Demagnetization Fault Detection for Double-Sided Permanent Magnet Linear Motor Based on Three-Line Magnetic Signal Signature Analysis

Abstract

This article investigates a novel demagnetization fault detection method for double-sided permanent magnet synchronous linear motors (DS-PMSLM) that are applied in laser cutting machines. First, a finite-element analysis model of DS-PMSLM is built to preset different demagnetization faults of the permanent magnets (PMs). Second, magnetic intensity distribution signals in air gap region are selected as fault signal for the first time to detect demagnetization faults in DS-PMSLM. Three-line method coupled with self-designed connection structure is originally proposed to obtain and analyze three-line magnetic signals. Simulation comparison with back electromotive force signal validates that three-line magnetic signals are more effective to solve the problem for this double-sided structure PMSLM. Then, continuous wavelet transform (CWT) is innovatively introduced to conduct peak feature extraction for magnetic signals. Comparison experiments between CWT, direct locating method, and polynomial fitting method certify the superiority of CWT. By peak feature analysis, the positions, sides, and severity degrees of PMs' demagnetization faults can be detected accurately. Finally, a prototype is manufactured, and a Gauss meter experimental platform is established to verify the validity and advantages of this method.