The use of self organizing neural network for vibration unbalance assessment

In the paper the possibility of using the self-organizing (Self Organizing Maps - SOM) neural network as a model for the assessment of vibration causes and respectively the assessment of vibration unbalance is presented. For a successful assessment of vibration unbalance the clustering of measurement data from a frequency spectrum (Hz, mm/s) is included in the model. The presented model of self-organizing neural network may be used as a template for the assessment, i.e. identification of other types of vibration loadings and also its integration in a CMMS system (Computerized Maintenance Management System) as an expert system for a decision making support is proposed.