Intelligent fire extinguishing system for wind turbine generator

Abstract

The invention provides an intelligent fire-fighting system for a wind turbine generator, and belongs to the technical field of fire-fighting control, and the system comprises a device module which carries out the arrangement of a fire extinguishing device for the wind turbine generator based on the equipment function, environment data and historical fire data of the wind turbine generator; the monitoring module is used for monitoring the interior of the wind turbine generator in real time and determining a real-time video stream and real-time temperature data of the wind turbine generator; the analysis module is used for generating a real-time fire report; according to the strategy module, a fire extinguishing system formulates a fire early warning strategy or a fire extinguishing control strategy; according to the control module, the fire extinguishing system optimizes the wind turbine generator based on the fire early warning strategy, and the fire extinguishing system remotely controls the fire extinguishing device based on the fire extinguishing control strategy. The system can flexibly formulate an early warning or fire extinguishing strategy, automatically responds to the fire, improves the customization of a fire extinguishing system, enhances the precision and coverage of fire prevention and control, improves the fire prevention and control efficiency and emergency response speed of the wind turbine generator, and remarkably reduces the fire risk and loss.

Description

Technical Field

[0001] The present invention relates to the technical field of fire control, and in particular to an intelligent fire protection system for a wind turbine generator set. Background Art

[0002] In the fire safety management of wind turbines, traditional fire protection systems mainly rely on single detection methods such as smoke detectors and temperature sensors. However, wind turbines are located at high altitudes and in flammable environments, and the response speed and accuracy of traditional methods are often insufficient. In addition, wind turbines have complex structures, and conventional equipment is difficult to fully monitor all internal areas. There are problems such as monitoring blind spots, false alarms or missed alarms, and there is a lack of predictions for different fire types and targeted firefighting strategies, which affects the effectiveness of emergency response.

[0003] Therefore, the present invention provides an intelligent fire fighting system for a wind turbine generator set. Summary of the invention

[0004] The present invention provides an intelligent fire fighting system for a wind turbine generator set. The fire fighting device of the wind turbine generator set is arranged through the equipment function, environmental data and historical fire data of the wind turbine generator set, and the real-time video stream and real-time temperature data of real-time monitoring are analyzed to generate a real-time fire report. According to the real-time fire report, the equipment characteristics of the wind turbine generator set and the fire fighting device, a fire warning strategy or a fire fighting control strategy is formulated and optimized and controlled. The warning or fire fighting strategy can be flexibly formulated, the fire situation is automatically responded to, and the efficient prevention of fire is realized. The customization of the fire fighting system is improved, the accuracy and coverage of fire prevention and control are enhanced, the fire prevention and control efficiency and the emergency response speed of the wind turbine generator set are improved, and the fire risk and loss are significantly reduced.

[0005] The present invention provides an intelligent fire fighting system for a wind turbine generator set, comprising:

[0006] Device module: Determine the predicted fire type of the wind turbine based on the equipment function, environmental data and historical fire data of the wind turbine, and arrange the fire extinguishing device for the wind turbine based on the predicted fire type;

[0007] Monitoring module: Real-time monitoring of the interior of the wind turbine, determining the real-time video stream and real-time temperature data of the wind turbine;

[0008] Analysis module: The fire protection system receives and analyzes real-time video streams and real-time temperature data to generate real-time fire reports;

[0009] Strategy module: The fire protection system formulates fire warning strategies or fire extinguishing control strategies based on real-time fire reports, wind turbine equipment characteristics and fire extinguishing devices;

[0010] Control module: The fire protection system optimizes the wind turbines based on the fire warning strategy, and remotely controls the fire extinguishing devices based on the fire extinguishing control strategy.

[0011] According to an intelligent fire fighting system for a wind turbine generator system provided by the present invention, the environmental data includes operating environment data of all equipment and climate environment data of all external areas.

[0012] According to an intelligent fire fighting system for a wind turbine generator system provided by the present invention, the device module comprises:

[0013] The first feature vector unit: extracts features from the equipment function, operating environment data and corresponding historical fire data of each equipment in the wind turbine generator set, and determines the first feature vector of each equipment;

[0014] The second feature vector unit is used to extract features from the climate environment data of each external area of ​​the wind turbine and the corresponding historical fire data, and determine the second feature vector of each external area;

[0015] First area unit: based on the first characteristic vectors of all devices and the second characteristic vectors of all external areas, the wind turbine group is divided into regions to determine a first area, wherein the first area includes a plurality of first sub-areas, and the first sub-area includes a plurality of devices and a plurality of external areas;

[0016] A fitting unit: determining a fitting first feature vector of each first sub-region based on first feature vectors of all devices in each first sub-region in the first region, and determining a fitting second feature vector of each first sub-region based on second feature vectors of all external regions of each first sub-region in the first region;

[0017] A fire type prediction unit: determining a predicted fire subtype of each first subregion based on a fitted first eigenvector and a fitted second eigenvector of each first subregion in the first region, and determining a predicted fire type based on the predicted fire subtypes of all first subregions;

[0018] Fire extinguishing device unit: arranges sub-fire extinguishing devices for each first sub-area based on the predicted fire sub-type of each first sub-area, and determines the fire extinguishing device for the first area based on the arranged sub-fire extinguishing devices of all the first sub-areas.

[0019] According to an intelligent fire fighting system for a wind turbine generator system provided by the present invention, the monitoring module comprises:

[0020] Real-time video stream unit: based on the visual camera group, real-time monitoring of the first area is performed, obtaining a real-time sub-video stream of each first sub-area, and determining a real-time video stream based on the real-time sub-video streams of all the first sub-areas;

[0021] The real-time temperature data unit monitors the first area in real time based on the infrared camera group, obtains the real-time sub-temperature data of each first sub-area, and determines the real-time temperature data based on the real-time sub-temperature data of all the first sub-areas.

[0022] According to an intelligent fire fighting system for a wind turbine generator system provided by the present invention, the analysis module comprises:

[0023] Image frame unit: extract continuous image frames from each real-time sub-video stream in the real-time video stream, pre-process all image frames extracted from each real-time sub-video stream, and annotate each pre-processed image frame with a sequence label;

[0024] Image sub-data unit: based on all pre-processed image frames in the real-time sub-video stream and the sequence label of each image frame, determine the image sub-data of the first sub-region corresponding to the real-time sub-video stream;

[0025] Smoke image unit: convert each image frame in the image sub-data of each first sub-region into a grayscale image, extract texture features and judge grayscale values ​​of all grayscale images in the first sub-region, and determine that a grayscale image whose texture feature value is less than a texture threshold and whose grayscale value of any pixel is less than the grayscale threshold is a smoke image;

[0026] Smoke data unit: if there is a smoke image, determine the smoke sub-data of each first sub-region containing the smoke image based on all the smoke images containing each first sub-region of the smoke image, determine the real-time smoke data of the wind turbine generator set based on the smoke sub-data of all the first sub-regions containing the smoke image, and if there is no smoke image, determine that the real-time smoke data of the wind turbine generator set is empty;

[0027] Abnormal temperature data unit: analyzes the real-time temperature data, and if there is abnormal temperature, determines the real-time abnormal temperature data of the wind turbine generator set; if there is no abnormal temperature, determines that the real-time abnormal temperature data of the wind turbine generator set is empty;

[0028] Flame data unit: if the real-time smoke data is not empty or the real-time abnormal temperature data is not empty, determine the flame sub-data of each first sub-area based on the image sub-data of each first sub-area, and determine the real-time flame data of the wind turbine generator set; otherwise, determine that the real-time flame data of the wind turbine generator set is empty;

[0029] Real-time fire reporting unit: Generates real-time fire reports based on real-time flame data, real-time smoke data, and real-time abnormal temperature data.

[0030] According to an intelligent fire fighting system for a wind turbine generator system provided by the present invention, the flame data unit comprises:

[0031] Color image subunit: converting each image frame in the image sub-data of each first sub-region into an image in the HSV color space, and determining a color image of each image frame;

[0032] Flame image subunit: judging the set hue threshold range, the set saturation threshold range and the set brightness threshold range of each pixel point of all color images in each first sub-region, determining that a pixel point whose hue value, saturation value or brightness value of any pixel point in the color image is not within the corresponding threshold range is a flame pixel point, and determining that a color image containing one or more flame pixels in the color image is a flame image;

[0033] The flame sub-data sub-unit: based on the sequence labels of the image frames corresponding to the color images corresponding to all the flame images in each first sub-region containing the flame image, first sorting all the flame images in each first sub-region containing the flame image, and determining the flame sub-data of each first sub-region containing the flame image based on all the flame images after the first sorting;

[0034] The flame spread value subunit: determines the flame spread value of each first sub-region including the flame image based on all flame images in the flame sub-data of each first sub-region including the flame image;

[0035] The flame data subunit determines the real-time flame data based on the flame sub-data of all the first sub-areas including the flame image and the flame spread value.

[0036] According to an intelligent fire fighting system for a wind turbine generator system provided by the present invention, the flame spread value subunit comprises:

[0037]

[0038] Among them, FD i represents the flame spread value of the first sub-area of ​​the ith region, S ij represents the flame area of ​​the jth flame image of the i-th first sub-region, S ij-1 represents the flame area of ​​the j-1th flame image of the i-th first sub-region, Bc j represents the flame change value of all pixels of the jth flame image in the i-th first sub-region, UC j represents the centroid change value of the jth flame image in the i-th first sub-region, O j , O j-1 denote the sequence labels of the image frames corresponding to the j-th flame image and the j-1-th flame image of the i-th first sub-region, respectively; w1, w2, and w3 denote the first weight, the second weight, and the third weight of the flame diffusion value, respectively; X j , X j-1represents the centroid hue value of the flame area of ​​the jth flame image and the j-1th flame image of the i-th first sub-area, respectively, j , Y j-1 Respectively represent the centroid saturation value of the flame area of ​​the jth flame image and the j-1th flame image of the i-th first sub-area, Z j , Z j-1 represents the centroid brightness value of the flame area of ​​the jth flame image and the j-1th flame image of the i-th first sub-area, respectively; w4, w5, and w6 represent the fourth weight, fifth weight, and sixth weight of the centroid change value, respectively; B ij (x k1 ,y k2 ,z k3 ) represents the value of the pixel point of the k1th hue value, k2th saturation value and k3th brightness value of the jth flame image in the i-th first sub-region in the corresponding binary image, B ij-1 (x k1 ,y k2 ,z k3 ) represents the values ​​of the pixel point of the k1th hue value, k2th saturation value and k3th brightness value of the j-1th flame image of the i-th first sub-region in the corresponding binary image, I j (x k1 ,y k2 ,z k3 ) represents the pixel point with the k1th hue value, k2th saturation value and k3th brightness value of the jth flame image in the i-th first sub-region, I j-1 (x k1 ,y k2 ,z k3 ) represents the k1th hue value, k2th saturation value and k3th brightness value of the pixel of the j-1th flame image of the ith first sub-region, T represents the pixel difference threshold; 1≤k1≤iN1, 1≤k2≤iN2, 1≤k3≤iN3, kN1, kN2, kN3 represent the number of hue values, the number of saturation values ​​and the number of brightness values ​​of the ith first sub-region, respectively.

[0039] According to an intelligent fire fighting system for a wind turbine generator system provided by the present invention, the strategy module comprises:

[0040] Fire extinguishing control strategy unit: if the real-time flame data in the real-time fire report is not empty, the fire protection system determines the fire extinguishing agent based on the real-time fire report and the characteristics of the wind turbine equipment, and the fire protection system formulates a fire extinguishing control strategy based on the real-time fire report, the fire extinguishing agent and the fire extinguishing device;

[0041] Fire warning strategy unit: If the real-time flame data in the real-time fire report is empty, the fire protection system formulates a fire warning strategy based on the real-time fire report and the characteristics of the wind turbine equipment.

[0042] Compared with the prior art, the present invention has the following beneficial effects:

[0043] Fire extinguishing devices are arranged on wind turbines based on the equipment functions, environmental data and historical fire data of wind turbines. Real-time video streams and real-time temperature data of real-time monitoring are analyzed to generate real-time fire reports. Fire warning strategies or fire extinguishing control strategies are formulated and optimized based on real-time fire reports, wind turbine equipment characteristics and fire extinguishing devices. Early warning or fire extinguishing strategies can be flexibly formulated, fire conditions can be automatically responded to, and efficient fire prevention can be achieved. The customization of fire protection systems is improved, the accuracy and coverage of fire prevention and control are enhanced, the fire prevention and control efficiency and emergency response speed of wind turbines are improved, and fire risks and losses are significantly reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0045] Figure 1 The invention is a schematic structural diagram of an intelligent fire fighting system for a wind turbine generator system provided in an embodiment of the invention. DETAILED DESCRIPTION

[0046] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0047] Embodiment 1:

[0048] The embodiment of the present invention provides an intelligent fire fighting system for a wind turbine generator system. Figure 1 As shown, including:

[0049] Device module: Determine the predicted fire type of the wind turbine based on the equipment function, environmental data and historical fire data of the wind turbine, and arrange the fire extinguishing device for the wind turbine based on the predicted fire type;

[0050] Monitoring module: Real-time monitoring of the interior of the wind turbine, determining the real-time video stream and real-time temperature data of the wind turbine;

[0051] Analysis module: The fire protection system receives and analyzes real-time video streams and real-time temperature data to generate real-time fire reports;

[0052] Strategy module: The fire protection system formulates fire warning strategies or fire extinguishing control strategies based on real-time fire reports, wind turbine equipment characteristics and fire extinguishing devices;

[0053] Control module: The fire protection system optimizes the wind turbines based on the fire warning strategy, and remotely controls the fire extinguishing devices based on the fire extinguishing control strategy.

[0054] In this embodiment, the wind turbine is a device that uses wind energy to generate electricity, and is mainly composed of an impeller, a generator, a tower and other parts.

[0055] In this embodiment, by analyzing the equipment functions of the wind turbine (such as the operating characteristics of equipment such as the turbine, gearbox, generator, etc.), environmental data (such as external conditions such as temperature, humidity, wind speed, etc.), and historical fire data (types, causes, and patterns of fires that have occurred in the past), the system can predict potential fire types, such as electrical fires, mechanical friction fires, etc.

[0056] In this embodiment, fire extinguishing devices are arranged inside the wind turbine according to the predicted fire type, such as configuring carbon dioxide fire extinguishers near high-temperature components and arranging dry powder fire extinguishers around electrical equipment, so as to quickly respond to different fire types.

[0057] In this embodiment, a response plan is formulated based on the real-time fire report generated by the analysis module, combined with the equipment characteristics of the wind turbine and the deployed fire extinguishing devices. The module can generate a fire warning strategy (when the flame data in the report is empty) or a fire extinguishing control strategy (when a fire is detected).

[0058] In this embodiment, the early warning strategy may include strengthening monitoring, issuing alarm signals, issuing early warning information to maintenance personnel, etc.; the fire extinguishing control strategy may include a specific fire extinguishing equipment use plan and its execution method.

[0059] In this embodiment, the formulated strategies are executed. For the fire warning strategy, the control module will perform some optimization measures on the wind turbine, such as adjusting the equipment operation mode, starting an additional cooling system, etc.; for the fire extinguishing control strategy, the control module will start the fire extinguishing equipment through the remote control device to perform fire extinguishing actions, such as releasing carbon dioxide or dry powder fire extinguishing agent, shutting down specific equipment to prevent the spread of fire, etc.

[0060] The beneficial effects of the above technical solution are as follows: the fire extinguishing devices of the wind turbines are arranged according to the equipment functions, environmental data and historical fire data of the wind turbines, and real-time fire reports are generated by analyzing the real-time video streams and real-time temperature data of real-time monitoring. According to the real-time fire reports, the characteristics of the wind turbine equipment and the fire extinguishing devices, fire warning strategies or fire extinguishing control strategies are formulated and optimized and controlled. Early warning or fire extinguishing strategies can be flexibly formulated, and fire conditions can be automatically responded to to achieve efficient fire prevention. The customization of the fire protection system is improved, the accuracy and coverage of fire prevention and control are enhanced, the fire prevention and control efficiency and emergency response speed of the wind turbines are improved, and the fire risks and losses are significantly reduced.

[0061] Embodiment 2:

[0062] An embodiment of the present invention provides an intelligent fire fighting system for a wind turbine generator set, wherein the environmental data includes operating environment data of all equipment and climate environment data of all external areas.

[0063] In this embodiment, the environmental data includes the operating environment data of all devices and the climate environment data of all external areas.

[0064] In this embodiment, the equipment operating environment data records the operating conditions of each equipment inside the wind turbine, such as factors such as temperature, humidity and airflow, which may affect the working state and fire risk of the equipment.

[0065] In this embodiment, the climate environment data refers to the meteorological condition data of the area outside the wind turbine generator set, such as temperature, humidity, wind speed, rainfall, etc. These data affect the safety of the wind turbine equipment and the possibility of fire.

[0066] The beneficial effects of the above technical solution are: by determining the environmental data, the data accuracy can be guaranteed, and the data basis for generating real-time fire reports can be improved.

[0067] Embodiment 3:

[0068] An embodiment of the present invention provides an intelligent fire fighting system for a wind turbine generator set, wherein the device module comprises:

[0069] The first feature vector unit: extracts features from the equipment function, operating environment data and corresponding historical fire data of each equipment in the wind turbine generator set, and determines the first feature vector of each equipment;

[0070] The second feature vector unit is used to extract features from the climate environment data of each external area of ​​the wind turbine and the corresponding historical fire data, and determine the second feature vector of each external area;

[0071] First area unit: based on the first characteristic vectors of all devices and the second characteristic vectors of all external areas, the wind turbine group is divided into regions to determine a first area, wherein the first area includes a plurality of first sub-areas, and the first sub-area includes a plurality of devices and a plurality of external areas;

[0072] A fitting unit: determining a fitting first feature vector of each first sub-region based on first feature vectors of all devices in each first sub-region in the first region, and determining a fitting second feature vector of each first sub-region based on second feature vectors of all external regions of each first sub-region in the first region;

[0073] A fire type prediction unit: determining a predicted fire subtype of each first subregion based on a fitted first eigenvector and a fitted second eigenvector of each first subregion in the first region, and determining a predicted fire type based on the predicted fire subtypes of all first subregions;

[0074] Fire extinguishing device unit: arranges sub-fire extinguishing devices for each first sub-area based on the predicted fire sub-type of each first sub-area, and determines the fire extinguishing device for the first area based on the arranged sub-fire extinguishing devices of all the first sub-areas.

[0075] In this embodiment, the function, operating environment and historical fire data of each device in the wind turbine are analyzed, the main features that can affect the occurrence of fire are extracted, and a corresponding feature vector is generated for each device.

[0076] In this embodiment, the climate environment and historical fire data of each area outside the wind turbine are analyzed, the characteristics of external factors related to fire are extracted, and a feature vector is generated for each external area.

[0077] In this embodiment, the first sub-area may be a connected whole area or multiple unconnected sub-areas, that is, all devices included in the first sub-area and all external areas may be in one whole area or multiple unconnected sub-areas.

[0078] In this embodiment, the first feature vectors of all devices included in the same first sub-area are similar, and the second feature vectors of all external areas are similar.

[0079] In this embodiment, feature fitting is performed on the equipment in each first sub-area and the external area to generate a "fitted first feature vector" and a "fitted second feature vector" for more accurately describing the fire features of the sub-area.

[0080] The fitted first eigenvector and the fitted second eigenvector are respectively vectors generated by comprehensively calculating the first eigenvectors of all devices in the corresponding first sub-area and the second eigenvectors of all external environments, and are used to describe the characteristics of the entire area.

[0081] In this embodiment, the subtype of the fire is predicted based on the fitted feature vector of each first sub-region, and then the predicted subtypes of all sub-regions are integrated to determine the overall fire type of the first region.

[0082] In this embodiment, customized fire extinguishing device layout is performed according to the predicted fire subtype of each sub-area, and finally the fire extinguishing devices of all sub-areas are combined to form an overall fire extinguishing system for the first area.

[0083] The beneficial effects of the above technical solution are as follows: the predicted fire type of the wind turbine is determined according to the equipment function, environmental data and historical fire data of the wind turbine, and the fire extinguishing device is arranged for the wind turbine according to the predicted fire type. The fire type of each sub-area can be accurately predicted, and personalized fire extinguishing device arrangement is achieved for each sub-area, thereby improving the customization and efficiency of the fire protection system, enhancing the accuracy and coverage of fire prevention and control, and optimizing the overall fire response capability of the wind turbine.

[0084] Embodiment 4:

[0085] An embodiment of the present invention provides an intelligent fire protection system for a wind turbine generator set, wherein the monitoring module comprises:

[0086] Real-time video stream unit: based on the visual camera group, real-time monitoring of the first area is performed, obtaining a real-time sub-video stream of each first sub-area, and determining a real-time video stream based on the real-time sub-video streams of all the first sub-areas;

[0087] The real-time temperature data unit monitors the first area in real time based on the infrared camera group, obtains the real-time sub-temperature data of each first sub-area, and determines the real-time temperature data based on the real-time sub-temperature data of all the first sub-areas.

[0088] In this embodiment, a visual camera device is used to perform real-time monitoring of the first area of ​​the wind turbine generator set. Each first sub-area has an independent real-time video stream. The images captured by the visual camera device are finally integrated with the video streams of each sub-area to form a real-time video stream covering the entire first area, ensuring that the monitoring image is comprehensive and has no blind spots.

[0089] In this embodiment, an infrared camera is used to monitor the temperature of the first area. The infrared camera can detect temperature changes and generate independent temperature data in each first sub-area. These data can help the system identify high temperature areas or potential fire risk points. Finally, the real-time temperature data of each sub-area are integrated to obtain the overall temperature data of the first area.

[0090] The beneficial effects of the above technical solution are: real-time monitoring of the interior of the wind turbine, determining the real-time video stream and real-time temperature data of the wind turbine, which can improve the data basis for generating real-time fire reports, so as to achieve precise fire warning and positioning, and improve the intelligent level of wind turbine fire protection.

[0091] Embodiment 5:

[0092] An embodiment of the present invention provides an intelligent fire protection system for a wind turbine generator set, wherein the analysis module comprises:

[0093] Image frame unit: extract continuous image frames from each real-time sub-video stream in the real-time video stream, pre-process all image frames extracted from each real-time sub-video stream, and annotate each pre-processed image frame with a sequence label;

[0094] Image sub-data unit: based on all pre-processed image frames in the real-time sub-video stream and the sequence label of each image frame, determine the image sub-data of the first sub-region corresponding to the real-time sub-video stream;

[0095] Smoke image unit: convert each image frame in the image sub-data of each first sub-region into a grayscale image, extract texture features and judge grayscale values ​​of all grayscale images in the first sub-region, and determine that a grayscale image whose texture feature value is less than a texture threshold and whose grayscale value of any pixel is less than the grayscale threshold is a smoke image;

[0096] Smoke data unit: if there is a smoke image, determine the smoke sub-data of each first sub-region containing the smoke image based on all the smoke images containing each first sub-region of the smoke image, determine the real-time smoke data of the wind turbine generator set based on the smoke sub-data of all the first sub-regions containing the smoke image, and if there is no smoke image, determine that the real-time smoke data of the wind turbine generator set is empty;

[0097] Abnormal temperature data unit: analyzes the real-time temperature data, and if there is abnormal temperature, determines the real-time abnormal temperature data of the wind turbine generator set; if there is no abnormal temperature, determines that the real-time abnormal temperature data of the wind turbine generator set is empty;

[0098] Flame data unit: if the real-time smoke data is not empty or the real-time abnormal temperature data is not empty, determine the flame sub-data of each first sub-area based on the image sub-data of each first sub-area, and determine the real-time flame data of the wind turbine generator set; otherwise, determine that the real-time flame data of the wind turbine generator set is empty;

[0099] Real-time fire reporting unit: Generates real-time fire reports based on real-time flame data, real-time smoke data, and real-time abnormal temperature data.

[0100] In this embodiment, continuous image frames (ie, each static frame of the video) are extracted from the real-time video stream, and these image frames are pre-processed (such as noise reduction, size adjustment, etc.).

[0101] In this embodiment, the image frame represents a single static image separated from a video and used for analyzing video content, and continuous image frames are combined to form a video stream.

[0102] In this embodiment, the sequence tag is annotation information added to the image frame, and is used to mark the time sequence of the image frame in the video.

[0103] In this embodiment, the grayscale image means that a color image is converted into an image containing only brightness information, and the pixel value represents the brightness level, and the range is usually 0 (black) to 255 (white).

[0104] In this embodiment, the texture feature describes the properties of the image surface structure, such as the degree of roughness, smoothness or pattern.

[0105] In this embodiment, the grayscale value represents the brightness value of each pixel in the grayscale image, and is used to determine the brightness and darkness characteristics of the image.

[0106] In this embodiment, the smoke image represents an image frame containing smoke features determined by grayscale value and texture features.

[0107] In this embodiment, the abnormal temperature refers to a temperature value exceeding the normal operating range, which may be a precursor to equipment overheating or environmental fire.

[0108] The beneficial effects of the above technical solution are as follows: the fire protection system receives and analyzes real-time video streams and real-time temperature data, generates real-time fire reports, realizes accurate detection of smoke and flames, improves the accuracy of fire identification, significantly speeds up early warning, and provides efficient and comprehensive fire monitoring and early warning support for wind turbines.

[0109] Embodiment 6:

[0110] An embodiment of the present invention provides an intelligent fire protection system for a wind turbine generator set, wherein the flame data unit comprises:

[0111] Color image subunit: converting each image frame in the image sub-data of each first sub-region into an image in the HSV color space, and determining a color image of each image frame;

[0112] Flame image subunit: judging the set hue threshold range, the set saturation threshold range and the set brightness threshold range of each pixel point of all color images in each first sub-region, determining that a pixel point whose hue value, saturation value or brightness value of any pixel point in the color image is not within the corresponding threshold range is a flame pixel point, and determining that a color image containing one or more flame pixels in the color image is a flame image;

[0113] The flame sub-data sub-unit: based on the sequence labels of the image frames corresponding to the color images corresponding to all the flame images in each first sub-region containing the flame image, first sorting all the flame images in each first sub-region containing the flame image, and determining the flame sub-data of each first sub-region containing the flame image based on all the flame images after the first sorting;

[0114] The flame spread value subunit: determines the flame spread value of each first sub-region including the flame image based on all flame images in the flame sub-data of each first sub-region including the flame image;

[0115] The flame data subunit determines the real-time flame data based on the flame sub-data of all the first sub-areas including the flame image and the flame spread value.

[0116] In this embodiment, the HSV color space is a color representation method that divides color into three elements: hue, saturation, and value. It is suitable for flame detection because the color characteristics of flames are more obvious.

[0117] In this embodiment, the color image refers to an image obtained by converting the image frame into the HSV color space, which facilitates separation and detection of color features.

[0118] In this embodiment, the flame pixel point is a pixel point having flame characteristics determined in the color image by hue, saturation and brightness.

[0119] In this embodiment, the hue threshold, the saturation threshold, and the brightness threshold may be determined based on all images of the first sub-region converted into the HSV color space, or may be determined based on historical fire data, etc.

[0120] In this embodiment, the flame image represents a color image in which at least one flame pixel is detected in the image, indicating that the image may contain flames.

[0121] In this embodiment, the flame sub-data represents data formed by sorting out the flame image in the first sub-region and its time sequence.

[0122] In this embodiment, the real-time flame data is real-time information formed by integrating the flame sub-data of all sub-areas and the flame spread value.

[0123] The beneficial effects of the above technical solution are: determining the real-time flame data of the wind turbine set, and can determine high-precision flame data, providing strong support for fire warning and control, and effectively improving the response speed and accuracy of wind turbine set fire monitoring.

[0124] Embodiment 7:

[0125] An embodiment of the present invention provides an intelligent fire protection system for a wind turbine generator set, wherein the flame spread value subunit comprises:

[0126]

[0127] Among them, FD i represents the flame spread value of the first sub-area of ​​the ith region, S ij represents the flame area of ​​the jth flame image of the i-th first sub-region, S ij-1 represents the flame area of ​​the j-1th flame image of the i-th first sub-region, Bc j represents the flame change value of all pixels of the jth flame image in the i-th first sub-region, UC j represents the centroid change value of the jth flame image in the i-th first sub-region, O j , O j-1 denote the sequence labels of the image frames corresponding to the j-th flame image and the j-1-th flame image of the i-th first sub-region, respectively; w1, w2, and w3 denote the first weight, the second weight, and the third weight of the flame diffusion value, respectively; X j , X j-1 represents the centroid hue value of the flame area of ​​the jth flame image and the j-1th flame image of the i-th first sub-area, respectively, and Y j , Y j-1 Respectively represent the centroid saturation value of the flame area of ​​the jth flame image and the j-1th flame image of the i-th first sub-area, Z j , Z j-1 represents the centroid brightness value of the flame area of ​​the jth flame image and the j-1th flame image of the i-th first sub-area, respectively; w4, w5, and w6 represent the fourth weight, fifth weight, and sixth weight of the centroid change value, respectively; B ij (x k1 ,y k2 ,z k3 ) represents the value of the pixel point of the k1th hue value, k2th saturation value and k3th brightness value of the jth flame image in the i-th first sub-region in the corresponding binary image, B ij-1 (x k1 ,y k2 ,z k3) represents the values ​​of the pixel point of the k1th hue value, k2th saturation value and k3th brightness value of the j-1th flame image of the i-th first sub-region in the corresponding binary image, I j (x k1 ,y k2 ,z k3 ) represents the pixel point with the k1th hue value, k2th saturation value and k3th brightness value of the jth flame image in the i-th first sub-region, I j-1 (x k1 ,y k2 ,z k3 ) represents the k1th hue value, k2th saturation value and k3th brightness value of the pixel of the j-1th flame image of the ith first sub-region, T represents the pixel difference threshold; 1≤k1≤iN1, 1≤k2≤iN2, 1≤k3≤iN3, kN1, kN2, kN3 represent the number of hue values, the number of saturation values ​​and the number of brightness values ​​of the ith first sub-region, respectively.

[0128] In this embodiment, the flame spread value reflects the degree of flame spread in the corresponding first sub-region, and describes the development speed and spreading trend of the flame.

[0129] In this embodiment, The relative change of the flame area of ​​the j-th flame image of the i-th first sub-region relative to the flame area of ​​the j-1-th flame image is used to determine whether the flame is spreading. The greater the area change, the higher the possibility of flame expansion.

[0130] In this embodiment, the flame change value Bc of all pixels of the jth flame image in the i-th first sub-area is j Capture the details of the flame spread of the j-th flame image. The greater the boundary change, the more intense the flame activity.

[0131] In this embodiment, the centroid change value UC of the jth flame image of the i-th first sub-region j Measures the fluctuation or expansion of the flame's center of mass in terms of hue, saturation, and brightness. If the center of mass has a large displacement in a certain direction, it means that the flame may be expanding or jumping.

[0132] In this embodiment, The flame expansion speed is estimated by the area change rate of the flame region and the sequence label. If the flame expands rapidly in a short period of time, it means that the flame is very active and the degree of jumping or diffusion is high.

[0133] The beneficial effects of the above technical solution are as follows: based on all flame images in the flame sub-data of each first sub-area containing the flame image, the flame spread value of each first sub-area containing the flame image is determined, the flame spread trend is quantified, and the accuracy of wind turbine fire monitoring is improved.

[0134] Embodiment 8:

[0135] An embodiment of the present invention provides an intelligent fire protection system for a wind turbine generator set, wherein the strategy module comprises:

[0136] Fire extinguishing control strategy unit: if the real-time flame data in the real-time fire report is not empty, the fire protection system determines the fire extinguishing agent based on the real-time fire report and the characteristics of the wind turbine equipment, and the fire protection system formulates a fire extinguishing control strategy based on the real-time fire report, the fire extinguishing agent and the fire extinguishing device;

[0137] Fire warning strategy unit: If the real-time flame data in the real-time fire report is empty, the fire protection system formulates a fire warning strategy based on the real-time fire report and the characteristics of the wind turbine equipment.

[0138] In this embodiment, the core function of the real-time fire report is to provide detailed information of the current fire, including data such as flame distribution, temperature, and smoke.

[0139] In this embodiment, when the report shows that the real-time flame data is not empty (i.e., the flame does exist), the system determines that fire extinguishing measures need to be taken. At this time, the fire protection system will select the most appropriate fire extinguishing agent based on the equipment characteristics of the wind turbine (such as equipment structure, material properties, possible fire sources, etc.). For example, if the fire source may involve electrical equipment, the fire protection system may select carbon dioxide fire extinguishing agent to prevent conductivity.

[0140] In this embodiment, the fire protection system integrates real-time fire reports, selected fire extinguishing agents and fire extinguishing devices installed on wind turbines to formulate a set of fire extinguishing control strategies. This strategy includes the activation of fire extinguishing devices, the release amount of fire extinguishing agents and the release location and time, etc., to ensure effective and safe control of the fire.

[0141] In this embodiment, if the real-time flame data in the real-time fire report is empty (i.e. no flame is detected), the fire protection system believes that there is no clear fire phenomenon at present, but there may still be a fire risk. In this case, the fire protection system will formulate a fire warning strategy based on other risk indicators in the real-time fire report (such as abnormal temperature, smoke distribution) and the specific situation of the wind turbine equipment. The strategy may include strengthening the monitoring frequency, early warning prompts, and inspection guidance of staff so that timely handling can be carried out when a fire may occur to prevent the occurrence or spread of a fire.

[0142] In this embodiment, the fire extinguishing agent is a substance used to put out the fire source. The characteristics of the wind turbine equipment are taken into consideration when selecting the fire extinguishing agent. Common fire extinguishing agents include water, foam, dry powder, carbon dioxide, etc.

[0143] The beneficial effects of the above technical solutions are as follows: the fire protection system optimizes the wind turbines based on the fire warning strategy, and the fire protection system remotely controls the fire extinguishing devices based on the fire extinguishing control strategy. It can flexibly formulate warning or fire extinguishing strategies, automatically respond to fire conditions, and achieve efficient fire prevention. It improves the customization of the fire protection system, enhances the accuracy and coverage of fire prevention and control, improves the fire prevention and control efficiency and emergency response speed of wind turbines, and significantly reduces fire risks and losses.

[0144] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0145] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0146] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An intelligent fire fighting system for a wind turbine generator system, characterized in that: include: Device module: Determine the predicted fire type of the wind turbine based on the equipment function, environmental data and historical fire data of the wind turbine, and arrange the fire extinguishing device for the wind turbine based on the predicted fire type; Monitoring module: Real-time monitoring of the interior of the wind turbine, determining the real-time video stream and real-time temperature data of the wind turbine; Analysis module: The fire protection system receives and analyzes real-time video streams and real-time temperature data to generate real-time fire reports; Strategy module: The fire protection system formulates fire warning strategies or fire extinguishing control strategies based on real-time fire reports, wind turbine equipment characteristics and fire extinguishing devices; Control module: The fire protection system optimizes the wind turbines based on the fire warning strategy, and remotely controls the fire extinguishing devices based on the fire extinguishing control strategy.

2. The intelligent fire fighting system for wind turbines according to claim 1, characterized in that: The environmental data includes the operating environment data of all devices and the climate environment data of all external areas.

3. The intelligent fire fighting system for wind turbines according to claim 2, characterized in that: The device module comprises: The first feature vector unit: extracts features from the equipment function, operating environment data and corresponding historical fire data of each equipment in the wind turbine generator set, and determines the first feature vector of each equipment; The second feature vector unit is used to extract features from the climate environment data of each external area of ​​the wind turbine and the corresponding historical fire data, and determine the second feature vector of each external area; First area unit: based on the first characteristic vectors of all devices and the second characteristic vectors of all external areas, the wind turbine group is divided into regions to determine a first area, wherein the first area includes a plurality of first sub-areas, and the first sub-area includes a plurality of devices and a plurality of external areas; A fitting unit: determining a fitting first feature vector of each first sub-region based on first feature vectors of all devices in each first sub-region in the first region, and determining a fitting second feature vector of each first sub-region based on second feature vectors of all external regions of each first sub-region in the first region; A fire type prediction unit: determining a predicted fire subtype of each first subregion based on a fitted first eigenvector and a fitted second eigenvector of each first subregion in the first region, and determining a predicted fire type based on the predicted fire subtypes of all first subregions; Fire extinguishing device unit: arranges sub-fire extinguishing devices for each first sub-area based on the predicted fire sub-type of each first sub-area, and determines the fire extinguishing device for the first area based on the arranged sub-fire extinguishing devices of all the first sub-areas.

4. The intelligent fire fighting system for wind turbines according to claim 3, characterized in that: The monitoring module comprises: Real-time video stream unit: based on the visual camera group, real-time monitoring of the first area is performed, obtaining a real-time sub-video stream of each first sub-area, and determining a real-time video stream based on the real-time sub-video streams of all the first sub-areas; The real-time temperature data unit monitors the first area in real time based on the infrared camera group, obtains the real-time sub-temperature data of each first sub-area, and determines the real-time temperature data based on the real-time sub-temperature data of all the first sub-areas.

5. The intelligent fire fighting system for wind turbines according to claim 1, characterized in that: The analysis module comprises: Image frame unit: extract continuous image frames from each real-time sub-video stream in the real-time video stream, pre-process all image frames extracted from each real-time sub-video stream, and annotate each pre-processed image frame with a sequence label; Image sub-data unit: based on all pre-processed image frames in the real-time sub-video stream and the sequence label of each image frame, determine the image sub-data of the first sub-region corresponding to the real-time sub-video stream; Smoke image unit: convert each image frame in the image sub-data of each first sub-region into a grayscale image, extract texture features and judge grayscale values ​​of all grayscale images in the first sub-region, and determine that a grayscale image whose texture feature value is less than a texture threshold and whose grayscale value of any pixel is less than the grayscale threshold is a smoke image; Smoke data unit: if there is a smoke image, determine the smoke sub-data of each first sub-region containing the smoke image based on all the smoke images containing each first sub-region of the smoke image, determine the real-time smoke data of the wind turbine generator set based on the smoke sub-data of all the first sub-regions containing the smoke image, and if there is no smoke image, determine that the real-time smoke data of the wind turbine generator set is empty; Abnormal temperature data unit: analyzes the real-time temperature data, and if there is abnormal temperature, determines the real-time abnormal temperature data of the wind turbine generator set; if there is no abnormal temperature, determines that the real-time abnormal temperature data of the wind turbine generator set is empty; Flame data unit: if the real-time smoke data is not empty or the real-time abnormal temperature data is not empty, determine the flame sub-data of each first sub-area based on the image sub-data of each first sub-area, and determine the real-time flame data of the wind turbine generator set; otherwise, determine that the real-time flame data of the wind turbine generator set is empty; Real-time fire reporting unit: Generates real-time fire reports based on real-time flame data, real-time smoke data, and real-time abnormal temperature data.

6. The intelligent fire fighting system for wind turbines according to claim 5, characterized in that: The flame data unit comprises: Color image subunit: converting each image frame in the image sub-data of each first sub-region into an image in the HSV color space, and determining a color image of each image frame; Flame image subunit: judging the set hue threshold range, the set saturation threshold range and the set brightness threshold range of each pixel point of all color images in each first sub-region, determining that a pixel point whose hue value, saturation value or brightness value of any pixel point in the color image is not within the corresponding threshold range is a flame pixel point, and determining that a color image containing one or more flame pixels in the color image is a flame image; The flame sub-data sub-unit: based on the sequence labels of the image frames corresponding to the color images corresponding to all the flame images in each first sub-region containing the flame image, first sorting all the flame images in each first sub-region containing the flame image, and determining the flame sub-data of each first sub-region containing the flame image based on all the flame images after the first sorting; The flame spread value subunit: determines the flame spread value of each first sub-region including the flame image based on all flame images in the flame sub-data of each first sub-region including the flame image; The flame data subunit determines the real-time flame data based on the flame sub-data of all the first sub-areas including the flame image and the flame spread value.

7. The intelligent fire fighting system for wind turbines according to claim 6, characterized in that: The flame spread value subunit comprises: Among them, FD i represents the flame spread value of the first sub-area of ​​the ith region, S ij represents the flame area of ​​the jth flame image of the i-th first sub-region, S ij-1 represents the flame area of ​​the j-1th flame image of the i-th first sub-region, Bc j represents the flame change value of all pixels of the jth flame image in the i-th first sub-region, UC j represents the centroid change value of the jth flame image in the i-th first sub-region, O j , O j-1 denote the sequence labels of the image frames corresponding to the j-th flame image and the j-1-th flame image of the i-th first sub-region, respectively; w1, w2, and w3 denote the first weight, the second weight, and the third weight of the flame diffusion value, respectively; X j , X j-1 represents the centroid hue value of the flame area of ​​the jth flame image and the j-1th flame image of the i-th first sub-area, respectively, and Y j , Y j-1 Respectively represent the centroid saturation value of the flame area of ​​the jth flame image and the j-1th flame image of the i-th first sub-area, Z j , Z j-1 represents the centroid brightness value of the flame area of ​​the jth flame image and the j-1th flame image of the i-th first sub-area, respectively; w4, w5, and w6 represent the fourth weight, fifth weight, and sixth weight of the centroid change value, respectively; B ij (x k1 ,y k2 ,z k3 ) represents the value of the pixel point of the k1th hue value, k2th saturation value and k3th brightness value of the jth flame image in the i-th first sub-region in the corresponding binary image, B ij-1 (x k1 ,y k2 ,z k3 ) represents the values ​​of the pixel point of the k1th hue value, k2th saturation value and k3th brightness value of the j-1th flame image of the i-th first sub-region in the corresponding binary image, I j (x k1 ,y k2 ,z k3 ) represents the pixel point with the k1th hue value, k2th saturation value and k3th brightness value of the jth flame image in the i-th first sub-region, I j-1 (x k1 ,y k2 ,z k3 ) represents the k1th hue value, k2th saturation value and k3th brightness value of the pixel of the j-1th flame image of the ith first sub-region, T represents the pixel difference threshold; 1≤k1≤iN1, 1≤k2≤iN2, 1≤k3≤iN3, kN1, kN2, kN3 represent the number of hue values, the number of saturation values ​​and the number of brightness values ​​of the ith first sub-region, respectively.

8. The intelligent fire fighting system for wind turbines according to claim 1, characterized in that: The policy module comprises: Fire extinguishing control strategy unit: if the real-time flame data in the real-time fire report is not empty, the fire protection system determines the fire extinguishing agent based on the real-time fire report and the characteristics of the wind turbine equipment, and the fire protection system formulates a fire extinguishing control strategy based on the real-time fire report, the fire extinguishing agent and the fire extinguishing device; Fire warning strategy unit: If the real-time flame data in the real-time fire report is empty, the fire protection system formulates a fire warning strategy based on the real-time fire report and the characteristics of the wind turbine equipment.

Images