On-line monitoring induction motors by fuzzy logic: A study for predictive maintenance operators

The monitoring of induction motors through predictive techniques and artificial intelligence has grown considerably in recent years. These techniques allow the detection of a defect in its early stages, consequently allow maintenance personnel to schedule the intervention, working within the concept of planned corrective maintenance, avoiding catastrophic failures on the production line. Among these techniques, there is the Fuzzy Logic. The motor operational conditions are described by using fuzzy linguistic variables in an effective monitoring program that acquire, analyze and present the results. The knowledge base, comprising fuzzy rules and databases, was built to support the fuzzy inference process to analyze the data processing. The experimental results shown the efficiency of the vibration sensor developed and the strategies for detection diagnosis, and on-line monitoring tasks. The results were undoubtedly impressive and in a near future the system developed can be adapted and used in real predictive maintenance programs in industries.