Design Of Volume Monitoring System For Coal Stacking In Transmission Belt Based On Vision

Conveyor belt of coal accident is in coal-fired power plant coal conveying equipment in the process of running one of the most common fault,if can detect in the conveyor belt of coal was carried out on the pile of coal volume estimated at the same time,the volume is too big,can prevent the accumulation of coal cause transmission belt skid,tear,and prevent the happening of the transmission belt of coal accidents,ensure the safe operation of power system,the existing pile of coal inspection robot sensors,explosion-proof camera monitoring and intelligent intelligent detection error rate is high,the deficiencies and the difficulty to estimate the pile of coal volume,etc.In view of the disadvantages existing in the detection of coal stacking accident in the transmission belt above,this paper proposes two design schemes of the volume monitoring system of coal stacking in the transmission belt based on vision based on the intelligent inspection robot of the coal-carrying belt in orbit.Scheme 1 uses the Yolov4 transmission belt coal stacking volume monitoring system,while Scheme 2 uses Yolact transmission belt coal stacking volume monitoring system.In both schemes,a single camera is installed on the inspection robot for inspection at a fixed speed.When coal stacking occurs in the transmission belt,image data is saved at a fixed time interval and transmitted to the upper computer.Then,the volume of coal stacking is obtained by obtaining three-dimensional coordinate points of coal stacking through SGBM binocular vision algorithm.In order to improve the working efficiency of the whole system,Yolov4 and Yolact algorithms are implemented on the embedded system to detect the coal stacking in the transmission belt.By building an experimental system,the influence of the two detection algorithms on the working efficiency and accuracy of the coal stacking volume measurement in the system is studied and the feasibility of the two system schemes is verified.The experimental results show that both of the two schemes can detect the coal pile in time and estimate the volume of coal pile effectively,thus preventing the coal pile accidents in the transmission belt.In Scheme 1,the system works more efficiently because the detection speed of coal pile is faster.Scheme 2 has higher precision because of more accurate edge extraction of coal pile.