A gas leakage concentration monitoring and early warning method and system

Abstract

The present application relates to gas leakage monitoring technical field, especially to a kind of gas leakage concentration monitoring and early warning method and system, the steps of the method include: the infrared image, acoustic emission signal and wind speed and direction information of specific space are collected by sensor;According to the layout optimization strategy of gas concentration sensor, the distribution position of gas concentration sensor is determined, and the layout optimization strategy includes closed space layout optimization strategy and open space layout optimization strategy;Gas leakage detection network is constructed, the input acoustic emission signal and infrared image are detected by acoustic emission signal processing unit and infrared image processing unit, and gas leakage area is obtained;The gas concentration sensor of gas leakage area is opened, and gas leakage concentration distribution is obtained;Combined with gas leakage concentration distribution and wind speed and direction information, output early warning information.The present application realizes the monitoring and early warning of gas leakage concentration by optimizing the layout of sensor and constructing gas leakage detection network.

Description

Technical Field

[0001] This invention relates to the field of gas leak monitoring technology, and in particular to a method and system for monitoring and early warning of gas leak concentration. Background Technology

[0002] During the long-term transportation of urban gas pipelines, the transported gas often carries inherent risks. Pipelines themselves are susceptible to corrosion or human-caused damage, potentially leading to accidents such as ruptures, perforations, and leaks. Furthermore, due to the flammable and explosive nature of natural gas, leaks can easily trigger fires and explosions, posing a serious threat to people's lives and property. However, existing gas leak detection technologies fail to optimize the placement of gas concentration sensors for specific environments and neglect the impact of ambient wind speed and direction on gas leak concentration detection.

[0003] For example, Chinese patent CN110107815B discloses a method and device for detecting leaks in gas pipelines, which relates to the field of gas pipeline leak monitoring technology. The method includes the following steps: detecting the concentration of combustible gas in underground spaces around multiple gas pipelines; determining whether the concentration of combustible gas meets the leakage conditions; if the leakage conditions are met, determining that the gas pipeline is leaking, and obtaining and reporting the current leak location.

[0004] A Chinese patent with authorization announcement number CN111076096B discloses a method for identifying gas pipeline leaks, including the following steps: S1, collecting gas leak concentration data, time data, and spatial data of each pipe section in the gas pipeline network; S2, setting initial parameters, including a gas leak concentration threshold, leak level classification, time window, and spatial radius; S3, based on the initial parameters, performing spatiotemporal clustering analysis on the gas leak concentration data, time data, and spatial data of each pipe section in the gas pipeline network using the DBSCAN algorithm; S4, calculating the average leak level of leak points in the gas pipeline network based on the clustering analysis results to identify high-risk areas; S5, optimizing the initial parameters based on the given average leak level, and repeating steps S3 to S5 after optimization; S6, outputting the identification result until the calculated average leak level meets the given average leak level. All of the above patents suffer from the problem raised in this background: existing gas leak monitoring technologies fail to optimize the layout of gas concentration sensors according to specific environments and ignore the influence of ambient wind speed and direction on gas leak concentration detection. Summary of the Invention

[0005] In order to overcome the defects and shortcomings of the existing technology, the present invention provides a gas leak concentration monitoring and early warning method and system. By optimizing the layout of gas concentration sensors and constructing a gas leak detection network, the gas leak concentration can be monitored and warned.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] This invention provides a method for monitoring and early warning of gas leak concentration, comprising the following steps:

[0008] Infrared images, acoustic emission signals, and wind speed and direction information are collected in a specific space using sensors.

[0009] The distribution location of gas concentration sensors is determined based on the layout optimization strategy of the gas concentration sensors. The layout optimization strategy includes the closed space layout optimization strategy and the open space layout optimization strategy.

[0010] A gas leak detection network is constructed, and the input acoustic emission signals and infrared images are detected by an acoustic emission signal processing unit and an infrared image processing unit to obtain the gas leak area;

[0011] Turn on the gas concentration sensor in the gas leak area to obtain the gas leak concentration distribution;

[0012] By combining information on gas leak concentration distribution and wind speed and direction, early warning information is output.

[0013] As a preferred technical solution, the specific space includes enclosed space and open space.

[0014] As a preferred technical solution, the enclosed space gas concentration sensor layout optimization strategy is used to optimize the layout of gas concentration sensors in an enclosed space, and the specific steps include:

[0015] Treating gas concentration sensor nodes as charged particles, there is mutual interaction between gas concentration sensor nodes.

[0016] The force is expressed by the following formula:

[0017]

[0018] In the formula F ij S represents the i-th gas concentration sensor node. i and the j-th gas concentration sensor node S j The interaction force between them, d ij S represents the i-th gas concentration sensor node. i and the j-th gas concentration sensor node S j The distance between them, d th Represents the distance threshold, α ij Indicates gas concentration sensor node S i To gas concentration sensor node S j The direction angle of the line segment, ω A The gravitational parameter representing the virtual force, ωR Repulsive force parameters representing virtual forces;

[0019] Calculate the resultant force of the virtual forces acting on node Si of the gas concentration sensor. Expressed as follows:

[0020]

[0021] In the formula, k represents the number of gas concentration sensor nodes;

[0022] Update gas concentration sensor node S i The position is represented by the following formula:

[0023]

[0024] In the formula (x old ,y old ) represents sensor node S i The position coordinates before the update, (x new ,y new ) represents the updated position coordinates of the sensor node, and MaxStep represents the sensor node S. i Maximum step size, F x Indicates sensor node S i The resultant force of virtual forces The component on the x-axis, F y Indicates sensor node S i The resultant force of virtual forces The component on the y-axis.

[0025] As a preferred technical solution, the open space gas concentration sensor layout optimization strategy incorporates wind speed and direction information based on the closed space gas concentration sensor layout optimization strategy. Specific steps include:

[0026] The monitoring area was divided into 16 wind direction angle sector zones, and the wind direction probability was calculated. The wind direction probability formula is:

[0027]

[0028] In the formula, k represents the 16 wind direction angle sector areas where the monitoring area is located, 1≤k≤16, n k f(θ) represents the number of times each wind direction occurs. k ) represents the probability of wind direction with wind angle θ;

[0029] The wind direction with the highest probability is used as the average wind direction. If multiple wind directions have equal probabilities, the wind direction corresponding to the maximum wind speed is used as the average wind direction.

[0030] The gas concentration sensor layout optimization strategy for the enclosed space is implemented within the fan-shaped area of ​​the average wind direction in the monitoring area to determine the location of the gas concentration sensor.

[0031] As a preferred technical solution, the acoustic emission signal processing unit takes the acoustic emission signal as input. The acoustic emission signal processing unit includes 256 3×3 convolutional layers, 128 5×5 convolutional layers, 128 4×4 convolutional layers, 3 pooling layers, 2 Dropout layers with a dropout rate of 0.2, 1 LSTM layer, 1 Flatten layer, and 1 fully connected layer with a sigmoid activation function. The output of the fully connected layer is the leakage detection result.

[0032] As a preferred technical solution, the infrared image processing unit uses adjacent frames of acquired infrared images as input, and the specific steps include:

[0033] The first frame of infrared image acquired is used as the infrared template image, and subsequent adjacent frames of infrared image are used as the infrared search image;

[0034] The AlexNet network was used as a feature extractor to extract features from the infrared template image and the infrared search image, resulting in an infrared template feature map of size 6×6×128 and an infrared search feature map of size 22×22×128.

[0035] The infrared template feature map and the infrared search feature map are input into the similarity comparison function to obtain the response feature map. The position with the highest value in the response feature map is the location of the gas leak area.

[0036] As a preferred technical solution, the similarity comparison function is expressed by the following formula:

[0037]

[0038] In the formula, z represents the infrared template image, and x represents the infrared search image. denoted as the feature extractor, * denotes cross-correlation operation, and b represents the value at each position in the response feature map. The position with the highest value in the response feature map is the location of the gas leak area.

[0039] The present invention also provides a gas leak concentration monitoring and early warning system, comprising:

[0040] The information acquisition module is used to acquire infrared images, acoustic emission signals, and wind speed and direction information of a specific space through sensors;

[0041] The sensor layout optimization module is used to determine the distribution location of the gas concentration sensors according to the layout optimization strategy of the gas concentration sensors. The layout optimization strategy includes the closed space layout optimization strategy and the open space layout optimization strategy.

[0042] The gas leak detection network construction module is used to detect the input acoustic emission signal and infrared image through the acoustic emission signal processing unit and the infrared image processing unit to obtain the gas leak area; the gas leak concentration distribution module is used to activate the gas concentration sensor in the gas leak area to obtain the gas leak concentration distribution.

[0043] The early warning information output module is used to output early warning information by combining the gas leak concentration distribution and wind speed and direction information.

[0044] The present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a method for monitoring and early warning of gas leak concentration.

[0045] The present invention provides a controller comprising a memory and a processor, wherein the memory is used to store a computer program, and the processor is used to execute the computer program to implement a gas leak concentration monitoring and early warning method.

[0046] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0047] (1) This invention optimizes the layout of gas concentration sensors in different spaces by adopting a closed space gas concentration sensor layout optimization strategy and an open space gas concentration sensor layout optimization strategy, thereby improving the accuracy of gas leak concentration monitoring.

[0048] (2) This invention constructs a gas leak detection network to extract features from the collected infrared images and acoustic emission signals to obtain the gas leak area, and then activates the gas concentration sensor in the gas leak area to monitor the gas leak concentration, which effectively reduces the load on the gas concentration sensor. Attached Figure Description

[0049] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0050] Figure 1 This is a schematic diagram of the overall process of a gas leak concentration monitoring and early warning method according to the present invention;

[0051] Figure 2 This is a schematic diagram of the acoustic emission signal processing unit in a gas leak concentration monitoring and early warning method of the present invention;

[0052] Figure 3 This is a schematic diagram of the structure of a gas leak concentration monitoring and early warning system according to the present invention. Detailed Implementation

[0053] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present invention and the specific features in the embodiments are detailed descriptions of the technical solution of the present invention, rather than limitations thereof. In the absence of conflict, the embodiments of the present invention and the technical features in the embodiments can be combined with each other.

[0054] Example 1

[0055] like Figure 1 As shown, this embodiment provides a method for monitoring and early warning of gas leak concentration, which specifically includes the following steps:

[0056] S1: Collect infrared images, acoustic emission signals, and wind speed and direction information of a specific space through sensors. The specific space includes enclosed spaces and open spaces.

[0057] S2: Determine the distribution location of the gas concentration sensors based on the layout optimization strategy of the gas concentration sensors. The layout optimization strategy includes the closed space layout optimization strategy and the open space layout optimization strategy.

[0058] S21: The strategy for optimizing the layout of gas concentration sensors in enclosed spaces is used to optimize the layout of gas concentration sensors in enclosed spaces. Specific steps include:

[0059] Treating gas concentration sensor nodes as charged particles, there are interaction forces between them, expressed by the following formula:

[0060]

[0061] In the formula F ij S represents the i-th gas concentration sensor node. i and the j-th gas concentration sensor node S j The interaction force between them, d ij S represents the i-th gas concentration sensor node. i and the j-th gas concentration sensor node S j The distance between them, d th Represents the distance threshold, α ij Indicates gas concentration sensor node S i To gas concentration sensor node S j The direction angle of the line segment, ω A The gravitational parameter representing the virtual force, ω R Repulsive force parameters representing virtual forces;

[0062] Calculate the resultant force of the virtual forces acting on node Si of the gas concentration sensor. Expressed as follows:

[0063]

[0064] In the formula, k represents the number of gas concentration sensor nodes;

[0065] Update gas concentration sensor node S i The position is represented by the following formula:

[0066]

[0067] In the formula (x old ,y old ) represents sensor node S i The position coordinates before the update, (x new ,y new ) represents the updated position coordinates of the sensor node, and MaxStep represents the sensor node S. i Maximum step size, F x Indicates sensor node S i The resultant force of virtual forces The component on the x-axis, F y Indicates sensor node S i The resultant force of virtual forces The component on the y-axis.

[0068] S22: Open Space Gas Concentration Sensor Layout Optimization Strategy. Based on the closed space gas concentration sensor layout optimization strategy, wind speed and direction information are introduced. Specific steps include:

[0069] The monitoring area was divided into 16 wind direction angle sector zones, and the wind direction probability was calculated. The wind direction probability formula is:

[0070]

[0071] In the formula, k represents the 16 wind direction angle sector areas where the monitoring area is located, 1≤k≤16, n k f(θ) represents the number of times each wind direction occurs. k ) represents the probability of wind direction with wind angle θ;

[0072] The wind direction with the highest probability is used as the average wind direction. If multiple wind directions have equal probabilities, the wind direction corresponding to the maximum wind speed is used as the average wind direction.

[0073] The gas concentration sensor layout optimization strategy for the enclosed space is implemented within the fan-shaped area of ​​the average wind direction in the monitoring area to determine the location of the gas concentration sensor.

[0074] S3: Construct a gas leak detection network, and use an acoustic emission signal processing unit and an infrared image processing unit to detect the input acoustic emission signal and infrared image to obtain the gas leak area;

[0075] like Figure 2 As shown, the acoustic emission signal processing unit takes the acoustic emission signal as input. The acoustic emission signal processing unit includes 256 3×3 convolutional layers, 128 5×5 convolutional layers, 128 4×4 convolutional layers, 3 pooling layers, 2 Dropout layers with a dropout rate of 0.2, 1 LSTM layer, 1 Flatten layer, and 1 fully connected layer with a sigmoid activation function. The output of the fully connected layer is the leakage detection result.

[0076] The infrared image processing unit takes adjacent frames of acquired infrared images as input, and the specific steps include:

[0077] The first frame of infrared image acquired is used as the infrared template image, and subsequent adjacent frames of infrared image are used as the infrared search image;

[0078] The AlexNet network was used as a feature extractor to extract features from the infrared template image and the infrared search image, resulting in an infrared template feature map of size 6×6×128 and an infrared search feature map of size 22×22×128.

[0079] The infrared template feature map and the infrared search feature map are input into the similarity comparison function to obtain the response feature map. The position with the highest value in the response feature map is the location of the gas leak area.

[0080] The similarity comparison function is expressed by the following formula:

[0081]

[0082] In the formula, z represents the infrared template image, and x represents the infrared search image. denoted as the feature extractor, * denotes cross-correlation operation, and b represents the value at each position in the response feature map. The position with the highest value in the response feature map is the location of the gas leak area.

[0083] S4: Activate the gas concentration sensor in the gas leak area to obtain the gas leak concentration distribution.

[0084] S5: Combines gas leak concentration distribution and wind speed and direction information to output early warning information.

[0085] Example 2

[0086] like Figure 3 As shown, this embodiment provides a gas leak concentration monitoring and early warning system 20, including: an information acquisition module 21, used to acquire infrared images, acoustic emission signals and wind speed and direction information of a specific space through sensors;

[0087] The sensor layout optimization module 22 is used to determine the distribution position of the gas concentration sensor according to the layout optimization strategy of the gas concentration sensor. The layout optimization strategy includes the closed space layout optimization strategy and the open space layout optimization strategy.

[0088] The gas leak detection network construction module 23 is used to detect the input acoustic emission signal and infrared image through the acoustic emission signal processing unit and the infrared image processing unit to obtain the gas leak area;

[0089] The gas leak concentration distribution module 24 is used to activate the gas concentration sensor in the gas leak area to obtain the gas leak concentration distribution.

[0090] The early warning information output module 25 is used to output early warning information by combining the gas leak concentration distribution and wind speed and direction information.

[0091] In this embodiment, the information acquisition module 21 is used to acquire infrared images, acoustic emission signals and wind speed and direction information of a specific space through sensors. The specific space includes enclosed spaces and open spaces.

[0092] In this embodiment, the sensor layout optimization module 22 is used to determine the distribution position of the gas concentration sensor according to the layout optimization strategy of the gas concentration sensor. The layout optimization strategy includes a closed space layout optimization strategy and an open space layout optimization strategy.

[0093] The strategy for optimizing the layout of gas concentration sensors in enclosed spaces involves several steps, including:

[0094] Treating gas concentration sensor nodes as charged particles, there are interaction forces between them, expressed by the following formula:

[0095]

[0096] In the formula F ij S represents the i-th gas concentration sensor node. i and the j-th gas concentration sensor node S j The interaction force between them, d ij S represents the i-th gas concentration sensor node. i and the j-th gas concentration sensor node S j The distance between them, d th Represents the distance threshold, α ij Indicates gas concentration sensor node S i To gas concentration sensor node S j The direction angle of the line segment, ω A The gravitational parameter representing the virtual force, ω R Repulsive force parameters representing virtual forces;

[0097] Calculate the resultant force of the virtual forces acting on node Si of the gas concentration sensor. Expressed as follows:

[0098]

[0099] In the formula, k represents the number of gas concentration sensor nodes;

[0100] Update gas concentration sensor node S i The position is represented by the following formula:

[0101]

[0102] In the formula (x old ,y old ) represents sensor node S i The position coordinates before the update, (x new ,y new ) represents the updated position coordinates of the sensor node, and MaxStep represents the sensor node S. i Maximum step size, F x Indicates sensor node S i The resultant force of virtual forces The component on the x-axis, F y Indicates sensor node S i The resultant force of virtual forces The component on the y-axis.

[0103] The open space gas concentration sensor layout optimization strategy incorporates wind speed and direction information based on the enclosed space gas concentration sensor layout optimization strategy. Specific steps include:

[0104] The monitoring area was divided into 16 wind direction angle sector zones, and the wind direction probability was calculated. The wind direction probability formula is:

[0105]

[0106] In the formula, k represents the 16 wind direction angle sector areas where the monitoring area is located, 1≤k≤16, n k f(θ) represents the number of times each wind direction occurs. k ) represents the probability of wind direction with wind angle θ;

[0107] The wind direction with the highest probability is used as the average wind direction. If multiple wind directions have equal probabilities, the wind direction corresponding to the maximum wind speed is used as the average wind direction.

[0108] The gas concentration sensor layout optimization strategy for the enclosed space is implemented within the fan-shaped area of ​​the average wind direction in the monitoring area to determine the location of the gas concentration sensor.

[0109] In this embodiment, the gas leak detection network construction module 23 is used to construct a gas leak detection network. The input acoustic emission signal and infrared image are detected by the acoustic emission signal processing unit and the infrared image processing unit to obtain the gas leak area.

[0110] The gas leak detection network includes an acoustic emission signal processing unit and an infrared image processing unit. The acoustic emission signal processing unit takes the acoustic emission signal as input and includes 256 3×3 convolutional layers, 128 5×5 convolutional layers, 128 4×4 convolutional layers, 3 pooling layers, 2 Dropout layers with a dropout rate of 0.2, 1 LSTM layer, 1 Flatten layer, and 1 fully connected layer with a sigmoid activation function. The output of the fully connected layer is the leak detection result.

[0111] The infrared image processing unit takes adjacent frames of acquired infrared images as input, and the specific steps include:

[0112] The first frame of infrared image acquired is used as the infrared template image, and subsequent adjacent frames of infrared image are used as the infrared search image;

[0113] The AlexNet network was used as a feature extractor to extract features from the infrared template image and the infrared search image, resulting in an infrared template feature map of size 6×6×128 and an infrared search feature map of size 22×22×128.

[0114] The infrared template feature map and the infrared search feature map are input into the similarity comparison function to obtain the response feature map. The position with the highest value in the response feature map is the location of the gas leak area.

[0115] The similarity comparison function is expressed by the following formula:

[0116]

[0117] In the formula, z represents the infrared template image, and x represents the infrared search image. denoted as the feature extractor, * denotes cross-correlation operation, and b represents the value at each position in the response feature map. The position with the highest value in the response feature map is the location of the gas leak area.

[0118] In this embodiment, the gas leak concentration distribution module 24 is used to activate the gas concentration sensor in the gas leak area to obtain the gas leak concentration distribution.

[0119] In this embodiment, it is used to combine gas leak concentration distribution and wind speed and direction information to output early warning information.

[0120] The parameters and steps of each unit module in the gas leak concentration monitoring and early warning system of the present invention described above can be referred to the parameters and steps in the embodiments of the gas leak concentration monitoring and early warning method described above, and will not be repeated here.

[0121] Example 3

[0122] This invention discloses a computer-readable storage medium storing a computer program. When executed by a processor, the computer program implements a gas leak concentration monitoring and early warning method as described above. It should be noted that all computer programs for the gas leak concentration monitoring and early warning method are implemented using Python. The sensor layout optimization module, gas leak detection network construction module, gas leak concentration distribution module, and early warning information output module are all controlled by a remote server. The remote server has an Intel Xeon Gold 5215 CPU, an NVIDIA GTX2080Ti 11GB GPU, an Ubuntu 18.04.2 operating system, a PyTorch 1.7.0 deep learning framework, CUDA version 10.2, and uses cuDNN 7.5.3 for accelerated inference. The Intel Xeon Gold 5215 includes a memory and a processor. The memory stores the computer program, and the processor executes the computer program, causing the Intel Xeon Gold 5215 to implement the gas leak concentration monitoring and early warning method. Those skilled in the art will understand that this invention can be implemented as a system, method, or computer program product.

[0123] Therefore, this disclosure can be implemented in the following forms: it can be entirely hardware, entirely software (including firmware, resident software, microcode, etc.), or a combination of hardware and software, generally referred to herein as a "circuit," "module," or "system." Furthermore, in some embodiments, the invention can also be implemented as a computer program product in one or more computer-readable media, the computer-readable medium containing computer-readable program code.

[0124] Any combination of one or more computer-readable media may be used. A computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. A computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in connection with an instruction execution system, apparatus, or device.

[0125] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for monitoring and early warning of gas leak concentration, characterized in that, The process includes the following steps: acquiring infrared images, acoustic emission signals, and wind speed and direction information for a specific space using sensors; determining the distribution locations of gas concentration sensors based on a layout optimization strategy, which includes both enclosed space and open space layout optimization strategies; constructing a gas leak detection network, and detecting the input acoustic emission signals and infrared images using an acoustic emission signal processing unit and an infrared image processing unit to obtain the gas leak area; activating the gas concentration sensors in the gas leak area to obtain the gas leak concentration distribution; and combining the gas leak concentration distribution with wind speed and direction information to output a warning message. The specific steps of the gas concentration sensor layout optimization strategy include: (1) If the specific space is a closed space, the gas concentration sensor nodes are regarded as charged particles, and there is an interaction force between the gas concentration sensor nodes, which is expressed by the following formula: In the formula F ij S represents the i-th gas concentration sensor node. i and the j-th gas concentration sensor node S j The interaction force between them, d ij S represents the i-th gas concentration sensor node. i and the j-th gas concentration sensor node S j The distance between them, d th Represents the distance threshold, a ij Indicates gas concentration sensor node S i To gas concentration sensor node S j The direction angle of the line segment, ω A The gravitational parameter representing the virtual force, ω R Repulsive force parameters representing virtual forces; Calculate the resultant force of the virtual forces acting on node Si of the gas concentration sensor. Expressed as follows: In the formula, k represents the number of gas concentration sensor nodes; Update gas concentration sensor node S i The position is represented by the following formula: In the formula (x old ,y old ) represents sensor node S i The position coordinates before the update, (x new ,y new ) represents the updated position coordinates of the sensor node, and MaxStep represents the sensor node S. i Maximum step size, F x Indicates sensor node S i The resultant force of virtual forces The component on the x-axis, F y Indicates sensor node S i The resultant force of virtual forces Components on the y-axis; (2) If the specific space is an open space, wind speed and direction information are introduced based on the optimized layout strategy of gas concentration sensors in enclosed spaces. The specific steps include: The monitoring area was divided into 16 wind direction angle sector zones, and the wind direction probability was calculated. The wind direction probability formula is: In the formula, k represents the 16 wind direction angle sector areas where the monitoring area is located, 1≤k≤16, n k f(θ) represents the number of times each wind direction occurs. k ) represents the probability of wind direction with wind angle θ; The wind direction with the highest probability is used as the average wind direction. If multiple wind directions have equal probabilities, the wind direction corresponding to the maximum wind speed is used as the average wind direction. The gas concentration sensor layout optimization strategy for the enclosed space is implemented within the fan-shaped area of ​​the average wind direction in the monitoring area to determine the location of the gas concentration sensor.

2. The gas leak concentration monitoring and early warning method according to claim 1, characterized in that, The acoustic emission signal processing unit takes the acoustic emission signal as input and includes 256 3×3 convolutional layers, 128 5×5 convolutional layers, 128 4×4 convolutional layers, 3 pooling layers, 2 Dropout layers with a dropout rate of 0.2, 1 LSTM layer, 1 Flatten layer, and 1 fully connected layer with a sigmoid activation function. The output of the fully connected layer is the leakage detection result.

3. The gas leak concentration monitoring and early warning method according to claim 1, characterized in that, The infrared image processing unit takes adjacent frames of acquired infrared images as input, and the specific steps include: The first frame of infrared image acquired is used as the infrared template image, and subsequent adjacent frames of infrared image are used as the infrared search image; The AlexNet network was used as a feature extractor to extract features from the infrared template image and the infrared search image, resulting in infrared template feature map and infrared search feature map. The infrared template feature map and the infrared search feature map are input into the similarity comparison function to obtain the response feature map. The position with the highest value in the response feature map is the location of the gas leak area.

4. The gas leak concentration monitoring and early warning method according to claim 3, characterized in that, The similarity comparison function is expressed by the following formula: In the formula, z represents the infrared template image, and x represents the infrared search image. 'b' represents the feature extractor, '*' represents the cross-correlation operation, and 'b' represents the value at each position in the response feature map. The position with the highest value in the response feature map is the location of the gas leak area.

5. A gas leak concentration monitoring and early warning system, implemented based on the gas leak concentration monitoring and early warning method according to any one of claims 1-4, characterized in that, The system includes: The information acquisition module is used to acquire infrared images, acoustic emission signals, and wind speed and direction information of a specific space through sensors; the sensor layout optimization module is used to determine the distribution location of gas concentration sensors according to the layout optimization strategy of gas concentration sensors. The layout optimization strategy includes closed space layout optimization strategy and open space layout optimization strategy. The gas leak detection network construction module is used to detect the input acoustic emission signal and infrared image through the acoustic emission signal processing unit and the infrared image processing unit to obtain the gas leak area; The gas leak concentration distribution module is used to activate the gas concentration sensor in the gas leak area and obtain the gas leak concentration distribution. The early warning information output module is used to output early warning information by combining the gas leak concentration distribution and wind speed and direction information.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements a gas leak concentration monitoring and early warning method according to any one of claims 1-4.

7. A controller comprising a memory and a processor, the memory for storing a computer program, characterized in that, The processor is used to execute the computer program to implement the gas leak concentration monitoring and early warning method according to any one of claims 1-4.

Images