论文部分内容阅读
为了提高煤矿瓦斯爆炸灾害风险识别能力,提出通过典型样本和模式识别(PR)模型构建识别库进行风险识别的方法。在以径向基函数(RBF)为核函数的支持向量机(SVM)模型基础上,应用粒子群优化算法(PSO)优化模型参数,建立瓦斯爆炸灾害风险识别的PSO-SVM模型;构建风险识别指标体系,确定风险模式类别,并以指标风险类别分界点为基础,提出新的数据规范方法;用典型样本训练和测试PSO-SVM模型,样本识别率为100%,表明以典型样本和PSO-SVM模型构建的识别库对瓦斯爆炸灾害风险有较强的识别能力,同时指出典型样本库应不断补充完善以增强其适应能力。
In order to improve the ability of gas explosion disaster risk identification in coal mine, a method of risk identification based on typical samples and pattern recognition (PR) model is proposed. Based on Support Vector Machine (SVM) model with radial basis function (RBF) as kernel function, particle swarm optimization (PSO) is used to optimize model parameters to establish PSO-SVM model of gas explosion disaster risk identification. Index system to determine the risk model category, and based on the index risk category cut-off point, a new data normalization method is proposed; the typical sample training and testing PSO-SVM model, the sample identification rate of 100%, indicating that the typical sample and PSO- The recognition library built by SVM model has a strong ability to identify the risk of gas explosion disasters. At the same time, it is pointed out that the typical sample library should be constantly supplemented to enhance its adaptability.