粗糙集与支持向量机在采空区自然发火预测中的应用研究

Research on Application of Rough Set and Support Vector Machine in Prediction of Spontaneous Combustion in Caving Zone

针对煤矿采空区煤自燃的特点,提出一种基于粗糙集与支持向量机的预测模型,用于采空区自然发火的预测;在支持向量机对原始样本数据进行处理之前,用粗糙集作为前端预处理器对数据进行约简,剔除冗余信息,简化样本数据空间的维数,从而加快支持向量机的训练速度,提高模型的预测精度;实验结果表明,该方法预测准确,为采空区自然发火的防治提供了一个新思路。

Aiming at the characteristics of spontaneous combustion in ceving zone, a prediction model based on rough set and support vec tor machine is proposed for the prediction of spontaneous combustion in caving zone. Rough set is set as a front-end processor to remove re-dundant information and reduce the dimensionality of data before that the original sample data is processed by support vector machine , in or-der to speed up the training rate of support vector machine and improve the prediction precision of prediction model. The experimental results show, the method has achieved good effect.