Fault Diagnosis for Marine Engine System Based on Complex Network Cluster Method

Abstract

Fault diagnosis can be achieved by the cluster analysis process. We give a complete similarity measurement method and a cluster criterion function defined by modularity increment in complex network community detection. An agglomerative cluster method is proposed and the diagnosis rules are extracted to fault diagnosis for the marine engine system. Using the samples collected from self-manufactured marine engine room simulator, fault diagnosis simulation experiment is carried out to verify the algorithm performance. The results show that the method is accurate and less time-consuming, and is able to recognize the fault pattern which does not exist in the fault history data.

Fault Diagnosis Method for Marine Engine System Combined with Multiple Clusters Using Label Propagation

“…1. Data preprocessing includes signal preprocessing, maximum -minimum standardized data conversion, and dimensionality reduction process [6]. 4.…”

Complex networks in advanced manufacturing systems

“…Known from the selected literatures, it also has the specific applications based on complex networks in different industries. For example, automotive industry [146][147][148], aircraft and aerospace industry [149][150], electronics and semiconductor industry [151][152], home appliance manufacturing industry [153], shipping manufacturing [154], ceramic industry [155] and others [156][157][158]. These results are helpful for the decision support for enterprises and governments both in micro level and in macro level to respond to the dynamic competitions.…”

“…Finding: In total, there are 15 papers (10% of selected papers) [144][145][146][147][148][149][150][151][152][153][154][155][156][157][158] on organization structure. It is helpful for the architecture improvement of enterprises, alliances and industry clusters.…”

Fault Diagnosis for Marine Diesel Engine Based on Class Label Probability and Random Walk

“…In a random order, for each sample i, calculate the label probabilty vector, find out C imax ,and update it in C; 6. If i ∀ , we have C i =C imax , set C(t)=C and go to (7). Otherwise, set k=k+1, go to (5);…”

“…It is difficult to obtain the samples on any fault pattern from a real ship because of the limited experiment conditions, so the samples we use in simulation experiments are collected from a self-manufactured engine room simulator whose simulation object is a ship with a MAN B&W 6S50MC main diesel engine and a TCA66 turbocharger [6,7]. There are many kinds of working conditions and we give the fault diagnosis application under a load of 90% as a simple example here.…”