A method for gas distribution detection for confined space work safety

Abstract

The application provides a gas distribution detection method for limited space operation safety, and belongs to the technical field of limited space gas detection. The gas distribution detection method for limited space operation safety comprises the following steps: collecting the three-dimensional structure of the current limited space; placing a plurality of collection terminals according to the three-dimensional structure of the current limited space; obtaining current collection data of the plurality of collection terminals; and establishing a gas distribution model in the limited space, analyzing the current collection data, and judging the safety of the limited space operation. The application can solve the problem that the position of the gas detection terminal is unclear in the prior art, so that more accurate gas distribution data for safety can be obtained.

Description

Technical Field

[0001] This invention belongs to the field of confined space gas detection technology, and more specifically, relates to a gas distribution detection method for safety in confined space operations. Background Technology

[0002] Many cities, large power plants, and industrial and mining enterprises in China have adopted underground cable trenches to supply electricity. However, underground cable trenches are enclosed or partially enclosed, with narrow and limited entrances and exits. They are not designed as fixed workplaces, have poor natural ventilation, and are prone to accumulating toxic, harmful, flammable, and explosive substances or creating confined spaces with insufficient oxygen content. To ensure the safety of workers entering these confined spaces, various protective tests and regular inspections are required to prevent accidents.

[0003] In recent years, accidents such as poisoning from harmful gases and asphyxiation due to lack of oxygen have occurred frequently in various industries in China during confined space operations, and the safety situation of confined space operations has become extremely serious. Currently, when workers enter confined spaces, there is a problem that they cannot determine the gas distribution within the space, and rashly entering can easily lead to safety accidents.

[0004] The existing technology does not clearly define how to place gas detection terminals to obtain more accurate gas distribution data for safety purposes. Summary of the Invention

[0005] In view of this, the present invention provides a gas distribution detection method for safety in confined space operations, which can solve the problem in the prior art that it is unclear how to place the gas detection terminal to obtain more accurate gas distribution data for safety.

[0006] This invention is implemented as follows:

[0007] This invention provides a gas distribution detection method for safety in confined space operations, comprising the following steps:

[0008] S10: Acquire the three-dimensional structure of the currently defined finite space;

[0009] S20: Place multiple acquisition terminals based on the current three-dimensional structure of the limited space and similar historical spatial structures;

[0010] S30: Obtain the current data collected by the plurality of data acquisition terminals;

[0011] S40: Establish a gas distribution model within a confined space, analyze the currently collected data, and output the analysis results to the user.

[0012] Based on the above technical solution, the gas distribution detection method for confined space operation safety of the present invention can be further improved as follows:

[0013] The step of placing multiple acquisition terminals according to the three-dimensional structure of the current limited space specifically includes:

[0014] Historical data is acquired, which is an array. Each element of the array is a completed gas distribution detection task in a limited space. The gas distribution detection task includes the three-dimensional spatial structure, the number and location of the acquisition terminals, and the gas distribution data.

[0015] Calculate the first similarity between the spatial three-dimensional structure of each element in the historical data and the current three-dimensional structure of the finite space;

[0016] The element with the highest first similarity in the historical data is used as the reference task;

[0017] The number and location of the acquisition terminals for the reference task are processed to obtain the number and location of the acquisition terminals placed within the current limited space;

[0018] Multiple acquisition terminals are placed according to the number and location of acquisition terminals within the limited space.

[0019] Furthermore, the step of using the element with the highest first similarity in the historical data as a reference task; and processing the number and location of the acquisition terminals for the reference task to obtain the number and location of the acquisition terminals placed in the current limited space specifically includes:

[0020] The spatial three-dimensional structure of the reference task is meshed to obtain a reference mesh, wherein the reference mesh includes multiple consecutive small cubes with a specified length, width and height of h, and the location of the acquisition terminal of the reference task is described by multiple consecutive green small cubes.

[0021] The three-dimensional structure of the current finite space is meshed to obtain the target mesh, which includes multiple continuous small cubes with a specified length, width and height of h;

[0022] Establish a mapping relationship between the reference mesh and the target mesh, and based on the mapping relationship, set the green cube in the reference mesh to red in the target mesh;

[0023] The positions of consecutive green cubes in the target grid are used as the placement positions of the acquisition terminals, with each placement position corresponding to one acquisition terminal.

[0024] Furthermore, the step of meshing the spatial three-dimensional structure of the reference task to obtain a reference mesh specifically includes:

[0025] With the geometric center of the reference space as the origin, establish a first coordinate system with the vertical direction, the NS direction, and the EW direction;

[0026] With side length h and the origin of the first coordinate system as the center, establish a first initial small cube, wherein the normal direction of one face of the first initial small cube is the vertical direction;

[0027] In the first coordinate system, multiple consecutive small cubes are established sequentially with the first initial small cube as the center to form a first set of small cubes. The multiple small cubes in the first set of small cubes cover the entire space of the reference task, and each small cube in the first set of small cubes intersects with the space of the reference task.

[0028] The specific steps for meshing the current three-dimensional structure of the finite space to obtain the target mesh include:

[0029] Using the geometric center of the finite space as the origin, establish a second coordinate system with the vertical direction, the NS direction, and the EW direction;

[0030] With side length h and the origin of the second coordinate system as the center, establish a second initial small cube, wherein the normal direction of one face of the second initial small cube is the vertical direction;

[0031] In the second coordinate system, multiple consecutive small cubes are established sequentially with the second initial small cube as the center to form a second set of small cubes. The multiple small cubes in the second set of small cubes cover the entire current finite space, and each small cube in the second set of small cubes intersects with the current finite space.

[0032] Furthermore, the step of establishing a mapping relationship between the reference mesh and the target mesh, and setting the small cube mapped from the green cube in the reference mesh to red in the target mesh according to the mapping relationship, specifically includes:

[0033] Establish a three-dimensional projective space for the reference mesh, denoted as the reference projective space; establish a three-dimensional projective space for the target mesh, denoted as the target projective space.

[0034] A pixel remapping method is used to map all points corresponding to the green cubes in the reference projective space to the target projective space. If a point corresponding to a green cube in the reference projective space intersects with a point corresponding to a target grid cube in the target projective space, the target grid cube corresponding to the intersection is marked in red. If a point corresponding to a green cube in the reference projective space does not intersect with a point corresponding to a target grid cube in the target projective space, a neighboring point corresponding to the green cube is found in the target projective space, and the neighboring point corresponding to the target grid cube is set to red.

[0035] The steps involved in establishing a gas distribution model within a confined space and analyzing the currently collected data include:

[0036] Establish a training dataset: The input to the training dataset is the location of each acquisition terminal of the historical data and the gas data collected, and the output of the training dataset is the gas distribution data of the historical data.

[0037] A gas distribution model in a finite space is established using a convolutional neural network, and the model is trained using a training dataset.

[0038] The gas distribution data of the current finite space is obtained by using the gas distribution model in the finite space to calculate the current collected data.

[0039] Furthermore, the method for calculating the first similarity between the spatial three-dimensional structure of each element in the historical data and the current three-dimensional structure of the finite space is: point cloud similarity calculation method.

[0040] Furthermore, the value of h ranges from [1cm, 20cm].

[0041] Furthermore, h = 10 cm.

[0042] Compared with existing technologies, the beneficial effects of the gas distribution detection method for safety in confined space operations provided by this invention are as follows: by processing the confined space into a three-dimensional mesh, the gas distribution in each small cube of the confined space can be matched with historical data for similarity. The location of the acquisition terminal can be determined by using historical data, and the current gas distribution in the confined space can be detected. This solves the technical problem in existing technologies where it is unclear how to place the gas detection terminal, thus obtaining more accurate gas distribution data for safety. Attached Figure Description

[0043] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 This is a flowchart illustrating the steps for ensuring safety in confined space operations.

[0045] Figure 2 A schematic diagram of the coordinate system for the 3D mesh is provided. Detailed Implementation

[0046] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0047] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0048] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0049] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0050] like Figure 1 The diagram shown is a flowchart of a procedure for ensuring safety in confined space operations provided by the present invention, including the following steps:

[0051] S10: Collect the three-dimensional structure of the current finite space;

[0052] S20: Place multiple acquisition terminals based on the current three-dimensional structure of the limited space and based on similar historical spatial structures;

[0053] S30: Acquire the current data collected by multiple acquisition terminals;

[0054] S40: Establish a gas distribution model within a confined space, analyze the currently collected data, and output the analysis results to the user.

[0055] In this invention, the data acquisition terminal uses a ZCE04B four-in-one gas sensor module, capable of detecting the concentrations of multiple gases including O2, H2S, CO, and CH4. It can also be used in factories housing equipment containing the harmful gas SF6, such as factories storing GIS equipment. It can be applied to various electrical or non-electrical spaces requiring gas release monitoring, such as cable trenches.

[0056] Alternatively, the acquisition terminal can use arrays of various gas sensors, such as monitoring the oxygen content in the channel, or, depending on the actual situation of harmful gases, using methane (CH4), carbon monoxide (CO), carbon dioxide (CO2), and hydrogen sulfide (H2S) as the main monitoring targets of the gas sensor array in the toxic and harmful gas monitoring field.

[0057] In the above technical solution, the step of placing multiple acquisition terminals based on the current three-dimensional structure of the limited space and based on similar historical spatial structures specifically includes:

[0058] Acquire historical data, which is an array. Each element of the array represents a completed gas distribution detection task within a finite space. The gas distribution detection task includes the three-dimensional spatial structure, the number and location of the acquisition terminals, and the gas distribution data.

[0059] Calculate the first similarity between the spatial three-dimensional structure of each element in the historical data and the current three-dimensional structure in the finite space;

[0060] The element with the highest similarity in the historical data is used as the reference task;

[0061] The number and location of the acquisition terminals for the reference task are processed to obtain the number and location of the acquisition terminals placed in the current limited space;

[0062] Multiple data acquisition terminals are placed according to the number and location of the acquisition terminals within the limited space.

[0063] It should be noted that this invention uses historical data from gas distribution detection tasks to perform similarity matching with the current three-dimensional structure of a confined space, predicts the current gas distribution, and adjusts the location of the data acquisition terminal accordingly. Examples include trenches and factory spaces of the same specifications.

[0064] Preferably, after obtaining the analysis results, the user can, based on experience, further adjust the actual position of each of the multiple data acquisition terminals. During the monitoring test, the content of the sensitive gas detected by the sensor is released at random locations to detect the alarm response time. Whether the time can be met or shortened can be determined after user adjustments, and whether the response time can be further reduced.

[0065] If the alarm's response time can be shortened, this approach should be included as part of the historical data for the detection task.

[0066] Based on the historical data of the gas distribution detection task with the highest similarity, the number and location of the acquisition terminals in the current limited space are compared and adjusted to be the same as the number and location of the acquisition terminals in the historical data of the gas distribution detection task with the highest similarity.

[0067] For example, taking cable trenches as an example, historical data shows that there have been instances where monitoring results were good, such as timely alarms and positive user feedback regarding gas sensitivity. Alternatively, there is monitoring data from GIS equipment in factory buildings. The locations of these terrain-based spatial acquisition terminals are sometimes evenly spaced, sometimes added at high-frequency leakage locations, and sometimes placed at locations with a high predicted probability.

[0068] Furthermore, in the above technical solution, the element in the historical data with the highest first similarity is used as the reference task; the step of processing the number and location of the acquisition terminals of the reference task to obtain the number and location of the acquisition terminals placed in the current limited space specifically includes:

[0069] The spatial three-dimensional structure of the reference task is meshed to obtain a reference mesh, which consists of multiple consecutive small cubes with a specified length, width and height of h. The location of the acquisition terminal of the reference task is described by multiple consecutive green small cubes.

[0070] The current three-dimensional structure in the finite space is meshed to obtain the target mesh, which consists of multiple continuous small cubes with a specified length, width and height of h.

[0071] Establish a mapping relationship between the reference mesh and the target mesh, and based on the mapping relationship, set the small green cube in the reference mesh to red in the target mesh;

[0072] The positions of consecutive green cubes in the target grid are used as the placement positions of the acquisition terminals, with each placement position corresponding to one acquisition terminal.

[0073] It should be noted that by processing the spatial three-dimensional structure of the reference task into a mesh, the deployment positions of the acquisition terminals in the historical data can be accurately marked. After similarity matching is completed, the precise position of the acquisition terminal in the current limited space can be mapped, which facilitates the adjustment of the acquisition terminal position.

[0074] Furthermore, in the above technical solution, the step of meshing the spatial three-dimensional structure of the reference task to obtain a reference mesh specifically includes:

[0075] With the geometric center of the reference space as the origin, establish a first coordinate system with the vertical direction, the NS direction, and the EW direction;

[0076] It should be noted that NS is the north-south direction, and EW is the east-west direction.

[0077] With side length h and the origin of the first coordinate system as the center, establish a first initial small cube, wherein the normal direction of one face of the first initial small cube is the vertical direction;

[0078] In the first coordinate system, multiple consecutive small cubes are established sequentially with the first initial small cube as the center to form the first set of small cubes. The multiple small cubes in the first set of small cubes cover the entire space of the reference task, and each small cube in the first set of small cubes intersects with the space of the reference task.

[0079] The specific steps to mesh the current finite three-dimensional structure to obtain the target mesh include:

[0080] Using the geometric center of the current finite space as the origin, establish a second coordinate system with the vertical direction, the NS direction, and the EW direction;

[0081] With side length h and the origin of the second coordinate system as the center, establish a second initial small cube, wherein the normal direction of one face of the second initial small cube is the vertical direction;

[0082] In the second coordinate system, multiple consecutive small cubes are established with the second initial small cube as the center, forming a second set of small cubes. The multiple small cubes in the second set of small cubes cover the entire current finite space, and each small cube in the second set of small cubes intersects with the current finite space.

[0083] Furthermore, in the above technical solution, the step of establishing a mapping relationship between the reference mesh and the target mesh, and setting the small cube mapped from the green cube in the reference mesh to red in the target mesh according to the mapping relationship, specifically includes:

[0084] Establish a three-dimensional projective space for the reference mesh, denoted as the reference projective space; establish a three-dimensional projective space for the target mesh, denoted as the target projective space.

[0085] A pixel remapping method is used to map all points corresponding to the green cubes in the reference projective space to the target projective space. If a point corresponding to a green cube in the reference projective space intersects with a point corresponding to a target grid cube in the target projective space, the target grid cube corresponding to the intersection is marked in red. If a point corresponding to a green cube in the reference projective space does not intersect with a point corresponding to a target grid cube in the target projective space, a neighboring point corresponding to the green cube is found in the target projective space, and the neighboring point corresponding to the target grid cube is set to red.

[0086] It should be noted that three-dimensional projective space refers to the set of all points in three-dimensional projective geometry. It is called the extended space after adding a plane (i.e., the plane at infinity, which is composed of all points at infinity on all lines) to three-dimensional Euclidean space (or affine space). Ordinary elements (points, lines, and planes) in the extended space and elements at infinity are treated equally without distinction. This space becomes a model of three-dimensional projective space.

[0087] In the above technical solution, the steps for establishing a gas distribution model within a confined space and analyzing the currently collected data include:

[0088] Establish a training dataset: The input to the training dataset is the location of each acquisition terminal in the historical data and the gas data collected, and the output of the training dataset is the gas distribution data of the historical data.

[0089] A gas distribution model in a finite space is established using a convolutional neural network, and the model is trained using a training dataset.

[0090] The gas distribution data of the current finite space is obtained by using the gas distribution model in the finite space to calculate the gas distribution data of the current finite space based on the collected data.

[0091] The steps for calculating the first similarity between the spatial three-dimensional structure of each element in the historical data and the current finite spatial three-dimensional structure are as follows:

[0092] Obtain the first orthographic projection of the spatial three-dimensional structure of each element in the historical data to form the first orthographic projection set;

[0093] Calculate the second orthographic projection of the current 3D structure in the finite space;

[0094] The similarity between each first orthographic projection contour and the second orthographic projection contour in the first orthographic projection set is calculated using OpenCV's matchShapes function and denoted as the first similarity.

[0095] The steps for obtaining the orthographic projection of a three-dimensional spatial structure are as follows:

[0096] With the geometric center of the three-dimensional spatial structure as the center, establish the circumsphere of the three-dimensional spatial structure;

[0097] 125 points are uniformly selected on the circumscribed sphere, and each point is connected to its adjacent points to form 246 minimum triangles;

[0098] The three-dimensional spatial structure is projected onto the plane containing each smallest triangle, and the projection with the largest area is selected as the orthographic projection.

[0099] Furthermore, in the above technical solution, the value range of h is [1cm, 20cm].

[0100] By using a mapping method with multiple segmented cubes, more accurate model matching can be achieved. Users can simply place sensors at recommended locations, saving time and improving efficiency.

[0101] like Figure 2 As shown, the first embodiment of meshing the spatial three-dimensional structure of the reference task is as follows:

[0102] A closed factory building for SF6 was selected within the limited space. The factory building is a cuboid structure with a length of 10 meters, a width of 10 meters, and a height of 3 meters. The geometric center of the cuboid is selected as the origin O, and a cube with h of 1 cm is selected. A total of 1000 cubes are built in the north-south direction, 300 cubes are built in the vertical direction, and 1000 cubes are built in the east-west direction. The three-dimensional grid has a specification of 1000×1000×300. The corresponding positions of the data acquisition terminal are marked with color.

[0103] The second embodiment of meshing the spatial three-dimensional structure of the reference task is as follows:

[0104] The difference from the first embodiment is that h is 10cm and the three-dimensional grid has a size of 100×100×30.

[0105] The third embodiment of meshing the spatial three-dimensional structure of the reference mission is as follows:

[0106] The difference from the first embodiment is that h is 20cm and the three-dimensional grid is 50×50×15.

[0107] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A gas distribution detection method for safety in confined space operations, characterized in that, Includes the following steps: S10: Acquire the three-dimensional structure of the currently defined finite space; S20: Place multiple acquisition terminals based on the current three-dimensional structure of the limited space and similar historical spatial structures; S30: Obtain the current data collected by the plurality of data acquisition terminals; S40: Establish a gas distribution model within a confined space, analyze the currently collected data, and output the analysis results to the user; The step of placing multiple acquisition terminals based on the current three-dimensional structure of the limited space and similar historical spatial structures specifically includes: Historical data is acquired, which is an array. Each element of the array is a completed gas distribution detection task in a limited space. The gas distribution detection task includes the three-dimensional spatial structure, the number and location of the acquisition terminals, and the gas distribution data. Calculate the first similarity between the spatial three-dimensional structure of each element in the historical data and the current three-dimensional structure of the finite space; The element with the highest first similarity in the historical data is used as the reference task; The number and location of the acquisition terminals for the reference task are processed to obtain the number and location of the acquisition terminals placed within the current limited space; Multiple data acquisition terminals are placed according to the number and location of the acquisition terminals within the current limited space; The step of using the element with the highest first similarity in the historical data as a reference task and processing the number and location of the acquisition terminals for the reference task to obtain the number and location of the acquisition terminals placed in the current limited space specifically includes: The spatial three-dimensional structure of the reference task is meshed to obtain a reference mesh, wherein the reference mesh includes multiple consecutive small cubes with a specified length, width and height of h, and the location of the acquisition terminal of the reference task is described by multiple consecutive green small cubes. The three-dimensional structure of the current finite space is meshed to obtain the target mesh, which includes multiple continuous small cubes with a specified length, width and height of h; Establish a mapping relationship between the reference mesh and the target mesh, and based on the mapping relationship, set the green cube in the reference mesh to red in the target mesh; The positions of consecutive green cubes in the target grid are used as the placement positions of the acquisition terminals, with each placement position corresponding to one acquisition terminal.

2. The gas distribution detection method for safety in confined space operations according to claim 1, characterized in that, The specific steps of meshing the spatial three-dimensional structure of the reference task to obtain the reference mesh include: With the geometric center of the reference space as the origin, establish a first coordinate system with the vertical direction, the NS direction, and the EW direction; With side length h and the origin of the first coordinate system as the center, establish a first initial small cube, wherein the normal direction of one face of the first initial small cube is the vertical direction; In the first coordinate system, multiple consecutive small cubes are established sequentially with the first initial small cube as the center to form a first set of small cubes. The multiple small cubes in the first set of small cubes cover the entire space of the reference task, and each small cube in the first set of small cubes intersects with the space of the reference task. The specific steps for meshing the current three-dimensional structure of the finite space to obtain the target mesh include: Using the geometric center of the finite space as the origin, establish a second coordinate system with the vertical direction, the NS direction, and the EW direction; With side length h and the origin of the second coordinate system as the center, establish a second initial small cube, wherein the normal direction of one face of the second initial small cube is the vertical direction; In the second coordinate system, multiple consecutive small cubes are established sequentially with the second initial small cube as the center to form a second set of small cubes. The multiple small cubes in the second set of small cubes cover the entire current finite space, and each small cube in the second set of small cubes intersects with the current finite space.

3. The gas distribution detection method for safety in confined space operations according to claim 1, characterized in that, The steps of establishing a mapping relationship between a reference mesh and a target mesh, and setting the green cube in the reference mesh to red in the target mesh according to the mapping relationship, specifically include: Establish a three-dimensional projective space for the reference mesh, denoted as the reference projective space; establish a three-dimensional projective space for the target mesh, denoted as the target projective space. A pixel remapping method is used to map all points corresponding to the green cubes in the reference projective space to the target projective space. If a point corresponding to a green cube in the reference projective space intersects with a point corresponding to a target grid cube in the target projective space, the target grid cube corresponding to the intersection is marked in red. If a point corresponding to a green cube in the reference projective space does not intersect with a point corresponding to a target grid cube in the target projective space, a neighboring point corresponding to the green cube is found in the target projective space, and the neighboring point corresponding to the target grid cube is set to red.

4. The gas distribution detection method for safety in confined space operations according to claim 1, characterized in that, The steps for establishing a gas distribution model within a confined space and analyzing the currently collected data include: Establish a training dataset: The input to the training dataset is the location of each acquisition terminal of the historical data and the gas data collected, and the output of the training dataset is the gas distribution data of the historical data. A gas distribution model in a finite space is established using a convolutional neural network, and the model is trained using a training dataset. The gas distribution data of the current finite space is obtained by using the gas distribution model in the finite space to calculate the current collected data.

5. A gas distribution detection method for safety in confined space operations according to claim 1, characterized in that, The method for calculating the first similarity between the spatial three-dimensional structure of each element in the historical data and the current three-dimensional structure of the finite space is: point cloud similarity calculation method.

6. A gas distribution detection method for safety in confined space operations according to claim 1, characterized in that, The value of h ranges from [1cm, 20cm].

7. A gas distribution detection method for safety in confined space operations according to claim 6, characterized in that, h=10cm.

Images