| With the development of railway, highway, city subway and tunnel construction, tunnel boring machine (TBM) is used more and more widely. Tunnel boring machine is large-scale, complicated equipment, integrates electronic, hydraulic, laser technology.Once a failure may cause great economic losses and casualties. Therefore, how to ensure the reliable operation of TBM and how to timely and effective analysis fault diagnoses are very necessary. There are vast amounts of data stored in the data acquisition system.A lot of valuable knowledge hidden behind of the data is not understood and used effectively. These information in the hidden potential rules shall be mining. Data mining can discover valuable information and knowledge from a large amount of data, therefore, this paper applies data mining to fault diagnosis of TBM, to solve identification problem of fault rule.On the basis of theoretical analysis, main structure and common faults about TBM, this paper focuses on the research of rough set and decision tree in data mining. The essence of fault diagnosis based on data mining is the classification rules of fault signal. C4.5 can quickly generate comprehensible rules. In addition, the number of fault feature vectors of TBM is huge, a large number of redundant information can't be identified, so before using decision tree, this paper adopts rough set. The combination of two kinds of data mining algorithm reduces the complexity of the calculation, greatly reducing the workload of data acquisition and diagnostic time, by the diagnosis, rules are more concise.Finally, the paper proposes a fault diagnosis system based on Web,which introduces the system of database management system detailedly, user management system and fault diagnosis system and so on, realizes the system database management diagnostic function.This paper expounds the remote fault diagnosis system. System uses B/S architecture, combined with SQL Server 2000,VB and dreamweaver8.0, develops the fault diagnosis system. |
PDF Full Text Request