Application To Fault Diagnosis Of Bearings Of Vibrating Screen Based On Blind Source Separation

With the further development of Chinese scientific research and economicconstruction, the areas and application related to screening equipment have becamemore and more extensive, in the field of raw material production and application, wecan see the mechanical screening equipments. During these mechanical screeningequipments, the most common equipment is vibrating screening. In the coal industrydepartment, water conservancy and hydropower department, the transportation industry,chemical industry department, even the sanitation department have applied to vibratingscreen. Obviously, in each industry department, the vibrating screen has played a vitalrole. The bearing system for the normal work of vibrating screen has vital function. Itsworking condition not only affects the safety and stable operation of equipment, but alsohas a direct influence on continued production. The fault may cause serious economiclosses, and even disastrous accident, so the fault diagnosis and analysis technology tobearing is more urgent.Fault diagnosis technology is new development of scientific fields, and have notformed a relatively complete system of science. For the purpose of the research, thecontent of the category, usually have great difference on engineering applicationbackground, even the major of technical personnel, so there is some shortcomings andproblems in the existing fault theory and method. It would be safe to say that featureextraction is a bottleneck problem, which affects accuracy of fault diagnosis andreliability of early forecast. The blind source separation technology provides a positivemethod for vibration signal processing, recognition of fault diagnosis.But the same to other algorithm, it has its own limitation that observed numbermust be greater than vibration sources number, if it can’t satisfy with the precondition, separation will eventually result in failure. In view of this limit, this paper putsforward the blind source separation based on ensemble average empirical modedecomposition (EEMD-BSS), this algorithm can be better refrain from this restriction,when the observed numbers are less than the number of vibration sources,thealgorithm can also separate the fault signature, so as to achieve the purpose ofseparation.At last, this paper use the traditional blind source separation algorithm andImproved EEMD-BSS algorithms for separation of bearing’ fault data, including themulti-channel and single channel, which has separate the fault signature better, provedefficiency of the algorithm.