Research On Structural Health Monitoring Of Large Scale Electric Shovel Bucket

The electric shovel is widely used in open-pit mining.In the mining process,the bucket especially the bucket teeth,which mounted on the front end of the bucket,directly contacts with rock,ore and other materials.The bucket teeth often wear,fracture,and shedding in this situation.The failure of the bucket structure not only increases the frictional force,but also affects the mining efficiency.What's the worse,the scrape of the broken or shedding bucket tooth will be mixed with the ore material and send into the crushing machine.The crushing machine can be damaged as the result of the hard tooth and then lead to the failure of the whole mining-crushing production line.This kind of failure can produce huge economic losses of the society.In order to solve this problem,this paper introduced a method that monitor the status of the electric shovel bucket by using infrared vision system,and combined with the machine vision technology to study the detecting algorithm to detect whether the tooth is broken.Firstly,this article analyzed the structure and common fault types of the shovel bucket.According to the working environment and conditions of the electric shovel,the infrared monitoring system is selected to monitor the shovel bucket.A complete structural health monitoring system is designed according to the requirements of the system.Secondly,study on fault classification of electric shovel bucket.Convolution neural network is one of the algorithms in the field of deep learning,and it has good classification performance.Therefore,this paper using the convolution neural network to classify the faults of the shovel bucket from two aspects.On the one hand,classification according to the type of failure;on the other hand,classification according to the location of the failure.Then,research on the position identification and recognition of bucket tooth.Combining the image matching theory,matching the model image and bucket image by using the SIFT feature matching method to locate the bucket tooth position accurately.According to the theory of image counting,the number of tooth can be counted and the area of each tooth can be calculated.Therefore,it can easy determine the bucket structure is healthy or not.The reason behind that is the the number of the bucket teeth is constant and the area of each bucket teeth is same.Finally,the visual interface of the monitoring system is designed.The visual interface of the bucket structure health monitoring system can be designed based on the MATLAB GUI platform.The functions of this system have lots of applications such as the image reading,the bucket tooth location,the counting of the bucket tooth,the fault diagnosis and the fault alarm.