Fire early warning system for wind power generator cabin

Abstract

The invention provides a fire early warning system for a wind power generator cabin, and relates to the technical field of fan fire protection, and the system comprises a visual smoke AI recognition module which is used for obtaining the smoke concentration and diffusion velocity of each segmented area of the wind power generator cabin in each monitoring period; the temperature acquisition and analysis unit is used for acquiring the temperature of each segmented area of the wind power generator cabin in each monitoring period and calculating the temperature rising speed of the wind power generator cabin in each monitoring period; the fire early warning module is used for judging whether a fire early warning signal is sent to the control center or not based on the fire risk assessment coefficient of each segmented area of the wind power generator cabin; the cabin point cloud model construction module is used for performing color marking on each coordinate point of the cabin original point cloud model to form a cabin real-time fire condition point cloud model; and the automatic fire-fighting module is used for starting fire-fighting equipment to carry out targeted fire fighting. According to the invention, fire risk assessment can be carried out before a fire occurs, so that fire-fighting equipment is started in advance to control fire spreading to the greatest extent.

Description

Technical Field

[0001] The invention belongs to the technical field of wind turbine fire protection, and in particular relates to a fire warning system for a wind turbine generator cabin. Background Art

[0002] With the rapid development of new energy construction, wind power generation, as an important renewable energy source, has seen its installed capacity continue to grow. However, wind turbines are usually built far away from cities, mountaintops, sea surfaces, etc. Once a fire occurs, it is difficult to achieve immediate and effective firefighting and rescue, which can easily cause great property losses. At present, wind power generation technology is widely used, and the unit capacity and cost of wind turbines are constantly increasing. At the same time, the development of offshore wind power has also put forward higher requirements for fire protection systems; The airflow speed in the cabin of a wind turbine is fast and there are many fire hazards. When a fire occurs, it spreads quickly. The existing wind turbine fire-fighting system can only send out a fire alarm when a fire occurs, but cannot conduct a fire risk assessment before the fire occurs, thereby preventing the fire from spreading by activating fire-fighting equipment in advance. Summary of the invention

[0003] The present invention provides a fire warning system for a wind turbine cabin, so as to solve at least one of the above-mentioned technical problems.

[0004] In order to solve the above technical problems, the present invention discloses a fire warning system for a wind turbine cabin, comprising: The visual smoke AI recognition module is used to intelligently identify and analyze the surveillance video images in the wind turbine cabin, and obtain the concentration and diffusion speed of smoke in each segmented area of ​​the wind turbine cabin during each monitoring period; A temperature collection and analysis unit, used to collect the temperature of each segmented area of ​​the wind turbine generator cabin in each monitoring period and calculate the temperature increase rate of the wind turbine generator cabin in each monitoring period; The fire warning module is used to perform fire risk assessment based on the current smoke concentration, diffusion speed, cabin temperature, and temperature rise speed, obtain the fire risk assessment coefficient of each segmented area of ​​the wind turbine cabin, and send a fire warning signal to the control center based on the fire risk assessment coefficient of each segmented area of ​​the wind turbine cabin to obtain the fire fighting strategy for each segmented area of ​​the wind turbine cabin; The cabin point cloud model building module is used to build the cabin original point cloud model based on the monitoring video in the wind turbine cabin without abnormal conditions, and to color-mark each coordinate point of the cabin original point cloud model based on the fire risk assessment results of each segmented area of ​​the wind turbine cabin in each monitoring period within the total monitoring time, so as to form a cabin real-time fire status point cloud model and transmit it to the control center; The automatic fire-fighting module is used to start the corresponding fire-fighting equipment for targeted fire-fighting based on the real-time fire situation point cloud model of the cabin and the fire-fighting strategy given by the control center.

[0005] Preferably, the visual smoke AI recognition module includes: The optimal analysis image acquisition submodule is used to collect the monitoring video in the wind turbine cabin in each monitoring period, and obtain the optimal analysis image of each segmented area of ​​the wind turbine cabin based on the monitoring video in the wind turbine cabin in each monitoring period; The optimal analysis image preprocessing submodule is used to perform image preprocessing on the optimal analysis image of each segmented area of ​​the wind turbine cabin in each monitoring period; The identification and analysis submodule is used to analyze whether there is smoke in each segmented area based on the best analysis image of each segmented area of ​​the wind turbine cabin in each monitoring period, and to obtain the current concentration and diffusion speed of the smoke in each segmented area.

[0006] Preferably, the analysis and identification submodule includes: The smoke recognition unit is used to input the best analysis image of each segmented area of ​​the wind turbine cabin in each monitoring period into the trained smoke recognition model to determine whether there is smoke in the current best analysis image; The smoke analysis unit is used to input the best analysis image of each segmented area of ​​the wind turbine cabin in each monitoring period into the trained smoke analysis model, and analyze the concentration and diffusion speed of the smoke in the current best analysis image; The identification and analysis result output unit is used to transmit the smoke identification and analysis results to the fire warning module.

[0007] Preferably, the temperature rise rate of the wind turbine cabin in each monitoring cycle is calculated: ;in, is the cabin temperature increase rate of the i-th monitoring cycle in the j-th partition area, is the cabin temperature of the i-th monitoring period in the j-th segmentation area, is the cabin temperature of the jth partition area in the i+1th monitoring period, is the total duration of the i-th monitoring cycle.

[0008] Preferably, the fire warning module includes: A data receiving submodule is used to receive the concentration, diffusion rate, temperature, and temperature rise rate of each segmented area of ​​the wind turbine cabin in each monitoring period; The fire risk assessment submodule is used to calculate the fire risk assessment coefficient of each segmented area of ​​the wind turbine cabin within the current total monitoring time based on the smoke concentration, smoke diffusion speed and temperature, and temperature rise speed of each segmented area of ​​the wind turbine cabin in each monitoring period within the total monitoring time; The alarm decision submodule sends a fire warning signal to the control center when the fire risk assessment coefficient of the corresponding segmented area within the current total monitoring time is greater than the benchmark fire risk assessment coefficient. The control center then gives a fire fighting strategy for each segmented area in the wind turbine cabin. The control center provides fire fighting strategies for each partition area of ​​the wind turbine cabin, including: When the fire is small, choose the hot aerosol fire-fighting equipment closest to the partition area to extinguish the fire. Otherwise, choose the ultra-fine dry powder fire-fighting equipment closest to the partition area to extinguish the fire.

[0009] Preferably, based on the smoke concentration, smoke diffusion speed and temperature, and temperature rise speed of each segmented area of ​​the wind turbine cabin in each monitoring period within the total monitoring time, the fire risk assessment coefficient of each segmented area of ​​the wind turbine cabin within the current total monitoring time is calculated, including: ;in, is the fire risk assessment coefficient of the jth segmented area of ​​the wind turbine cabin within the current monitoring time, is the number of monitoring cycles within the total monitoring duration, is the weight value of smoke density, is the smoke concentration of the jth segmentation area in the ith monitoring period, is the weight value of cabin temperature, is the cabin temperature of the i-th monitoring period in the j-th segmentation area, is the weight value of smoke diffusion speed, is the smoke diffusion speed of the j-th segmentation area in the i-th monitoring period, is the weight value of the cabin temperature increase rate, is the cabin temperature increase rate of the i-th monitoring cycle in the j-th block segmentation area.

[0010] Preferably, the cabin point cloud model building module includes: The cabin original point cloud model construction submodule is used to create the cabin original point cloud model based on the monitoring video inside the wind turbine cabin when there is no abnormality; The submodule for creating a point cloud model of the real-time fire status in the cabin is used to color-mark the corresponding areas on the original point cloud model of the cabin based on the fire risk assessment coefficient of each segmented area of ​​the wind turbine cabin in each monitoring period within the total monitoring time and the benchmark fire risk assessment coefficient, to form a point cloud model of the real-time fire status in the cabin.

[0011] Preferably, the cabin real-time fire status point cloud model creation submodule includes: A color marking unit is used to compare the fire risk assessment coefficient of each segmented area of ​​the wind turbine cabin in each monitoring period with the benchmark fire risk assessment coefficient. If the fire risk assessment coefficient of the segmented area is greater than the benchmark fire risk assessment coefficient, the corresponding segmented area is color marked once; The color elimination unit is used to count the number of color markings on each segmented area. If the number of color markings is zero for x consecutive monitoring cycles, the corresponding segmented area will be color eliminated once. After the total monitoring time is over, a point cloud model of the real-time fire status in the cabin is formed.

[0012] Preferably, the fire warning module further includes a fire spread degree estimation submodule, and the fire spread degree estimation submodule includes: A fire risk assessment coefficient prediction unit is used to construct a fire risk assessment coefficient prediction matrix based on the fire risk assessment coefficient of each segmented area of ​​the wind turbine generator cabin in each monitoring period, and predict the fire risk assessment coefficient of the corresponding segmented area of ​​the wind turbine generator cabin in the next monitoring period; The fire spread coefficient calculation unit is used to calculate the quotient of the difference between the fire risk assessment coefficient of the divided area corresponding to the wind turbine generator cabin in the next monitoring period and the fire risk assessment coefficient of the divided area corresponding to the wind turbine generator cabin in the current monitoring period and the preset fire risk assessment coefficient difference, so as to obtain the fire spread coefficient of the current divided area; The fire isolation assessment unit, if the fire spread coefficient of the current segmented area is a negative number, there is no need to isolate the fire; if the fire spread coefficient of the current segmented area is a positive number, fire isolation is required, and the difference between the product of the fire spread coefficient of the current segmented area and the area of ​​the current segmented area and the area of ​​the current segmented area is calculated to obtain the area of ​​the area that needs to be isolated from the fire, and the area data of the area that needs to be isolated from the fire is sent to the control center, which determines the range of fire isolation based on the area of ​​the area that needs to be isolated from the fire.

[0013] Preferably, the fire risk assessment coefficient prediction matrix is: ;in, is the fire risk assessment coefficient prediction matrix of the j-th segmentation area of ​​the wind turbine cabin in the i-th monitoring period, is the fire risk assessment coefficient of the jth segmented area in the first monitoring cycle of the wind turbine cabin, is the fire risk assessment coefficient of the jth segmented area in the second monitoring cycle of the wind turbine cabin, is the fire risk assessment coefficient of the j-th segmented area in the third monitoring cycle of the wind turbine cabin, is the fire risk assessment coefficient of the jth segmented area in the fourth monitoring cycle of the wind turbine cabin, For wind turbine cabin The fire risk assessment coefficient of the jth segmented area in the monitoring period is: For wind turbine cabin Fire risk assessment coefficient of the jth segmented area in a monitoring period; The fire risk assessment coefficient of the jth segmentation area in the wind turbine cabin in the i+1th monitoring cycle: Obtain the mean of all matrix elements in any row except the first row in the fire risk assessment coefficient prediction matrix of the j-th segmented area of ​​the wind turbine cabin in the ith monitoring period, and set the value of the last column in the fire risk assessment coefficient prediction matrix of the j-th segmented area of ​​the wind turbine cabin in the ith monitoring period to a value equal to the mean, thereby obtaining a new matrix, and taking the rank of the new matrix as the fire risk assessment coefficient of the j-th segmented area of ​​the wind turbine cabin in the i+1-th monitoring period.

[0014] Compared with the prior art, the present invention has the following beneficial effects: The present invention can detect smoke concentration and diffusion speed more accurately through intelligent recognition and analysis of monitoring video images. The fire warning module can quickly assess the fire risk. When the fire risk assessment coefficient is greater than the benchmark fire risk assessment coefficient, a fire warning signal is sent to the control center so that the control center can take measures quickly. The cabin point cloud model construction module makes the fire risk situation clear at a glance, which is convenient for management and control. The automatic fire fighting module can start the corresponding fire extinguishing equipment according to the real-time fire situation and fire fighting strategy, thereby improving the fire fighting efficiency. The present invention can conduct a fire risk assessment before a fire occurs, thereby activating fire-fighting equipment in advance to control the spread of fire to the greatest extent. The fire situation in each area can be dynamically updated through the real-time fire situation point cloud model of the cabin, thereby realizing dynamic monitoring of the fire-fighting effect and facilitating remote panoramic control of the fire in the entire cabin. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings: Figure 1 The present invention is a schematic diagram of a fire warning system for a wind turbine cabin. DETAILED DESCRIPTION

[0016] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0017] In addition, in the present invention, the descriptions of "first", "second", etc. are only used for descriptive purposes, and do not specifically refer to the order or sequence, nor are they used to limit the present invention. They are only used to distinguish components or operations described with the same technical terms, and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions and technical features between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0018] The present invention provides the following embodiments Example 1 The embodiment of the present invention provides a fire warning system for a wind turbine cabin, such as Figure 1 As shown, including: The visual smoke AI recognition module is used to intelligently identify and analyze the surveillance video images in the wind turbine cabin, and obtain the concentration and diffusion speed of smoke in each segmented area of ​​the wind turbine cabin during each monitoring period; A temperature collection and analysis unit, used to collect the temperature of each segmented area of ​​the wind turbine generator cabin in each monitoring period and calculate the temperature increase rate of the wind turbine generator cabin in each monitoring period; The fire warning module is used to perform fire risk assessment based on the current smoke concentration, diffusion speed, cabin temperature, and temperature rise speed, obtain the fire risk assessment coefficient of each segmented area of ​​the wind turbine cabin, and send a fire warning signal to the control center based on the fire risk assessment coefficient of each segmented area of ​​the wind turbine cabin to obtain the fire fighting strategy for each segmented area of ​​the wind turbine cabin; The cabin point cloud model building module is used to build the cabin original point cloud model based on the monitoring video in the wind turbine cabin without abnormal conditions, and to color-mark each coordinate point of the cabin original point cloud model based on the fire risk assessment results of each segmented area of ​​the wind turbine cabin in each monitoring period within the total monitoring time, so as to form a cabin real-time fire status point cloud model and transmit it to the control center; The automatic fire-fighting module is used to start the corresponding fire-fighting equipment for targeted fire-fighting based on the real-time fire situation point cloud model of the cabin and the fire-fighting strategy given by the control center.

[0019] In this embodiment, the fire-fighting equipment in the cabin is arranged in different areas, and the divided areas are areas obtained by dividing the interior area of ​​the cabin according to the blocks where the fire-fighting equipment is arranged.

[0020] In this embodiment, a fire warning signal is sent to the control center based on the fire risk assessment coefficient of each divided area of ​​the wind turbine cabin, including when the fire risk assessment coefficient of the divided area is greater than the benchmark judgment fire risk assessment coefficient, it proves that there is a fire risk, and a fire warning signal is sent to the control center.

[0021] In this embodiment, the fire risk assessment coefficient is a numerical value used to assess the risk of fire occurrence and the degree of fire in the divided area.

[0022] In this embodiment, the fire fighting strategy includes selecting the hot aerosol fire fighting equipment closest to the partitioned area to extinguish the fire when the fire is small, and otherwise selecting the ultra-fine dry powder fire fighting equipment closest to the partitioned area to extinguish the fire.

[0023] In this embodiment, the original point cloud model of the nacelle is a digital model of the wind turbine nacelle without abnormalities, which is composed of a large number of three-dimensional coordinate points.

[0024] In this embodiment, the real-time fire status point cloud model of the cabin is a point cloud model obtained by color marking each coordinate point of the original point cloud model of the cabin based on the fire risk assessment results of each segmented area of ​​the wind turbine cabin in each monitoring period within the total monitoring time.

[0025] In this embodiment, based on the real-time fire situation point cloud model of the cabin and the fire fighting strategy given by the control center, the corresponding fire fighting equipment is activated to carry out targeted fire fighting, including activating the fire fighting equipment in the corresponding segmented areas to extinguish or isolate the fire in the segmented areas with different color depths.

[0026] Beneficial effects of the above technical solution: the present invention can detect smoke concentration and diffusion speed more accurately through intelligent recognition and analysis of monitoring video images. The fire warning module can quickly assess the fire risk. When the fire risk assessment coefficient is greater than the benchmark fire risk assessment coefficient, a fire warning signal is sent to the control center so that the control center can take measures quickly. The cabin point cloud model construction module makes the fire risk situation clear at a glance, which is convenient for management and control. The automatic fire fighting module can start the corresponding fire fighting equipment according to the real-time fire situation and fire fighting strategy, thereby improving the fire fighting efficiency. The present invention can perform fire risk assessment before a fire occurs, thereby activating fire-fighting equipment in advance to control the spread of fire to the greatest extent.

[0027] Example 2 Based on Example 1, the visual smoke AI recognition module includes: The optimal analysis image acquisition submodule is used to collect the monitoring video in the wind turbine cabin in each monitoring period, and obtain the optimal analysis image of each segmented area of ​​the wind turbine cabin based on the monitoring video in the wind turbine cabin in each monitoring period; The optimal analysis image preprocessing submodule is used to perform image preprocessing on the optimal analysis image of each segmented area of ​​the wind turbine cabin in each monitoring period; The identification and analysis submodule is used to analyze whether there is smoke in each segmented area based on the best analysis image of each segmented area of ​​the wind turbine cabin in each monitoring period, and to obtain the current concentration and diffusion speed of the smoke in each segmented area.

[0028] In this embodiment, the best analysis image is an image with the best image quality among a plurality of video frame images in the monitoring video inside the wind turbine cabin in each monitoring cycle.

[0029] In this embodiment, image preprocessing includes denoising, edge monitoring, threshold processing and feature extraction.

[0030] The beneficial effects of the above technical solution are: through optimal analysis of image acquisition and preprocessing, the accuracy of image recognition is improved, the effect of real-time monitoring of the smoke concentration and diffusion speed of each segmented area is achieved, the response speed of the early warning system is improved, and the possibility of false alarms is reduced by preprocessing the image.

[0031] Example 3 Based on Example 2, the analysis and identification submodule includes: The smoke recognition unit is used to input the best analysis image of each segmented area of ​​the wind turbine cabin in each monitoring period into the trained smoke recognition model to determine whether there is smoke in the current best analysis image; The smoke analysis unit is used to input the best analysis image of each segmented area of ​​the wind turbine cabin in each monitoring period into the trained smoke analysis model, and analyze the concentration and diffusion speed of the smoke in the current best analysis image; The identification and analysis result output unit is used to transmit the smoke identification and analysis results to the fire warning module.

[0032] In this embodiment, the smoke recognition model is a model obtained by training a neural network model using cabin images with or without smoke as input and judgment results of whether or not smoke exists as output.

[0033] In this embodiment, the smoke analysis model is a model obtained by training a neural network model using a cabin image with a large amount of smoke as input and the smoke concentration and diffusion speed corresponding to the cabin image as output.

[0034] The beneficial effects of the above technical solution are: by training the smoke recognition model and the smoke analysis model, the recognition accuracy is improved, the presence of smoke, its concentration and diffusion rate can be quickly identified, the response speed of the early warning system is improved, the degree of automation is high, and the need for manual intervention is reduced.

[0035] Example 4 Based on Example 1, the temperature rise rate of the wind turbine cabin in each monitoring cycle is calculated: ;in, is the cabin temperature increase rate of the i-th monitoring period in the j-th partition area, is the cabin temperature of the i-th monitoring period in the j-th segmentation area, is the cabin temperature of the jth partition area in the i+1th monitoring period, is the total duration of the i-th monitoring cycle.

[0036] The beneficial effects of the above technical solution are: it can monitor the temperature changes in each monitoring cycle in real time, and the calculation of the temperature rise rate helps to more accurately assess the fire risk. By monitoring the temperature changes in real time, timely measures can be taken to prevent fires.

[0037] Example 5 Based on Example 1, the fire warning module includes: A data receiving submodule is used to receive the concentration, diffusion rate, temperature, and temperature rise rate of each segmented area of ​​the wind turbine cabin in each monitoring period; The fire risk assessment submodule is used to calculate the fire risk assessment coefficient of each segmented area of ​​the wind turbine cabin within the current total monitoring time based on the smoke concentration, smoke diffusion speed and temperature, and temperature rise speed of each segmented area of ​​the wind turbine cabin in each monitoring period within the total monitoring time; The alarm decision submodule sends a fire warning signal to the control center when the fire risk assessment coefficient of the corresponding segmented area within the current total monitoring time is greater than the benchmark fire risk assessment coefficient. The control center then gives a fire fighting strategy for each segmented area in the wind turbine cabin. The control center provides fire fighting strategies for each partition area of ​​the wind turbine cabin, including: When the fire is small, choose the hot aerosol fire-fighting equipment closest to the partition area to extinguish the fire. Otherwise, choose the ultra-fine dry powder fire-fighting equipment closest to the partition area to extinguish the fire.

[0038] The beneficial effects of the above technical solution are: by comprehensively evaluating the fire risk by multiple parameters, the accuracy of the early warning is improved, and the control center can judge the size of the fire based on the gap between the fire risk assessment coefficient and the benchmark fire risk assessment coefficient, and when the fire is small, the hot aerosol fire-fighting equipment closest to the divided area is selected for fire extinguishing, otherwise the ultra-fine dry powder fire-fighting equipment closest to the divided area is selected for fire extinguishing, thereby rationally utilizing the fire-fighting equipment to provide the best fire-fighting strategy.

[0039] Example 6 On the basis of Example 5, based on the smoke concentration, smoke diffusion speed and temperature, and temperature rise rate of each segmented area of ​​the wind turbine cabin in each monitoring period within the total monitoring time, the fire risk assessment coefficient of each segmented area of ​​the wind turbine cabin within the current total monitoring time is calculated, including: ;in, is the fire risk assessment coefficient of the jth segmented area of ​​the wind turbine cabin within the current monitoring time, is the number of monitoring cycles within the total monitoring duration, is the weight value of smoke density, is the smoke concentration of the jth segmentation area in the ith monitoring period, is the weight value of cabin temperature, is the cabin temperature of the i-th monitoring period in the j-th segmentation area, is the weight value of smoke diffusion speed, is the smoke diffusion speed of the j-th segmentation area in the i-th monitoring period, is the weight value of the cabin temperature increase rate, is the cabin temperature increase rate of the i-th monitoring cycle in the j-th block segmentation area.

[0040] The beneficial effects of the above technical solution are: by comprehensively considering multiple parameters (smoke concentration, temperature, smoke diffusion rate and temperature rise rate), the accuracy of fire risk assessment is improved, the weight value can be adjusted to adapt to different environmental conditions, the flexibility of the early warning system is improved, and the fire risk assessment is carried out by integrating multiple factors, thereby improving the reliability of the early warning.

[0041] Example 7 Based on Example 1, the cabin point cloud model construction module includes: The cabin original point cloud model construction submodule is used to create the cabin original point cloud model based on the monitoring video inside the wind turbine cabin when there is no abnormality; The submodule for creating a point cloud model of the real-time fire status in the cabin is used to color-mark the corresponding areas on the original point cloud model of the cabin based on the fire risk assessment coefficient of each segmented area of ​​the wind turbine cabin in each monitoring period within the total monitoring time and the benchmark fire risk assessment coefficient, to form a point cloud model of the real-time fire status in the cabin.

[0042] In this embodiment, creating the original point cloud model of the cabin specifically includes obtaining the best analysis image of each segmented area of ​​the wind turbine cabin based on the surveillance video in the wind turbine cabin during each monitoring period, that is, taking the video frame with the highest clarity of each segmented area in the surveillance video in the wind turbine cabin during each monitoring period as the best analysis image of each segmented area, performing image preprocessing and depth information extraction on the best analysis image of each segmented area, obtaining the three-dimensional coordinate information of each pixel point in the best analysis image of each segmented area, constructing the original point cloud model of the cabin based on the three-dimensional coordinate information of each pixel point in the best analysis image of each segmented area, and optimizing the original point cloud model of the cabin.

[0043] In this embodiment, based on the fire risk assessment coefficient of each segmented area of ​​the wind turbine cabin in each monitoring period within the total monitoring time and the benchmark fire risk assessment coefficient, the corresponding areas on the original point cloud model of the cabin are color marked, including real-time deepening or weakening of the color of each segmented area on the original point cloud model of the cabin. The higher the color saturation, the more serious the fire.

[0044] The beneficial effects of the above technical solution are: through color marking, the fire risk situation of each area is intuitively displayed, and the point cloud model of the real-time fire situation in the cabin makes the fire risk situation clear at a glance, which is convenient for management and control. Through the point cloud model, the visualization of the early warning system is improved, which facilitates panoramic control of the fire in the entire cabin.

[0045] Example 8 Based on Example 7, the cabin real-time fire situation point cloud model creation submodule includes: A color marking unit is used to compare the fire risk assessment coefficient of each segmented area of ​​the wind turbine cabin in each monitoring period with the benchmark fire risk assessment coefficient. If the fire risk assessment coefficient of the segmented area is greater than the benchmark fire risk assessment coefficient, the corresponding segmented area is color marked once; The color elimination unit is used to count the number of color markings on each segmented area. If the number of color markings is zero for x consecutive monitoring cycles, the corresponding segmented area will be color eliminated once. After the total monitoring time is over, a point cloud model of the real-time fire status in the cabin is formed.

[0046] The beneficial effects of the above technical solution are: through color marking and elimination, the fire situation in each area can be dynamically updated, dynamic monitoring of fire fighting effects can be achieved, and remote panoramic control of the fire in the entire cabin is convenient.

[0047] Example 9 Based on Example 1, The fire warning module also includes a fire spread estimation submodule, which includes: A fire risk assessment coefficient prediction unit is used to construct a fire risk assessment coefficient prediction matrix based on the fire risk assessment coefficient of each segmented area of ​​the wind turbine generator cabin in each monitoring period, and predict the fire risk assessment coefficient of the corresponding segmented area of ​​the wind turbine generator cabin in the next monitoring period; The fire spread coefficient calculation unit is used to calculate the quotient of the difference between the fire risk assessment coefficient of the divided area corresponding to the wind turbine generator cabin in the next monitoring period and the fire risk assessment coefficient of the divided area corresponding to the wind turbine generator cabin in the current monitoring period and the preset fire risk assessment coefficient difference, so as to obtain the fire spread coefficient of the current divided area; The fire isolation assessment unit, if the fire spread coefficient of the current segmented area is a negative number, then there is no need to isolate the fire; if the fire spread coefficient of the current segmented area is a positive number, then it is necessary to isolate the fire, and calculate the difference between the product of the fire spread coefficient of the current segmented area and the area of ​​the current segmented area and the area of ​​the current segmented area to obtain the area of ​​the area that needs to be isolated from the fire, and transmit the area data of the area that needs to be isolated from the fire to the control center, and the control center determines the fire isolation range based on the area of ​​the area that needs to be isolated from the fire; Fire risk assessment coefficient prediction matrix: ;in, is the fire risk assessment coefficient prediction matrix of the j-th segmentation area of ​​the wind turbine cabin in the i-th monitoring period, is the fire risk assessment coefficient of the jth segmented area in the first monitoring cycle of the wind turbine cabin, is the fire risk assessment coefficient of the jth segmented area in the second monitoring period of the wind turbine cabin, is the fire risk assessment coefficient of the j-th segmented area in the third monitoring cycle of the wind turbine cabin, is the fire risk assessment coefficient of the jth segmented area in the fourth monitoring cycle of the wind turbine cabin, For wind turbine cabin The fire risk assessment coefficient of the jth segmented area in the monitoring period is: For wind turbine cabin Fire risk assessment coefficient of the jth segmented area in a monitoring period; The fire risk assessment coefficient of the jth segmentation area in the wind turbine cabin in the i+1th monitoring cycle: Obtain the mean of all matrix elements in any row except the first row in the fire risk assessment coefficient prediction matrix of the j-th segmented area of ​​the wind turbine cabin in the ith monitoring period, and set the value of the last column in the fire risk assessment coefficient prediction matrix of the j-th segmented area of ​​the wind turbine cabin in the ith monitoring period to a value equal to the mean, thereby obtaining a new matrix, and taking the rank of the new matrix as the fire risk assessment coefficient of the j-th segmented area of ​​the wind turbine cabin in the i+1-th monitoring period.

[0048] The beneficial effects of the above technical solution are: for areas with a positive fire spread coefficient, the area that needs to be isolated can be determined based on the product of the spread coefficient and the area and the difference between the area, so that a targeted fire isolation strategy can be formulated to effectively control the spread of the fire. By updating the fire risk assessment coefficient and the fire spread coefficient in real time, dynamic monitoring and management of fire risks can be achieved. Based on these data, the control center can flexibly adjust the fire fighting strategy and improve the efficiency of emergency response. At the same time, by scientifically evaluating the degree of fire spread, the allocation of fire fighting resources can be optimized.

[0049] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A fire warning system for a wind turbine cabin, characterized in that: include: The visual smoke AI recognition module is used to intelligently identify and analyze the surveillance video images in the wind turbine cabin, and obtain the concentration and diffusion speed of smoke in each segmented area of ​​the wind turbine cabin during each monitoring period; A temperature collection and analysis unit, used to collect the temperature of each segmented area of ​​the wind turbine generator cabin in each monitoring period and calculate the temperature increase rate of the wind turbine generator cabin in each monitoring period; The fire warning module is used to perform fire risk assessment based on the current smoke concentration, diffusion speed, cabin temperature, and temperature rise speed, obtain the fire risk assessment coefficient of each segmented area of ​​the wind turbine cabin, and determine whether to send a fire warning signal to the control center based on the fire risk assessment coefficient of each segmented area of ​​the wind turbine cabin, and obtain the fire fighting strategy for each segmented area of ​​the wind turbine cabin; The cabin point cloud model building module is used to build the cabin original point cloud model based on the monitoring video in the wind turbine cabin without abnormal conditions, and to color-mark each coordinate point of the cabin original point cloud model based on the fire risk assessment results of each segmented area of ​​the wind turbine cabin in each monitoring period within the total monitoring time, so as to form a cabin real-time fire status point cloud model and transmit it to the control center; The automatic fire-fighting module is used to start the corresponding fire-fighting equipment for targeted fire-fighting based on the real-time fire situation point cloud model of the cabin and the fire-fighting strategy given by the control center.

2. A fire warning system for a wind turbine cabin according to claim 1, characterized in that: The visual smoke AI recognition module includes: The optimal analysis image acquisition submodule is used to collect the monitoring video in the wind turbine cabin in each monitoring period, and obtain the optimal analysis image of each segmented area of ​​the wind turbine cabin based on the monitoring video in the wind turbine cabin in each monitoring period; The optimal analysis image preprocessing submodule is used to perform image preprocessing on the optimal analysis image of each segmented area of ​​the wind turbine cabin in each monitoring period; The identification and analysis submodule is used to analyze whether there is smoke in each segmented area based on the best analysis image of each segmented area of ​​the wind turbine cabin in each monitoring period, and to obtain the current concentration and diffusion speed of the smoke in each segmented area.

3. A fire warning system for a wind turbine cabin according to claim 2, characterized in that: The analysis and identification submodules include: The smoke recognition unit is used to input the best analysis image of each segmented area of ​​the wind turbine cabin in each monitoring period into the trained smoke recognition model to determine whether there is smoke in the current best analysis image; The smoke analysis unit is used to input the best analysis image of each segmented area of ​​the wind turbine cabin in each monitoring period into the trained smoke analysis model, and analyze the concentration and diffusion speed of the smoke in the current best analysis image; The identification and analysis result output unit is used to transmit the smoke identification and analysis results to the fire warning module.

4. A fire warning system for a wind turbine cabin according to claim 1, characterized in that: Calculate the temperature rise rate of the wind turbine cabin in each monitoring cycle: ;in, is the cabin temperature increase rate of the i-th monitoring period in the j-th partition area, is the cabin temperature of the i-th monitoring period in the j-th segmentation area, is the cabin temperature of the jth partition area in the i+1th monitoring period, is the total duration of the i-th monitoring cycle.

5. The fire warning system for a wind turbine cabin according to claim 1, characterized in that: The fire warning module includes: A data receiving submodule is used to receive the concentration, diffusion rate, temperature, and temperature rise rate of each segmented area of ​​the wind turbine cabin in each monitoring period; The fire risk assessment submodule is used to calculate the fire risk assessment coefficient of each segmented area of ​​the wind turbine cabin within the current total monitoring time based on the smoke concentration, smoke diffusion speed and temperature, and temperature rise speed of each segmented area of ​​the wind turbine cabin in each monitoring period within the total monitoring time; The alarm decision submodule sends a fire warning signal to the control center when the fire risk assessment coefficient of the corresponding segmented area within the current total monitoring time is greater than the benchmark fire risk assessment coefficient. The control center then gives a fire fighting strategy for each segmented area in the wind turbine cabin. The control center provides fire fighting strategies for each partition area of ​​the wind turbine cabin, including: When the fire is small, choose the hot aerosol fire-fighting equipment closest to the partition area to extinguish the fire. Otherwise, choose the ultra-fine dry powder fire-fighting equipment closest to the partition area to extinguish the fire.

6. A fire warning system for a wind turbine cabin according to claim 5, characterized in that: Based on the smoke concentration, smoke diffusion speed and temperature, and temperature rise rate of each segmented area of ​​the wind turbine cabin in each monitoring cycle within the total monitoring time, the fire risk assessment coefficient of each segmented area of ​​the wind turbine cabin within the current total monitoring time is calculated, including: ;in, is the fire risk assessment coefficient of the jth segmented area of ​​the wind turbine cabin within the current monitoring time, is the number of monitoring cycles within the total monitoring duration, is the weight value of smoke density, is the smoke concentration of the jth segmentation area in the ith monitoring period, is the weight value of cabin temperature, is the cabin temperature of the i-th monitoring period in the j-th segmentation area, is the weight value of smoke diffusion speed, is the smoke diffusion speed of the j-th segmentation area in the i-th monitoring period, is the weight value of the cabin temperature increase rate, is the cabin temperature increase rate of the i-th monitoring cycle in the j-th block segmentation area.

7. A fire warning system for a wind turbine cabin according to claim 1, characterized in that: The cabin point cloud model building modules include: The cabin original point cloud model construction submodule is used to create the cabin original point cloud model based on the monitoring video inside the wind turbine cabin when there is no abnormality; The submodule for creating a point cloud model of the real-time fire status in the cabin is used to color-mark the corresponding areas on the original point cloud model of the cabin based on the fire risk assessment coefficient of each segmented area of ​​the wind turbine cabin in each monitoring period within the total monitoring time and the benchmark fire risk assessment coefficient, to form a point cloud model of the real-time fire status in the cabin.

8. A fire warning system for a wind turbine cabin according to claim 7, characterized in that: The submodules for creating the point cloud model of the real-time fire situation in the cabin include: A color marking unit is used to compare the fire risk assessment coefficient of each segmented area of ​​the wind turbine cabin in each monitoring period with the benchmark fire risk assessment coefficient. If the fire risk assessment coefficient of the segmented area is greater than the benchmark fire risk assessment coefficient, the corresponding segmented area is color marked once; The color elimination unit is used to count the number of color markings on each segmented area. If the number of color markings is zero for x consecutive monitoring cycles, the corresponding segmented area will be color eliminated once. After the total monitoring time is over, a point cloud model of the real-time fire status in the cabin is formed.

9. A fire warning system for a wind turbine cabin according to claim 1, characterized in that: The fire warning module also includes a fire spread estimation submodule, which includes: A fire risk assessment coefficient prediction unit is used to construct a fire risk assessment coefficient prediction matrix based on the fire risk assessment coefficient of each segmented area of ​​the wind turbine generator cabin in each monitoring period, and predict the fire risk assessment coefficient of the corresponding segmented area of ​​the wind turbine generator cabin in the next monitoring period; The fire spread coefficient calculation unit is used to calculate the quotient of the difference between the fire risk assessment coefficient of the divided area corresponding to the wind turbine generator cabin in the next monitoring period and the fire risk assessment coefficient of the divided area corresponding to the wind turbine generator cabin in the current monitoring period and the preset fire risk assessment coefficient difference, so as to obtain the fire spread coefficient of the current divided area; The fire isolation assessment unit, if the fire spread coefficient of the current segmented area is a negative number, there is no need to isolate the fire; if the fire spread coefficient of the current segmented area is a positive number, fire isolation is required, and the difference between the product of the fire spread coefficient of the current segmented area and the area of ​​the current segmented area and the area of ​​the current segmented area is calculated to obtain the area of ​​the area that needs to be isolated from the fire, and the area data of the area that needs to be isolated from the fire is sent to the control center, which determines the range of fire isolation based on the area of ​​the area that needs to be isolated from the fire.

10. A fire warning system for a wind turbine cabin according to claim 9, characterized in that: Fire risk assessment coefficient prediction matrix: ;in, is the fire risk assessment coefficient prediction matrix of the j-th segmentation area of ​​the wind turbine cabin in the i-th monitoring period, is the fire risk assessment coefficient of the jth segmented area in the first monitoring cycle of the wind turbine cabin, is the fire risk assessment coefficient of the jth segmented area in the second monitoring period of the wind turbine cabin, is the fire risk assessment coefficient of the j-th segmented area in the third monitoring cycle of the wind turbine cabin, is the fire risk assessment coefficient of the jth segmented area in the fourth monitoring cycle of the wind turbine cabin, For wind turbine cabin The fire risk assessment coefficient of the jth segmented area in the monitoring period is: For wind turbine cabin Fire risk assessment coefficient of the jth segmented area in a monitoring period; The fire risk assessment coefficient of the jth segmentation area in the wind turbine cabin in the i+1th monitoring cycle: Obtain the mean of all matrix elements in any row except the first row in the fire risk assessment coefficient prediction matrix of the j-th segmented area of ​​the wind turbine cabin in the ith monitoring period, and set the value of the last column in the fire risk assessment coefficient prediction matrix of the j-th segmented area of ​​the wind turbine cabin in the ith monitoring period to a value equal to the mean, thereby obtaining a new matrix, and taking the rank of the new matrix as the fire risk assessment coefficient of the j-th segmented area of ​​the wind turbine cabin in the i+1-th monitoring period.

Images