球形支持向量机在煤自燃预测中的应用

APPLICATION OF SPHERICAL-SVM IN FORECASTING SPONTANEOUS COMBUSTION OF COAL

引入模糊隶属度和最小二乘思想,采用邻域粗糙集方法对煤自燃预测的输入向量进行维数约简和粒子群优化(PSO)算法优化支持向量机模型中的参数,提出一种模糊最小二乘球形支持向量机(FLHSSVM),并用序贯最小化(SMO)算法求解FLHSSVM中的二次规划问题,建立煤自燃预测模型。实验结果表明,该方法有效简化了训练样本,提高了FLHSSVM训练速度,且分类精度良好,有很好的泛化能力。

In this paper we introduce fuzzy membership and least squares method, adopt neighbourhood rough set method to reduce the dimensions of input vectors of the coal spontaneous combustion, and use particle swarm optimisation （PSO） algorithm to optimise the parameters of support vector machine （SVM） model. Then we present a fuzzy least square spherical support vector machine （ FLHSSVM）, use sequential minimise optimisation （SMO） method to solve the quadratic programming problem in FLHSSVM, and establish a coal spontaneous combustion forecast model. Experimental results show that this method effectively simplifies the training sample, enhances the training speed of FLHSSVM, and has refined classification accuracy, it well proves the generalisation capability.