煤与瓦斯突出强度的FOA-SVM预测模型与应用

Coal and Gas Outburst Intensity Prediction of FOA-SVM Model and Application

为了能够对煤与瓦斯突出进行准确的辨识,本文提出将果蝇算法(FOA)与支持向量机(SVM)相结合的预测方法。首先通过Karhunen-Loève变换(K-L变换)进行特征提取,降低特征向量的维数,减小运算量;然后将经过K-L变换的样本作为FOA-SVM模型输入,通过果蝇算法全局寻优,自动搜索符合本预测模型最佳参数组合。通过对预测模型的训练与仿真表明:本文提出的方法具有设计实现简单,辨识精度高、推广能力强的特点,为煤矿灾害预测提供理论支持。

In order to accurately identify the coal and gas outburst,this paper proposes a prediction method based on the combination of the fruit fly optimization algorithm（FOA）and the support vector machine（SVM）.Firstly,through Karhunen-Loève transform（K-L transform）to extract features,this can reduce the dimension of feature vectors and the amount of computation;the samples through K-L transform are the FOA-SVM model＇s input,by FOA globally searching the best parameter combination for the prediction model. The training and simulation results show that the proposed method has the characteristics of simple design,high accuracy and strong generalization ability,which provides theoretical support for the coal mine disaster prediction.