一种新型煤灰分双能量γ射线检测方法  

Dual-energy γ-ray determination of ash in coal based on chaos least squares support vector machines

作　　者：程栋[1] 滕召胜[1] 黎福海[1] 代扬[1] 

机构地区：[1]湖南大学电气与信息工程学院,湖南长沙410082

出　　处：《中南大学学报（自然科学版）》2014年第5期1510-1515,共6页Journal of Central South University:Science and Technology

基　　金：国家科技支撑计划项目(2012BAJ24B00)

摘　　要：针对传统方法对煤灰分检测误差大的问题,提出基于混沌最小二乘支持向量机(chaos-LSSVM)的煤灰分双能量γ射线检测方法。其中,双能量γ射线透射法可减小煤炭形状、厚度、粒度、堆密度等因素引入的检测误差,而最小二乘支持向量机算法可减小标定误差,混沌算法可优化最小二乘支持向量机计算进程中惩罚系数g和核函数宽度参数δ。通过241Am和137Cs作为低能和中能γ射线源煤灰分的实验验证,Chao-LSSVM检测方法灰分平均相对误差可达到0.80%,与传统标定方法(直线逼近、最小二乘逼近方法)煤灰分检测的平均相对误差2.22%和3.19%相比,本文提出的方法具有优化的煤灰分检测准确度。To decrease the error occurring in the ash determination of coal using traditional technologies, a new method was proposed in which dual-energy γ-ray and chaos least squares support vector machine（chaos-LSSVM） were applied. The error induced by curve calibration and the shape, thickness, particle size and bulk density of coal can be respectively decreased by chaos-LSSVM and dual-energy γ-ray. The efficiency of the method was examined by the experimental system with 241Am and 137Cs as the resource of low energy and medium energy γ-ray. The results show that the average relative error of ash determination in coal is 0.80% for the method, while there are respectively 2.22% and 3.19% for line-approaching method and the least squares approaching method, which shows that the proposed method has excellent performance.

关 键 词：煤灰分检测 混沌最小二乘支持向量机 双能量γ射线 

分 类 号：TH83[机械工程—仪器科学与技术]