Application of Signal Feature Extraction of Double Cavity Jaw Crusher Based on DEPSO

摘要

—The sparse decomposition of vibration signal is the important part of the fault diagnosis of Double Cavity Jaw Crusher. But the calculation count of sparse decomposition is very large, and it is difficult to fulfill signal processing. After analyzing characteristics of Double Cavity Jaw Crusher, this paper proposes applying the hybrid algorithm, DEPSO which mixed the characteristics of particle swarm optimization (PSO) and difference evolution (DE) algorithm to extracting signal feature of Double Cavity Jaw Crusher and using it to complete signal decomposition of the best atoms search. With the combination of PSO and DE, this method avoids falling into the partial optimal solution. Besides, after the algorithm import the chiasma or variation operations?, the adaptability of the algorithm has made a lot of improvement. The result shows that applying DEPSO to extracting signal feature of Double Cavity Jaw Crusher greatly improves the searching speed, efficiency and accuracy of decomposition, and the calculation has also dropped down dramatically.