【作者】

Takashi Fujii

【摘要】

该论文已在芬兰赫尔辛基举行的第28届CIMAC大会上发表。论文的版权归CIMAC所有。The recent advancement in computer performances and the developed machine learning algorithm have been bringing us to the more sophisticated abnormality detection and condition diagnosis based on big data analysis. These results have also been required to seek safer and more economical operations. However, when it comes to the machineries in the engine room in a ship, the practical diagnostic system that gives sufficient economic effects does not seem to prevail because: • The engine room consists of a number of machineries of different manufacturers, each of which must be equipped with its own diagnostic system; • The diagnostic system for smaller machineries in particular is less likely to be developed due to a lack of expected cost effectiveness; or • The measurement items for the engine system lack unified standards for data format, impeding the development of multi-purpose systems. To find a solution to the abovementioned problems, the mechanism equipped with the following functions has been developed: • A diagnostic platform applicable to diverse machineries and equipped with the support functions for the automatic abnormality detection, automatic condition diagnosis and automatic troubleshooting; • The application of the automatic abnormality detection algorithm based on machine learning and the establishment of the optimal method to apply the algorithm to each vessel machinery; • The integration of the measured data and diagnostic results of each machinery with maintenance and management; and • The function for the uniform management and effective use of the aforementioned information. A field test has been conducted witnessed by multiple machinery manufacturers for the purpose of diagnosing the all engine room machineries. The test has been indicating the originally anticipated effects. The previous system has been applied to the practical ships in the last few years, with the majority of problems already solved indicating the effectiveness of this concept. Machinery manufacturers can apply this platform to those machineries for which no diagnostic system was developed for the cost-effectiveness reasons. The platform also helps users uniformly control a number of different machineries through a single system and easily have an advanced diagnosis from the specialist of machinery manufacturers. This study aims to further increase the accuracy in data analysis and support safer and more economical operations using diagnostic results. The study can also develop the platform such that it will be applicable not only to the engine room machineries but to the ship as a whole, rather than showing versatile diagnostic functions.

【会议名称】

第28届CIMAC会议

【会议地点】

芬兰 赫尔辛基

【下载次数】

2

返回