加权时间序列预测模型及其在矿业安全系统中的应用

依据安全系统的特征属性,探讨了典型的指数时间序列预测模型加权参数的估计方法,并构建了加权指数时间序列预测模型。引入修正指数权重函数,通过其特征参数调整,能较好的体现研究系统时间序列的时间属性。对我国矿业安全事故数序列进行了加权预测研究,通过比较常规预测模型和加权预测模型的预测残差平方和,表明加权预测模型可较大幅度的降低预测模型近期的预测残差,从而保证了预测的可靠性。加权时间序列预测模型具有较好的系统适应性,可推广应用于其他各类时间序列的系统预测。

Weighted time series prediction model and its application in mining safety system

Based on safety system's attribution,the research has discussed the measures of estimating weighted parameters on the typical model of exponential time series prediction,and built a exponential weighted time series predicting model.Adjusting the parameters of revised exponential weighted function can reflect the time effect exactly in time series of the researched system.Comparing with the traditional prediction model,the short-term residual error of the weighted predicting model is lower.It proves that the...