Applying Data Mining Technologies to Diagnosis of Automotive Engines Efficiency

• Main Authors: Yang-Ting Jeng, 鄭仰廷
• Other Authors: Mu-Der Jeng
• Format: Others
• Language: zh-TW
• Published: 2011
• Online Access: http://ndltd.ncl.edu.tw/handle/89722503483179861724

碩士 === 國立臺灣海洋大學 === 電機工程學系 === 99 === In the conventional vehicle diagnostics, a skilled technician determined the abnormality or noise of the certain parts based on his experience. For examples : a cylinder misfire, the brake pads are too thin, the noise caused by the worn bearings of tires and so on. However, a well experienced and skilled technician is not always available. With the rapid development of vehicle electronics, vehicles have become the complex of high technology. And the computer technology is updated rapidly, if it could be used on the vehicle diagnosis system, we could use the diagnostic system to monitor, record and collect a variety of data in the vehicle operation, and stored it in the database. Therefore, how to complete the vehicle health diagnosis by vehicle electronic and computer technology is the main topic of this thesis to be discussed. For this topic, this thesis applied the Grey Relational Analysis (GRA) in the health diagnosis of vehicles and infer the failure symptom of vehicles. And it will form a complete and efficient fault diagnosis system. Furthermore, this thesis proposes a modified formula by Data Mining technology to modify the formula of predictive fuel consumption proposed in the " AVR-Based Fuel Consumption Gauge ". And there are smaller error value and more accurate predictive value in this modified formula.