Unmanned aerial vehicle automatic coal inventory system and method suitable for gas film closed coal yard

Abstract

The invention discloses an unmanned aerial vehicle automatic coal inventory system and method suitable for a gas film closed coal yard, and the system comprises an unmanned aerial vehicle, a feature device, a server, a remote control device, a WIFI system, and a bucket wheel machine full-automatic control system, and the unmanned aerial vehicle comprises an unmanned aerial vehicle body, a control device, an airborne PC, an IMU, and a laser radar. The server and the bucket wheel machine full-automatic control system are connected to the local area network through network cables, the WIFI system is connected to the local area network, the airborne PC is provided with a WIFI module and is connected to the WIFI system through a WIFI network, and the remote control device is in low-rate data communication with the control device in a wireless mode. The method has the following advantages and effects that in the gas film closed coal yard, the unmanned aerial vehicle is automatically driven to fly, the laser scanning data of the coal yard is obtained, the three-dimensional model of the coal yard is established, the amount of each coal pile is calculated, the coal stocktaking efficiency is improved, the investment and maintenance cost of a coal stocktaking system is reduced, and the time for coal stocktaking personnel to make contact with the severe environment of the closed coal yard is shortened.

Description

Technical Field

[0001] The present disclosure relates to the technical field of scene reconstruction, and the present invention relates to an automatic coal handling system and method of an unmanned aerial vehicle applicable to an air-film-enclosed coal yard. Background Art

[0002] Coal is the most important raw material for thermal power plants, accounting for about 70% to 80% of the cost of thermal power plants. Under the current coal market, the stacking and storage management of coal yards is of great significance for the blending of coal in thermal power plants, improving the utilization rate of coal yard inventory, and reducing costs and increasing efficiency. With the gradual increase in environmental awareness, the closure of coal yards has gradually become the mainstream direction of thermal power plants. The less-manned and unmanned operation of equipment and processes in closed coal yards has become the transformation direction of closed coal yards. At present, the inventory of coal yards in thermal power plants mainly includes three aspects: volume, density, and quality.

[0003] At present, some coal-fired power plants have built fully automatic coal counting systems in closed coal yards. The mainstream coal counting systems in air membrane closed coal yards include manual counting, laser coal counting instrument counting on the top of bucket wheel machine, drone counting, slide rail coal counting and other methods. Among them, manual counting has the disadvantages of long time, high labor intensity, low accuracy and long-term exposure of coal counting personnel to harsh environment; laser coal counting instrument counting on the top of bucket wheel machine adopts the method of installing a laser coal counting instrument on the top of bucket wheel machine and performing coal counting operation by walking bucket wheel machine, but this method has the disadvantages of large blind area, low measurement accuracy and affecting the normal stacking and reclaiming operation of bucket wheel machine; drone counting has positioning difficulties such as low positioning accuracy due to weak GPS signal, difficult installation of UWB base station and difficulty in full coverage of signal, so it can only be carried out by manual remote control, and all data of laser point cloud need to be globally spliced, and the coal counting software algorithm is difficult and time-consuming; and the suspended slide rail coal counting device in the shed not only has high investment cost, but also has a series of problems such as large track positioning error, easy damage of mobile trolley and high maintenance cost. Summary of the invention

[0004] In view of the shortcomings of the prior art, the present invention provides a UAV automatic coal handling system and method suitable for an air-membrane closed coal yard, which can automatically drive the UAV to fly in the air-membrane closed coal yard, obtain laser scanning data of the coal yard, establish a three-dimensional model of the coal yard, and calculate the coal quantity of each coal pile, so as to improve the coal handling efficiency, reduce the investment and maintenance costs of the coal handling system, and reduce the time that the coal handling personnel are exposed to the harsh environment of the closed coal yard.

[0005] To achieve the above-mentioned purpose, the present invention provides the following technical solutions: an unmanned aerial vehicle automatic coal handling system suitable for an air-film-enclosed coal yard, comprising an unmanned aerial vehicle, a feature device, a server, a remote control device, a WIFI system and a bucket wheel excavator fully automatic control system, the unmanned aerial vehicle comprising an unmanned aerial vehicle body, a control device, an airborne PC, an IMU, and a laser radar, the server and the bucket wheel excavator fully automatic control system are connected to a local area network via a network cable, the WIFI system is connected to the local area network, the airborne PC is equipped with a WIFI module, and is connected to the WIFI system via a WIFI network, and the remote control device performs low-speed data communication with the control device via wireless.

[0006] The present invention is further configured as follows: the bucket wheel machine fully automatic control system obtains bucket wheel machine positioning data in real time, and the positioning data includes the trolley track position, the cantilever pitch angle, and the cantilever rotation angle.

[0007] The present invention is further configured as follows: the server receives positioning data of the bucket wheel machine's fully automatic control system and sends it to the onboard PC of the drone; and receives coal yard laser scanning data from the onboard PC of the drone and generates a three-dimensional model of the coal yard, calculates the coal quantity and outputs a coal pan report.

[0008] The present invention is further configured as follows: the control device performs data communication with the remote control device, receives instructions from the remote control device and the onboard PC, and controls the UAV body to fly; The IMU inertial measurement unit outputs IMU data to an onboard PC; The laser radar outputs laser scanning data to an onboard PC; The onboard PC receives IMU and lidar data, calculates the drone position data, generates a flight map, and automatically outputs the coal yard laser scanning point cloud data through the WIFI AP.

[0009] The present invention is further configured as follows: the feature device is a structural device with specific geometric features, which is arranged on the coal retaining wall outside the air-film enclosed coal yard and on the coal retaining walls on both sides of the bucket wheel excavator track, and is used by the onboard PC to identify feature points using laser scanning point cloud data obtained by a laser radar.

[0010] An automatic coal handling method of a drone applicable to an air film closed coal yard, comprising the above-mentioned automatic coal handling system of a drone applicable to an air film closed coal yard, Step 1: Preparation before coal handling: In the air film coal yard, arrange characteristic devices along the outer side and the coal retaining walls on both sides of the track; calibrate the coordinates of the UAV IMU and laser radar; establish the initial coal yard map; Step 2: The server receives the bucket wheel machine operation information and sets a flight route map to avoid the bucket wheel machine operation area; Step 3: The drone flies and uses the SLAM algorithm to achieve positioning, navigation, and global point cloud mapping; Step 4: After the flight, the server receives the UAV IMU positioning data and lidar point cloud data; the server builds a global three-dimensional model of the coal yard, calculates the coal quantity, and outputs a coal counting report.

[0011] The present invention is further configured as follows: in the step 1, the flying height of the UAV is H, the height of the coal retaining wall is h, the scanning angle of the laser radar is α, the width of the coal pile is W, the distance between the coal pile and the coal retaining wall is l, and the spacing between the characteristic devices satisfies: (1) , where R = H × tanα; (2) , where r = (Hh) × tanα.

[0012] The present invention is further configured as follows: the SLAM establishment steps in step 3 are as follows: 3.1. Coordinate transformation: convert the IMU data and the laser radar point cloud data frame into the site coordinate system. The laser radar point cloud data frame is transformed by the IMU data. The site coordinate system takes the site length direction as the X-axis, the site width direction as the Y-axis, and the site height direction as the Z-axis. 3.2. Since the laser radar moves during a scanning process, motion distortion correction is performed on the frame data in each point cloud data frame; a frame data output only contains one time point; within a laser radar scanning cycle, the IMU outputs a number of position data; a point cloud frame data calculates the true coordinates of each scanning point in the frame based on the IMU position data in the frame to achieve motion distortion correction; 3.3. Remove useless noise points according to the characteristics of the coal yard, including point clouds below the height of the characteristic device (h-δ) and other abnormal points; δ is a threshold related to the geometric size of the characteristic device to avoid the relevant point clouds being mistakenly deleted when the characteristic device changes its posture (such as tilt); 3.4. Select the point cloud with a height of (h±δ) in the point cloud frame and perform feature analysis. When the point cloud feature meets the geometric characteristics of the feature device, identify the geometric center point of the feature device; 3.5. When the X coordinate value of the IMU data is equal to the X coordinate of the geometric center point of the feature device, the point cloud frame is recorded and set as (i=1,2,…), the corresponding IMU coordinates are =( , , ), the geometric center coordinates of the feature device are =( , , ); 3.6. Using ICP algorithm, (i=1,2,…) for point cloud stitching, where The initial frame becomes the initial map of the flight, and the subsequent point clouds are stitched together to gradually generate a new map; 3.7. During the flight, according to the new attitude data updated by the bucket wheel machine, the no-fly zone in the map is updated, and the flight route is automatically modified to avoid the no-fly zone; 3.8. Based on the map generated in 3.6, get the latest map The geometric center coordinates of the feature device corresponding to the frame =( , , ), and the coordinates recorded in 3.5 generate the observation-estimation residual Δ ; Using Δ Correct IMU coordinates for ; 3.9, use replace Process the IMU data and lidar point cloud frame data in subsequent flights, that is, repeat steps 3.1-3.8.

[0013] The present invention is further configured as follows: the method for reasoning about the global three-dimensional model of the coal yard in step 4 is as follows: 4.1 Perform coordinate transformation on the initial IMU and LiDAR point cloud data frames; perform motion distortion correction on the frame data in each point cloud data frame; remove abnormal points according to the characteristics of the coal yard; 4.2 Select the point cloud with a height of (h±δ) in the point cloud frame and perform feature analysis. When the point cloud feature meets the geometric characteristics of the feature device, identify the geometric center point of the feature device; 4.3 Assume that the point cloud frame dataset of the geometric center point of the feature device is ={point cloud frame|contains the geometric center point of the i-th feature device}, using the ICP algorithm, (i=1,2,…) stitch point clouds one by one; 4.4. When the X coordinate value of the IMU data is equal to the X coordinate of the geometric center point of the feature device, the point cloud frame is recorded and set as (i=1,2,…), the corresponding IMU coordinates are =( , , ), the geometric center coordinates of the feature device are =( , , ); 4.5. Based on the point cloud set generated in 4.3, obtain The geometric center coordinates of the feature device corresponding to the frame =( , , ), and the coordinates in 4.4 generate the observation-estimation residual Δ , using Δ Correct IMU coordinates for ; 4.6. Utilization , perform motion distortion correction on the subsequent sequence of point cloud data frames, and repeat the process of 4.1-4.5 until all point cloud frames are processed; 4.7. The point cloud set finally generated is gridded according to the XY axis, a three-dimensional model is constructed, and the volume of each coal pile is calculated.

[0014] The present invention has the following effects: 1. A method for high-precision positioning and coal yard mapping of UAVs that integrates laser SLAM (Simultaneous Localization and Mapping) and IMU (Inertial Measurement Unit) technologies is proposed; Feature devices are arranged along the two sides of the coal yard. During the flight of the UAV, IMU (inertial measurement unit) and laser scanning data are obtained simultaneously. The UAV is equipped with a lidar sensor and an IMU sensor, making full use of the high frequency of the IMU sensor and the high precision of the lidar. By adopting the extended Kalman filter and graph optimization loop detection technology, real-time mapping of the closed coal yard is achieved, and autonomous flying coal panning is realized without human intervention in the middle; 2. Develop an obstacle avoidance control technology for UAV flight in a closed coal yard; On the map constructed by SLAM, the flight route of the drone is automatically planned. When SLAM mapping of the coal yard is carried out during flight, the laser radar is used to perceive the obstacles of mechanical equipment such as bucket wheel machines on the flight path in real time, and the dynamic window local planning technology is used to re-plan the flight route and flight speed of the drone to ensure safe avoidance of obstacles, so as to achieve the simultaneous operation of bucket wheel machine stacker and reclaimer and drone coal panning operation; 3. Propose an autonomous and all-round inventory method for closed coal yards; Using high-precision UAV flight positioning and mapping technology, we can obtain all-round laser scanning data of the coal yard. Using ICP (Inductively Coupled Plasma) technology (mainly including ICP emission spectroscopy technology and ICP-MS (Inductively Coupled Plasma Mass Spectrometry) technology), we can merge and splice the acquired laser scanning data clusters. Using TIN algorithm, we can achieve three-dimensional modeling, correct IMU measurement deviation, calculate the volume of each coal pile in the coal yard, and realize all-round inventory in the closed coal yard, eliminating the blind spots inherent in other automatic coal counting methods, and improving the running speed of coal counting software and the accuracy of coal counting. Utilizing this system, an autonomous inventory system for a closed coal yard with an autonomous flying drone can achieve one-click and fully autonomous operation of the coal yard inventory process, improve inventory efficiency and accuracy, and reduce the repeatability error of inventory measurement. The inventory process can be operated with one click, the drone can fly autonomously, and no human intervention is required, and the time required to process inventory data and produce reports is much shorter. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 is a system block diagram of the present invention; Figure 2 It is a top view schematic diagram of the coal yard layout in the present invention; Figure 3 It is a side view schematic diagram of the coal yard arrangement in the present invention; Figure 4 It is the SLAM flow chart of the present invention; Figure 5 This is a flow chart of the coal pan software processing in the present invention; Figure 6 This is a real-life modeling picture after being processed by the coal pan software in the present invention. DETAILED DESCRIPTION

[0016] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0017] In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and "third" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance.

[0018] Reference Figures 1 to 6 The embodiments of the present invention are further described.

[0019] An automatic coal handling system using drones suitable for air film enclosed coal yards (see attached Figure 1 ), including a drone, a feature device, a server, a remote control device, a WIFI system and a bucket wheel machine full-automatic control system. The drone includes a drone body, a control device, an onboard PC (onboard computer), an IMU (inertial measurement unit), and a laser radar.

[0020] The remote control device, server and drone hangar constitute the ground station.

[0021] The ground station server and bucket wheel machine automatic control system are connected to the local area network through network cables, the WIFI system is connected to the local area network, and the drone onboard PC is equipped with a WIFI module, which is connected to the WIFI system through the WIFI network. The remote control device of the ground station communicates with the drone control device at a low rate via wireless.

[0022] The bucket wheel machine's fully automatic control system obtains the bucket wheel machine's positioning data in real time. The positioning data includes the trolley track position, cantilever pitch angle, and cantilever rotation angle.

[0023] The ground station server receives the positioning data of the bucket wheel machine's fully automatic control system and sends it to the drone's onboard PC; it receives the coal yard laser scanning data from the drone's onboard PC and generates a three-dimensional model of the coal yard, calculates the coal quantity, and outputs a coal pan report.

[0024] The drone hangar is used for parking and charging drones. In order to ensure stable flight of drones in the air-membrane enclosed coal yard, the following design can be adopted: 3D laser radar positioning, which enables stable flight in the presence of strong magnetic interference; It uses four axes and eight propellers, with an outer contour of 600mm. The exterior is fully protected by a carbon fiber spherical cover to ensure that when the drone collides with an obstacle, the lightweight spherical cover can rebound or rotate 360 ​​degrees to protect the drone and surrounding buildings such as the air membrane. The maximum pitch angle is 30°, the maximum ascent speed is 5 m / s (non-SLAM mode), the maximum descent speed is 3 m / s, the maximum horizontal flight speed is 16 m / s, the maximum wind resistance speed under level 5 wind is 10 m / s, the hovering accuracy is ±0.1 m, and in automatic mode, the entire flight process does not require manual intervention, and the deviation between the actual flight waypoint and the set waypoint position is less than 20 cm.

[0025] The drone control device communicates data with the remote control device of the ground station, receives instructions from the remote control device and the onboard PC, and controls the drone body to fly. The IMU inertial measurement unit outputs IMU data to the onboard PC; the laser radar outputs laser scanning data to the onboard PC. The onboard PC receives IMU (inertial measurement unit) and laser radar data, calculates the drone's position data, generates a flight map, and automatically outputs the coal yard laser scanning point cloud data through WIFI AP (abbreviation for Wireless Fidelity Access Point).

[0026] The LiDAR uses a solid-state LiDAR with a circular scanning area.

[0027] The feature device is a structural device with specific geometric features, which is arranged on the outer coal retaining wall of the air film closed coal yard and the coal retaining wall on both sides of the bucket wheel machine track, and is used by the onboard PC to identify the feature points through the laser scanning point cloud data obtained by the laser radar. The beneficial effect is that after the coal pile is taken out, especially after the site is cleared and leveled, there are few site features in the laser scanning point cloud data, and the point cloud data of different frames cannot be spliced.

[0028] An automatic coal handling method using a drone suitable for an air-film-enclosed coal yard comprises the above-mentioned automatic coal handling system using a drone suitable for an air-film-enclosed coal yard.

[0029] Step 1: Preparation before coal handling: In the air film coal yard, arrange characteristic devices along the outer side and the coal retaining walls on both sides of the track; calibrate the coordinates of the UAV IMU (inertial measurement unit) and laser radar; establish an initial coal yard map; the detailed steps are as follows: 1.1. The characteristic device adopts a geometric shape that is easily identifiable from the coal yard, such as a cross star.

[0030] 1.2. The characteristic devices are arranged on the coal retaining walls outside the coal yard and on both sides of the track ( Figure 2 , Figure 3 ), and calibrate the position information of the coal yard coordinate system. To ensure that the front and rear point cloud frames cover the characteristic device and the coal pile, the UAV flight altitude H, the coal retaining wall height h, the laser radar scanning angle α, the coal pile width W, the distance between the coal pile and the coal retaining wall l, and the spacing between the characteristic devices satisfy: (1) , where R = H × tanα; (2) , where r = (Hh) × tanα.

[0031] 1.3. According to the installation geometric relationship of the lidar and IMU (inertial measurement unit) on the UAV, the conversion relationship between the coordinate systems of the lidar, IMU (inertial measurement unit) and coal yard.

[0032] 1.4. Use manual methods to control the flight of the drone. After the server collects data, it calibrates the position of each feature device, establishes an initial coal yard map, manually determines the initial flight route, and publishes the initial coal yard map and initial flight route to the onboard PC.

[0033] 1.5. The bucket wheel excavator body geometry is published to the onboard PC.

[0034] 1.6. IMU (Inertial Measurement Unit), LiDAR and onboard PC use clock synchronization protocol to maintain real-time synchronization.

[0035] Step 2: The server receives the bucket wheel machine operation information and sets the flight route map to avoid the bucket wheel machine operation area; the detailed steps are as follows: 2.1. The server receives the bucket wheel machine position data from the bucket wheel machine automatic control system in real time, including the trolley track position, boom pitch angle, and boom rotation angle, and converts them into coal yard coordinate system parameters. The bucket wheel machine coordinate system parameters are sent to the onboard PC in real time.

[0036] 2.2. The onboard PC designates the area where the bucket wheel machine is located as a no-fly zone, modifies the flight route according to the original flight route, and automatically starts the flight. (The original flight route is shown in Figure 2 middle arrow).

[0037] Step 3: The drone is flying, using the SLAM (Simultaneous Localization and Mapping) algorithm to achieve positioning, navigation, and global point cloud mapping; During the flight, the drone maintains a stable and safe altitude H, which is set according to the coal yard. The onboard PC uses the SLAM (Simultaneous Localization and Mapping) algorithm to achieve positioning, navigation, and global point cloud mapping. The steps for establishing SLAM are as follows: (Refer to the attached Figure 4 ) 3.1. Coordinate transformation: transform the IMU data and LiDAR point cloud data frame into the site coordinate system, where the LiDAR point cloud data frame is transformed by the IMU data. The site coordinate system takes the site length direction as the X-axis, the site width direction as the Y-axis, and the site height direction as the Z-axis.

[0038] 3.2 Since the laser radar moves during a scan, motion distortion correction is performed on the frame data in each point cloud data frame. A frame data output only contains one time point; in a laser radar scan cycle, the IMU outputs several position data. A point cloud frame data calculates the true coordinates of each scanning point in the frame based on the IMU position data in the frame to achieve motion distortion correction.

[0039] 3.3. Remove useless noise points according to the characteristics of the coal yard, including point clouds below the height of the feature device (h-δ) and other abnormal points. δ is a threshold related to the geometric size of the feature device to avoid the relevant point clouds being accidentally deleted when the feature device changes its posture (such as tilt).

[0040] 3.4. Select the point cloud with a height of (h±δ) in the point cloud frame and perform feature analysis. When the point cloud feature meets the geometric characteristics of the feature device, identify the geometric center point of the feature device.

[0041] 3.5. When the X coordinate value of the IMU data is equal to the X coordinate of the geometric center point of the feature device, the point cloud frame is recorded and set as (i=1,2,…), the corresponding IMU coordinates are =( , , ), the geometric center coordinates of the feature device are =( , , ); 3.6. Using ICP algorithm, (i=1,2,…) for point cloud stitching, where The initial frame becomes the initial map of the flight, and the subsequent point clouds are stitched together to gradually generate a new map; 3.7. During the flight, according to the new attitude data updated by the bucket wheel machine, the no-fly zone in the map is updated, and the flight route is automatically modified to avoid the no-fly zone; 3.8. Based on the map generated in 3.6, get the latest map The geometric center coordinates of the feature device corresponding to the frame =( , , ), and the coordinates recorded in 3.5 generate the observation-estimation residual Δ ; Using Δ Correct IMU coordinates for ; 3.9, use replace Process the IMU data and lidar point cloud frame data in subsequent flights, that is, repeat steps 3.1-3.8.

[0042] When the bucket wheel machine is released in the no-fly zone, fly again and repeat 3.1-3.9 to obtain a new set of data.

[0043] Among them, the fusion algorithm design of laser SLAM (localization and mapping) and IMU (inertial measurement unit); During the flight of the drone, IMU and laser scanning data are obtained simultaneously. The drone is equipped with a lidar sensor and an IMU sensor to fully utilize the high frequency of the IMU sensor and the high precision of the lidar. By adopting the extended Kalman filter and graph optimization loop detection technology, real-time mapping of the closed coal yard is achieved. After the mapping is completed, it is uploaded to the server, and the personnel conduct the final verification of the mapping effect.

[0044] The main process of data processing is as follows: (1) Use IMU (inertial measurement unit) technology to calculate the spatial position of the drone, process the corresponding laser scanning data, and generate a single-frame local map.

[0045] (2) Use ICP analysis to match the local maps of adjacent frames, calculate the position relationship of the acquisition points of adjacent frames, and calibrate with the corresponding IMU data to correct the IMU data.

[0046] (3) After the IMU data is corrected, the local maps of each frame are recalculated and spliced ​​into a global map using ICP technology.

[0047] About navigation and flight obstacle avoidance design: The navigation and flight obstacle avoidance design mainly realizes the autonomous flight and one-key flight function of the drone in indoor environment, without human intervention. It mainly includes the following processes: (1) Using the geometric data of the coal shed structure, a three-dimensional model of the coal shed is constructed.

[0048] (2) Synthesize the SLAM map and the 3D model of the coal shed to generate a 3D spatial model of the coal yard.

[0049] (3) Based on the analysis of available flight space, the flight route file is automatically planned and generated and transmitted to the drone.

[0050] When SLAM mapping of the coal yard is carried out during flight, laser radar is used to perceive mechanical obstacles such as bucket wheel machines on the flight path in real time, and dynamic window local planning technology is used to re-plan the flight route and flight speed of the drone to ensure safe avoidance of obstacles. At the same time, humans can monitor the flight situation through video signals, and use the remote control handle to intervene in time when abnormalities occur to prevent accidents such as collisions.

[0051] Step 4: After the flight, the server receives the drone IMU positioning data and the laser radar point cloud data; the server builds a global 3D model of the coal yard, calculates the coal volume, and outputs a coal counting report. The specific steps are as follows: (Refer to Attachment Figure 5 ) After the flight, the onboard PC uploads 3.2 including point cloud frame data and IMU data to the ground station server using WIFI and LAN communication.

[0052] 4.1. Perform coordinate transformation on the initial IMU and LiDAR point cloud data frames; perform motion distortion correction on the frame data in each point cloud data frame; and remove abnormal points based on the characteristics of the coal yard.

[0053] 4.2. Select the point cloud with a height of (h±δ) in the point cloud frame and perform feature analysis. When the point cloud feature meets the geometric characteristics of the feature device, identify the geometric center point of the feature device.

[0054] 4.3 Assume that the point cloud frame dataset of the geometric center point of the feature device is ={point cloud frame|contains the geometric center point of the i-th feature device}, using the ICP algorithm, (i=1,2,…) stitch point clouds one by one; 4.4. When the X coordinate value of the IMU data is equal to the X coordinate of the geometric center point of the feature device, the point cloud frame is recorded and set as (i=1,2,…), the corresponding IMU coordinates are =( , , ), the geometric center coordinates of the feature device are =( , , ); 4.5. Based on the point cloud set generated in 4.3, obtain The geometric center coordinates of the feature device corresponding to the frame =( , , ), and the coordinates in 4.4 generate the observation-estimation residual Δ , using Δ Correct IMU coordinates for ; 4.6. Utilization , perform motion distortion correction on the subsequent sequence of point cloud data frames, and repeat the process of 4.1-4.5 until all point cloud frames are processed; 4.7. The point cloud set generated at last is processed by XY-axis gridding to build a three-dimensional model and calculate the volume of each coal pile. According to the known stacking density of each coal pile, the coal quantity of each coal pile is calculated.

[0055] 4.8. Output inventory report, including at least the location, volume and weight information of coal pile.

[0056] When there is data obtained from two flights, repeat 4.1-4.6 respectively to generate two point cloud sets, replace the no-fly zone data in the first point cloud set with the corresponding interval point cloud set in the second point cloud set, and then proceed to 4.7 and 4.8.

[0057] After the drone flight is over, the laser scanning data of the lidar is transmitted to the ground server for processing. Using ICP technology, the acquired laser scanning data clusters are merged and spliced, and the TIN algorithm is used to achieve three-dimensional modeling, calculate the volume of each coal pile in the coal yard, and realize the inventory without dead corners in the closed coal yard. In addition, the inventory measurement repeatability error of this system is no more than 0.5%; the inventory process is one-button operation, the drone flies autonomously, and no human intervention is required; the inventory data processing report time is no more than 20 minutes.

[0058] The equipment in the present invention can adopt the following equipment models: Drone: DJI series; IMU: Epson M-G370PDT0 (X2G000221000200) inertial measurement module; LiDAR: Epson M-G370PDT0 (X2G000221000200) inertial measurement module; Drone control device: Pixhawk mini4; Onboard PC: NUC Tiger Canyon (the commercial version of the mini PC in Intel's 11th generation NUC). ‌ This specific embodiment is merely an explanation of the present invention and is not a limitation of the present invention. After reading this specification, those skilled in the art may make non-creative modifications to the present embodiment as needed. However, as long as they are within the scope of the claims of the present invention, they are protected by the patent law.

Claims

1. An automatic coal handling system using drones suitable for air-film enclosed coal yards, characterized by: It includes a drone, a feature device, a server, a remote control device, a WIFI system and a bucket wheel excavator automatic control system. The drone includes a drone body, a control device, an airborne PC, an IMU, and a laser radar. The server and the bucket wheel excavator automatic control system are connected to a local area network via a network cable, the WIFI system is connected to the local area network, the airborne PC has a WIFI module and is connected to the WIFI system via a WIFI network, and the remote control device performs low-speed data communication with the control device via wireless.

2. The unmanned aerial vehicle automatic coal handling system applicable to the air film closed coal yard according to claim 1 is characterized by: The bucket wheel machine fully automatic control system obtains bucket wheel machine positioning data in real time, and the positioning data includes the trolley track position, the cantilever pitch angle, and the cantilever rotation angle.

3. The unmanned aerial vehicle automatic coal handling system applicable to the air film closed coal yard according to claim 1 is characterized by: The server receives the positioning data of the bucket wheel machine's fully automatic control system and sends it to the onboard PC of the drone; and receives the coal yard laser scanning data of the onboard PC of the drone and generates a three-dimensional model of the coal yard, calculates the coal quantity and outputs a coal pan report.

4. The unmanned aerial vehicle automatic coal handling system applicable to the air film closed coal yard according to claim 1 is characterized by: The control device communicates data with the remote control device, receives instructions from the remote control device and the onboard PC, and controls the UAV body to fly; The IMU inertial measurement unit outputs IMU data to an onboard PC; The laser radar outputs laser scanning data to an onboard PC; The onboard PC receives IMU and lidar data, calculates the drone position data, generates a flight map, and automatically outputs the coal yard laser scanning point cloud data via WIFIAP.

5. The unmanned aerial vehicle automatic coal handling system applicable to the air film closed coal yard according to claim 1 is characterized by: The feature device is a structural device with specific geometric features, which is arranged on the coal retaining wall outside the air-film enclosed coal yard and on the coal retaining walls on both sides of the bucket wheel machine track, and is used by the onboard PC to identify feature points using laser scanning point cloud data obtained by the laser radar.

6. An automatic coal handling method using a drone in a gas-film-enclosed coal yard, comprising an automatic coal handling system using a drone in a gas-film-enclosed coal yard as claimed in any one of claims 1 to 5, characterized in that: Step 1: Preparation before coal handling: In the air film coal yard, arrange characteristic devices along the outer side and the coal retaining walls on both sides of the track; calibrate the coordinates of the UAV IMU and laser radar; establish the initial coal yard map; Step 2: The server receives the bucket wheel machine operation information and sets a flight route map to avoid the bucket wheel machine operation area; Step 3: The drone flies and uses the SLAM algorithm to achieve positioning, navigation, and global point cloud mapping; Step 4: After the flight, the server receives the UAV IMU positioning data and lidar point cloud data; the server builds a global three-dimensional model of the coal yard, calculates the coal quantity, and outputs a coal counting report.

7. The method for automatic coal handling by drone in an air-film-enclosed coal yard according to claim 6 is characterized by: In step 1, the flying height of the drone is H, the height of the coal retaining wall is h, the scanning angle of the laser radar is α, the width of the coal pile is W, the distance between the coal pile and the coal retaining wall is l, and the spacing between the characteristic devices satisfies: (1) , where R = H × tanα; (2) , where r = (Hh) × tanα.

8. The method for automatic coal handling by drone in an air-film-enclosed coal yard according to claim 6 is characterized by: The SLAM establishment steps in step 3 are as follows: 3.

1. Coordinate transformation: convert the IMU data and the laser radar point cloud data frame into the site coordinate system. The laser radar point cloud data frame is transformed by the IMU data. The site coordinate system takes the site length direction as the X-axis, the site width direction as the Y-axis, and the site height direction as the Z-axis. 3.

2. Since the laser radar moves during a scanning process, motion distortion correction is performed on the frame data in each point cloud data frame; a frame data output only contains one time point; within a laser radar scanning cycle, the IMU outputs a number of position data; a point cloud frame data calculates the true coordinates of each scanning point in the frame based on the IMU position data in the frame to achieve motion distortion correction; 3.

3. Remove useless noise points according to the characteristics of the coal yard, including point clouds below the height of the characteristic device (h-δ) and other abnormal points; δ is a threshold related to the geometric size of the characteristic device to avoid the relevant point clouds being mistakenly deleted when the characteristic device changes its posture (such as tilt); 3.

4. Select the point cloud with a height of (h±δ) in the point cloud frame and perform feature analysis. When the point cloud feature meets the geometric characteristics of the feature device, identify the geometric center point of the feature device; 3.

5. When the X coordinate value of the IMU data is equal to the X coordinate of the geometric center point of the feature device, the point cloud frame is recorded and set as (i=1,2,…), the corresponding IMU coordinates are =( , , ), the geometric center coordinates of the feature device are =( , , ); 3.

6. Using ICP algorithm, (i=1,2,…) for point cloud stitching, where The initial frame becomes the initial map of the flight, and the subsequent point clouds are stitched together to gradually generate a new map; 3.

7. During the flight, according to the new attitude data updated by the bucket wheel machine, the no-fly zone in the map is updated, and the flight route is automatically modified to avoid the no-fly zone; 3.

8. Based on the map generated in 3.6, get the latest map The geometric center coordinates of the feature device corresponding to the frame =( , , ), and the coordinates recorded in 3.5 generate the observation-estimation residual Δ ; Using Δ Correct IMU coordinates for ; 3.9, use replace Process the IMU data and lidar point cloud frame data in subsequent flights, that is, repeat steps 3.1-3.

8.

9. The method for automatic coal handling by drone in an air-film-enclosed coal yard according to claim 6 is characterized by: The method for explaining the global three-dimensional model of the coal yard in step 4 is as follows: 4.1 Perform coordinate transformation on the initial IMU and LiDAR point cloud data frames; perform motion distortion correction on the frame data in each point cloud data frame; remove abnormal points according to the characteristics of the coal yard; 4.2 Select the point cloud with a height of (h±δ) in the point cloud frame and perform feature analysis. When the point cloud feature meets the geometric characteristics of the feature device, identify the geometric center point of the feature device; 4.3 Assume that the point cloud frame dataset of the geometric center point of the feature device is ={point cloud frame|contains the geometric center point of the i-th feature device}, using the ICP algorithm, (i=1,2,…) stitch point clouds one by one; 4.

4. When the X coordinate value of the IMU data is equal to the X coordinate of the geometric center point of the feature device, the point cloud frame is recorded and set as (i=1,2,…), the corresponding IMU coordinates are =( , , ), the geometric center coordinates of the feature device are =( , , ); 4.

5. Based on the point cloud set generated in 4.3, obtain The geometric center coordinates of the feature device corresponding to the frame =( , , ), and the coordinates in 4.4 generate the observation-estimation residual Δ , using Δ Correct IMU coordinates for ; 4.

6. Utilization , perform motion distortion correction on the subsequent sequence of point cloud data frames, and repeat the process of 4.1-4.5 until all point cloud frames are processed; 4.

7. The point cloud set finally generated is gridded according to the XY axis, a three-dimensional model is constructed, and the volume of each coal pile is calculated.

Images