基于BP神经网络的含氧油气浓度预测模型研究

Research on the Prediction Model of Oxygen-bearing Oil and Gas Concentration Based on BP Neural Network

针对储罐清洗中油气遇到明火、静电等可能会引发火灾、爆炸等现象,提出一种基于BP神经网络的含氧油气浓度预测模型,通过预测油罐内含氧油气浓度以确保储罐清洗工作环境的安全。利用Matlab软件和仿真数据建立了基于BP神经网络的油气浓度预测模型,通过LabVIEW软件中Matlab script节点调用预测模型,自主完成储油罐内油气浓度的预测并进行误差分析。结果表明,基于BP神经网络建立的含氧油气浓度预测模型,当隐藏层节点数为8时,均方根误差(RMSE)为0.000058,回归系数(R^(2))为99.314%,能够准确地预测储罐内含氧油气浓度。

In order to ensure the safety of the tank cleaning work environment by predicting the concentration of oxygenated oil and gas in the tank,a BP neural network-based prediction model is proposed for the fire and explosion that may be caused by oil and gas encountering open flame and static electricity in tank cleaning.The prediction model of oil and gas concentration based on BP neural network is established by using Matlab software and simulation data.The prediction model is called by Matlab script node in LabVIEW software,and the prediction of oil and gas concentration in the storage tank is completed independently with error analysis being performed.Results show that the oxygenated oil and gas concentration in the storage tank can be accurately predicted by the oxygenated oil and gas concentration prediction model based on BP neural network,with the number of hidden layer nodes of 8,a root mean square error(RMSE)of 0.000058,and a regression coefficient(R^(2))of 99.314%.