No abstract

As mining depth gradually increases, the complex and changeable behavior state of mine production systems leads to the increasingly prominent problem of ''planning is difficult to control''. The safety production situation and the difficulty of emergency management are gradually upgraded. However, the theoretical study on the behavioral trend of complex mine production systems is still an immature field. This paper proposes the scientific problem of ''theory and method of time-varying computational experiments for fully mechanized mining processes in artificial system environments''. Taking the typical fully mechanized mining process in the Yushen mining area in northern Shaanxi, China, as the research object, through computer modeling, simulation and use of multiagent system theory, APSM (Agent Publish-Subscribe Model) coordination technology, multiagent cross-emergence and multilayer learning networks, the artificial fully mechanized mining system modeling, sequential mining process deduction and state transfer theory are systematically studied. First, an artificial system model equivalent to the function of the actual fully mechanized mining system is constructed. Then, under the artificial system environment, the time-varying computational experiments of the fully mechanized mining process are realized through the autonomous deduction of the fully mechanized mining agent based on a multilayer neural network and the emergence of multiagent interactions based on subscription perception; this approach aims to solve the problem of determining the overall behavior trend of the mine under the condition of ''long time and large space'' and to provide intellectual support and scientific basis for the ''first experiment and then produce'' technological model of intelligent mining. INDEX TERMS Mine, fully mechanized mining process, artificial system, time-varying computational experiments.

In order to solve the problems of a slow solving speed and easily falling into the local optimization of an ore-blending process model (of polymetallic multiobjective open-pit mines), an efficient ore-blending scheduling optimization method based on multiagent deep reinforcement learning is proposed. Firstly, according to the actual production situation of the mine, the optimal control model for ore blending was established with the goal of minimizing deviations in ore grade and lithology. Secondly, the open-pit ore-matching problem was transformed into a partially observable Markov decision process, and the ore supply strategy was continuously optimized according to the feedback of the environmental indicators to obtain the optimal decision-making sequence. Thirdly, a multiagent deep reinforcement learning algorithm was introduced, which was trained continuously and modeled the environment to obtain the optimal strategy. Finally, taking a large open-pit metal mine as an example, the trained multiagent depth reinforcement learning algorithm model was verified via experiments, with the optimal training model displayed on the graphical interface. The experimental results show that the ore-blending optimization model constructed is more in line with the actual production requirements of a mine. When compared with the traditional multiobjective optimization algorithm, the efficiency and accuracy of the solution have been greatly improved, and the calculation results can be obtained in real-time.