RS-SVM组合模型下煤矿安全风险预测

发布时间：2018-06-04 04:25

本文选题：粗集 + 遗传算法　；参考：《中国矿业大学学报》2017年02期

【摘要】：针对煤矿安全风险预测的可靠性与效率问题,构建了基于粗集-支持向量机(RS-SVM)的煤矿安全风险预测模型.在不改变样本分类质量的条件下,运用RS方法作为预处理器约简特征参数,然后基于SVM方法进行分类建模.以煤矿瓦斯爆炸风险预测为例,从人员、环境、设备及管理4方面建立煤矿安全生产预警指标体系.利用遗传算法进行RS粗糙集属性约简,将初始风险因子由31项剔除为5项.将得到的约简集作为新的论域,并基于约简集利用SVM进行样本训练,通过回判法对30个训练样本的计算结果进行验证,得出SVM模型精确可靠的结论.进而利用5个预测样本得到其2015年瓦斯爆炸的风险预测结果,与实际情况完全相符.表明本文建立的RS-SVM组合预测模型对煤矿安全风险预测具有良好的指导作用.
[Abstract]:Aiming at the reliability and efficiency of coal mine safety risk prediction, a coal mine safety risk prediction model based on rough set support vector machine (SVM) is established. Without changing the quality of the sample classification, RS method is used as the feature parameter of preprocessor reduction, and then the classification model is built based on SVM method. Taking the coal mine gas explosion risk prediction as an example, the early warning index system of coal mine safety production is established from the aspects of personnel, environment, equipment and management. The attribute reduction of RS rough set is carried out by genetic algorithm, and the initial risk factor is eliminated from 31 to 5. The reduced set is regarded as a new domain, and based on the reduction set, SVM is used to train the samples. The calculation results of 30 training samples are verified by the method of back judgment, and the conclusion that the SVM model is accurate and reliable is obtained. Furthermore, the risk prediction results of the gas explosion in 2015 are obtained by using five predicted samples, which are in good agreement with the actual situation. The results show that the RS-SVM combined forecasting model established in this paper has a good guiding effect on coal mine safety risk prediction.
【作者单位】：中国矿业大学(北京)管理学院;
【基金】：国家自然科学基金项目(61471362)
【分类号】：TD79;TP18

【相似文献】