Research On Fault Diagnosis Of Diesel Engine Based On Rough Set Theory

As the developing of the internal diesel engine, its basic structure has been fixed. But the implement of the function of different parts has been being rapidly improved. And the technology applied has a trend of more complication, more electrization and diversification. It makes a higher requirement of the safe and reliable operation and fault monitoring for machinery equipment. If there is potential failure, it should be discovered promptly as early as possible to avoid a major accident, taking effective measures. So the fault diagnosis for locomotive diesel has a positive, practical significance and economic benefits.On the basis of summing up and learning from the research results achieved by the scholars at home and abroad, with the actual requirements of subject and on the premise of combining the theories of Rough Sets and Wavelet Analysis, this subject provides a set of the diesel engine fault diagnostic system design based on S3C2410 embedded hardware platforms, by monitoring functional relationship among cylinder pressure and crank angle.Based on the mechanism of diesel engine fault diagnosis, this paper analyzes that it is absolutely feasible to do the fault diagnosis without disintegration by the use of diesel engine cylinder pressure signal. Then does a more detailed presentation for the key fault technologies of Rough Sets, emphatically analyzes its guiding significance to the diesel engine fault diagnosis system. This paper further proposes the establishment design of the hardware system for diesel engine fault diagnosis which meets the needs of theory. The main function of the design includes fault data acquisition, data recording, data analysis, and data communications and so on. In the process of cylinder pressure signal acquisition, designs a balanced high-temperature pressure sensor and takes the method of signal the crank angle and pressure signals simultaneous acquisition for the interception of pressure signals, which provides experimental data and distinguish basis for the state identification. For the fault feature information of diesel engine cylinder pressure signal acquisition, the system takes the 16 bits A/D converter chip to guarantee the accuracy of the sample time. The paper focuses on the soft-hardware realization of the front-end data acquisition modules, which take ARM as the major processor, and the putting up for the application functional modules based on the S3C2410 hardware platform. The system takes fully into account the diagnosis of real-time, reliability, advanced and practical, and effective to achieve the status identification and fault diagnosis of the diesel engine.