摘要
为了监测捣固车液压系统的工作状态,提出了采用完全总体经验模态分解和模糊熵理论相结合的方法提取振动信号特征。将采集到的捣固车液压系统振动信号进行完全总体经验模态分解,选取前四个主要的固有模态函数分别计算模糊熵值构成高维特征向量,最后采用支持向量机进行故障状态的分类识别。实验结果表明,采用此方法提取的故障特征,故障识别率可以达到90%,证明了该特征提取算法的有效性。
To monitor the working condition of hydraulic system of tamping machine, a novel methods for feature ex-traction is proposed by combination of complete ensemble empirical mode decomposition (CEEMD) and fuzzy entropy theory. After CEEMD, collected vibration signals from tamping machine hydraulic system have been decomposed trinsic mode functions (IMFs),and the first four main fuzzy entropies of these IMFs constitute a high dimensional feature vector. Finally,the support vector machine is used to identify the fault conditions based on the feature vectors.The experi-mental resutt shows that the recognition rate is 90% when the extracted fault features by this method are used.Therefore,this feature extraction method is effective for tamping machine hydraulic system fault signals.
出处
《计算机与数字工程》
2016年第12期2407-2410,2418,共5页
Computer & Digital Engineering
基金
国家自然基金项目(编号:61263023)资助
关键词
捣固车
液压系统
完全总体经验模态分解
模糊熵
特征提取
tamping machine,hydraulic system,complete ensemble empirical mode decomposition (CEEMD),fuzzy entropy,feature extraction