FOA optimization based GRNN fault forecasting method for rotary machine

Abstract

The invention discloses an FOA optimization based GRNN (General Regression Neural Network) fault forecasting method for a rotary machine and relates to a technical field of rotary machine fault forecasting methods. The method includes steps of (1), adopting Cincinnati antifriction bearing experiment data; (2) dividing discrete data of collection signals in step (1) into two parts including training data and predication data, putting the training data into the FOA optimization based GRNN for forecasting model training; (3) performing fault forecasting on forecasting data by using the forecasting model trained in step (2), calculating the root-mean-square error and forecasting time and acquiring a comparison diagram of a simulation forecasting curve and an actual curve; (4), determining fault generation time and fault types according to results obtained in step (3). According to the invention, through optimization on GRNN smoothing factors by FOA, the FOA optimization based GRNN fault forecasting model is established, the GRNN optimal model establishing time is reduced and influence on forecasting results by human factors is reduced effectively.

Description

Technical field: [0001] The invention relates to a FOA-optimized GRNN fault prediction method for rotating machinery, and belongs to the technical field of fault prediction methods for rotating machinery. Background technique: [0002] Rotating machinery is widely used in the mechanical equipment industry, and its operation safety is directly related to the entire production process. If an accident occurs, the entire production line will break down, which will bring serious economic losses to the enterprise and even endanger the lives of surrounding staff. Therefore, effective monitoring of the operation status of mechanical equipment has become a key measure to ensure the safe production of enterprises and improve economic benefits. Fault prediction is based on the current use status of the target equipment, combined with its structural characteristics, operating environment and historical data, using a reasonable model algorithm to make judgments on possible future faults,...

AI Technical Summary

Problems solved by technology

[0002] Rotating machinery is widely used in the mechanical equipment industry, and its operation safety is directly related to the entire production process. If an accident occurs, the entire production line will break down, which will bring serious economic losses to the enterprise and even endanger the lives of surrounding staff.

Images