|
Fault diagnosis technology development for decades , has grown from a simple observation , data processing to the sensor and dynamic testing techniques as a means to signal processing and modeling process based on modern diagnostic techniques , and now to a new stage - intelligent diagnostic technology stage. This thesis is based on the typical rotating machinery - the fans as the research object , building aspects in the diagnostic system to do the research . The focus of this study is to rough set reduction allows the diagnosis process faster and more accurate ; build a diagnostic platform , so that it can identify typical failure , this being to build a knowledge base and design a strategy of knowledge reasoning . System , we do the research and discussion of the following areas : 1) extract fault symptoms and fault reasoning method . 2 ) using Access2003 and the framework of the theory , the fan signs database , then establish connection with VB6.0 database combining forward reasoning strategy to build a diagnostic system . 3) to extract the characteristics of the common fan failure , then follow signs order order , one-to-one correspondence digitized to form a feature vector , the use of the theory of rough set and reduction , with the results of MATLAB, prove reduction after diagnosis , faster , more comprehensive. 4) the use of a fuzzy inference theory , diagnostic discriminate on the basis of building a knowledge base . Fuzzy inference is a fuzzy knowledge uncertainty reasoning , fuzzy reasoning theory is based on the fuzzy set theory and fuzzy logic developed on this basis , it handled things vague , applies to the fault knowledge identification , simulation of diagnosis can also be established to illustrate this point . 5 ) use VB6.0 build fault diagnosis system , the when complex logic level , inevitable to face the problem of complex knowledge search , knowledge organization and reasoning mechanisms need to consider improvements , this system has to constantly improve function, the knowledge base can be continually adding and maintenance, this fault diagnosis with the use of time , the diagnosis result can be more and more accurate .
|