METHOD OF THE MODES CONTROL BASED ON THE NEURAL NETWORK DIAGNOSTICS OF FAULTS AND TECHNICAL CONDITION OF THE ELECTRIC-DRIVE GAS-COMPRESSOR UNIT Russian patent published in 2018 - IPC G05B13/02 G06N3/02 G05B19/00 

FIELD: electrical engineering; test technology.

SUBSTANCE: invention refers to the diagnosis of the state of electric drive devices. Method of controlling the modes, which is based on the neural network diagnostic troubleshooting and on the technical state of the electric drive gas-compressor unit includes measurement of parameters, collection of information and verification of its reliability, filtering of measurements, checking of measurements, correction of readings, reconciliation with restrictions, calculation of signs of operational diagnosis, fault recognition and calculation of deviations. Calculation of the signs of operational diagnosis and recognition of faults is performed taking into account the algorithms, which are based on the operation of two neural networks of the Kohonen type, on the basis of which they calculate and evaluate the failure rates for the following subsystems: lubricating, supercharger, stator winding and mechanical defects of the motor. Technical condition is then evaluated based on the obtained coefficient values and the diagnostic and control mode is selected.

EFFECT: improved accuracy of diagnosis.

1 cl, 9 dwg