The Fault Diagnosis System For Diesel Engine

Since the advent of diesel engine, it has been applied in marine transportation and played a role of vessels’core power for its inherent advantages. Its performance directly affects the safety and efficiency of the operation of vessels. The work conditions of diesel engines are quite severe and changeable, so the probability of failure is extremely high, Furthermore, the diesel engine has a multilevel and complicated system, which result in the complexity of the process of monitoring its performance. When it breaks down, it is quite difficult and time-consuming to find out the failure cause, and also the repair work is urgent and cumbersome. Once the repairment and maintenance are delayed or incorrectly done, it would trigger a marine accident which could cause the marine pollution as well as the economic loss of the goods and even danger the sailors’lives. Therefore, only when the diesel engine fault diagnosis can be well managed, the personnel and the goods on the ship can be better ensured and the pollution to the ocean can be reduced. At present, most of the ship managements adopt traditional fault diagnosis methods and expected maintenance plans on a regular basis. However, with the development of diesel engine technology, the traditional fault diagnosis methods are difficult to meet the present demands. With the development of artificial intelligence technology, marine diesel engine fault diagnosis technology has opened up a new road. Using the the diagnosis method of BP neural network of artificial intelligence computer is an important direction of diesel engine fault diagnosis researches.This article proceeded from the reality based on the mature data acquisition system of the ship. I have designed a set of vessel’s diesel engine fault diagnosis system in which the BP neural network acts as the core. This system has realized multiple functions of data acquisition, status monitoring, signal analysis and fault diagnosis etc. The system has been developed under the circumstance of visual studio 2013, using MATLAB neural network toolbox to realize BP neural network and to analyze the collected data signals and using MySQL database to set up the fault database and to supervise the the data in the background. This software has a simple and practical human-machine interface and uses modular programming which can be easily added and extended and it establishes a foundation for future practical application.