METHOD AND SYSTEM FOR PLANNING PREVENTIVE MAINTENANCE AND REPAIR OF PROCESS EQUIPMENT BASED ON ACOUSTIC DIAGNOSTICS USING NEURAL NETWORKS Russian patent published in 2022 - IPC G01M7/02 G01N29/04 

FIELD: maintenance service.

SUBSTANCE: invention can be used for planning preventive maintenance and repair of process equipment based on the analysis of noise vibration and ultrasound patterns. The essence of the invention lies in the fact that acoustic data of process equipment is obtained simultaneously using one or more noise meters, one or more ultrasound meters and one or more vibration meters; with the help of one or more automated workstations (AWSs) for pattern marking, the received acoustic data is processed using augmentation and correction of the received noise-vibration and ultrasonic pattern, the corrected acoustic data dataset is marked up and the marked acoustic data is obtained; the marked-up acoustic data is transmitted to one or more intelligent modules, and the intelligent module processes the received acoustic data in real time and analyzes the processed acoustic data using one or more neural networks, and based on the results of the analysis, receives data on the planning of preventive maintenance and repair of process equipment; transmits data on the planning of preventive maintenance and repair of process equipment to one or more operator’s workstations for display.

EFFECT: ensuring the possibility of timely preventive maintenance and repair of process equipment.

17 cl, 4 dwg