A mine environment investigation method based on unmanned aerial vehicle investigation

Abstract

This invention provides a method for mine environmental surveying based on unmanned aerial vehicle (UAV) investigation, belonging to the field of mine environmental survey technology. The method involves dividing the mine area and determining UAV routes, using the UAV to acquire mine environmental data and build a 3D model, and further acquiring specific information about the mine environment as it changes. This invention solves the problem that current mine environmental surveys mostly rely on manual investigation, which is time-consuming, labor-intensive, and unsafe, and has the advantage of preventing mine accidents.

Description

Technical Field

[0001] This invention relates to the field of mine environmental survey technology, specifically to a mine environmental survey method based on unmanned aerial vehicle (UAV) investigation. Background Technology

[0002] A mine refers to an independent production and operation unit that extracts ore within a defined mining boundary. A mine mainly includes one or more mining workshops (also known as pitheads, mine shafts, open-pit mines, etc.) and some auxiliary workshops. Most mines also include ore dressing plants (coal washing plants).

[0003] During the construction and production of mines, due to complex natural conditions, poor working environment, and insufficient understanding of the objective laws of mine disasters, people may sometimes be negligent, violate regulations, or give orders that could lead to certain disasters.

[0004] To prevent mining accidents, investigations of the mining environment are essential. The mining environment includes meteorological and hydrological conditions, topography, stratigraphy, geological structure, neotectonic movements, hydrogeology, engineering geology, and environmental geology. However, due to the complexity of the mining environment, investigations are currently often conducted manually, which is time-consuming, labor-intensive, and has low safety standards. Summary of the Invention

[0005] The technical problem solved by this invention is that currently, most mine environmental surveys rely on manual investigation, which is time-consuming, labor-intensive, and has low safety.

[0006] To solve the above problems, the technical solution of the present invention is as follows:

[0007] A method for mine environmental survey based on drone inspection includes the following steps:

[0008] S1. Confirm the location of the target mine, delineate the mine area according to the mine area division rules, establish the drone route according to the mine area, and obtain the point cloud data set of the mine through the drone.

[0009] S2. Construct a three-dimensional model of the overall mine environment and three-dimensional models of each mine area using the point cloud data set acquired by the UAV, as the initial environmental data of the mine;

[0010] S3. The drone periodically acquires point cloud data sets of the mine according to the set route, and constructs a three-dimensional model of the overall environment of the mine and a three-dimensional model of each mining area through the point cloud data sets acquired by the drone, as mine environmental survey data;

[0011] S4. Compare and judge the initial environmental data of the mine with the environmental survey data of the mine. When it is judged that there is an environmental change, obtain the location information of the environmental change.

[0012] S5. The drone equipped with a recognition model, camera and gas sensor reaches the location of environmental change to collect low-level images and transmits the captured images back in real time. The recognition model identifies the captured images and returns the identified object types in the environment and the image screenshots to the terminal. At the same time, the gas sensor returns the acquired ambient gas concentration to the terminal.

[0013] S6. Based on the images returned by the drone, the types of objects in the environment obtained from the recognition model, the screenshots, the ambient gas concentration, and the mining progress, the results of the mine environment survey are obtained.

[0014] Furthermore, the mining area division rules are as follows: the area is divided according to the edge of the mine and the location of each mine entrance, and each drone route is taken by 5 to 10 drones to acquire scene photos.

[0015] Note: During the mining process, changes in the underground structure of the mine are inevitably linked to the mining operations. At the same time, changes in the mine environment will inevitably affect the environment of other land adjacent to the mine. Therefore, dividing the area based on the edge of the mine and the location of each mine entrance is the most effective way to reflect changes in the mine environment. This will facilitate the rapid capture of changes in the landform when conducting mine environmental surveys using drones.

[0016] Furthermore, the UAV in steps S2 and S3 is equipped with UAV-borne radar, GPS unit, and inertial measurement unit.

[0017] Note: The UAV-borne radar can acquire point cloud data of the mining environment during the UAV's flight. The GPS unit and inertial measurement unit can acquire the corresponding GPS data and IMU data of the point cloud data, which is beneficial for the establishment of a three-dimensional model of the mining environment and facilitates the subsequent judgment of changes in the terrain.

[0018] Furthermore, constructing a 3D model of the overall mine environment and 3D models of each mine area using point cloud data acquired by drones includes the following steps:

[0019] SA1. Based on the size of the mining area, set up 8 to 10 elevation benchmarks outside the mining area, obtain the latitude and longitude coordinates and elevation of the benchmarks, and set up GPS receivers within the mining area.

[0020] SA2. After the drone is started, it flies along the drone route. During the flight, the drone collects point cloud data, GPS data and IMU data of the mining area and transmits them back to the control terminal.

[0021] SA3. The point cloud data acquired by the UAV-borne radar, the GPS data acquired by the GPS unit, and the IMU data acquired by the inertial measurement unit are synchronously solved to generate point cloud data with absolute coordinates and stitched together to obtain a point cloud data set. The point cloud data set is then used to generate a triangular mesh model for each mining area, i.e., a three-dimensional model of each mining area.

[0022] SA4. Combine the three-dimensional models of all mining areas according to their positional relationships to obtain a three-dimensional model of the overall mining environment.

[0023] Note: The control terminal mentioned above is equipped with software for synchronously processing point cloud data, GPS data, and IMU data, as well as software for stitching together point cloud data sets and software for creating triangular mesh models, thus enabling the construction of a three-dimensional model of the mining environment. The control terminal can be a local server or a cloud server.

[0024] Furthermore, starting up the drone includes: activating the inertial measurement unit, the drone-borne radar, and the GPS, and setting the drone-borne radar to enter data acquisition mode.

[0025] Note: The drone has the ability to simultaneously collect point cloud data, GPS data, and IMU data, which is beneficial for building 3D models of the mining environment.

[0026] Preferably, step S4 includes:

[0027] By comparing and analyzing the initial environmental data and the environmental survey data of the mine, the distance difference between the point clouds at the same latitude and longitude is obtained. When the difference is less than 1% of the initial data, it is determined that the mine environment has not been deformed; otherwise, it is determined that the mine environment has been deformed. The latitude and longitude of the deformed point cloud and the corresponding environmental change type are output. After comparing all the point cloud data, a three-dimensional terrain change map of the mine environment is obtained.

[0028] Preferably, the types of environmental changes include: ground collapse and ground uplift.

[0029] Note: Ground collapse and ground uplift are both caused by changes in the underground structure of the mine. When changes are found in the underground structure of the mine, they should be promptly detected and relevant personnel should be alerted to avoid mine accidents.

[0030] Preferably, the gas sensor includes: a carbon monoxide sensor, a nitrogen dioxide sensor, and a hydrogen sulfide sensor.

[0031] Note: During low-altitude flight, the drone's gas sensors acquire the concentration of harmful gases above the location of terrain changes, which helps relevant personnel to determine the specific underground situation in conjunction with the progress of mining operations.

[0032] More preferably, the recognition model is trained based on the YOLOv5 algorithm, and the types of objects in the environment include: humans, mining equipment, trees, and vegetation.

[0033] Note: The YOLOv5 algorithm has the advantages of fast model training and the ability to produce real-time results when processing inference data in batches.

[0034] The beneficial effects of this invention are:

[0035] (1) This invention divides the mine into regions based on the location of the mine entrance and the edge area. Based on the regional division, the UAV route is designed to ensure that the UAV can establish a relatively accurate three-dimensional model of the key areas of the mine during its flight over the mine. On this basis, by regularly performing three-dimensional modeling of the mine surface environment, the two are compared to obtain the surface changes, so that relevant personnel can obtain the current status of landform changes in a timely manner, which helps to prevent mine accidents.

[0036] (2) Based on the judgment of landform changes, the present invention obtains images of the landform change location in real time through drones. Based on this, it provides the type of objects appearing in the image and takes screenshots. At the same time, it obtains the concentration of harmful gases in the air above the landform change location, which helps relevant personnel to judge the specific underground situation in conjunction with the mining progress. Attached Figure Description

[0037] Figure 1 This is a flowchart of a mine environment survey method based on drone investigation, as described in Example 1. Detailed Implementation

[0038] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0039] The terminology used in the embodiments of this invention is for the purpose of describing particular embodiments only and is not intended to limit the invention. The singular forms “a,” “the,” and “the” as used in the embodiments of this invention and the appended claims are also intended to include the plural forms, and “multiple” generally includes at least two unless the context clearly indicates otherwise.

[0040] Glossary

[0041] Point cloud data: Point cloud data refers to a collection of vectors in a three-dimensional coordinate system. Scanned data is recorded in the form of points, each containing three-dimensional coordinates, and some may contain color information (RGB) or reflectance intensity information.

[0042] An inertial measurement unit (IMU) is a device that measures the three-axis attitude angles (or angular rates) and acceleration of an object. Gyroscopes and accelerometers are the main components of the IMU, and their accuracy directly affects the accuracy of the inertial system. In practical operation, various unavoidable interference factors cause errors in the gyroscopes and accelerometers. From the initial alignment, the navigation error increases over time, especially the position error, which is a major drawback of inertial navigation systems. Therefore, external information is needed to assist in integrated navigation, effectively reducing the problem of error accumulation over time. To improve reliability, more sensors can be equipped for each axis. Generally, the IMU is mounted at the center of gravity of the object being measured.

[0043] Triangular mesh: A polygonal mesh composed entirely of triangles; a triangular mesh model is a graphic constructed from triangular meshes.

[0044] Unmanned Aerial Vehicle (UAV) Radar: UAV-borne radar refers to radar installed on unmanned aerial vehicles (UAVs). There are two types of UAV-borne radar: The first type is a search radar with moving target indication (MTI) operation, primarily employing look-down search and search-while-tracking (SLS) modes. It can detect moving targets on land or sea within a 100-kilometer range and can simultaneously search and track hundreds of targets. The second type is a synthetic aperture radar (SAR) / ground moving target indication radar, which provides continuous and clear target images through real-time high-resolution imaging. It can detect both stationary military targets and long-range moving targets on land (or water). This invention employs the second type of UAV-borne radar.

[0045] YOLOv5 is a series of object detection architectures and models pre-trained on the COCO dataset. It is an extension of the YOLO series, and its network structure consists of four modules: input, backbone, neck, and head. YOLOv5 modifies all four parts of the YOLOv4 network and achieves significant improvements. On the input side, it uses Mosaic data augmentation, adaptive anchor box calculation, and adaptive image scaling; on the backbone side, it uses Focus and CSP structures; on the neck side, it adds an FPN+PAN structure; and on the head side, it improves the loss function during training by using GIOU_Loss and DIOU_nms for predictive box filtering.

[0046] Example 1

[0047] This embodiment is a method for mine environmental survey based on drone investigation, including the following steps:

[0048] S1. Confirm the location of the target mine, delineate the mine area according to the mine area division rules, and establish drone routes based on the mine area. Use drones to acquire point cloud data sets of the mine. The mine area division rules are: divide the area according to the edge of the mine and the location of each mine entrance. Each drone route is acquired by 5 drones to obtain scene photos.

[0049] S2. Construct a three-dimensional model of the overall mine environment and three-dimensional models of each mine area using the point cloud data set acquired by the UAV as the initial environmental data of the mine. The UAV is equipped with UAV-borne radar, GPS unit and inertial measurement unit.

[0050] S3. Every half month, the drone will collect point cloud data sets of the mine according to the set route, and construct a three-dimensional model of the overall environment of the mine and a three-dimensional model of each mining area using the point cloud data sets acquired by the drone, as mine environmental survey data.

[0051] In steps S2 and S3, constructing a 3D model of the overall mine environment and 3D models of each mine area using the point cloud data set acquired by the UAV includes the following steps:

[0052] SA1. Based on the size of the mining area, set up 8 to 10 elevation benchmarks outside the mining area, obtain the latitude and longitude coordinates and elevation of the benchmarks, and set up GPS receivers within the mining area.

[0053] SA2. After the drone is started, it flies along the drone route. During the flight, the drone collects point cloud data, GPS data and IMU data of the mining area and transmits them back to the control terminal.

[0054] SA3. The point cloud data acquired by the UAV-borne radar, the GPS data acquired by the GPS unit, and the IMU data acquired by the inertial measurement unit are synchronously solved to generate point cloud data with absolute coordinates and stitched together to obtain a point cloud data set. The point cloud data set is then used to generate a triangular mesh model for each mining area, i.e., a three-dimensional model of each mining area.

[0055] SA4. Combine the three-dimensional models of all mining areas according to their positional relationships to obtain a three-dimensional model of the overall mining environment.

[0056] In this embodiment, step SA2, starting the UAV includes: starting the inertial measurement unit, the UAV-borne radar, and the GPS, and setting the UAV-borne radar to enter the data acquisition state.

[0057] In this embodiment, the software used for synchronous calculation in step SA3 is Inertial Explorer and ScanLookExpor, the software for point cloud data set is MicroStation V8i, and the triangular mesh model is constructed using a TIN-based thinning algorithm. All of the above processing procedures are existing technologies.

[0058] S4. Compare and judge the initial environmental data of the mine with the environmental survey data of the mine. When it is determined that there is an environmental change, obtain the location information of the environmental change, including the following:

[0059] By comparing and analyzing the initial environmental data and the environmental survey data of the mine, the distance difference between the point clouds at the same latitude and longitude is obtained. When the difference is less than 1% of the initial data, it is determined that the mine environment has not been deformed; otherwise, it is determined that the mine environment has been deformed. The latitude and longitude of the deformed point cloud and the corresponding environmental change type are output. After comparing all the point cloud data, a three-dimensional terrain change map of the mine environment is obtained.

[0060] In this embodiment, the types of environmental changes include: ground collapse and ground uplift.

[0061] S5. The drone, equipped with a recognition model, camera, and gas sensor, reaches the location of environmental changes to collect low-level images and transmits the captured images back in real time. The recognition model identifies the captured images and returns the identified object types and screenshots to the terminal. At the same time, the gas sensor returns the acquired ambient gas concentration to the terminal.

[0062] In this embodiment, the gas sensors include: a carbon monoxide sensor, a nitrogen dioxide sensor, and a hydrogen sulfide sensor.

[0063] S6. Based on the images returned by the drone, the types of objects in the environment obtained from the recognition model, the screenshots, the ambient gas concentration, and the mining progress, the results of the mine environment survey are obtained.

[0064] In this embodiment, the recognition model is trained based on the YOLOv5 algorithm, and the types of objects in the environment include: humans, mining equipment, trees, and vegetation.

[0065] The YOLOv5 algorithm boasts advantages such as rapid model training and the ability to generate real-time results when batch processing inference data. Therefore, in this embodiment, the recognition model can identify the types of objects appearing in the captured footage in real time. This is particularly useful in the event of a ground collapse, enabling timely detection of survivors.

[0066] Example 2

[0067] This embodiment is a method for mine environmental survey based on drone investigation. The difference between this embodiment and embodiment 1 is that:

[0068] In step S1, each drone route is captured by 8 drones taking scene photos.

[0069] In step S3, the drone acquires a set of point cloud data of the mine each month according to a set route.

[0070] Example 3

[0071] This embodiment is a method for mine environmental survey based on drone investigation. The difference between this embodiment and embodiment 1 is that:

[0072] In step S1, each drone route is captured by 10 drones taking scene photos.

[0073] In step S3, the drone acquires a set of point cloud data of the mine every three months according to a set route.

Claims

1. A method for mine environmental survey based on unmanned aerial vehicle (UAV) inspection, characterized in that, Includes the following steps: S1. Confirm the location of the target mine, delineate the mine area according to the mine area division rules, establish the drone route according to the mine area, and obtain the point cloud data set of the mine through the drone. S2. Construct a 3D model of the overall mine environment and 3D models of each mine area using the point cloud data set acquired by the UAV, as the initial environmental data of the mine; the construction of the 3D model of the overall mine environment and 3D models of each mine area using the point cloud data set acquired by the UAV includes the following steps: SA1. Based on the size of the mining area, set up 8 to 10 elevation benchmarks outside the mining area, obtain the latitude and longitude coordinates and elevation of the benchmarks, and set up GPS receivers within the mining area. SA2. After the drone is started, it flies along the drone route. During the flight, the drone collects point cloud data, GPS data and IMU data of the mining area and transmits them back to the control terminal. SA3. The point cloud data acquired by the UAV-borne radar, the GPS data acquired by the GPS unit, and the IMU data acquired by the inertial measurement unit are synchronously solved to generate point cloud data with absolute coordinates and stitched together to obtain a point cloud data set. The point cloud data set is then used to generate a triangular mesh model for each mining area, i.e., a three-dimensional model of each mining area. SA4. Combine the three-dimensional models of all mining areas according to their positional relationships to obtain a three-dimensional model of the overall mining environment; S3. The drone periodically acquires point cloud data sets of the mine according to the set route, and constructs a three-dimensional model of the overall environment of the mine and a three-dimensional model of each mining area through the point cloud data sets acquired by the drone, as mine environmental survey data; S4. Compare and judge the initial environmental data of the mine with the environmental survey data of the mine. When it is judged that there is an environmental change, obtain the location information of the environmental change. S5. The drone equipped with a recognition model, camera and gas sensor reaches the location of environmental change to collect low-level images and transmits the captured images back in real time. The recognition model identifies the captured images and returns the identified object types in the environment and the image screenshots to the terminal. At the same time, the gas sensor returns the acquired ambient gas concentration to the terminal. S6. Based on the images returned by the drone, the types of objects in the environment obtained from the recognition model, the screenshots, the concentration of ambient gases, and the progress of mining, the results of the mine environment survey are obtained.

2. The method for mine environmental survey based on unmanned aerial vehicle (UAV) investigation as described in claim 1, characterized in that, The mining area division rule is as follows: the area is divided according to the edge of the mine and the location of each mine entrance, and each drone route is taken by 5 to 10 drones to acquire scene photos.

3. The method for mine environmental survey based on unmanned aerial vehicle (UAV) investigation as described in claim 1, characterized in that, The UAVs in steps S2 and S3 are equipped with UAV-borne radar, GPS unit, and inertial measurement unit.

4. The method for mine environmental survey based on unmanned aerial vehicle (UAV) investigation as described in claim 1, characterized in that, The process of starting the drone includes: starting the inertial measurement unit, the drone-borne radar, and the GPS, and setting the drone-borne radar to enter the data acquisition state.

5. A method for mine environmental survey based on unmanned aerial vehicle (UAV) investigation as described in claim 1, characterized in that, Step S4 includes: By comparing and analyzing the initial environmental data and the environmental survey data of the mine, the distance difference between the point clouds at the same latitude and longitude is obtained. When the difference is less than 1% of the initial data, it is determined that the mine environment has not been deformed; otherwise, it is determined that the mine environment has been deformed. The latitude and longitude of the deformed point cloud and the corresponding environmental change type are output. After comparing all the point cloud data, a three-dimensional terrain change map of the mine environment is obtained.

6. A method for mine environmental survey based on unmanned aerial vehicle (UAV) investigation as described in claim 5, characterized in that, The types of environmental changes include: ground collapse and ground uplift.

7. A method for mine environmental survey based on unmanned aerial vehicle (UAV) investigation as described in claim 1, characterized in that, The gas sensors include: a carbon monoxide sensor, a nitrogen dioxide sensor, and a hydrogen sulfide sensor.

8. A method for mine environmental survey based on unmanned aerial vehicle (UAV) investigation as described in claim 1, characterized in that, The recognition model is trained based on the YOLOv5 algorithm, and the types of objects in the environment include: humans, mining equipment, and vegetation.

Images