Simulation experiment on single-phase fault diagnosis of a five-phase motor drive system based on current signature analysis
[Objective]The widespread integration of motor drive devices as the core of transportation,aerospace,and various key applications has garnered widespread attention from academia toward the development of fault diagnosis algorithms for drive systems.Presently,a multitude of viable diagnostic algorithms exists to address the single-phase open-circuit fault in five-phase permanent magnet synchronous motor drive systems.However,the complexity of model analysis and uncertainties surrounding intelligent algorithms have relegated these methods to the domain of theoretical research,hindering their broad practical application and posing significant challenges for integration into university teaching experiments.Despite the favorability of diagnostic methods based on signal characterization due to their simplicity and efficiency,these methods are limited by harmonic contamination and short-term diagnosis failure.Addressing these limitations,this study develops a fault diagnosis and localization method based on current characterization,with the aim of mitigating these issues by considering different characteristics,such as sudden amplitude and phase changes in the third harmonic current following the fault.[Methods]The proposed fault diagnosis and localization method includes several key steps.First,the phase current sampling signal undergoes filtering to remove high harmonic components,followed by the calculation using Clarke's coordinate transformation matrix to determine iα3 and iβ3 in the orthogonal stationary coordinate system.Then,the fault judgment signal G0 is derived through root mean square calculations,marking the completion of the signal preprocessing process.Next,the study sets a reasonable threshold to determine the fault occurrence based on the amplitude change shown by the G0 signal post-fault.Additionally,the implementation of self-increment or zero-setting operations,achieved by manipulation of the counting variable h,serves to mitigate the influence of external interference on the motor under specific working conditions.Upon verifying fault occurrence,the study proceeds to calculate the phase relationship angle amplitude between the two using the inverse tangent function with the third harmonic current effective value as the object.The positive and negative complementary phase angle sign information of the product of the two is then collected through sliding average sampling to prevent diagnostic failures of AC signals in the calculation.Finally,the fault location table is employed to determine the fault location and conduct the necessary fault removal operation.To prevent secondary diagnostic failures caused by delayed access to the system by the subsequent fault-tolerant system,an additional signal-locked self-resection design is incorporated.[Results]The analysis of the MATLAB/Simulink simulation modeling results discloses the excellent efficacy of the improved diagnostic algorithm.It not only realizes diagnostic isolation within half a current cycle following the fault but also effectively alleviates diagnostic failures resulting from the AC signal in the phase angle calculations.In addition,the diagnostic system shows robust characteristics under demanding working conditions,such as continuous variable load and speed.Moreover,simulation data reveal the fault localization situation without the self-resection design,highlighting the essential and rational nature of incorporating this design.[Conclusions]The approach of separating fault localization angle amplitude calculations from symbol information acquisition improves the operational reliability of the diagnostic system and contributes to improved diagnostic accuracy.Beyond its technical contributions,the study shows significant theoretical and practical value for teaching experimental courses in colleges and universities,in particular within areas such as electrical transmission and motor control.Its practical significance in cultivating the dynamics modeling and analysis abilities of students further showcases its broader educational impact.Therefore,this study not only offers an advanced fault diagnosis and localization method but also shows substantial promise for theoretical research and practical applications,providing valuable insights for educational curricula and highlighting its relevance in developing the expertise of students in dynamics modeling and analysis.
five permant magnet synchronous motorfault diagnosisanalysis of current characteristicsMATLAB/Simulink simulation
NETL
NSTL
万方数据
