基于变精度粗糙集的煤与瓦斯突出预测  

Prediction of Coal and GasOutburst Based on Variable Precision Rough Set

作　　者：臧大进[1,2] 王耀才[2] 刘增良[1] 

机构地区：[1]铜陵学院电气工程系,安徽铜陵244000 [2]中国矿业大学,江苏徐州221008

出　　处：《控制工程》2012年第1期13-16,47,共5页Control Engineering of China

基　　金：安徽省自然资金立项资助(90416246)

摘　　要：煤与瓦斯突出是受诸多因素影响的复杂问题,为了提高其预测的准确性,本文以变精度粗糙集理论中的决策表为主要工具,首先将影响煤与瓦斯突出的检测信号作为预测的条件属性集,煤与瓦斯突出量作为对预测的决策属性,建立决策表,然后利用小生境遗传算法适合于进行多峰值函数优化的特点,提出了一种基于小生境遗传算法的粗糙集属性约简方法,用于求解决策表的多个约简,进而进行值约简后抽取出预测规则。算例结果说明了本算法的正确性和可行性。The coal and gas outburst is a complicated problem since plenty of the cases influence it. In order to enhance its predictable veracity, we propose a a reduction approach in rough set based on Niche Genetic Algorithm. At first the decision table is set up by using coal and gas outburst detection signal as the condition attribute set of the prediction, and coal and gas outburst quantity as decision attribute set of the prediction. At last, applying Niche Genetic Algorithm to solve Muhimodal Function Optimizationtake, this approach can be used to solve several reductions in decision table, carry on value reduction and extract the prediction rule. The example results prove the accuracy and feasibility of this algorithm.

关 键 词：变精度粗糙集 煤与瓦斯突出 预测 遗传算法 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]