Knowledge base construction method oriented to fault diagnosis and fault prediction of numerical control machine tool

The present invention relates to a method for constructing the knowledge base for fault diagnosis and fault prediction CNC machine, the following steps: a step, by the remote monitoring device on-line monitoring of high-grade turning center, to obtain a plurality of sets of data Xj vibration representing different fault types (t), j is the number of collected vibration data group, n is a positive integer; step two, state coarse wavelet packet analysis processing of the real time monitoring of the plurality of sets of vibration data Xj (t) sequentially obtain energy feature vector T 'property as a condition to fault type decision attribute, build knowledge original fault decision table; step three, the knowledge of the original fault fault decision table attribute reduction characteristic difference matrix generation rule based knowledge base configuration; step 4 using the credibility of the rules as an evaluation of the final rule to measure and evaluate. The present invention can provide effective protection for the fault diagnosis and fault prediction, the present invention can be widely used as additives in the turning center.