基于PSO-SVM的矿井输送带火灾危险程度预测模型研究

Prediction Model of Mine Tape Fire Hazard Based on PSO-SVM

矿井带式输送机输送带火灾严重威胁煤矿安全生产。矿井输送带火灾风险监测预警方面存在指标内在关联性差、火灾不同发展阶段的标志性气体和特征温度不明确、火灾风险预警指标缺失等难题。因此,采用热重红外联用仪和锥型量热仪装置,研究了矿井输送带火灾热解燃烧的特性,热解初始温度段,首先生成较多的CO_(2)、H2O以及少量CO、HCl,温度升至252℃时,HCl生成量开始迅速增加,298℃时HCl的生成量达到峰值1.26%;CO在416℃时生成量开始迅速增加,485℃时CO的生成量达到峰值0.29%;CO_(2)在整个热解过程中产生量最大,总结出CO、CO_(2)、HCl作为输送带火灾监测预警指标。建立了基于粒子群算法优化支持向量机的模型,并对模型进行了最优参数与效果的研究分析。为了验证模型的准确性和可靠性,采用最小二乘误差指标将PSOSVM与SVM预测结果进行对比。

The coal mine belt conveyor belt fire seriously threatens the safety of coal mine production.There are problems in the monitoring and warning of coal mine belt fire risks,such as poor internal correlation of indicators,unclear landmark gases and characteristic temperatures at different stages of fire development,and missing fire risk warning indicators.Therefore,the characteristics of coal mine belt fire pyrolysis and combustion were studied using a thermogravimetric infrared spectrometer and a cone calorimeter device.In the initial temperature range of pyrolysis,a large amount of CO_(2),H2O,and a small amount of CO and HCl were first generated.When the temperature rose to 252℃,the amount of HCl generated began to rapidly increase,and at 298℃,the peak amount of HCl generated reached 1.26%;The production of CO began to rapidly increase at 416℃,and reached a peak of 0.29%at 485℃;CO_(2) is generated the most during the entire pyrolysis process,and CO,CO_(2),and HCl are summarized as indicators for monitoring and warning of tape fires.A model based on particle swarm optimization for support vector machine optimization was established,and the optimal parameters and effectiveness of the model were studied and analyzed.In order to verify the accuracy and reliability of the model,the least squares error index was used to compare the prediction results of PSO-SVM and SVM.