Wind power drop prediction method and device and readable storage medium
By using the wind power descent gradient probability distribution prediction method, the problems of high computational complexity and high cost in existing technologies are solved, and low-complexity wind power drop prediction is achieved, ensuring continuous operation of production equipment and reducing carbon emissions.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-12
- Publication Date
- 2026-04-03
AI Technical Summary
Existing wind power drop prediction methods require a large amount of external data, resulting in high computational complexity and cost, making it difficult to meet the needs of zero-carbon and low-carbon production under high wind power consumption.
The probability distribution of wind power descent gradient is used for prediction. By predicting the probability distribution of wind power descent gradient within the target time period, it can be determined whether the production equipment can continue to operate, thus avoiding the need to acquire meteorological data and traverse scenarios, and reducing computational complexity.
It achieves low-cost and low-complexity wind power drop prediction, and can provide early warning of significant wind power drops, avoiding production interruptions and reducing carbon emissions.
Smart Images
Figure CN121787758A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of wind farm technology, and in particular to a method, device and readable storage medium for predicting wind power drop. Background Technology
[0002] Wind energy, as a clean and renewable energy source, is receiving increasing attention from countries around the world. A wind turbine, or simply a wind power unit, is a device that uses wind energy to generate electricity.
[0003] Wind power refers to the effective net power that can be transmitted to the power grid after wind turbines capture wind energy, convert it into electrical energy through a conversion system, and after deducting losses from the wind turbines themselves, conversion losses, and line losses. It is usually measured in kilowatts (kW) or megawatts (MW). Wind power drop prediction refers to the ability to predict and issue early warnings about potential significant and rapid drops in wind power by monitoring and analyzing relevant data from wind farms. Common methods for wind power drop prediction include acquiring external data such as meteorological data and wind turbine operating data, and then using this data to predict wind power drops. Meteorological data includes wind speed, wind direction, and turbulence intensity.
[0004] The aforementioned wind power drop prediction method requires a large amount of external data, resulting in high computational complexity and cost. Summary of the Invention
[0005] This application provides a wind power drop prediction method, device, and readable storage medium. The electronic device predicts wind power drop based on the probability distribution of wind power descent gradient, without the need to acquire meteorological data, etc., and has low computational complexity, thereby reducing the cost of wind power drop prediction.
[0006] In a first aspect, this application provides a method for predicting wind power drop, including: Predict the probability distribution of wind power decline gradient of target object within a target time period. The probability distribution of wind power decline gradient is used to indicate the probability statistics of different wind power decline gradients of target object within the target time period. The wind power decline gradient is used to indicate the magnitude of wind power decline of target object per unit time. Based on the probability distribution of the wind power descent gradient, it is determined whether the production equipment can produce continuously within the target time period. The production equipment is the equipment that uses at least the electrical energy provided by the target object to produce within the target time period. When the production equipment cannot produce continuously during the target time period, it is determined that the target object has experienced a drop in wind power during the target time period.
[0007] Secondly, this application provides a wind power drop prediction device, comprising: The prediction module is used to predict the probability distribution of the wind power decline gradient of the target object within a target time period. The probability distribution of the wind power decline gradient is used to indicate the probability statistics of different wind power decline gradients of the target object within the target time period. The wind power decline gradient is used to indicate the magnitude of the decline of the wind power of the target object per unit time. The processing module is used to determine whether the production equipment can continue to produce within the target time period based on the probability distribution of the wind power decrease gradient, wherein the production equipment is the equipment that uses at least the electrical energy provided by the target object to produce within the target time period; The determination module is used to determine that when the production equipment cannot produce continuously during the target time period, the target object experiences a drop in wind power during the target time period.
[0008] Thirdly, this application provides an electronic device, including: a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the electronic device implements the method described in the first aspect or various possible implementations of the first aspect.
[0009] Fourthly, this application provides a computer-readable storage medium storing computer instructions that, when executed by a processor, are used to implement the method described in the first aspect or various possible implementations of the first aspect.
[0010] Fifthly, this application provides a computer program product comprising a computing program, which, when executed by a processor, implements the method described in the first aspect or various possible implementations of the first aspect.
[0011] This application provides a wind power drop prediction method, device, and readable storage medium. The electronic device predicts the probability distribution of wind power descent gradient for a target object within a target time period. Based on this probability distribution, it determines whether production equipment can operate continuously within the target time period. Production equipment is defined as equipment that utilizes at least the electrical energy provided by the target object for production within the target time period. When production equipment cannot operate continuously within the target time period, it is determined that a wind power drop has occurred for the target object during that period. Using this approach, the electronic device predicts wind power drop based on the probability distribution of wind power descent gradient, eliminating the need to acquire meteorological data or traverse numerous scenarios, resulting in low computational complexity and reducing the cost of wind power drop prediction. Attached Figure Description
[0012] To more clearly illustrate the technical solutions in this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0013] Figure 1 This is a flowchart of the wind power drop prediction method provided in this application; Figure 2 This is a flowchart illustrating the method for predicting wind power drop provided in this application, which determines the moment when production equipment cannot operate continuously. Figure 3 This is another flowchart of the wind power drop prediction method provided in this application; Figure 4 A schematic diagram of the wind power drop prediction device provided in this application; Figure 5 A schematic diagram of the structure of the electronic device provided in this application. Detailed Implementation
[0014] Production equipment, also known as load equipment, refers to core electrical equipment directly related to production, including but not limited to submerged arc furnaces and compressors. In practice, production equipment is required to operate continuously, meaning that the production process is uninterrupted, the equipment runs continuously, and the energy supply remains stable for a given period. Taking a manganese-silicon submerged arc furnace as an example, continuous production refers to a production mode that aims to "continuously and stably produce manganese-silicon alloy," minimizing planned downtime and avoiding unplanned interruptions over a longer period.
[0015] Research has found that traditional grid connection and energy storage control strategies are not entirely applicable to the actual situation of low-carbon production under high wind power consumption. This is because traditional solutions ignore the need for continuous production, which seriously affects the production capacity and technology promotion of zero-carbon and low-carbon alloys.
[0016] Wind energy is characterized by randomness and volatility, making the output power of wind farms unstable and significantly impacting the normal operation and dispatch of the power system. To achieve zero-carbon and low-carbon production, wind power drop early warning systems can predict periods of significant wind power decline in advance, allowing for proactive adjustments to grid power and energy storage equipment to minimize downtime and reduce additional carbon emissions.
[0017] Traditional wind power drop prediction methods include prediction based on external data and prediction based on models. The latter involves acquiring large amounts of external data, such as meteorological data and wind turbine operating data, and then using this data to predict wind power drop. This method requires a large amount of external data, resulting in high computational complexity and cost.
[0018] In model-based prediction methods, a model considering the uncertainty of wind power output is pre-created. This model simulates the fluctuation patterns of wind power output and outputs data on wind power output fluctuations. Then, various large-scale scenarios are generated based on the wind power output fluctuation data to cover various combinations of uncertainties on both the source and load sides. When a wind power drop warning is issued, all scenarios are traversed to determine whether a wind power drop warning is necessary. This method requires extensive scenario traversal and relies on complex control logic.
[0019] Based on this, this application provides a wind power drop prediction method, device and readable storage medium. The electronic device predicts wind power drop based on the probability distribution of wind power descent gradient, without the need to acquire meteorological data, without traversing a large number of scenarios, and with low computational complexity, thereby reducing the cost of wind power drop prediction.
[0020] The solution of this application is applied to electronic devices, such as servers, desktop computers, laptops, etc. The electronic devices can predict the probability distribution of wind power decline gradient of a target object within a target time period, and make wind power drop prediction based on the probability distribution of wind power decline gradient.
[0021] In this application, the target object can be a wind farm, a wind turbine generator, all wind farms in a region, or some wind turbine generators in a wind farm, etc.
[0022] In this application, the production equipment is high-energy-consuming equipment, including but not limited to data centers, electric arc furnaces, refrigeration equipment, pumps, boilers, etc., and the electric arc furnaces include but are not limited to manganese silicon electric arc furnaces, ferrosilicon electric arc furnaces, etc.
[0023] Figure 1 This is a flowchart of the wind power drop prediction method provided in this application. This embodiment includes: 101. Predict the probability distribution of wind power descent gradient of the target object within the target time period. The probability distribution of wind power descent gradient is used to indicate the probability statistics of different wind power descent gradients of the target object within the target time period. The wind power descent gradient is used to indicate the magnitude of the decrease in wind power of the target object per unit time.
[0024] Based on the current time, the target time period is a future time period. The starting time of the target time period can be the current time or a future time after the current time. For example, if the current time is 9:00 AM, the target time period is the next 12 hours, 24 hours, or 48 hours starting from 9:00 AM. Similarly, if the current time is 8:55 AM, the target time period is the next 6 hours, 24 hours, or 48 hours starting from 9:00 AM. Correspondingly, the historical time period is a period preceding the current time. The end time of the historical time period coincides with the start time of the target time period, or the duration between them is less than a preset duration, such as 5 minutes, 3 minutes, or 2 minutes; this application does not limit this.
[0025] In this application, the target object can be a wind farm, a wind turbine generator, or all wind farms within a region. The wind power descent gradient is used to indicate the rate of decrease in wind power of the target object per unit time. Taking a wind farm as the target object and a unit time of 1 minute as an example, if the wind power of the wind farm decreases from 20 megawatts (MW) to 15 MW within 10 minutes, then the wind power descent gradient is: (20-15) ÷ 10 = 0.5 MW / min; if the wind power decreases from 20 megawatts (MW) to 15 MW within 1 hour, then the wind power descent gradient is: (20-15) ÷ 60 = 0.083 MW / min.
[0026] The probability distribution of wind power descent gradients refers to the statistical probability of different magnitudes of wind power descent gradients occurring within a target time period. For example, if a total of 1000 wind power descents are predicted within the target time period, the number of wind power descent gradients between 0 and 0.1 MW / zone is 520 (52% probability), indicating a gradual decrease in wind power within this range; the number of wind power descent gradients between 0.1 and 0.2 MW / zone is 310 (31% probability), also indicating a relatively gradual decrease in wind power within this range; the number of wind power descent gradients between 0.2 and 0.3 MW / zone is 120 (12% probability), indicating a relatively rapid decrease in wind power within this range; the number of wind power descent gradients between 0.3 and 0.5 MW / zone is 40 (4% probability), indicating a very rapid decrease in wind power within this range; and the number of wind power descent gradients greater than 0.5 MW / zone is 10 (1% probability), indicating a sharp drop in wind power within this range.
[0027] 102. Determine whether the production equipment can produce continuously within the target time period based on the probability distribution of the wind power descent gradient. If the production equipment cannot produce continuously within the target time period, proceed to step 103; if the production equipment can produce continuously within the target time period, proceed to step 104.
[0028] In this application, continuous production refers to the uninterrupted operation of production equipment and the absence of drastic fluctuations in energy supply within a target time period. Electronic equipment determines whether continuous production is possible within the target time period based on the probability distribution of wind power descent gradients. For example, the electronic equipment determines the probability that the wind power descent gradient is greater than 0.5 MW / zone based on the probability distribution of wind power descent gradients. If this probability is greater than a preset probability, it indicates that there are many instances of sudden wind power drops within the target time period, leading to drastic fluctuations in energy supply and consequently preventing continuous production.
[0029] 103. Determine that the target object experiences a drop in wind power within the target time period.
[0030] 104. Determine that the target object will not experience a drop in wind power during the target time period.
[0031] If the electronic equipment determines that the production equipment cannot operate continuously within the target time period, then it determines that the target object experiences a drop in wind power within the target time period. If the electronic equipment determines that the production equipment can operate continuously within the target time period, then it determines that the target object will not experience a drop in wind power within the target time period.
[0032] The wind power drop prediction method provided in this application involves an electronic device predicting the probability distribution of wind power descent gradient for a target object within a target time period. Based on this probability distribution, it determines whether production equipment can operate continuously within the target time period. Production equipment is defined as equipment that utilizes at least the electrical energy provided by the target object for production within the target time period. When production equipment cannot operate continuously within the target time period, it is determined that a wind power drop has occurred at the target object. Using this method, the electronic device predicts wind power drop based on the probability distribution of wind power descent gradient, eliminating the need to acquire meteorological data or traverse numerous scenarios, resulting in low computational complexity and reducing the cost of wind power drop prediction.
[0033] Optionally, in the above embodiments, during the process of the electronic device determining whether the production equipment can operate continuously within the target time period based on the wind power descent gradient probability distribution, the electronic device first determines the state of charge (SOC) of the energy storage device, whereby the SOC indicates the ratio of the energy storage device's current remaining power to its rated capacity. Then, the electronic device determines a scheduling strategy within the target time period based on the wind power descent gradient probability distribution, the SOC, future load production plans, and grid power arrangements. If the scheduling strategy is determined, it is determined that the production equipment can operate continuously within the target time period; otherwise, it is determined that the production equipment cannot operate continuously within the target time period.
[0034] In this application, the energy suppliers for the production equipment include, but are not limited to, the target device, grid power, and energy storage devices. The energy storage devices are primarily used to fill the gap caused by the decrease in wind power at the target device, and the State of Charge (SOC) directly determines the energy storage device's replenishment capability. A higher SOC indicates that the energy storage device has sufficient remaining power and a large discharge capacity; a lower SOC indicates that the energy storage device has less remaining power and a limited discharge capacity.
[0035] In this application, the future load production plan indicates the planned power of production equipment at each time point within the target time period; the grid power arrangement indicates the grid power constraints at each time point within the target time period; the scheduling strategy indicates the output sequence of various energy suppliers within the target time period; the output sequence indicates the power consumed by energy suppliers at each time point within the target time period; and the power consumed indicates the power used by production equipment from the power supplied by energy suppliers. If the scheduling strategy can be determined, it means that at each time point within the target time period, the sum of the power consumed by all energy suppliers equals the sum of the planned power of the production equipment and the demand power of other energy demanders. Other energy demanders include, but are not limited to, auxiliary equipment of the production equipment.
[0036] Electronic equipment determines its scheduling strategy for a target time period based on the probability distribution of wind power descent gradient, state of charge, future load production plans, and grid power arrangements. In other words, by combining these factors, the electronic equipment determines the power consumption of each energy supplier at each point in time within the target time period, thus obtaining the power output sequence of various energy suppliers during that period. For example, if the target time period is 12 hours, it can be divided into 72 10-minute intervals. The power output sequence of the target entity indicates the power consumption of the target entity in each of these 72 10-minute intervals. Assuming that the target entity can provide 6MW of power at each point in a 10-minute interval, with 5MW used for production equipment, then the power consumption of the target entity at each point in that 10-minute interval is 5MW.
[0037] Based on the probability distribution of wind power descent gradient, state of charge (SOC), future load production plans, and grid power arrangements, electronic equipment can determine whether a scheduling strategy can be established for the target time period. If a scheduling strategy can be established, it indicates that multi-source collaboration can enable the production equipment to operate continuously within the target time period. If a scheduling strategy cannot be established, it indicates that even multi-source collaboration cannot ensure continuous production within the target time period, for reasons including but not limited to: a high probability of a large wind power descent gradient, excessively low SOC of energy storage devices, and insufficient power supplied by the grid.
[0038] This approach combines the probability distribution of wind power descent gradient, state of charge, future load production plans, and grid power arrangements to determine the scheduling strategy for the target time period. Based on whether the scheduling strategy can be determined, it can be determined whether the production equipment can continue to produce within the target time period. This method is fast and highly accurate.
[0039] The following section provides a detailed explanation of how electronic devices determine the scheduling strategy for the target time period based on the probability distribution of the wind power decline gradient, the state of charge, future load production plans, and grid power arrangements.
[0040] In one approach, a solver, also known as an energy storage grid power dispatch solver, is pre-deployed on the electronic equipment. This solver takes the probability distribution of wind power descent gradient, state of charge, future load production plans, and grid power arrangements as known conditions, and treats the power consumption of each energy supplier at each time point within the target time period as unknowns, then solves for these unknowns. If all unknowns can be solved, the power output time series is generated based on the solved unknowns, thus obtaining the dispatch strategy for the target time period. If all unknowns cannot be solved, it indicates that the production equipment cannot meet continuous production demands in the future target time period.
[0041] In another approach, the electronic device determines a target descent gradient based on the wind power descent gradient probability distribution and the state of charge, wherein the target descent gradient is the wind power descent gradient in the wind power descent gradient probability distribution that matches the magnitude of the state of charge. Then, the electronic device determines a scheduling strategy for the target time period based on the target descent gradient, the future load production plan, and the grid power arrangement.
[0042] Unlike the previous method, where the solver's inputs included the wind power descent gradient probability distribution, state of charge (SOC), future load production plans, and grid power arrangements, this method uses electronic equipment to determine the target descent gradient based on the wind power descent gradient probability distribution and SOC. For example, if the wind power descent gradient probability distribution shows the target object's wind power descent gradient range as 0.5 MW / min to 0.7 MW / min, and the SOC is relatively high, then the target descent gradient is the wind power descent gradient with a 70% probability, such as 0.55 MW / min, meaning a relatively gentle wind power descent gradient is selected. A relatively high SOC is also referred to as a high water level, such as SOC equal to 80% or 90%.
[0043] For example, if the wind power descent gradient probability distribution shows that the target object's wind power descent gradient ranges from 0.5 MW / min to 0.7 MW / min, and the SOC is relatively low, then the target descent gradient corresponds to the wind power descent gradient with a 10% probability, such as 0.7 MW / min. This selects the wind power descent gradient under extreme conditions, facilitating timely early warning. A relatively low SOC is also referred to as a relatively low water level, such as SOC equal to 20%, SOC equal to 30%, etc.
[0044] Understandably, since energy storage devices are also energy suppliers, they will provide power to production equipment at certain points within the target time period, causing the State of Charge (SOC) to continuously decrease. As the SOC decreases, more extreme scenarios need to be considered. That is, the lower the SOC, the earlier the warning should be issued to ensure continuous production; the higher the SOC, the better the energy storage devices can cope even with a significant drop in wind power. After a period of time, the SOC decreases, and once it reaches a certain level, extreme scenarios must be considered. This approach encourages production when the SOC is high and provides timely warnings when the SOC is low to protect production.
[0045] After the electronic equipment determines the target descent gradient, it inputs this gradient into the solver. The solver uses the target descent gradient, future load production plans, and grid power arrangements as known conditions, and treats the power consumption of each energy supplier at each time point within the target time period as unknowns, then solves for the unknowns. If all unknowns can be solved, the power output time series is generated based on the solved unknowns, thus obtaining the scheduling strategy within the target time period; if all unknowns cannot be solved, it indicates that the production equipment cannot meet the continuous production demand in the future target time period.
[0046] This approach allows electronic equipment to determine the target descent gradient based on the state of charge and the probability distribution of wind power descent gradient. Then, based on the target descent gradient, future load production plans, and grid power arrangements, it determines the scheduling strategy for the target time period. The solution does not require using the entire probability distribution of wind power descent gradient, reducing computational load and improving the efficiency of wind power drop prediction. Furthermore, it enables dynamic adjustment of the state of charge, preventing insufficient energy storage during the target time period due to sudden wind speed drops.
[0047] Optionally, in the above embodiments, during the process of the electronic device determining the scheduling strategy for the target time period based on the wind power descent gradient probability distribution, the state of charge, future load production plans, and grid power arrangements, the electronic device first determines a set of constraints. Then, under the constraints of each constraint in the set of constraints, the electronic device determines the scheduling strategy for the target time period based on the wind power descent gradient probability distribution, the state of charge, the future load production plans, and the grid power arrangements. Here, the energy storage device and the grid power supply are the energy suppliers providing power to the production equipment during the target time period.
[0048] The various constraints mentioned above will be explained in detail below.
[0049] A. First constraint.
[0050] In this application, the first constraint is used to indicate the constraint conditions of the energy storage device. This is shown in equations (1.1) and (1.2) below: Formula (1.1) Formula (1.2) in, This represents the energy storage power of the energy storage device at time t within the target time period. This indicates the upper limit of the discharge power of the energy storage device. This indicates the upper limit of the charging power of the energy storage device. This represents the state of charge of the energy storage device at time t. This indicates the lowest state of charge of the energy storage device. This represents the highest state of charge of the energy storage device. According to formulas (1.1) and (1.2), the energy storage power of the energy storage device at time t within the target time period must be between the upper limit of the discharge power and the upper limit of the charging power, that is, the energy storage power cannot exceed the upper limit of the charging and discharging power specified by the energy storage device, and the state of charge must be between the lowest state of charge and the highest state of charge of the energy storage device.
[0051] B. Second constraint.
[0052] In this application, the second constraint is used to indicate the constraint conditions for continuous production. This is shown in the following formulas (2.1), (2.2), and (2.3): Formula (2.1) = Formula (2.2) ∆ Formula (2.3) in, This represents the planned power of the production equipment at time t. This indicates the minimum continuous production power of the production equipment. This indicates the percentage of power that can be reduced. This indicates the rated production power of the production equipment. ∆ represents the planned power of the production equipment at time t-1. This indicates the maximum rate of load reduction. For example, if the rated power of a production equipment is 10MW, that's the rated power; the percentage of power that can be reduced is the rated power. If it is 10%, then the minimum continuous production power is 9MW.
[0053] C. Third constraint.
[0054] In this application, the third constraint is used to indicate the constraint conditions of the grid power, as shown in the following formula (3): Formula (3) in, This represents the length of the target time period, and t represents time t within the target time period. Indicates the power consumption of the grid. This represents the total power supplied by various energy providers. This represents the power consumption of the grid at time t. It represents the total power supplied by various energy providers at time t.
[0055] After determining the set of constraints, the electronic equipment creates an objective function. The optimization direction of the objective function can be to maximize the proportion of green electricity, maximize production capacity, etc., and this application does not limit this. The form of the objective function can be flexibly set. Subsequently, under the constraints of various conditions in the set of constraints, the electronic equipment solves the objective function based on the probability distribution of wind power descent gradient, state of charge, future load production plans, and grid power arrangements to determine the scheduling strategy within the target time period.
[0056] By adopting this approach, electronic devices consider various constraints when determining scheduling strategies, making the scheduling strategies more aligned with actual needs and avoiding theoretically feasible but unrealistic scheduling strategies, thereby improving the quality of scheduling strategies.
[0057] Optionally, in the above embodiments, when the production equipment cannot produce continuously within the target time period, after the electronic device determines that the wind power of the target object has dropped within the target time period, it also determines the moment when the production equipment cannot produce continuously. For example, please refer to... Figure 2 , Figure 2 This is a flowchart illustrating the method for predicting wind power drop provided in this application, specifically the process for determining when production equipment cannot operate continuously. This embodiment includes: 201. The electronic equipment determines that the target object experiences a drop in wind power within the target time period.
[0058] 202. The electronic device determines whether the scheduling strategy for the first sub-time period can be determined based on the probability distribution of the wind power descent gradient. If the scheduling strategy for the first sub-time period can be determined based on the probability distribution of the wind power descent gradient, then proceed to step 203; if the scheduling strategy for the first sub-time period cannot be determined based on the probability distribution of the wind power descent gradient, then proceed to step 206.
[0059] In this application, the target time period is pre-divided into multiple sub-time periods according to a preset duration, and these sub-time periods are sequentially adjacent. For example, if the target time period is 6 hours long, it is divided into 6 adjacent sub-time periods, each with a length of 1 hour. Or, if the target time period is 4 hours long, it is divided into 8 adjacent sub-time periods, each with a length of half an hour. The first and second sub-time periods are two adjacent sub-time periods within the multiple sub-time periods.
[0060] After dividing the time period into sub-time periods, the electronic equipment sequentially determines whether it can determine the scheduling strategy for each sub-time period based on the probability distribution of the wind power descent gradient. If the scheduling strategy for the first sub-time period can be determined based on the probability distribution of the wind power descent gradient, it means that the production equipment can produce continuously during the first sub-time period; if the scheduling strategy for the first sub-time period cannot be determined based on the probability distribution of the wind power descent gradient, it means that the production equipment cannot produce continuously during the first sub-time period.
[0061] 203. Determine whether the scheduling strategy for the second sub-time period can be determined based on the probability distribution of the decreasing wind power gradient. If the scheduling strategy for the second sub-time period can be determined based on the probability distribution of the decreasing wind power gradient, proceed to step 204. If the scheduling strategy for the second sub-time period cannot be determined based on the probability distribution of the decreasing wind power gradient, proceed to step 205.
[0062] If the scheduling strategy for the second sub-time period can be determined based on the probability distribution of the decreasing wind power gradient, it means that the production equipment can operate continuously during the second sub-time period; if the scheduling strategy for the second sub-time period cannot be determined based on the probability distribution of the decreasing wind power gradient, it means that the production equipment cannot operate continuously during the second sub-time period.
[0063] It should be noted that although in the above embodiments, the electronic device determines whether a scheduling strategy can be determined for each sub-time period, this application is not limited to this. For example, if the scheduling strategy for the first sub-time period can be determined, the electronic device then determines whether a scheduling strategy for the total time period can be determined based on the probability distribution of the wind power decrease gradient. This total time period is the sum of the first and second sub-time periods. When the first sub-time period is not the first sub-time period in the target time period, the total time period includes all time periods from the first sub-time period to the current sub-time period. For example, the target time period is 4 hours long and is divided into 8 sub-time periods, numbered sequentially as sub-time period 1, sub-time period 2... sub-time period 8. The electronic device can determine a scheduling strategy for sub-time periods 1 to 3. Then, the electronic device determines whether a scheduling strategy can be determined for sub-time periods 1 to 4.
[0064] 204. Traverse the remaining sub-time periods.
[0065] After judging the current sub-time period, if a scheduling strategy can be determined for the current sub-time period, then continue to judge the next sub-time period until all sub-time periods have been traversed.
[0066] 205. Determine that the production equipment cannot operate continuously from the start time of the second sub-time period within the target time period.
[0067] Continuing with the example above, the target time period is 4 hours long and divided into 8 sub-time periods. If, for sub-time periods 1 through 4, the electronic equipment cannot determine a scheduling strategy, then production cannot continue from sub-time period 4 onwards. If the target time period is from 8:00 AM to 12:00 PM on November 8th, with sub-time period 1 being 8:00 AM to 8:30 AM, sub-time period 2 being 8:30 AM to 9:00 AM, and so on, then production cannot continue from 9:30 AM onwards.
[0068] Understandably, the shorter the sub-time period, the more accurate the determination of the moment when the production equipment cannot continue production.
[0069] 206. Determine that the production equipment cannot produce continuously from the start time of the first sub-time period within the target time period.
[0070] In this approach, after the electronic equipment predicts a drop in wind power at the target location, it divides the target time period into multiple sub-time periods. It then sequentially determines whether a scheduling strategy can be established for each sub-time period to identify the moments when production equipment cannot operate continuously. This facilitates precise changes to grid power arrangements and improves the quality of energy dispatch.
[0071] Optionally, in the above embodiments, when the production equipment cannot produce continuously within the target time period, after the electronic device determines that the target object has experienced a drop in wind power within the target time period, it also outputs indication information. The indication information is used to indicate the updated grid power arrangement, and the grid power arrangement is used to indicate the grid power constraints at each time point within the target time period.
[0072] For example, after predicting a power drop at a target location, the electronic device further generates and outputs indication information. For instance, the original grid power arrangement was to provide 3 megawatts of power. The updated grid power arrangement is to provide 5 megawatts of power.
[0073] Using this approach, after the electronic equipment predicts a drop in wind power at the target location, it generates and outputs a grid power arrangement to ensure continuous production of the equipment, facilitating accurate energy dispatching by staff and improving energy dispatching efficiency.
[0074] The following section provides a detailed explanation of how electronic devices predict the probability distribution of the wind power descent gradient of a target object within a target time period.
[0075] In one approach, an electronic device acquires historical wind power time series data and inputs the historical wind power time series data into a target model, so that the target model outputs the wind power descent gradient probability distribution. The historical wind power time series data is used to indicate the sequence data of the continuous change of wind power of the target object over time within a historical time period. The duration between the end time of the historical time period and the start time of the target time period is less than a preset duration.
[0076] In this method, a target model is pre-deployed on the electronic device. After obtaining the historical wind power time series of the target object, the electronic device performs low-pass filtering on the historical wind power time series to remove noise. For example, please refer to Table 1.
[0077] Table 1
[0078] Please refer to Table 1. Part of the historical time period is from 10:01 to 10:07. The wind power was recorded every minute to obtain the historical wind power time series. In this historical wind power time series, the 3.2MW at 10:04 and the 2.6MW at 10:06 are noise and need to be filtered out.
[0079] After noise is eliminated, a smoothed wind power time series is obtained. The electronic device determines the wind power descent gradient for historical time periods based on the smoothed wind power time series. Then, the electronic device inputs the smoothed wind power time series and the corresponding wind power descent gradient into the target model, so that the target model outputs the probability distribution of the wind power descent gradient of the target object within the target time period.
[0080] In another approach, an electronic device acquires the predicted wind power time series and inputs the predicted wind power time series into a target model so that the target model outputs the wind power descent gradient probability distribution. The predicted wind power time series is used to indicate the sequence data of the continuous change of wind power of the target object over time within the target time period.
[0081] Electronic equipment predicts wind power for a target time period, thus obtaining a predicted wind power time series. Prediction methods include, but are not limited to, wind power prediction based on numerical weather prediction and prediction based on multispectral images. After obtaining the predicted wind power time series for the target time period, the electronic equipment performs low-pass filtering on the predicted wind power time series to remove noise, and determines the wind power descent gradient based on the noise-removed predicted wind power time series. Then, the electronic equipment inputs the noise-removed predicted wind power time series and the wind power descent gradient into the target model, so that the target model outputs the probability distribution of the wind power descent gradient of the target object within the target time period.
[0082] Using this approach, the electronic device determines the probability distribution of the wind power descent gradient based on the historical or predicted wind power time series of the target object. It does not require access to or observation of external data such as meteorological data, and is based solely on the historical or predicted wind power time series, resulting in low computational complexity and low cost.
[0083] Optionally, in the above embodiments, the electronic device further acquires a sample set, wherein each pair of training samples in the sample set includes a sample wind power time series and a sample wind power descent gradient corresponding to the sample wind power time series. Then, the electronic device trains the target model using the training samples in the sample set.
[0084] For example, an electronic device trains an initial model based on a large number of training samples. During the training process, the parameters of the initial model are continuously adjusted until the value of the loss function is minimized. The model with the minimum value of the loss function is then used as the target model. The initial model can be, for example, a Convolutional Neural Network (CNN) or a Recurrent Neural Network (RNN), and this application is not limited to these types.
[0085] Using this approach, electronic devices train a target model and then use the model to predict the probability distribution of wind power decrease gradient of the target object within the target time period, which is fast and accurate.
[0086] Figure 3 This is another flowchart of the wind power drop prediction method provided in this application. This embodiment includes: 301. Obtain the historical wind power time series, future load production plan, grid power security and energy storage device status of the target object.
[0087] 302. Low-pass filtering is applied to the historical wind power time series to eliminate noise.
[0088] 303. Determine the wind power descent gradient based on the filtered historical wind power time series.
[0089] In this step, the electronic device determines the wind power descent gradient based on the filtered historical wind power time series, thereby obtaining the different wind power descent gradients that appear in the historical time period corresponding to the historical wind power time series.
[0090] 304. Predict the probability distribution of wind power descent gradient for the target time period based on the wind power descent gradient.
[0091] The electronic device inputs the historical wind power time series and wind power descent gradient into the target model, so that the target model outputs the probability distribution of wind power descent gradient for the target time period.
[0092] 305. Determine the scheduling strategy within the target time period based on the probability distribution of wind power decline gradient, future load production plans, grid power security, and the state of charge of energy storage devices.
[0093] 306. Determine whether the production equipment can produce continuously within the target time period in the future. If the production equipment can produce continuously within the target time period in the future, proceed to step 307; if the production equipment cannot produce continuously within the target time period in the future, proceed to step 308.
[0094] In this step, the electronic device determines whether a scheduling strategy has been generated in step 305. If a scheduling strategy has been generated in step 305, it means that the production equipment can produce continuously within the future target time period; if a scheduling strategy has not been generated in step 305, it means that the production equipment cannot produce continuously within the future target time period.
[0095] 307. Output scheduling strategy.
[0096] 308. Issue an early warning if wind power drops within a target time period.
[0097] 309. Determine the time points within the target time period when continuous production is not possible and output the indication information. The indication information is used to indicate the updated power grid arrangement, and the power grid arrangement is used to indicate the power grid constraints at each time point within the target time period.
[0098] The following are embodiments of the apparatus described in this application, which can be used to execute the embodiments of the method described in this application. For details not disclosed in the apparatus embodiments of this application, please refer to the embodiments of the method described in this application.
[0099] Figure 4This is a schematic diagram of the wind power drop prediction device provided in this application. The wind power drop prediction device 400 includes: a prediction module 41, a processing module 42, and a determination module 43.
[0100] Prediction module 41 is used to predict the probability distribution of wind power decline gradient of target object within a target time period. The probability distribution of wind power decline gradient is used to indicate the probability statistics of different wind power decline gradients of target object within the target time period. The wind power decline gradient is used to indicate the magnitude of wind power decline of target object per unit time. Processing module 42 is used to determine whether the production equipment can continue to produce within the target time period based on the probability distribution of the wind power decrease gradient, wherein the production equipment is the equipment that uses at least the electrical energy provided by the target object to produce within the target time period; The determination module 43 is used to determine that the target object experiences a drop in wind power during the target time period when the production equipment cannot produce continuously during the target time period.
[0101] In one feasible implementation, the processing module 42 is configured to determine the state of charge (SOC) of the energy storage device, wherein the SOC indicates the ratio of the current remaining power of the energy storage device to its rated capacity; and, based on the wind power descent gradient probability distribution, the SOC, future load production plans, and grid power arrangements, determine a scheduling strategy for the target time period, wherein the future load production plans indicate the planned power of the production device at each time point within the target time period, the grid power arrangements indicate the grid power constraints at each time point within the target time period, the scheduling strategy indicates the output sequence of various energy suppliers within the target time period, the output sequence indicates the power used by the energy suppliers at each time point within the target time period, and the power used indicates the power used by the production device from the power supplied by the energy suppliers; if the scheduling strategy is determined, it is determined that the production device can operate continuously within the target time period; if the scheduling strategy cannot be determined, it is determined that the production device cannot operate continuously within the target time period.
[0102] In one feasible implementation, when the processing module 42 determines the scheduling strategy for the target time period based on the wind power descent gradient probability distribution, the state of charge, the future load production plan, and the grid power arrangement, it is used to determine a target descent gradient based on the wind power descent gradient probability distribution and the state of charge, wherein the target descent gradient is the wind power descent gradient in the wind power descent gradient probability distribution that matches the magnitude of the state of charge; and to determine the scheduling strategy for the target time period based on the target descent gradient, the future load production plan, and the grid power arrangement.
[0103] In one feasible implementation, when the processing module 42 determines the scheduling strategy for the target time period based on the wind power descent gradient probability distribution, the state of charge, the future load production plan, and the grid power arrangement, it is used to determine a set of constraints. The set of constraints includes at least one of a first constraint, a second constraint, and a third constraint. The first constraint indicates the constraints of the energy storage device, the second constraint indicates the constraints of continuous production, and the third constraint indicates the constraints of the grid power. The energy storage device and the grid power are the energy suppliers that provide power to the production equipment during the target time period. Under the constraints of each constraint in the set of constraints, the scheduling strategy for the target time period is determined based on the wind power descent gradient probability distribution, the state of charge, the future load production plan, and the grid power arrangement.
[0104] In one feasible implementation, after the determining module 43 determines that the target object experiences a wind power drop within the target time period, the processing module 42 is further configured to determine whether a scheduling strategy for a first sub-time period can be determined based on the wind power drop gradient probability distribution, wherein the length of the first sub-time period is less than the length of the target time period; when the scheduling strategy for the first sub-time period is determined based on the wind power drop gradient probability distribution, it is further configured to determine whether a scheduling strategy for a second sub-time period can be determined based on the wind power drop gradient probability distribution, wherein the first sub-time period and the first sub-time period are adjacent time periods within the target time period, and the length of the second sub-time period is less than the length of the target time period; when the scheduling strategy for the second sub-time period cannot be determined based on the wind power drop gradient probability distribution, it is determined that the production equipment cannot continuously produce from the start time of the second sub-time period within the target time period.
[0105] In one feasible implementation, after the determining module 43 determines that the target object experiences a wind power drop within the target time period, the processing module 42 is further configured to output indication information, which is used to indicate the updated grid power arrangement, and the grid power arrangement is used to indicate the grid power constraints at each time point within the target time period.
[0106] In one feasible implementation, the prediction module 41 is used to acquire historical wind power time series and input the historical wind power time series into the target model so that the target model outputs the wind power descent gradient probability distribution. The historical wind power time series is used to indicate the sequence data of the continuous change of wind power of the target object over time within a historical time period, and the duration between the end time of the historical time period and the start time of the target time period is less than a preset duration; or, to acquire predicted wind power time series and input the predicted wind power time series into the target model so that the target model outputs the wind power descent gradient probability distribution. The predicted wind power time series is used to indicate the sequence data of the continuous change of wind power of the target object over time within the target time period.
[0107] In one feasible implementation, the processing module 42 is further configured to acquire a sample set, wherein each pair of training samples in the sample set includes a sample wind power time series and a sample wind power descent gradient corresponding to the sample wind power time series; and to train the target model using the training samples in the sample set.
[0108] The wind power drop prediction device provided in this application can perform the actions of the electronic devices in the above embodiments. Its implementation principle and technical effect are similar, and will not be described again here.
[0109] Figure 5 A schematic diagram of the structure of the electronic device provided in this application. The electronic device 500 includes: Processor 51 and memory 52; The memory 52 stores computer instructions and test data; The processor 51 executes the computer instructions stored in the memory 52, causing the processor 51 to perform the wind power drop prediction method as described above.
[0110] The specific implementation process of processor 51 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.
[0111] Optionally, the electronic device 500 also includes a communication component 53. The processor 51, memory 52, and communication component 53 can be connected via a bus 54.
[0112] This application also provides a computer-readable storage medium storing computer instructions that, when executed by a processor, are used to implement the wind power drop prediction method described above.
[0113] This application also provides a computer program product comprising a computer program that, when executed by a processor, implements the wind power drop prediction method as described above.
[0114] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the application disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and embodiments are to be considered exemplary only, and the true scope and spirit of this application are indicated by the following claims.
[0115] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.
Claims
1. A method for predicting wind power drop, characterized in that, include: Predict the probability distribution of wind power decline gradient of target object within a target time period. The probability distribution of wind power decline gradient is used to indicate the probability statistics of different wind power decline gradients of target object within the target time period. The wind power decline gradient is used to indicate the magnitude of wind power decline of target object per unit time. Based on the probability distribution of the wind power descent gradient, it is determined whether the production equipment can produce continuously within the target time period. The production equipment is the equipment that uses at least the electrical energy provided by the target object to produce within the target time period. When the production equipment cannot produce continuously during the target time period, it is determined that the target object has experienced a drop in wind power during the target time period.
2. The method according to claim 1, characterized in that, Determining whether the production equipment can operate continuously within the target time period based on the probability distribution of the wind power descent gradient includes: Determine the state of charge (SOC) of the energy storage device, wherein the SOC indicates the ratio of the current remaining charge of the energy storage device to its rated capacity; Based on the wind power descent gradient probability distribution, the state of charge, the future load production plan, and the grid power arrangement, a scheduling strategy is determined for the target time period. The future load production plan is used to indicate the planned power of the production equipment at each time point within the target time period. The grid power arrangement is used to indicate the grid power constraints at each time point within the target time period. The scheduling strategy is used to indicate the output sequence of various energy suppliers within the target time period. The output sequence is used to indicate the power used by the energy suppliers at each time point within the target time period. The power used is used to indicate the power used by the production equipment from the power supplied by the energy suppliers. Given the established scheduling strategy, it is determined that the production equipment can operate continuously within the target time period. If the scheduling strategy cannot be determined, it is determined that the production equipment cannot produce continuously during the target time period.
3. The method according to claim 2, characterized in that, The step of determining the scheduling strategy within the target time period based on the wind power decline gradient probability distribution, the state of charge, future load production plans, and grid power arrangements includes: Based on the wind power descent gradient probability distribution and the state of charge, a target descent gradient is determined, wherein the target descent gradient is the wind power descent gradient in the wind power descent gradient probability distribution that matches the magnitude of the state of charge; Based on the target descent gradient, the future load production plan, and the grid power arrangement, the scheduling strategy for the target time period is determined.
4. The method according to claim 2, characterized in that, The step of determining the scheduling strategy within the target time period based on the wind power decline gradient probability distribution, the state of charge, future load production plans, and grid power arrangements includes: A set of constraints is determined, the set of constraints including at least one of a first constraint, a second constraint and a third constraint, the first constraint indicating the constraints of the energy storage device, the second constraint indicating the constraints of continuous production, and the third constraint indicating the constraints of the grid power, wherein the energy storage device and the grid power are the energy suppliers that supply power to the production equipment during the target time period; Under the constraints of the set of constraints, the scheduling strategy for the target time period is determined based on the probability distribution of wind power descent gradient, the state of charge, the future load production plan, and the grid power arrangement.
5. The method according to any one of claims 1 to 4, characterized in that, When the production equipment cannot produce continuously during the target time period, after determining that the target object experiences a drop in wind power during the target time period, the method further includes: Determine whether a scheduling strategy for the first sub-time period can be determined based on the probability distribution of the wind power descent gradient, wherein the length of the first sub-time period is less than the length of the target time period; When the scheduling strategy for the first sub-time period is determined based on the probability distribution of the wind power descent gradient, it is determined whether the scheduling strategy for the second sub-time period can be determined based on the probability distribution of the wind power descent gradient. The first sub-time period and the first sub-time period are adjacent time periods in the target time period, and the length of the second sub-time period is less than the length of the target time period. When the scheduling strategy for the second sub-time period cannot be determined based on the probability distribution of the wind power descent gradient, it is determined that the production equipment cannot produce continuously from the start time of the second sub-time period within the target time period.
6. The method according to any one of claims 1 to 4, characterized in that, When the production equipment cannot produce continuously during the target time period, after determining that the target object experiences a drop in wind power during the target time period, the method further includes: Output indication information, which is used to indicate the updated power grid arrangement, and the power grid arrangement is used to indicate the power grid constraints at each time point within the target time period.
7. The method according to any one of claims 1 to 4, characterized in that, The predicted gradient probability distribution of wind power descent of the target object within the target time period includes: The historical wind power time series is obtained and input into the target model so that the target model outputs the wind power descent gradient probability distribution. The historical wind power time series is used to indicate the sequence data of the continuous change of wind power of the target object over time within a historical time period. The duration between the end time of the historical time period and the start time of the target time period is less than a preset duration. or, Obtain the predicted wind power time series and input the predicted wind power time series into the target model so that the target model outputs the wind power descent gradient probability distribution. The predicted wind power time series is used to indicate the sequence data of the continuous change of wind power of the target object over time within the target time period.
8. The method according to claim 7, characterized in that, Also includes: Obtain a sample set, wherein each pair of training samples in the sample set includes a sample wind power time series and a sample wind power descent gradient corresponding to the sample wind power time series; The target model is trained using the training samples in the sample set.
9. An electronic device comprising a processor, a memory, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it causes the electronic device to implement the method as described in any one of claims 1 to 8.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 8.