and second tier reserve capacity requirements to calculate the third tier reserve capacity requirement, which includes: Based on rolling weather forecast information, weather change events are identified, and combined with the range of new energy power plants covered by the weather change events and historical similar samples, the expected value of the power reduction of new energy caused by weather change at the target time is estimated. Calculate the difference between the expected power reduction and the sum of the Tier 1 and Tier 2 reserve capacity requirements; If the difference is greater than zero, the difference will be used as the third-level reserve capacity requirement; if the difference is less than or equal to zero, the third-level reserve capacity requirement will be set to zero.
7. The method for calculating the reserve capacity demand of a high-proportion new energy power system considering hierarchical sharing as described in claim 6, characterized in that, Before calculating the weighted distance, the historical sample database is also continuously updated, specifically as follows: Collect the day-ahead forecast power series and the day-ahead measured power series of new energy sources and loads over several historical days; According to the rules for dividing characteristic time periods, the predicted power sequence of the day and the measured power sequence of the day are cropped into time periods to construct a historical sample library corresponding to each characteristic time period; The historical sample databases are updated on a first-in, first-out (FIFO) basis to ensure that each database always retains historical samples from a fixed number of days prior to the evaluation date.
8. A system for calculating the reserve capacity demand of a high-proportion renewable energy power system considering hierarchical sharing, using the method described in any one of claims 1-7, characterized in that, include: The sample library construction module is used to acquire power grid operation load data, divide the characteristic time periods for reserve capacity demand calculation, and establish a sample library of wind and solar load prediction and actual measurement for the characteristic time periods. The clustering module is used to perform nearest neighbor search using a dynamic clustering algorithm based on the historical predicted power and historical measured power of new energy and load in each characteristic time period, to obtain the deviation of new energy predicted power and load predicted power in each characteristic time period of the evaluation day. The determination module is used to determine the target time with the largest actual net load within the assessment day based on the new energy predicted power deviation and load predicted power deviation corresponding to each characteristic time period, and to obtain the net load predicted negative deviation corresponding to the target time. The first calculation module is used to calculate the first-level reserve capacity requirement based on the net load prediction negative deviation and the preset maximum power deficit of unit components of different power types under unit component failure conditions. The second calculation module is used to calculate the second-level reserve capacity requirement based on the maximum power deficit of multi-element components of different power types under the preset multi-element fault conditions, and in combination with the sharing rules of the maximum power deficit of unit components in the first-level reserve capacity requirement. The third calculation module is used to estimate the expected power reduction of new energy sources at the target time under weather changes based on weather information, and to combine the expected power reduction with the first-level reserve capacity demand and the second-level reserve capacity demand to obtain the third-level reserve capacity demand.
9. A computer device, characterized in that, include: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, they implement the steps of the method for calculating the reserve capacity demand of a high proportion of new energy power system considering hierarchical sharing as described in any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that, It stores computer-executable instructions, which, when executed by a processor, implement the steps of the method for calculating the reserve capacity demand of a high-proportion new energy power system considering hierarchical sharing as described in any one of claims 1 to 7.
