Coal mine integrated management platform, internet of things, selforganizing map, Reinforcement learning, Data acquisition Abstract: To address the challenges of data acquisition in complex coal mine environments characterized by multiple edge devices and isolated sensor networks, this paper proposes a comprehensive coal mine management platform design scheme based on the Internet of Things (IoT) and SelfOrganizing Map (SOM) neural networks. The scheme constructs a twolayer collaborative optimization framework named TLO-QSOM. It dynamically selects sink nodes through upper-layer reinforcement learning (RL) and plans inspection paths via a lower-layer SOM neural network, achieving the synergistic goals of minimizing information age and optimizing system energy efficiency. Finally, a simulation environment containing 50-200 heterogeneous sensor nodes was constructed, simulating a 72-hour continuous operation scenario in an underground coal mine. Experimental results demonstrate that the TLOQSOM framework exhibits outstanding performance in key metrics: the gas over-limit response time is only 1.6 seconds; equipment fault localization accuracy reaches 96.8%; the multi-source data fusion consistency coefficient is 0.94; and the system's continuous operation time reaches 132 hours. Ablation studies further verified that the full TLO-QSOM version shows significant advantages across all evaluated metrics. https://mc.manuscriptcentral.com/jcmse Journal of Computational Methods in Science and Engineering For Peer Review Page 1 of 13 https://mc.manuscriptcentral.com/jcmse Journal of Computational Methods in Science and Engineering 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For Peer Review IoT-Enabled Integrated Coal Mine Management Platform Using Self-Organizing Map Neural Networks Abstract: To address the challenges of data acquisition in complex coal mine environments characterized by multiple edge devices and isolated sensor networks, this paper proposes a comprehensive coal mine management platform design scheme based on the Internet of Things (IoT) and Self-Organizing Map (SOM) neural networks. The scheme constructs a two-layer collaborative optimization framework named TLO-QSOM. It dynamically selects sink nodes through upper-layer reinforcement learning (RL) and plans inspection paths via a lower-layer SOM neural network, achieving the synergistic goals of minimizing information age and optimizing system energy efficiency. Finally, a simulation environment containing 50-200 heterogeneous sensor nodes was constructed, simulating a 72-hour continuous operation scenario in an underground coal mine. Experimental results demonstrate that the TLO-QSOM framework exhibits outstanding performance in key metrics: the gas over-limit response time is only 1.6 seconds; equipment fault localization accuracy reaches 96.8%; the multi-source data fusion consistency coefficient is 0.94; and the system's continuous operation time reaches 132 hours. Ablation studies further verified that the full TLO-QSOM version shows significant advantages across all evaluated metrics. Keywords: Coal mine integrated management platform; internet of things; selforganizing map; Reinforcement learning; Data acquisition 1.