Electr Eng DOI 10.1007/s00202-016-0404-3 ORIGINAL PAPER Analysis of statistical features based on start-up current envelope for broken rotor bar fault detection in line start permanent magnet synchronous motor Mohammad Rezazadeh Mehrjou 1,2 · Norman Mariun 1,2 · Norhisam Misron 1 · Mohd Amran Mohd Radzi 1,2 Received: 30 December 2015 / Accepted: 9 August 2016 © Springer-Verlag Berlin Heidelberg 2016 Abstract Line start-permanent magnet synchronous motor (LS-PMSM) is one of the modern high efficiency motor intro- duced to the industry, recently. Like other types of electrical machines, various failures may happen in LS-PMSM dur- ing its life time. Among different failures which may occur in the motors with squirrel cage rotor, rotor fault (broken rotor bar) is the most significant one. This paper investigates defect diagnosis of LS-PMSM, based on statistical analysis of the signals acquired from motor in time domain to find the performance feature for rotor faults detection. To extract the features related to broken rotor bar, a time domain analysis was applied to the transient current signal envelope. In this regards, thirteen statistical parameters were calculated and their reliability were compared for broken rotor bar detec- tion. ANOVA analysis was used to complete the information obtained from those parameters adjusted based on motor condition; starting load condition and their interaction. To perform a statistical comparison among the means for the selected features at different conditions, Tukey’s honest sig- nificant differences method was applied as post hoc testing procedure. Keywords Line start permanent magnet synchronous motor · Statistical features · Fault detection · Broken rotor bar · Current envelope signal B Mohammad Rezazadeh Mehrjou mehrjou.mo@gmail.com 1 Department of Electrical and Electronic Engineering, Universiti Putra Malaysia, 43400 Serdang, Selangor, Malaysia 2 Centre of Advanced Power and Energy Research (CAPER), Universiti Putra Malaysia, 43400 Serdang, Selangor, Malaysia 1 Introduction Electrical motors are those devices that convert electricity into mechanical energy where movement is needed. Motors account for two thirds of the total electricity used in industrial site. Since motor efficiency plays a critical role to reduce the amount of energy used by motors, a new standard, IEC/EN 60034-30:2008, which covers a much wider range of motor ratings was published. This standard that intends to be applied worldwide introduces four new IE (International Efficiency) classes -IE1 class for standard efficiency, IE2 class for high efficiency, IE3 class for premium efficiency and IE4 class for super premium efficiency class [1]. In an attempt to reach the high efficiency, there are some companies that design a motor that exceed IE4 efficiency levels and above. The per- manent magnet synchronous motors provide high efficiency with high power factor grade and efficiency that exceed from IE4; but, they need driver to provide the starting capability and operation which is non-economic for single speed appli- cations and also lack of high starting torque capability [2]. The LS-PMSMs have been introduced to fill this draw- back and also conform to the IE4. LS-PMSMs developed based on the hybrid rotor to provide the high starting torque like induction motor and high efficiency like permanent mag- net synchronous motors in addition to high efficiency range which is cost effective since there is no need to feed by inverter for single speed operation condition. Considering the mentioned advantages above, LS-PMSMs are which lead to grow the application of this motor in different aspects, recently. Accordingly, the LS-PMSM can be suggested as a high potential alternative to IMs and PMSM. Despite of its good characteristics, failures in motor cannot be prevented because it is mostly working in the industrial area and differ- ent stresses during the motor operation lead to stator or rotor failure [3]. In this regards, predicting and detection meth- 123