Identification of Load Categories in Rotor System Based on Vibration Analysis
Open Access
Abstract
Rotating machinery is often subjected to variable loads during operation. Thus, monitoring and identifying different load types is important. Here, five typical load types have been qualitatively studied for a rotor system. A novel load category identification method for rotor system based on vibration signals is proposed. This method is a combination of ensemble empirical mode decomposition (EEMD), energy feature extraction, and back propagation (BP) neural network. A dedicated load identification test bench for rotor system was developed. According to loads characteristics and test conditions, an experimental plan was formulated, and loading tests for five loads were conducted. Corresponding vibration signals of the rotor system were collected for each load condition via eddy current displacement sensor. Signals were reconstructed using EEMD, and then features were extracted followed by energy calculations. Finally, characteristics were input to the BP neural network, to identify different load types. Comparison and analysis of identifying data and test data revealed a general identification rate of 94.54%, achieving high identification accuracy and good robustness. This shows that the proposed method is feasible. Due to reliable and experimentally validated theoretical results, this method can be applied to load identification and fault diagnosis for rotor equipment used in engineering applications.Keywords
Funding Information
- National Natural Science Foundation of China (51475318)
This publication has 12 references indexed in Scilit:
- An imbalance fault detection method based on data normalization and EMD for marine current turbinesISA Transactions, 2017
- Solving metameric variable-length optimization problems using genetic algorithmsGenetic Programming and Evolvable Machines, 2016
- Optimization of Stripping Voltammetric Sensor by a Back Propagation Artificial Neural Network for the Accurate Determination of Pb(II) in the Presence of Cd(II)Sensors, 2016
- Detection of broken rotor bar faults in induction motor at low load using neural networkISA Transactions, 2016
- Nonlinear response and bifurcation analysis of a Duffing type rotor model under sine maneuver loadInternational Journal of Non-Linear Mechanics, 2016
- A New Method for Fault Diagnosis of Mine Hoist based on Manifold Learning and Genetic Algorithm Optimized Support Vector MachineElektronika ir Elektrotechnika, 2012
- A Time Domain Inverse Method for Identification of Random Acoustic Sources at Launch Vehicle Lift-OffJournal of Vibration and Acoustics, 2011
- ENSEMBLE EMPIRICAL MODE DECOMPOSITION: A NOISE-ASSISTED DATA ANALYSIS METHODAdvances in Adaptive Data Analysis, 2009
- Computer simulation of structural random loading identificationComputers & Structures, 2001
- Neural-network-based motor rolling bearing fault diagnosisIEEE Transactions on Industrial Electronics, 2000