Abstract:

The roller bearings of open pit belt conveyors face problems of low fault identification accuracy. To improve the accuracy and efficiency of fault diagnosis, a fault signal detection method of roller bearings with CPSO algorithm based on OVMD was proposed. Firstly, the excellent global optimization characteristics of CPSO were utilized, and the optimal parameter setting of the variational mode decomposition (VMD) algorithm was precisely locked to achieve effective parameter tuning of VMD. Then, VMD technology after parameter tuning was used to process the vibration data, and specific frequency band signal components were accurately extracted from the vibration data. Finally, the sparse maximum harmonic noise ratio deconvolution (SMHD) technology was used to purify the above frequency band signals, which significantly enhanced the identification accuracy of the fault characteristics of roller bearings of belt conveyors. The results show that CPSO has better performance for VMD improvement than other VMD optimization algorithms. The VMD algorithm after CPSO optimization combined with SMHD can successfully identify the specific fault points of the inner and outer rings of the rolling bearings under complex working conditions and determine the specific damage forms of the bearings.

Key words: optimized variational mode decomposition (OVMD), chaotic particle swarm optimization (CPSO), roller bearing, time domain and frequency domain, fault signal