Research On Visual Tracking Technology For Condition Monitoring And Fault Alarm Of Material Conveying Process

With the increasing level of industrial automation, material conveying system is becoming more and more automatic and intelligent, it has been widely used. But there are also some faults in the process of conveying, it will directly affect the automated production beats, reducing the production efficiency, causing economic losses if the fault is not timely solved. Therefore, we should not only maintain the material conveying equipment regularly, but also monitor the state of the process of material conveying and warn the fault in time. The traditional way of condition monitoring is internal sensors and manual inspection, but the information of internal sensors is insufficient and manual inspection is boring and non-persistent, resulting the bad monitoring. While the condition monitoring based on visual is non-contact, continue working and its effect is well. In this paper, we will visual track the moving material, build the condition monitoring and fault alarm system of the material conveying process to achieve a better condition monitoring.This article analyzes the visual tracking process of the material conveying and the primary mission of each key section, studies the movement characteristics of moving material and the environmental conditions in the process of material conveying and compares the commonly used algorithms of moving object detection. For the practical application of material conveying, this article proposes an integrated algorithm which combines three frames difference based on the Canny operator edge detection with the background subtraction, and uses mathematical morphology and area threshold for dealing with the result of moving material detection. At last, the experiment proved that my method achieved a good detection effect.In order to improve the real-time performance of tracking, combining with the movement characteristics of moving material, this article uses Kalman Filter to predict and track the moving material. We use centroid method and the minimum bounding rectangle to locate and track the moving material for improving the accuracy. For the sake of meeting the practical application, the multiple moving materials tracking method is studied. Through experiments, we realized the quickly and efficiently predictive tracking for both single moving material and multiple moving materials.In order to get the more smooth moving trajectory,this article makes use of cubic spline interpolation cardinal function to fit the moving trajectory of the moving material, analyzes the moving condition of the moving material and summarizes the fault types which maybe occur in the process of the material conveying. We put forward the method of fault alarm based on deviating from the threshold and elaborate the alarm principles of different fault types, laying a theoretical foundation for fault alarm.Finally, we build the experimental platform, establish the condition monitoring and fault alarm system of material conveying process, conduct the experiments which monitor the condition of the moving material in real time and the experiments of fault alarm by simulating each fault type artificially. Eventually, the experiments prove that this system can well realize the condition monitoring of the moving material and alarm timely when the fault occur.