关键词: 数字孪生, 煤矿, 综掘面, 智能调控

Abstract: The air duct air outlet parameters of fully mechanized excavation face cannot dynamically adjust the air flow state with the change of gas and dust concentration,and cannot realize the interconnection between the complex and harsh mine physical operation environment and the virtual model of intelligent coal mine development.For this reason,an intelligent control method for air flow in fully mechanized excavation face based on digital twin technology was proposed,and a dynamic virtual model of the air flow control device was established.The acquisition of air outlet parameter regulation rules,the gas dust concentration prediction based on Long Short-Term Memory (LSTM) network,the air outlet parameter decision that support virtual models to realize dynamic early warning and prediction based on Genetic Algorithm-Back Propagation (GA-BP) neural network were studied,and the key technologies of autonomous learning and decision-making functions were studied.The experiment showed that the digital twin system could reconstruct the working environment of the mine roadway with high fidelity,mix virtual model and actual reality,realize a new control mode of air flow in fully mechanized excavation face,effectively optimize the gas and dust distribution and finally improve the safety and intelligence of coal mining.

Key words: digital twin, coal mine, fully-mechanized caving face, intelligent control