Approach to Pre-Processing of Diagnostic Data Using Fuzzy Logic

Diagnostic incoming data from various equipment obtained as a result of experiments are usually unusable for further application without processing. This is due to their averaging, different formats of completion, as well as the presence of external noise and errors from the equipment. To solve this problem, it is necessary to carry out the procedure of data pre-processing. This paper proposes an approach to pre-processing of diagnostic data in the form of a generalized scheme with a detailed description of each stage to assess the technical state of electrical equipment (TS EE). It also presents an algorithm for building fuzzy triangular numbers using fuzzy logic. The novelty of pre-processing of diagnostic data is in taking into account the repeated parameter values, in presenting parameters in the form of fuzzy triangular numbers, as well as in forming vectors of importance weights for values of parameter term-sets. Calculation results with real data are presented at the end of the paper. The proposed processing of diagnostic data makes it possible to move from averaged parameter values to approximate real values under conditions of a variety of parameters and inaccuracy.