摘要
煤矿工作面瓦斯浓度时间序列预测是一个非线性、模糊数据处理问题。为了避免因具体预测值的不准确性导致的错误判断,提出了基于时间的模糊多项式预测方法。在建模过程中先将原始数据模糊化,再将模型模糊系数转化成线性规划求解。与其他预测模型对比分析,该方法的预测结果不是具体数值而是一个区间,有效地扩大了相关量的适用范围,使预测结果更趋于合理、科学和便于操作。
Time series prediction of gas concentration of coal working face is a nonlinear,fuzzy data processing problem. In order to avoid the misjudgment caused by the inaccuracy of the prediction value of the specific result,a fuzzy polynomial prediction based on time is proposed,the prediction results of this method is a specific value,rather than an interval,effectively expanding the scope of the relevant quantity,make predictions become more reasonable,scientific and convenient operation. In the process of modeling,the raw data is fuzzy,the model of fuzzy coefficient is transformed into solving linear programming. Through the comparative analysis,the model is better than the other forecasting model.
出处
《现代矿业》
CAS
2015年第1期21-24,共4页
Modern Mining
基金
甘肃省科技厅项目:石油化工企业应急演练系统(编号:1204GKCA004)
甘肃省财政厅专项资金立项资助(编号:甘财教[2013]116号)
关键词
瓦斯浓度
时间序列
三角模糊数
模糊多项式
预测区间
Gas concentration
Time series
Triangular fuzzy numbers
Fuzzy polynomia
Prediction interval
