Application of Human-Machine Collaboration Algorithm for Mine Pile Weight Estimation Based on Beidou High-Precision Location Service

摘要

The terminal yard is a site used to stack bulk mineral materials. When users need, they need to estimate the weight of the mineral material pile before loading. If there is too much or insufficient mineral material, short barge trucks need to be used for stacking, allocating, and clearing to increase the field capacity and transfer efficiency of the yard. Accurate pile quality estimation can effectively reduce the use times of short-distance trucks and reduce yard operating costs. Unlike traditional yard weight estimation, which requires accurate measurement of the volume of the pile, we use comprehensive processing technology based on Beidou high-precision positioning and UAV (Unmanned Aerial Vehicle) shooting high-precision image, combined with open source software (OpenDroneMap) to quickly obtain the rough volume. Then a SVM (Support Vector Machine) high-precision quality estimation model with 8 variables including volume will be established by manually inputting 7 variable factors of pile proportion, moisture content, ore pile height, iron ore form, month, and stacking time. Through actual data verification and analysis, the estimation error is less than 4.5%. Compared with the 20% error of manual experience, it greatly reduces the labor burden and the management cost caused by the inaccurate estimation. Besides, it also improves the management level of the material yard estimation.