基于ARIMA时间序列的瓦斯浓度预测研究

Prediction of Gas Concentration based on ARIMA Time Series

煤矿瓦斯浓度预测是避免煤矿瓦斯爆炸的有效手段,针对煤矿瓦斯浓度预测精度问题,提出一种基于ARIMA时间序列的瓦斯浓度预测模型。该模型首先通过差分法将原始瓦斯浓度数据处理为平稳性数据;然后利用ACF和PACF方法确定模型阶数,依据白噪声检验对模型进行残差分析;最终构建出对煤矿瓦斯浓度进行短期预测的ARIMA模型。通过实验结果表明,该模型实现了瓦斯浓度预测的可视化,并与SVR模型下的瓦斯浓度预测结果相比,其平均绝对误差比SVR模型平均绝对误差降低了13.27%,可为煤矿生产中瓦斯浓度预报提供良好的依据。

The prediction of gas concentration in coal mine is an effective means to avoid gas explosion in coal mine.Aiming at the prediction accuracy of gas concentration in coal mine,a prediction model of gas concentration based on ARIMA time series is proposed.Firstly,the original gas concentration data is processed into stationary data by difference method.Then ACF and PACF methods are used to determine the order of the model,and residual analysis of the model is carried out according to white noise test.Finally,the ARIMA model for short-term prediction of coal mine gas concentration is constructed.The experimental results show that the model can realize the visualization of gas concentration prediction,and the average absolute error of the model is 13.27%lower than that of SVR model,which can provide a good basis for gas concentration prediction in coal mine production.