模糊C-均值聚类算法及其在船舶故障诊断中的应用

船舶设备故障的早期诊断和预测,对船舶的安全运行具有非常重要的意义。由于船舶设备繁多,运行环境特殊,各种设备的故障症状与故障原因之间的关系十分复杂,致使传统诊断方法在实际应用中效果不理想。因此,研究采用模糊C-均值聚类算法来实现船舶故障的诊断乃是非常必要的。将被诊断对象间故障和症状的特征通过建立模糊关系矩阵进行故障分类,用当前所得的故障征兆群与过去该设备故障征兆结果相对照,找出最相似的结果,从而确定其故障。通过船舶主机轴系诊断的实例,证明了该方法的有效性。

The Fuzzy C-Means Clustering Algorithm and Its Application in the Fault Diagnosis of Ships

It is significant for a ship to diagnose and forecast the failures of facilities as early as possible.Because facilities in a ship are so many and the running conditions of them are so special,the traditional fault diagnosis methods are no more efficient in practice.A fuzzy C-means clustering algorithm is used in this paper and the features of faults and symptoms of the detected object are classified based on the established fuzzy connection matrix.The comparison between the fault symptom clusters collected from a facility recently and the previous outcomes of the fault symptoms of the facility is made,and then the closest outcomes are identified and the fault is spotted.A case of the recent fault detection for the shafting of main engine fully proves the effectiveness of the above mentioned method.