基于长短期记忆神经网络的燃气数据分析设计

Gas Data Analysis Based on LSTM Neural Network

目前城市天然气发展已经趋于成熟,不仅基础设施得以完善,而且燃气行业的数字化、信息化程度也在不断提高。针对当前燃气行业中数据应用存在的不足,以数据为中心,设计了集数据采集、传输、存储、查询、分析、预测及展示于一体的基于长短期记忆神经网络的燃气压力预警系统,使用长短期记忆神经网络对数据库中的数据进行分析;同时,通过分析预测未来数据能够对未来的事件做出预判,从而可以对可能发生的安全事故采取相应的事前预防措施。

At present,the degree of digitalization and informatization of gas industry is constantly improving.In view of the shortcomings of data application in the gas industry,this paper designs a gas pressure early warning system based on LSTM neural network,which integrates data collection,transmission,storage,query,analysis,prediction and display,and analyzes the data.This means that through the analysis and prediction of future data,we can make a prediction of future events,and take corresponding preventive measures for possible accidents.