基于Encoder-Decoder-ILSTM 模型的瓦斯浓度预测研究
陈小建
（山西新景矿煤业有限公司，山西 阳泉 045000）
摘要： 近年来，神经网络在各领域均发挥了巨大作用，同样在煤矿瓦斯浓度预测当中也有应用。为了提高模型的预
测精度和实时性，结合Encoder-Decoder 结构、长短期记忆形成、蛇优化算法提出了一种新的神经网络，为促进煤矿安
全生产提供了技术支持。
关键词： 神经网络；Encoder-Decoder；蛇优化算法；瓦斯浓度预测
中图分类号： TD712⁺.3      文献标志码： A      文章编号： 2095-0802-(2023)12-0102-04
Gas Concentration Prediction Based on Encoder-Decoder-ILSTM Modeling
CHEN Xiaojian
(Shanxi Xinjing Mining and Coal Industry Co., Ltd., Yangquan 045000, Shanxi, China)
Abstract: In recent years, neural networks have played a great role in various fields, so they are also applied to gas concentration
prediction. In order to improve the prediction accuracy and real-time performance of the model, a new neural network was
proposed by combining the Encoder-Decoder structure, long and short-term memory formation, and snake optimizer to provide
technical support for promoting coal mine safety production.

Key words: neural network; Encoder-Decoder; snake optimizer; gas concentration prediction