An Unmanned Intelligent Transportation Scheduling System for Open-pit Mine Vehicles Based on 5G and Big Data

Sai Zhang, Caiwu Lu, Song Jiang, Shan Lu, Neal N. Xiong
2020 IEEE Access  
situation assessment of multisource heterogeneous flow data of rock failure in underground metal mines"; 51774228, title: "Research on 5D refined mining production scheduling model and collaborative optimization method in metal open pit under constraints of Grade-Price-Cost"; 51864046, title: "Experimental theory and method of timevarying calculation for fully mechanized mining process under artificial system environment in Yushen mining area"), and the Natural Science Foundation of Shaanxi
more » ... ince (2019JLP-16, title: "Research and development of key technologies for intelligent production control and intelligent decision-making of open pit coal mine under cloud service"; 2020JC-44, title: "Integrated intelligent scheduling model of driverless multi vehicle cooperation in metal open pit under time and space road conditions"; 2019JM-492, title: "Research on deformation perception and situation prediction of open pit slope based on depth image recognition"). ABSTRACT With the maturity of the Internet of Things, 5G communication, big data and artificial intelligence technologies, open-pit mine intelligent transportation systems based on unmanned vehicles has become a trend in smart mine construction. Traditional open-pit mine transportation systems rely on human power for command, which often causes vehicle delay and congestion. The operation of unmanned vehicles in an open pit mine relies on many sensors. Using big data from the sensors, we optimize vehicle paths and build an efficient intelligent transportation system. Based on large amounts of data, such as unmanned vehicle GPS data, vehicle equipment information, production plan data, etc., with the goal of reducing vehicle transportation costs, total unmanned vehicle delay time, and ore content fluctuation rate, a multiobjective intelligent scheduling model for open-pit mine unmanned vehicles was established, and it is aligned with actual open pit mine production. Next, we use artificial intelligence algorithms to solve the scheduling problem. To improve the convergence, distribution and diversity of the classical fast nondominated sorting genetic algorithm (NSGA-II) to solve constrained high-dimensional multi-objective problems, we propose a decomposition-based constrained dominance principle genetic algorithm (DBCDP-NSGA-II), retaining feasible and non-feasible solutions in sparse areas, and compare it with four other commonly-used multi-objective optimization algorithms. Simulation analysis shows our algorithm provides the best overall performance results of the multi-objective models. Furthermore, we apply intelligent scheduling models and optimization algorithms to mining practice and obtain new truck operation routes and schedules, reducing truck operation costs by 18.2%, truck waiting time by 55.5%, and ore content fluctuation by 40.3%. For open-pit mine unmanned transportation, the approach provides a variety of optimized solutions for minimum transportation costs, minimum waiting time, minimum ore content fluctuation rate, and a balance of the three indicators. Through an artificial intelligence algorithm, this study realizes intelligent unmanned vehicle path planning and improves the operation efficiency of open-pit mine intelligent transportation systems.
doi:10.1109/access.2020.3011109 fatcat:ggkveo2ervbb7ejt4atiandslm