基于EEMD和AGA-SVM的截齿损耗诊断研究

Research on Pick Loss Diagnosis Based on EEMD and AGA-SVM

针对传统方法不易在线准确识别采掘设备截齿工作过程中的损耗状态问题,提出一种基于集合经验模态分解(EEMD)和自适应遗传算法-支持向量机(AGA-SVM)相结合的采掘进机械截齿损耗程度的诊断方法。首先,利用EEMD对截齿不同磨损程度状态下的振动及声发射信号展开分解,获得内禀模态函数(IMF);然后,将IMF分量作为特征向量输入IAGA-SVM诊断器;最后,优化核函数的参数及惩罚系数,并用所提模型对特征向量进行分类。结果表明,该方法可精准诊断采煤机截齿损耗程度状态,与SVM、GA-SVM相比,其有更优越的时效性和准确度。

Aiming at the difficulty of traditional methods to accurately identify the loss status of mining equipment picks in the working process,a coal mining based on the combination of ensemble empirical mode decomposition(EEMD)and adaptive genetic algorithm optimization support vector machine(AGA-SVM)is proposed Diagnosis method of pick loss degree of machine and roadheader.First,using EEMD to decompose the vibration and acoustic emission signals of the pick under different wear conditions to obtain the intrinsic mode function(IMF),and then input the IMF component as a feature vector into the AGA-SVM diagnostic device.Finally,the kernel function Optimize the parameters and penalty coefficients,and use the model proposed in this paper to classify the feature vectors.The results showed that this method can accurately diagnose the loss of shearer picks.Compared with SVM and GA-SVM,it has superior timeliness and accuracy.