The diagnostic classifier of the squirrel-cage induction motors with a minimal number of diagnostic premises

The problems of control and identification of failures of units that have a different degree of technological advancement are still important tasks to be addressed. A description of the dependence of complex production systems, regarding to the prediction of their faults and their consequences is also an important issue. The definition of state classifiers and the algorithm used for the graph searching are presented. A modified algorithm used in its principle to the needs of a solution of the stated problem leads to obtaining a path with the most probable states, defined on the basis of ordered series of relations (structural, reliability and failures forms). Assumed form of functional relationships of constituent components of distributed mechatronic systems allows for simplifying also the prediction process. The described advantages result from usage of computer algorithms (allowing for a mapping of the path identifying consequences of successive failures) and possibilities of representation of machines on different complexity levels (systems, subsystems, components etc.).