Motor Fault Detection by Displacement Detection and LSSVM
Journal Title: 河南科技大学学报（自然科学版） - Year 2018, Vol 39, Issue 2
Abstract
Displacements between stators and rotors were the main reasons for induction motor faults.8 detection coils were pre-placed into the notch grooves of stator core tooth for displacement detection, and they were configured into 6 groups.The differential output of each group was sensitive to the displacements between stator and rotor.The least squares support vector machine ( LSSVM) was combined with the detection coils for online fault detection of motors. A transformer pump motor was used for field test. The test results show that the combimation of displacement detection and LSSVM can detect the pump motor faults precisely. The proposed sparseness approach for LSSVM has the ability to keep a high fault detection rate and generalization performance while reducing the number of the support vectors and the online computation requirement. The sparse LSSVM models is suitable for online detection of motor fault.
Authors and Affiliations
Fangfang ZHANG, Shirong ZHANG, Qin CHENG
Keywords
Related Articles
- EP ID EP464871
- DOI -
- Views 87
- Downloads 0
Fractional Order Brain Emotional Learning Model Control Method of Ball Mill
Aiming to the characteristics of strong coupling and time variation of ball mill system,an improved control method of brain emotional learning( BEL) model was proposed. Fractional calculus was adopted to depict emotional...
Performance Test Bench of Vertical Axis Wind Turbines
In order to improve the control precision and the test automation of vertical axis wind turbine test data,a performance test system of the vertical axis wind turbine was developed. The system was divided into three subsy...
Blowup for Classical Solutions of n-Dimensional Euler Equations with Nonlinear Damping
The blowup of classical solutions for initial value of the isentropic Euler equations with nonlinear damping in n-dimensional space was studied. When the initial condition had compact support,by functional methods,the cl...
Modeling and Optimization of Non-ideal Litz Wire in Wireless Power Transfer System
In order to reduce the high-frequency losses of the Litz wire, finite element simulation combined with experimental methods were used to design different shapes of Litz wire joints. The distribution of the strands on the...
Motor Control of Pure Electric Vehicle Based on BP Fuzzy Neural Network Control
A proportional integral( PI) controller was designed based on back propagation( BP) fuzzy neural network to solve the problem that traditional PI control could not reach good control effect of the motor. Driver operation...