基于粒子群优化支持向量机的瓦斯突出层研究

Study on Gas Outburst Layer Based on Particle Swarm Optimization Algorithm Support Vector Machine

瓦斯突出是具有很强破坏性的地质灾害,是亟需解决的国际性技术难题。瓦斯突出与构造煤发育息息相关,通过识别构造煤的发育位置,制定针对措施,规避风险。粒子群算法可以将支持向量机参数选择最优化,以支持向量机理论、测井解释和岩心为基础,测井数据作为输入,构建预测模型,对鹤岗煤田煤层进行判别,将结果与取心情况进行对比,平均正确率可以达到90%以上。结果表明,粒子群优化的支持向量机方法具有速度快,精度高,操作简单的优势,将方法运用到构造煤的识别中具有较高的应用前景,为规避瓦斯突出灾害提供了理论依据并指明了方向,对安全生产具有指导性意义。

The gas outburst is a kind of very destructive geological hazard,thus is an international technical hard nut urgently to be solved.It is closely bound up with the tectonoclastic coal development.Targeted measures should be formulated through discrimination of tectonoclastic coal development position to evade risks.The particle swarm algorithm can optimize support vector machine parameters.Based on the support vector machine theory,well logging interpretation and rock cores,through logging data input established prediction model and carried out discrimination for coal seams in the Hegang coalfield.Comparing the discriminated result with the drilling coring,the average accuracy can be above 90%.The result has shown that the particle swarm optimization algorithm support vector machine method has advantages of fast speed,high accuracy and simple operation etc.Applied the method on tectonoclastic coal discrimination has higher application prospect,provided theoretical basis to evade gas outburst risks and pointed out orientation.Thus it has guiding significance in production safety.