Risk prediction method and apparatus for wind farm under extreme weather, medium, and device

Abstract

Disclosed in the present invention are a risk prediction method and apparatus for a wind farm under extreme weather, a medium, and a device. Video data, wind speed data, and wind direction data of each wind turbine are acquired, and a matching model is used to determine whether the rotation of a rotor matches the expectations of the current wind speed and wind direction, so as to quickly identify possible operation anomalies. When it is detected that the rotation of the rotor does not match the wind speed and the wind direction, the current weather condition is further analyzed, a future weather change is predicted, and on the basis of the data, a potential risk type is estimated. This solution realizes dynamic monitoring and accurate prediction of wind farms under extreme weather, can effectively avoid risks, and avoids the reduction of power generation efficiency due to the fact that faults are not detected in time, thereby ensuring safe and continuous operation of the wind farms.

Description

Risk prediction methods, devices, media and equipment for wind power plants under extreme weather conditions Technical Field

[0001] This invention relates to the field of risk prediction technology, and in particular to a method, apparatus, medium and equipment for risk prediction of wind power plants under extreme weather conditions. Background Technology

[0002] In severe weather conditions, the risk of failure at wind power plants increases significantly. Currently, risk management at wind farms still relies heavily on manual operation. Although monitoring devices collect data regularly, the analysis and judgment of this data typically depend on the experience of the staff to assess the risk of failure. Once a risk is identified, staff need to personally conduct on-site inspections.

[0003] However, due to the subjective factors involved in manual operation, this management method often makes it difficult to detect risks in a timely manner, which in turn leads to a decrease in the power generation efficiency of the wind farm and may even affect the safe operation of the entire wind power plant. Summary of the Invention

[0004] Therefore, it is necessary to provide risk prediction methods, devices, media and equipment for wind power plants under extreme weather conditions to solve the problem that existing technologies often fail to detect risks in a timely manner, which leads to a decrease in the power generation efficiency of wind farms and may even affect the safe operation of the entire wind power plant.

[0005] A risk prediction method for wind power plants under extreme weather conditions, wherein the wind power plant comprises multiple wind turbine generators, the method comprising:

[0006] Acquire video data, wind speed data, and wind direction data for each wind turbine generator set within a preset time period; wherein, the video data records the rotation of the wind turbine generator set's rotor.

[0007] The system uses a pre-defined matching model to determine whether the rotation of the wind turbine rotor of each wind turbine matches the wind speed and wind direction data.

[0008] If there is a mismatch between the rotation status of the wind turbine of the first wind turbine and the wind speed and direction data, the current weather conditions at the location of the first wind turbine are obtained from the video data, the future weather conditions are predicted based on the current weather conditions, and the risk type of the wind power plant is estimated based on the future weather conditions; wherein, the first wind turbine is any one of the plurality of wind turbines.

[0009] In one embodiment, determining whether the rotor rotation of each wind turbine matches the wind speed and wind direction data using a preset matching model includes:

[0010] The rotation of the wind turbine rotor of each wind turbine is divided into discrete rotor states using an optical flow algorithm, and the wind speed data and wind direction data of each wind turbine are also divided into discrete wind speed states and discrete wind direction states. The rotor states include low speed, medium speed and high speed, the wind speed states include low wind speed, medium wind speed and high wind speed, and the wind direction states include multiple wind direction angle ranges.

[0011] Obtain a preset set of state rules; wherein the set of state rules includes multiple different state rules, and each state rule is a combination of a wind turbine state, a wind speed state, and a wind direction state;

[0012] If the combination of the rotor state, wind speed state, and wind direction state of the first wind turbine belongs to the state rule set, then it is determined that the rotor rotation of the first wind turbine matches the wind speed data and wind direction data. If the combination of the rotor state, wind speed state, and wind direction state of the first wind turbine does not belong to the state rule set, then it is determined that the rotor rotation of the first wind turbine does not match the wind speed data and wind direction data.

[0013] In one embodiment, obtaining the preset set of state rules includes:

[0014] Multiple sets of sample data were obtained; each set of sample data included the rotor rotation, wind speed, and wind direction of a wind turbine under risk-free conditions.

[0015] The optical flow algorithm is used to divide the sample wind turbine rotation of each wind turbine into discrete wind turbine states, the sample wind speed data of each wind turbine into discrete wind speed states, and the sample wind direction data of each wind turbine into discrete wind direction states, so as to obtain multiple state sets.

[0016] Multiple i-th state sets are generated, the frequency proportion of each i-th state set is calculated, and state sets with a frequency proportion greater than a preset frequency proportion threshold are retained as candidate state rules; wherein, the i-th state set includes i wind turbine state, wind speed state and wind direction state, and the initial value of i is 1.

[0017] Let i = i + 1, return to the step of generating multiple i-th state sets and subsequent steps, until the number of candidate state rules no longer increases;

[0018] Each candidate state rule is split into a first sub-state set and a second sub-state set. Within each candidate state rule, the joint probability and correlation coefficient between the first sub-state set and the second sub-state set are calculated. All candidate state rules with a joint probability greater than a preset joint probability threshold and a correlation coefficient greater than a preset correlation coefficient threshold are retained as the state rule set.

[0019] In one embodiment, the calculation of the frequency proportion of each i-th state set is expressed as:

[0020] In the above formula, FR(x) represents the frequency proportion of state set x, and N x The total number of state sets x contained in the plurality of state sets is represented by T, where T represents the total number of all state sets.

[0021] The formula for calculating the joint probability is:

[0022] In the above formula, JP(ab) represents the joint probability between the first substate set a and the second substate set b;

[0023] The formula for calculating the correlation coefficient is:

[0024] In the above formula, AC(ab) represents the correlation coefficient between the first sub-state set a and the second sub-state set b.

[0025] In one embodiment, the step of predicting future weather conditions based on the current weather conditions and estimating the risk type of the wind power plant based on the future weather conditions includes:

[0026] The current weather conditions are input into an LSTM prediction model to predict future weather conditions, and a decision tree is used to classify the risk type of the wind power plant based on the future weather conditions.

[0027] In one embodiment, infrared sensors are installed around the wind power plant, and the method further includes:

[0028] If a warning is issued when a signal is received from the infrared sensor indicating that the wind power plant is at risk, an early warning will be provided.

[0029] In one embodiment, the method further includes:

[0030] Collect historical data; wherein, the historical data includes video data, wind speed data, and wind direction data of wind turbine generators under extreme weather conditions;

[0031] A risk model is established based on the historical data; wherein, the risk model defines the threshold range of normal operation and abnormal operation of wind turbine generator sets, and the risk types under different abnormal operation states;

[0032] The video data, wind speed data, and wind direction data of each wind turbine within a preset time period are compared with the threshold range defined in the risk model as real-time data. When the real-time data exceeds the threshold range, a risk warning signal is generated.

[0033] A risk prediction device for wind power plants under extreme weather conditions, the risk prediction device for wind power plants under extreme weather conditions includes:

[0034] The data acquisition module is used to acquire video data, wind speed data, and wind direction data for each wind turbine generator set within a preset time period; wherein, the video data records the rotation of the wind turbine generator set's rotor.

[0035] The data matching module is used to determine whether the rotation status of the wind turbine rotor of each wind turbine is matched with the wind speed and wind direction data through a preset matching model.

[0036] The risk assessment module is used to obtain the current weather conditions of the location of the first wind turbine from video data if there is a mismatch between the rotation status of the wind turbine and the wind speed and direction data; predict the future weather conditions based on the current weather conditions; and estimate the risk type of the wind power plant based on the future weather conditions; wherein the first wind turbine is any one of the plurality of wind turbines.

[0037] A computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the above-described risk prediction method for wind power plants under extreme weather conditions.

[0038] A terminal device includes a memory and a processor, the memory storing a computer program that, when executed by the processor, causes the processor to perform the steps of the above-described risk prediction method for wind power plants under extreme weather conditions.

[0039] This invention provides a method, apparatus, medium, and equipment for risk prediction of wind power plants under extreme weather conditions. By acquiring video, wind speed, and wind direction data for each wind turbine, a matching model is used to determine whether the rotor rotation matches the expected current wind speed and direction, thereby quickly identifying potential operational anomalies. When a mismatch between rotor rotation and wind speed and direction is detected, the current weather conditions are further analyzed and future weather changes are predicted. This data is then used to estimate potential risk types. This solution achieves dynamic monitoring and accurate prediction of wind farms under extreme weather conditions, effectively mitigating risks and avoiding reduced power generation efficiency due to failure to detect faults in a timely manner, thus ensuring the safe and continuous operation of wind farms. Attached Figure Description

[0040] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. 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.

[0041] in:

[0042] Figure 1 is a flowchart illustrating the risk prediction method for wind power plants under extreme weather conditions;

[0043] Figure 2 is a schematic diagram of the risk prediction device for wind power plants under extreme weather conditions.

[0044] Figure 3 is a structural block diagram of the terminal device. Detailed Implementation

[0045] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0046] The terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or apparatuses.

[0047] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0048] As shown in Figure 1, Figure 1 is a flowchart illustrating a risk prediction method for wind power plants under extreme weather conditions in one embodiment. The wind power plant includes multiple wind turbine generators located in specific geographical locations or within a defined area. The steps provided by the risk prediction method for wind power plants under extreme weather conditions in this embodiment include:

[0049] S101, acquire video data, wind speed data and wind direction data of each wind turbine within a preset time period.

[0050] The video data records the rotation of the wind turbine's rotor, including rotor speed and rotation duration. The wind speed data reflects the wind speed in the area where the wind turbine is located. The wind direction data reflects the wind direction in the area where the wind turbine is located.

[0051] For example, each wind turbine is equipped with an independent high-definition camera, wind speed sensor, and wind direction sensor, giving each unit its own data acquisition device. The camera records five minutes of image data of the turbine's rotation each time, while the wind speed sensor collects wind speed data and the wind direction sensor collects wind direction data. This data is then transmitted together to the wind farm's monitoring system for subsequent analysis.

[0052] S102, using a preset matching model, determines whether the rotation of the wind turbine rotor of each wind turbine matches the wind speed and wind direction data.

[0053] The preset matching model is a comparison model based on wind turbine rotation, wind speed and wind direction data, used to determine whether the wind turbine generator is operating normally.

[0054] Optionally, by analyzing historical data or designing standards, expected values ​​and allowable error ranges for wind turbine rotational speeds under different wind speeds and wind directions can be set. The current wind turbine rotational speed is compared with this threshold range to determine whether it meets the standard. For example, assuming a wind speed of 10 m / s and a wind direction of southwest (225°), the matching model expects the wind turbine rotational speed to be 20 ± 2 rpm. If the currently detected wind turbine rotational speed is between 18 and 22 rpm, it meets the threshold; otherwise, it is considered abnormal. This method is easy to implement, enabling monitoring of wind turbine generators through simple range judgment, effectively identifying deviations from normal rotational speeds, and thus detecting anomalies in a timely manner.

[0055] Optionally, a machine learning algorithm can be used to determine whether the current wind turbine rotation is consistent with the input real-time wind speed and direction data. For example, a support vector machine model can be trained to establish a reasonable range for wind turbine rotation under different wind speeds and directions. After real-time data is input, the model determines whether the wind turbine rotation speed deviates from the reasonable range; if it does, an alarm is triggered.

[0056] If there is a mismatch between the rotor rotation status of the first wind turbine and the wind speed and direction data, then execute S103. If the rotor rotation status of all first wind turbines matches the wind speed and direction data, then execute S104 and end. Here, "first wind turbine" refers to any one of multiple wind turbines.

[0057] S103. Obtain the current weather conditions at the location of the first wind turbine from the video data, predict the future weather conditions based on the current weather conditions, and estimate the risk type of the wind power plant based on the future weather conditions.

[0058] Current weather conditions refer to the real-time meteorological conditions at the location of the wind turbine, typically including wind speed, wind direction, precipitation (rain or snow), and fog / haze. Future weather conditions are predictions of the current weather conditions in the future. Risk type refers to the specific risk category that future weather conditions may pose to the wind turbine, such as rotor overload, icing damage, and structural damage. Different risk types have different impacts on wind turbines, requiring corresponding risk prevention measures.

[0059] Optionally, video image processing technology can be used to analyze the current weather conditions around the wind turbine generators in real time. For example, visibility, precipitation type, and intensity in the images can be used to determine the current weather status. If a significant decrease in visibility is detected around the wind turbine generators via video, combined with wind speed data indicating an increase in wind speed, the system can determine that the current weather may be caused by a blizzard, predict the risk of rotor icing, and propose countermeasures such as shutdown or preheating.

[0060] Optionally, in one specific embodiment, the current weather conditions are input into an LSTM prediction model to predict future weather conditions, and a decision tree is used to classify the risk type of the wind power plant based on the future weather conditions. For example, currently collected weather data such as wind speed, temperature, and humidity are input into the LSTM model to predict the future weather conditions for the next 30 minutes. The future weather conditions are then input into the decision tree. If the LSTM predicts a future wind speed exceeding 15 m / s, the decision tree determines that there may be a risk of "wind turbine overspeeding" and recommends taking deceleration measures for the wind turbine.

[0061] The aforementioned risk prediction method for wind power plants under extreme weather conditions acquires video, wind speed, and wind direction data for each wind turbine. It then uses a matching model to determine whether the rotor rotation matches the expected current wind speed and direction, thereby quickly identifying potential operational anomalies. When a mismatch between rotor rotation and wind speed / direction is detected, the current weather conditions are further analyzed, and future weather changes are predicted. This data is then used to estimate potential risk types. This approach enables dynamic monitoring and accurate prediction of wind farm conditions under extreme weather conditions, effectively mitigating risks and preventing reduced power generation efficiency due to untimely fault detection, thus ensuring the safe and continuous operation of the wind farm.

[0062] Optionally, S102 determines whether the rotor rotation of each wind turbine matches the wind speed and wind direction data using a preset matching model, specifically including the following steps:

[0063] A1. Using the optical flow algorithm, the rotation of the wind turbine rotor of each wind turbine generator set is divided into discrete wind turbine states, and the wind speed data of each wind turbine generator set is divided into discrete wind speed states, and the wind direction data of each wind turbine generator set is divided into discrete wind direction states.

[0064] Optical flow algorithm is used to analyze the motion of objects in video image sequences, identifying the direction and speed of movement of objects within the image. In video monitoring of wind turbine generators, optical flow algorithm can capture the rotational speed of the wind turbine blades. Wind turbine speeds are categorized as low, medium, and high. For example, a wind turbine speed of 10 revolutions per minute or less is defined as low speed; a speed between 10 and 20 revolutions per minute is defined as medium speed; and a speed of 20 revolutions per minute or more is defined as high speed.

[0065] Wind speed conditions include low wind speed, medium wind speed, and high wind speed. For example, wind speeds below 5 m / s are defined as "low wind speed", 5-15 m / s are defined as "medium wind speed", and above 15 m / s are defined as "high wind speed".

[0066] Wind direction status includes multiple wind direction angle ranges. For example, 0°-30° is defined as the first wind direction status, 30°-60° is defined as the second wind direction status, and so on.

[0067] A2. Obtain the preset set of state rules.

[0068] The state rule set includes multiple different state rules. Each state rule is a combination of a wind turbine state, a wind speed state, and a wind direction state, used to describe the reasonable operating state that the wind turbine should be in under different wind speed and wind direction conditions.

[0069] A3. If the combination of the wind turbine state, wind speed state, and wind direction state of the first wind turbine belongs to the state rule set, then it is determined that the wind turbine rotation status of the first wind turbine matches the wind speed data and wind direction data. If the combination of the wind turbine state, wind speed state, and wind direction state of the first wind turbine does not belong to the state rule set, then it is determined that the wind turbine rotation status of the first wind turbine does not match the wind speed data and wind direction data.

[0070] For example, suppose the state rule set includes the following rules: Rule 1: Wind turbine state is medium wind speed, wind speed state is medium wind speed, and wind direction state is 30°-60°. Rule 2: Wind turbine state is high wind speed, wind speed state is high wind speed, and wind direction state is 60°-90°. If the real-time data combination is: wind turbine state is medium wind speed, wind speed state is medium wind speed, and wind direction state is 45°, the system will match Rule 1, thus determining that the wind turbine is operating normally. If the real-time data combination is: wind turbine state is low wind speed, wind speed state is medium wind speed, and wind direction state is 45°, then it is not in the rule set, the system determines that the wind turbine is operating abnormally, and issues a warning.

[0071] This state combination matching can quickly identify abnormal situations where the rotor speed, wind speed, and wind direction do not match, which helps to locate potential mechanical faults or operational deviations.

[0072] Optionally, step A2, obtaining the preset state rule set, specifically includes the following steps:

[0073] A21. Obtain multiple sets of sample data.

[0074] Each set of sample data includes sample rotor rotation, sample wind speed, and sample wind direction data for a wind turbine under risk-free conditions. This sample data should cover rotor rotation, wind speed, and wind direction under various normal conditions to represent different operating states under risk-free conditions.

[0075] A22. Using the optical flow algorithm, the sample wind turbine rotation of each wind turbine is divided into discrete wind turbine states, the sample wind speed data of each wind turbine is divided into discrete wind speed states, and the sample wind direction data of each wind turbine is divided into discrete wind direction states, so as to obtain multiple state sets.

[0076] It is understandable that the logic of step A22 is basically the same as that of step A1, the only difference being that sample data is used here, so it will not be elaborated further.

[0077] A23. Generate multiple i-th state sets, calculate the frequency proportion of each i-th state set, and retain the state sets whose frequency proportion is greater than the preset frequency proportion threshold as candidate state rules.

[0078] The i-th state set includes i states from the wind turbine state, wind speed state, and wind direction state, with i initially set to 1. That is, the initial state set contains only one state (such as the wind turbine state or wind speed state). The frequency percentage represents the ratio of the frequency of each state set in the risk-free sample data. The frequency percentage is used to determine the representativeness of a certain state combination under risk-free conditions.

[0079] Optionally, the frequency proportion of each i-th state set is calculated, expressed as:

[0080] In the above formula, FR(x) represents the frequency proportion of state set x, and N x Let T represent the total number of state sets x contained in multiple state sets, and let T represent the total number of all state sets.

[0081] Furthermore, for each i-th state set, if its frequency proportion is greater than a preset frequency proportion threshold, it is retained as a candidate state rule. The frequency proportion threshold can be adjusted according to the actual situation of the wind farm, and representative data are usually selected from common state combinations.

[0082] A24. Let i = i + 1, return to execute A23 and subsequent steps until the number of candidate state rules no longer increases.

[0083] In this embodiment, let i = 3 to generate a ternary state combination (wind turbine state + wind speed state + wind direction state). After filtering, no new rules are generated. At this point, since new combinations no longer increase the number of candidate state rules, the iteration ends, and the final set of state rules is generated.

[0084] A25. Divide each candidate state rule into a first sub-state set and a second sub-state set. Within each candidate state rule, calculate the joint probability and correlation coefficient between the first sub-state set and the second sub-state set, and retain all candidate state rules whose joint probability is greater than a preset joint probability threshold and whose correlation coefficient is greater than a preset correlation coefficient threshold as the state rule set.

[0085] For example, suppose a candidate state rule is "Wind rotor state = high speed, wind speed state = high wind speed, wind direction state = first wind direction state", then it can be decomposed into: first sub-state set: "Wind rotor state = high speed, wind speed state = high wind speed" and second sub-state set: "Wind direction state = wind direction state". Of course, it can also be decomposed into other forms, and the same decomposition can be performed on other candidate state rules.

[0086] Next, the joint probability and correlation coefficient between the first and second sub-state sets are calculated. The joint probability refers to the probability that both the first and second sub-state sets will appear simultaneously in the risk-free sample data. The level of the joint probability reflects the coexistence relationship between these two sub-state sets under risk-free conditions. The correlation coefficient represents the strength of the association between the first and second sub-state sets. The correlation coefficient can be used to measure whether two sub-state sets have a significant correlation.

[0087] Optionally, the formula for calculating the joint probability is:

[0088] In the above formula, JP(ab) represents the joint probability between the first substate set a and the second substate set b;

[0089] The formula for calculating the correlation coefficient is:

[0090] In the above formula, AC(ab) represents the correlation coefficient between the first sub-state set a and the second sub-state set b.

[0091] The candidate state rule will only be retained when both the joint probability and the correlation coefficient are greater than the corresponding preset threshold, ensuring that the selected state rule set has strong representativeness.

[0092] Optionally, infrared sensors can be installed around the perimeter of the wind farm, ensuring coverage of critical entrances / exits or security areas near the turbines to detect potential external threats (such as personnel approaching equipment, animals or obstacles entering the risk zone). The method can also perform the following steps: when a sensor detects an abnormal heat source, it generates a trigger signal and sends a command to the monitoring system, indicating abnormal activity approaching the equipment. If the system has already detected other risks, the triggering of the infrared signal will further activate an early warning alert, notifying maintenance personnel.

[0093] Optionally, the above method may also perform the following steps:

[0094] B1. Collect historical data.

[0095] The historical data includes video data, wind speed data, and wind direction data of wind turbines under extreme weather conditions. This historical data serves as the foundational dataset for subsequent model building.

[0096] B2. Build risk models based on historical data.

[0097] The risk model defines the threshold ranges for normal and abnormal operating states of wind turbine generators, as well as the risk types under different abnormal operating states. Specifically, the risk model is formed by analyzing the data characteristics of wind turbine generators under normal and abnormal conditions based on collected historical data. The model defines the numerical thresholds for wind turbines under normal operating states and different abnormal states, and identifies the risk types corresponding to abnormal states, such as overspeed and mechanical failure.

[0098] B3. Compare the video data, wind speed data, and wind direction data of each wind turbine within a preset time period as real-time data with the threshold range defined in the risk model. When the real-time data exceeds the threshold range, a risk warning signal is generated.

[0099] In this way, risk models built using historical data can more accurately define normal and abnormal states, reducing false alarms caused by fluctuations in individual data.

[0100] In one embodiment, as shown in Figure 2, a risk prediction device for wind power plants under extreme weather conditions is proposed, the device comprising:

[0101] The data acquisition module 201 is used to acquire video data, wind speed data, and wind direction data for each wind turbine generator set within a preset time period; among which, the video data records the rotation of the wind turbine generator set's rotor.

[0102] The data matching module 202 is used to determine whether the rotation status of the wind turbine of each wind turbine is matched with the wind speed data and wind direction data through a preset matching model.

[0103] The risk assessment module 203 is used to obtain the current weather conditions of the location of the first wind turbine from the video data if there is a mismatch between the rotation status of the wind turbine and the wind speed and wind direction data; predict the future weather conditions based on the current weather conditions; and estimate the risk type of the wind power plant based on the future weather conditions; wherein the first wind turbine is any one of multiple wind turbines.

[0104] Figure 3 illustrates the internal structure of a terminal device in one embodiment. As shown in Figure 3, the terminal device includes a processor, a memory, and a network interface connected via a system bus. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and may also store a computer program. When executed by the processor, this computer program enables the processor to implement a risk prediction method for wind power plants under extreme weather conditions. The internal memory may also store a computer program, which, when executed by the processor, enables the processor to implement the risk prediction method for wind power plants under extreme weather conditions. Those skilled in the art will understand that the structure shown in Figure 3 is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the terminal device to which the present application is applied. A specific terminal device may include more or fewer components than shown in the figure, or combine certain components, or have different component arrangements.

[0105] A computer-readable storage medium storing a computer program, which, when executed by a processor, performs the following steps: acquiring video data, wind speed data, and wind direction data for each wind turbine generator set within a preset time period; wherein the video data records the rotation of the wind turbine generator set's rotor; determining whether the rotor rotation of each wind turbine generator set matches the wind speed data and wind direction data using a preset matching model; if there is a mismatch between the rotor rotation of a first wind turbine generator set and the wind speed data and wind direction data, then acquiring the current weather conditions at the location of the first wind turbine generator set from the video data, predicting future weather conditions based on the current weather conditions, and estimating the risk type of the wind power plant based on the future weather conditions; wherein the first wind turbine generator set is any one of a plurality of wind turbine generator sets.

[0106] A terminal device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it performs the following steps: acquiring video data, wind speed data, and wind direction data for each wind turbine generator set within a preset time period; wherein the video data records the rotation of the wind turbine generator set's rotor; determining whether the rotor rotation of each wind turbine generator set matches the wind speed data and wind direction data using a preset matching model; if there is a mismatch between the rotor rotation of a first wind turbine generator set and the wind speed data and wind direction data, then acquiring the current weather conditions at the location of the first wind turbine generator set from the video data, predicting future weather conditions based on the current weather conditions, and estimating the risk type of the wind power plant based on the future weather conditions; wherein the first wind turbine generator set is any one of a plurality of wind turbine generator sets.

[0107] It should be noted that the above-mentioned risk prediction method, device, equipment and computer-readable storage medium for wind power plants under extreme weather conditions belong to a general inventive concept, and the contents of the embodiments of the risk prediction method, device, equipment and computer-readable storage medium for wind power plants under extreme weather conditions are applicable to each other.

[0108] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. This program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM), etc.

[0109] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0110] The above embodiments merely illustrate several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A method for risk prediction of a wind power plant under extreme weather conditions, characterized in that, The wind power plant includes multiple wind turbine generators, and the method includes: Acquire video data, wind speed data, and wind direction data for each wind turbine generator set within a preset time period; wherein, the video data records the rotation of the wind turbine generator set's rotor. The system uses a pre-defined matching model to determine whether the rotation of the wind turbine rotor of each wind turbine matches the wind speed and wind direction data. If there is a mismatch between the rotation status of the wind turbine of the first wind turbine and the wind speed and direction data, the current weather conditions at the location of the first wind turbine are obtained from the video data, the future weather conditions are predicted based on the current weather conditions, and the risk type of the wind power plant is estimated based on the future weather conditions; wherein, the first wind turbine is any one of the plurality of wind turbines.

2. The method of claim 1, wherein, The step of determining whether the rotor rotation of each wind turbine matches the wind speed and wind direction data using a preset matching model includes: The rotation of the wind turbine rotor of each wind turbine is divided into discrete rotor states using an optical flow algorithm, and the wind speed data and wind direction data of each wind turbine are also divided into discrete wind speed states and discrete wind direction states. The rotor states include low speed, medium speed and high speed, the wind speed states include low wind speed, medium wind speed and high wind speed, and the wind direction states include multiple wind direction angle ranges. Obtain a preset set of state rules; wherein the set of state rules includes multiple different state rules, and each state rule is a combination of a wind turbine state, a wind speed state, and a wind direction state; If the combination of the rotor state, wind speed state, and wind direction state of the first wind turbine belongs to the state rule set, then it is determined that the rotor rotation of the first wind turbine matches the wind speed data and wind direction data. If the combination of the rotor state, wind speed state, and wind direction state of the first wind turbine does not belong to the state rule set, then it is determined that the rotor rotation of the first wind turbine does not match the wind speed data and wind direction data.

3. The method of claim 2, wherein, The acquisition of the preset state rule set includes: Multiple sets of sample data were obtained; each set of sample data included the rotor rotation, wind speed, and wind direction of a wind turbine under risk-free conditions. The optical flow algorithm is used to divide the sample wind turbine rotation of each wind turbine into discrete wind turbine states, the sample wind speed data of each wind turbine into discrete wind speed states, and the sample wind direction data of each wind turbine into discrete wind direction states, so as to obtain multiple state sets. Multiple i-th state sets are generated, the frequency proportion of each i-th state set is calculated, and state sets with a frequency proportion greater than a preset frequency proportion threshold are retained as candidate state rules; wherein, the i-th state set includes i wind turbine state, wind speed state and wind direction state, and the initial value of i is 1. Let i = i + 1, return to the step of generating multiple i-th state sets and subsequent steps, until the number of candidate state rules no longer increases; Each candidate state rule is split into a first sub-state set and a second sub-state set. Within each candidate state rule, the joint probability and correlation coefficient between the first sub-state set and the second sub-state set are calculated. All candidate state rules with a joint probability greater than a preset joint probability threshold and a correlation coefficient greater than a preset correlation coefficient threshold are retained as the state rule set.

4. The method of claim 3, wherein, The frequency proportion of each i-th state set is calculated, expressed as: In the above formula, FR(x) represents the frequency proportion of the state set x, N x represents the total number of state sets containing the state set x, and T represents the total number of all state sets. The calculation formula of the joint probability is: In the above formula, JP(ab) represents the joint probability between the first substate set a and the second substate set b; The formula for calculating the correlation coefficient is: In the above formula, AC(ab) represents the correlation coefficient between the first sub-state set a and the second sub-state set b.

5. The method of claim 1, wherein, The process of predicting future weather conditions based on the current weather conditions and estimating the risk type of wind power plants based on the future weather conditions includes: The current weather conditions are input into an LSTM prediction model to predict future weather conditions, and a decision tree is used to classify the risk type of the wind power plant based on the future weather conditions.

6. The method of claim 1, wherein, Infrared sensors are installed around the wind power plant, and the method further includes: If a warning is issued when a signal is received from the infrared sensor indicating that the wind power plant is at risk, an early warning will be provided.

7. The method of claim 1, wherein, The method further includes: Collect historical data; wherein, the historical data includes video data, wind speed data, and wind direction data of wind turbine generators under extreme weather conditions; A risk model is established based on the historical data; wherein, the risk model defines the threshold range of normal operation and abnormal operation of wind turbine generator sets, and the risk types under different abnormal operation states; The video data, wind speed data, and wind direction data of each wind turbine within a preset time period are compared with the threshold range defined in the risk model as real-time data. When the real-time data exceeds the threshold range, a risk warning signal is generated.

8. A device for risk prediction of a wind power plant under extreme weather conditions, characterized in that, The risk prediction device for wind power plants under extreme weather conditions includes: The data acquisition module is used to acquire video data, wind speed data, and wind direction data for each wind turbine generator set within a preset time period; wherein, the video data records the rotation of the wind turbine generator set's rotor. The data matching module is used to determine whether the rotation status of the wind turbine rotor of each wind turbine is matched with the wind speed and wind direction data through a preset matching model. The risk assessment module is used to obtain the current weather conditions of the location of the first wind turbine from video data if there is a mismatch between the rotation status of the wind turbine and the wind speed and direction data; predict the future weather conditions based on the current weather conditions; and estimate the risk type of the wind power plant based on the future weather conditions; wherein the first wind turbine is any one of the plurality of wind turbines.

9. A computer-readable storage medium, characterized in that, The device stores a computer program that, when executed by a processor, causes the processor to perform the steps of the method as described in any one of claims 1 to 7.

10. A terminal device, comprising: It includes a memory and a processor, the memory storing a computer program that, when executed by the processor, causes the processor to perform the steps of the method as described in any one of claims 1 to 7.

Images