TS-127
Maintenance Analysis due to Fault Diagnosis Success Ratio
Kenzo ISHIGE and Michio HORIGOME
ABSTRACT
Recently, the monitoring maintenance is well known as one of the most effective method to prevent the system from failure. If the distribution of system failure relatively might be represented by the exponential distribution, the periodic maintenance is not effective enough to prevent failure. In this paper, from the reliability improvement rate, maintenance cost and maintenance number, the characteristics of the monitoring maintenance are investigated using the success ratio of fault diagnosis when it is assumed that the system failure occurs through some malfunctions or faults.
Key Words: Success Ratio of Fault Diagnosis, Monitoring Maintenance, Reliability Improvement Rate, Maintenance Cost, Maintenance Number.
1. INTRODUCTION
This paper deals with the maintenance evaluation under the defect/failure diagnosis. We investigate how the success ratio of fault diagnosis provides effect on characteristics of CM (corrective maintenance) and PM (preventive maintenance).
Recently, it has been noticed that the monitored maintenance is a very useful method. The main reason is well known that the periodic maintenance method is not useful when the failure rate of the system is not IFR (increasing failure rate). But, the failure rate of the system with many elements has tendency to show constant [1], and therefore the effect of PM can not be expected even nowadays. Then in order to correspond such serious problems such as PM, the defect/failure diagnosis based PM has been said to be a very effective method.
In this paper it is assumed that the failure happens through the state of defect. That is, as shown in Fig.1, first defect occurs in the system, and then grows into the failure of the system. Consequently, we must diagnose the system when it remains between defect occurrence and failure. We call it the defect diagnosis success, that is, the diagnosis is finished before the defect reaches to the failure and also its ratio the fault diagnosis success ratio. [2][5]
Based on the diagnosis success ratio, we already investigated theoretically how it provides the effectiveness and the characteristics of system maintenance in detail [2][3]. In particular, we study two types of models: one is the instant diagnosis model which means diagnosis is finished instantly and another model which takes time in diagnosis.
Tokyo Univ. of Mercantile Marine
2-1-6 Etchujima, Koto-ku, Tokyo 135-8533 JAPAN
TEL:+81-3-5245-7424, E-mail: ishigek@otis.com
