CN-118807952-B - Slag detection system and method based on video detection device and three-dimensional laser radar

CN118807952BCN 118807952 BCN118807952 BCN 118807952BCN-118807952-B

Abstract

The invention discloses a slag detection system and method based on a video detection device and a three-dimensional laser radar, wherein a camera is used for collecting grid video data, a deep learning algorithm is used for identifying whether slag exists on a grid, if the slag does not exist, a hydraulic shut-off door is not needed for slag crushing, if the slag exists, three-dimensional laser point cloud data of the slag is obtained through the three-dimensional laser radar, the height of the slag is identified through a slag height detection module, whether slag crushing is needed is judged according to the height, if the slag is too high on the grid, the slag crushing is needed, if the slag height is judged, the slag position is judged through a slag position detection module, if the slag is judged to be not needed, the slag crushing is needed, and if the slag is judged to be tiled on the grid, the slag is needed to be crushed, so that the slag is prevented from being tiled on the grid to cause blockage. According to the invention, through detecting the grids, the deposition amount of the slag on the grids is identified, and then the hydraulic shut-off gate is controlled to crush the slag, so that the blockage of the steel belt machine is prevented, and the energy waste is avoided.

Inventors

JIA LU
MENG XIANGFENG
LI ZIHE
LIU XIN
ZHANG YANSONG
DONG JIANWEI
SU QING
LU JINLONG
WEI GUANGLU
TAO YE
SHA HAO
QU JINLONG

Assignees

滨沅国科(秦皇岛)智能科技股份有限公司

Dates

Publication Date: 20260505
Application Date: 20240704

Claims (3)

1. A slag detection method based on a video detection device and a three-dimensional laser radar is characterized by comprising the following steps: S1, collecting grating video data through a camera, identifying whether slag exists on a grating by using a deep learning algorithm, if the slag does not exist, not needing to hydraulically shut off a door to crush the slag, and if the slag exists, executing a step S2; S2, acquiring three-dimensional laser point cloud data of the material slag by a three-dimensional laser radar, further identifying the height of the material slag by a slag height detection module, judging whether slag crushing is needed according to the height, if the slag is too high on the grid, crushing the slag, if the slag height is judged, judging the slag position by a slag position detection module if the slag is found to be unnecessary to crush, and if the condition that the material slag is generated and paved on the grid is judged, crushing the slag is needed, and blocking caused by the fact that the slag is paved on the grid is prevented; The step S2 specifically comprises the following steps that the horizontal x-axis direction interval of two sets of three-dimensional laser radars is as follows The horizontal y-axis direction interval is The heights are respectively And The height of the grating is The distance of the grating in the y-axis direction is compared with the positions of the two sets of three-dimensional laser radars And The distance in the x-axis direction is And ; S21, acquiring and processing three-dimensional laser point cloud data; Firstly, utilizing a three-dimensional laser radar to scan a grating and a carrying space thereof to obtain three-dimensional laser point cloud data of slag, and further obtaining the distance from the three-dimensional laser radar to a slag target Angle of pitch Angle of deviation The horizontal distance between the slag point and the y axis of the three-dimensional laser radar is as follows: The horizontal distance between the slag point and the x axis of the three-dimensional laser radar is as follows: the vertical distance from the slag point to the three-dimensional laser radar is as follows: the data obtained by scanning the slag by the first three-dimensional laser radar is the distance from the radar to the target Angle of pitch Angle of deviation The slag position obtained through calculation by the first three-dimensional laser radar is that the horizontal distance between a slag point and the y axis of the first three-dimensional laser radar is as follows: the horizontal distance between the slag point and the x axis of the first three-dimensional laser radar is as follows: the vertical distance from the slag point to the first three-dimensional laser radar is as follows: The actual height of the slag point is The data obtained by scanning the slag by the second three-dimensional laser radar is the distance from the radar to the target Angle of pitch Angle of deviation The slag position obtained through calculation by the second three-dimensional laser radar is that the horizontal distance between the slag point and the y axis of the second three-dimensional laser radar is as follows: the horizontal distance between the slag point and the x axis of the second three-dimensional laser radar is as follows: the vertical distance from the slag point to the second three-dimensional laser radar is as follows: The actual height of the slag point is ; S22, detecting the height of the slag, judging whether slag crushing is needed, firstly judging the height of the slag, and setting the horizontal x-axis direction threshold of the grid as The grid horizontal y-axis direction threshold is Slag height threshold First three-dimensional laser radar point cloud data in the horizontal x-axis direction Direction of y axis Searching in range, selecting Maximum point The second three-dimensional laser radar point cloud data is at the highest point of the slag, and is in the horizontal x-axis direction Direction of y axis Searching in range, selecting Maximum point For the highest point of slag Or (b) Is greater than The waste residue is required to be crushed, otherwise, the slag is not required to be crushed; S23, detecting the position of the slag, judging whether slag is needed to be crushed according to the position of the slag if the slag is not needed to be crushed by judging the height of the slag, preventing the slag from being tiled on a grid to cause blockage, setting a height threshold value for judging the existence of the slag Selecting point cloud data of a first three-dimensional laser scanner in the horizontal x-axis direction Direction of y axis Searching in a range, searching the height of the slag To find the position interval in the x-axis direction Position interval in y-axis direction Selecting point cloud data of a second three-dimensional laser scanner in the horizontal x-axis direction Direction of y axis Searching in a range, searching the height of the slag To find the position interval in the x-axis direction Position interval in y-axis direction Setting a size threshold of the slag in the x-axis direction Size threshold in y-axis direction Determining the size of the slag according to the position interval of the two three-dimensional laser scanners, wherein the size of the slag in the x-axis direction is The first three-dimensional laser scanner data is calculated to be in the y-axis direction The size in the y-axis direction calculated by the data of the second three-dimensional laser scanner is If (if) Or (b) Or (b) Slag crushing is needed to prevent the slag from being tiled on the grid to cause blockage.
2. The slag detection method according to claim 1, wherein the deep learning algorithm is YoloV algorithm, wherein video data on a grid is continuously collected through a camera, and slag in a video frame is identified in real time by using a deep learning model so as to realize automatic slag detection; (1) The method comprises the steps of data collection and preprocessing, video data acquisition under the working state of a grid, video segmentation into single-frame images, preprocessing of the images and cutting of an interested region; (2) Marking the slag in the single frame image by using an image marking tool to generate label data required by training; (3) Training a model; the YoloV is selected as a basic model, and is adjusted to adapt to the specific requirements of slag identification; (4) The method comprises the steps of model deployment and application, deploying a trained model into an actual production environment, processing video data acquired by a camera in real time, identifying and classifying slag on a grid, and feeding back an identification result.
3. The slag detection system according to any one of claims 1-2, comprising a camera, two three-dimensional laser radars, a slag height detection module and a slag position detection module, wherein the camera, the three-dimensional laser radars, the slag height detection module and the slag position detection module are arranged on an observation window of a slag discharging system, the three-dimensional laser radars are respectively arranged in front of and behind the observation window, grid video data are collected through the camera, a deep learning algorithm YoloV is utilized to identify whether slag exists on a grid, if no slag exists, a hydraulic shut-off door is not needed for slag crushing, if the slag exists, three-dimensional laser point cloud data of the slag is obtained through the three-dimensional laser radars, the slag height detection module is utilized to identify the height of the slag, whether slag crushing is needed according to the height, if the slag is too high on the grid, slag crushing is needed, if the slag position detection module is found to be unnecessary to judge the slag position, if the slag is found to be required to be crushed, the slag crushing is needed to be paved on the grid, and blocking is prevented.

Description

Slag detection system and method based on video detection device and three-dimensional laser radar Technical Field The invention relates to intelligent identification of a slag discharging system, in particular to a slag detecting system and method based on a video detecting device and a three-dimensional laser radar. Background The air-cooled dry slag discharging system blocks slag blocks with larger sizes through the grids, so that the steel belt machine is prevented from being blocked. The slag is crushed and is a common treatment step of an air-cooled dry slag discharging system, and slag blocks with larger sizes on the grids are crushed through a hydraulic shut-off door on the grids, so that the slag falls into a steel belt machine below the grids, and the problems of equipment blockage and accumulation are reduced. At present, the waste residue is crushed mainly by setting a timer and opening and closing the hydraulic closing door at regular time, and in this way, when no residue blocks exist on the grille, the hydraulic closing door can be opened and closed, so that energy waste is caused. Accordingly, the prior art has drawbacks and needs improvement. Disclosure of Invention The invention aims to solve the technical problem of providing a slag detection system and method based on a video detection device and a three-dimensional laser radar aiming at the defects of the prior art. The technical scheme of the invention is as follows: A slag detection method based on a video detection device and a three-dimensional laser radar comprises the following steps: S1, collecting grating video data through a camera, identifying whether slag exists on a grating by using a deep learning algorithm, if the slag does not exist, not needing to hydraulically shut off a door to crush the slag, and if the slag exists, executing a step S2; S2, three-dimensional laser point cloud data of the material slag are obtained through a three-dimensional laser radar, the height of the material slag is further identified through a slag height detection module, whether slag crushing is needed is judged according to the height, if the slag is too high on the grid, slag crushing is needed, if the slag height is judged, the slag position is judged through a slag position detection module if the slag is found to be unnecessary to crush, if the condition that the material slag is spread on the grid is judged, slag crushing is needed, and the blocking caused by the fact that the material slag is spread on the grid is prevented. The step S2 specifically comprises the following steps: The two sets of three-dimensional laser radars are l x in the horizontal x-axis direction, l y in the horizontal y-axis direction, h d1 and h d2 in height and h in grid height respectively, wherein the distances of the grid in the y-axis direction are [ l 1dy1,l1dy2 ] and [ l 2dy1,l2dy2 ] and the distances of the grid in the x-axis direction are l 1dx and l 2dx compared with the positions of the two sets of three-dimensional laser radars; S21, three-dimensional laser point cloud data acquisition and processing Firstly, utilizing a three-dimensional laser radar to scan a grid and a carrying space thereof to obtain three-dimensional laser point cloud data of slag, and further obtaining a distance S i, a pitching angle alpha i and a deviating angle beta i of the three-dimensional laser radar to a slag target; the horizontal distance between the slag point and the y axis of the three-dimensional laser radar is S xi＝Si×cosαi×cosβi; the horizontal distance between the slag point and the x axis of the three-dimensional laser radar is S yi＝Si×cosαi×sinβi; the vertical distance from the slag point to the three-dimensional laser radar is S zi＝Si×sinαi; The data obtained by scanning the slag by the first three-dimensional laser radar is that the distance S 1i from the radar to the target, the pitching angle alpha 1i and the deviating angle beta 1i; the slag position obtained through calculation by the first three-dimensional laser radar is as follows: the horizontal distance between the slag point and the y axis of the first three-dimensional laser radar is S 1xi＝S1i×cosα1i×cosβ1i; The horizontal distance between the slag point and the x axis of the first three-dimensional laser radar is S 1yi＝S1i×cosα1i×sinβ1i; The vertical distance from the slag point to the first three-dimensional laser radar is S 1zi＝S1i×sinα1i; The actual height of the slag point is h q1＝S1zi+hd1; the data obtained by scanning the slag by the second three-dimensional laser radar is that the distance S 2i from the radar to the target, the pitching angle alpha 2i and the deviating angle beta 2i; the slag position obtained through calculation by the second three-dimensional laser radar is as follows: the horizontal distance between the slag point and the y axis of the second three-dimensional laser radar is S 2xi＝S2i×cosα2i×cosβ2i; The horizontal distance between the slag point and the x axis of the se