Urban underground comprehensive pipe gallery electrical fire monitoring and early warning method

Abstract

The invention discloses an urban underground comprehensive pipe gallery electrical fire monitoring and early warning method, which comprises the steps of 1, determining an index to be monitored, and deploying a multi-identification sensing node based on multi-identification wireless communication; the to-be-monitored indexes are acquired through a gas sensor, a temperature and humidity sensor, a voltage acquisition device and a near-infrared binocular camera, and are monitored and identified through a monitoring host. Step 2, based on the current electrical distribution condition, constructing an electrical fire monitoring and identification model; the method comprises the following steps: dividing test areas of a monitored building, setting a plurality of corresponding test nodes in each area, and mutually overlapping and associating the nodes to form a cyclic monitoring structure; and step 3, monitoring abnormal data acquisition and edge side data processing are carried out, and the ignition point is positioned based on binocular vision. According to the method, based on LoRa wireless communication and binocular vision image acquisition, fire hazard monitoring is carried out from multiple angles, the actual monitoring range is expanded, the fire hazard position is rapidly positioned, and the existing potential safety hazard is reduced.

Description

Technical Field

[0001] The invention relates to an electrical fire monitoring and early warning method for a city underground comprehensive pipe gallery, belonging to the technical field of electrical safety. Background Art

[0002] As an important type of underground public building with centralized functions, urban underground utility corridors are widely used all over the world. The underground utility corridors carry professional engineering pipelines for energy, water supply and drainage, communications, etc., and are equipped with maintenance, hoisting, monitoring and other facilities to ensure the normal operation of the city. All kinds of pipelines are uniformly designed and arranged in strict accordance with professional requirements.

[0003] The combustible materials in the city's underground comprehensive pipeline corridor mainly include cables, optical cables, pipelines, combustible gases, etc. Among them, the insulating sheath of my country's cables is mainly made of materials such as polyvinyl chloride and rubber, and the optical cable wrapping is mainly composed of plastic sheath, plastic protective tube sleeve or optical fiber, while the combustible gases mainly refer to methane and biogas. In the power and telecommunications compartment, the main causes of cable fires include short circuits between cables, short circuits to the ground, poor contact or line overload, which cause the temperature of the wires to increase sharply and eventually reach a common ignition point. In the gas compartment, the natural gas pipeline may leak flammable, explosive, toxic and other gases.

[0004] Since the internal structure of the integrated pipeline corridor is small and connected from front to back, the cables in the power compartment are laid in a concentrated manner. Therefore, once a fire occurs, the fire will quickly spread along the direction of the cables to other adjacent compartments; cable burning or flammable gas leakage will produce a large amount of toxic and harmful gases, making fire fighting difficult. After the high-voltage cable is powered off, it will generate instantaneous high pressure or residual pressure, which may cause a fire point hazard when extinguishing the fire in the power compartment. Summary of the invention

[0005] The present invention designs and develops an urban underground comprehensive pipe gallery electrical fire monitoring and early warning method, which is based on LoRa wireless communication and binocular visual image acquisition to monitor fire from multiple angles, expand the actual monitoring range, quickly locate the fire location, and reduce existing safety hazards.

[0006] The technical solution provided by the present invention is:

[0007] A method for monitoring and early warning of electrical fires in urban underground integrated pipe galleries, comprising:

[0008] Step 1: Determine the indicators to be monitored and deploy multi-identification sensing nodes based on multi-identification wireless communication;

[0009] Step 2: Based on the current electrical distribution situation, an electrical fire monitoring and identification model is constructed;

[0010] Step 3: Monitor abnormal data collection and edge-side data processing, and locate the fire point based on binocular vision.

[0011] Preferably, the step 1 comprises:

[0012] The indicators to be monitored are collected by: gas sensors, temperature and humidity sensors, voltage collection devices, and near-infrared binocular cameras, and are monitored and identified by a monitoring host.

[0013] Preferably, the step 2 comprises:

[0014] The monitored building is divided into test areas, and multiple corresponding test nodes are set in each area. The nodes are interconnected to form a cyclic monitoring structure;

[0015] Based on LoRa wireless communication, the LoRa module and base station gateway are constructed into a LoRa two-way monitoring transmission model structure;

[0016] Calculate the detection response time of the LoRa bidirectional monitoring transmission model and set it as the detection limit of the current model. The calculation formula for the response time is:

[0017] F=(ys×τ 2 )×ω;

[0018] Where F is the monitoring response time, y is the recognition range, s is the conversion monitoring mean, τ is the number of alarms, and ω is the detection delay.

[0019] Preferably, the current electrical fire edge monitoring limit is calculated:

[0020]

[0021] Where P is the electrical fire edge monitoring limit, x is the directional correction absolute value, τ is the monitoring cluster, e is the monitoring node, is the transmission range, and θ is the edge blur value.

[0022] Preferably, the step three comprises:

[0023] The monitoring area map is collected by setting up a near-infrared binocular camera;

[0024] Preprocess the image, perform threshold segmentation on the collected image, and obtain a binary image;

[0025] Perform flame recognition on the image, and if there is flame in the image monitoring area, an alarm will be issued;

[0026] If there is no flame in the image monitoring area, a judgment is made based on the measured values ​​of the gas sensor, temperature and humidity sensor, and voltage acquisition device as well as the fire edge monitoring limit;

[0027] Use near-infrared binocular cameras to measure the distance to or from the source of goods and locate the fire source in combination with the camera position;

[0028] Output the fire source location information and control the fire extinguishing device near the fire source to extinguish the fire.

[0029] Preferably, the flame identification process includes:

[0030] Identify the sharp corner features, roundness features and rectangularity features of the image;

[0031] In the binary image, the highest point of the target image is searched as the suspected flame vertex A, where A can be one pixel or multiple pixels M;

[0032] Calculate the target image widths N1, N2 and N3 of the H1th row, H2th row and H3th row from vertex A downward respectively;

[0033] Wherein, 1

[0034] If M<N1<N2<N3 is satisfied, it is determined to be a flame sharp angle;

[0035] When the circularity C≤K, the detection area is determined to be a flame;

[0036]

[0037] In the formula, C is the circularity, S is the area of ​​the monitoring area,

[0038] When the monitoring area is a rectangle, the rectangularity Take the maximum value 1.0. When the detection area is a circle, the rectangular The value is π / 4; when the detection area is flame, the rectangularity is;

[0039] In the formula, S t is the minimum circumscribed rectangle area of ​​the detection area.

[0040] Preferably, flame positioning comprises:

[0041] The left and right views captured by the binocular camera are monitored simultaneously. When a flame is detected, it is marked with a rectangle. The following formula is used to determine whether the flames detected in the left and right views are the same flame:

[0042]

[0043] In the formula, X cl is the horizontal coordinate of the center of the rectangular box in the left view, X cr is the horizontal coordinate of the center of the rectangle in the right view, For point (X cl , Y​cl ), where the disparity is the difference between the horizontal coordinates of the matching points in the left and right views, and a is the set threshold.

[0044] The beneficial effects described in the present invention: The urban underground comprehensive pipeline corridor electrical fire monitoring and early warning method provided by the present invention adopts a multi-target approach to break the limitation of the monitoring range, deploys LoRa wireless communication multi-target identification nodes, and on this basis, constructs a LoRa wireless communication fire monitoring model, adopts edge-assisted computing to realize fire monitoring and processing, and uses binocular vision to perceive and locate the fire source, conducts fire monitoring from multiple angles, expands the actual monitoring range, quickly locates the fire location, and reduces existing safety hazards. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 The present invention provides a flow chart of the urban underground comprehensive pipe gallery electrical fire monitoring and early warning method. DETAILED DESCRIPTION

[0046] The present invention is further described in detail below in conjunction with the accompanying drawings so that those skilled in the art can implement the invention with reference to the description.

[0047] like Figure 1 As shown, the present invention provides an urban underground integrated pipe gallery electrical fire monitoring and early warning method, which is based on LoRa wireless communication and binocular visual image acquisition, performs fire monitoring from multiple angles, expands the actual monitoring range, quickly locates the fire location, and reduces the existing safety hazards, including:

[0048] Step 1: Determine the indicators to be monitored and deploy multi-identification sensing nodes based on multi-identification wireless communication;

[0049] Step 2: Based on the current electrical distribution situation, an electrical fire monitoring and identification model is constructed;

[0050] Step 3: Monitor abnormal data collection and edge-side data processing, and locate the fire point based on binocular vision.

[0051] The electrical fire monitoring environment of the currently selected integrated pipe gallery building is constructed and set up. According to the electrical distribution and application type of the building target, the actual coverage of the test is clarified, and the target is divided into multiple test areas. Each area needs to set up corresponding test nodes, and the nodes are interconnected to form a cyclic detection structure. Each area is equipped with a corresponding electrical fire detection device, isolator, and monitoring host. The detection device includes multiple sensors: gas sensor, temperature and humidity sensor, voltage acquisition device, and near-infrared binocular camera for acquisition.

[0052] Through LoRa wireless communication technology, electrical fires in selected buildings are identified, monitored and processed.

[0053] Based on LoRa wireless communication, the LoRa module and base station gateway are constructed into a LoRa two-way monitoring transmission model structure;

[0054] Calculate the detection response time of the LoRa bidirectional monitoring transmission model and set it as the detection limit of the current model. The calculation formula for the response time is:

[0055] F=(ys×τ 2 )×ω;

[0056] Where F is the monitoring response time, y is the recognition range, s is the conversion monitoring mean, τ is the number of alarms, and ω is the detection delay.

[0057] Calculate the current electrical fire edge detection limit:

[0058]

[0059] Where P is the electrical fire edge monitoring limit, x is the directional correction absolute value, τ is the monitoring cluster, e is the monitoring node, is the transmission range, and θ is the edge blur value.

[0060] The obtained electrical fire edge monitoring limit is used as the standard for auxiliary monitoring. The sensors, detectors and LoRa wireless communication technology are comprehensively set, combined with actual needs, to adjust the coverage of fire monitoring, enhance the flexibility of edge-assisted fire detection calculation, and strengthen the quality effect of the designed detection method.

[0061] The fire is located based on binocular vision. In the present invention, as a preferred embodiment, a near-infrared binocular camera is selected.

[0062] The monitoring area map is collected by setting up a near-infrared binocular camera;

[0063] Preprocess the image, perform threshold segmentation on the collected image, and obtain a binary image;

[0064] Perform flame recognition on the image, and if there is flame in the image monitoring area, an alarm will be issued;

[0065] If there is no flame in the image monitoring area, a judgment is made based on the measured values ​​of the gas sensor, temperature and humidity sensor, and voltage acquisition device as well as the fire edge monitoring limit;

[0066] Use a near-infrared binocular camera to measure the distance to or fire source and locate the fire source in combination with the camera position;

[0067] Output the fire source location information and control the fire extinguishing device near the fire source to extinguish the fire.

[0068] In the present invention, as a preference, sharp corner features, circularity features and rectangularity features that do not require multiple frames of continuous images are selected for flame recognition.

[0069] The flame identification process includes:

[0070] Identify the sharp corner features, roundness features and rectangularity features of the image;

[0071] In the binary image, the highest point of the target image is searched as the suspected flame vertex A, where A can be one pixel or multiple pixels M;

[0072] Calculate the target image widths N1, N2 and N3 of the H1th row, H2th row and H3th row from vertex A downward respectively;

[0073] Wherein, 1

[0074] If M<N1<N2<N3 is satisfied, it is determined to be a flame sharp angle;

[0075] When the circularity C≤K, the detection area is determined to be a flame;

[0076]

[0077] In the formula, C is the circularity, S is the area of ​​the monitoring area,

[0078] When the monitoring area is a rectangle, the rectangularity Take the maximum value 1.0. When the detection area is a circle, the rectangular The value is π / 4; when the detection area is flame, the rectangularity is;

[0079] In the formula, S t is the minimum circumscribed rectangle area of ​​the detection area.

[0080] Flame positioning includes:

[0081] The left and right views captured by the binocular camera are monitored simultaneously. When a flame is detected, it is marked with a rectangle. The following formula is used to determine whether the flames detected in the left and right views are the same flame:

[0082]

[0083] In the formula, X cl is the horizontal coordinate of the center of the rectangular box in the left view, X cr is the horizontal coordinate of the center of the rectangle in the right view, For point (X cl , Y cl ), where the disparity is the difference between the horizontal coordinates of the matching points in the left and right views, and a is the set threshold. ​

[0084] In the present invention, near-infrared binocular cameras are set at multiple points to collect images of the monitoring area, and the images are pre-processed to reduce the influence of red, orange, and yellow on fire identification, and the image brightness information is used; the collected images are threshold segmented to obtain a binary image; flame identification is performed on the image, and when there is a flame in the image monitoring area, a fire alarm signal is issued, and combined with the gas concentration values ​​(including temperature, humidity, smoke, carbon dioxide, carbon monoxide, oxygen) and voltage values ​​detected at multiple points, fusion monitoring is performed to improve the accuracy of the alarm. The fire source is measured by a near-infrared visible light binocular camera, the fire source is located in combination with the camera position, the fire source position information is output, and the fire extinguishing device near the fire source is controlled to extinguish the fire.

[0085] The urban underground comprehensive pipe gallery electrical fire monitoring and early warning method provided by the present invention adopts a multi-target approach to break the limitation of the monitoring range, deploys LoRa wireless communication multi-target identification nodes, and on this basis, constructs a LoRa wireless communication fire monitoring model, adopts edge-assisted computing to realize fire monitoring and processing, and uses binocular vision to perceive and locate the fire source, monitors fire from multiple angles, expands the actual monitoring range, quickly locates the fire location, and reduces existing safety hazards.

[0086] Although the embodiments of the present invention have been disclosed as above, they are not limited to the applications listed in the specification and the implementation modes, and they can be fully applied to various fields suitable for the present invention. For those familiar with the art, additional modifications can be easily implemented. Therefore, without departing from the general concept defined by the claims and the scope of equivalents, the present invention is not limited to the specific details and the illustrations shown and described herein.

Claims

1. A method for monitoring and early warning of electrical fires in urban underground utility corridors, characterized in that: include: Step 1: Determine the indicators to be monitored and deploy multi-identification sensing nodes based on multi-identification wireless communication; Step 2: Based on the current electrical distribution situation, an electrical fire monitoring and identification model is constructed; Step 3: Monitor abnormal data collection and edge-side data processing, and locate the fire point based on binocular vision.

2. The urban underground utility corridor electrical fire monitoring and early warning method according to claim 1 is characterized in that: The step one comprises: The indicators to be monitored are collected by: gas sensors, temperature and humidity sensors, voltage collection devices, and near-infrared binocular cameras, and are monitored and identified by a monitoring host.

3. The urban underground utility corridor electrical fire monitoring and early warning method according to claim 2 is characterized in that: The second step comprises: The monitored building is divided into test areas, and multiple corresponding test nodes are set in each area. The nodes are interconnected to form a cyclic monitoring structure; Based on LoRa wireless communication, the LoRa module and base station gateway are constructed into a LoRa two-way monitoring transmission model structure; Calculate the detection response time of the LoRa two-way monitoring transmission model and set it as the detection limit of the current model. The calculation formula for the response time is: F=(y-s×τ 2 )×ω; Where F is the monitoring response time, y is the recognition range, s is the conversion monitoring mean, τ is the number of alarms, and ω is the detection delay.

4. The urban underground utility corridor electrical fire monitoring and early warning method according to claim 3 is characterized in that: Calculate the current electrical fire edge detection limit: Where P is the electrical fire edge monitoring limit, x is the directional correction absolute value, τ is the monitoring cluster, e is the monitoring node, is the transmission range, and θ is the edge blur value.

5. The urban underground utility corridor electrical fire monitoring and early warning method according to claim 4 is characterized in that: The step three comprises: The monitoring area map is collected by setting up a near-infrared binocular camera; Preprocess the image, perform threshold segmentation on the collected image, and obtain a binary image; Perform flame recognition on the image, and if there is flame in the image monitoring area, an alarm will be issued; If there is no flame in the image monitoring area, a judgment is made based on the measured values ​​of the gas sensor, temperature and humidity sensor, voltage acquisition device, and the fire edge monitoring limit; Use near-infrared binocular cameras to measure the distance to or from the source of goods and locate the fire source in combination with the camera position; Output the fire source location information and control the fire extinguishing device near the fire source to extinguish the fire.

6. The urban underground utility corridor electrical fire monitoring and early warning method according to claim 5 is characterized in that: The flame identification process includes: Identify the sharp corner features, roundness features and rectangularity features of the image; In the binary image, the highest point of the target image is searched as the suspected flame vertex A, where A can be one pixel or multiple pixels M; Calculate the target image widths N1, N2 and N3 of the H1th row, H2th row and H3th row from vertex A downward respectively; Wherein, 1<H1<H2<H3; If M<N1<N2<N3 is satisfied, it is determined to be a flame sharp angle; When the circularity C≤K, the detection area is determined to be a flame; In the formula, C is the circularity, S is the area of ​​the monitoring area, and L is the perimeter of the detection area; When the monitoring area is rectangular, the rectangularity Take the maximum value 1.

0. When the detection area is a circle, the rectangular The value is π / 4; when the detection area is flame, the rectangularity is; In the formula, S t is the minimum circumscribed rectangle area of ​​the detection area.

7. The urban underground utility corridor electrical fire monitoring and early warning method according to claim 6 is characterized in that: Flame positioning includes: The left and right views captured by the binocular camera are monitored simultaneously. When a flame is detected, it is marked with a rectangle. The following formula is used to determine whether the flames detected in the left and right views are the same flame: Where, X cl is the horizontal coordinate of the center of the rectangular box in the left view, X cr is the horizontal coordinate of the center of the rectangle in the right view, For point (X cl , Y cl ), where the disparity is the difference between the horizontal coordinates of the matching points in the left and right views, and a is the set threshold.

Images