Andrea Stefani

Doubly Fed Induction Machines Diagnosis Based on Signature Analysis of Rotor Modulating Signals

This paper introduces an advanced monitoring and diagnosis system for the detection of incipient electrical faults in doubly fed induction generators used in wind turbines. In this application, the rotor is supplied by a static converter for the control of active and reactive power flows from the generator to the electrical grid. A new diagnostic method based on the frequency analysis of the rotor modulating signals is proposed. Simulation and experimental results confirm that a convenient frequency analysis of these signals and a simple interpretation lead to an effective diagnostic procedure. The proposed system is suitable to be easily embedded in the power-converter digital control system at very low cost.

DTC Drives for Wide Speed Range Applications Using a Robust Flux-Weakening Algorithm

A control scheme for robust flux-weakening operation of direct-torque-control induction motor drive is proposed. The basic idea is to adjust the flux reference on the basis of the torque error, thus determining a spontaneous flux weakening. To exploit the maximum torque capability, it is necessary to estimate the maximum torque that the induction machine is able to generate at any speed. Initially, a basic version of the algorithm, requiring a simple off-line parameter tuning, is presented. Then, the algorithm is improved and completed with the online estimation of the maximum torque, hence avoiding the initial tuning process. The main features of the proposed methods are a little dependence on machine parameters and a smooth transition into and out of the flux-weakening operation mode. Experimental tests demonstrate the effectiveness of the control schemes.

A New Model-Based Technique for the Diagnosis of Rotor Faults in RFOC Induction Motor Drives

This paper proposes a new model-based diagnostic technique, which is the so-called virtual current technique (VCT), for the diagnosis of rotor faults in direct rotor field oriented controlled (DRFOC) induction motor drives. By measuring the oscillations at twice the slip frequency found in the rotor flux of the machine, and by conjugating this information with the knowledge of some motor parameters, as well as the parameters of the flux and current controllers, it is possible to generate a virtual magnetizing current which, after normalization, allows the detection and quantification of the extension of the fault. The proposed method allows one to overcome the major difficulties usually found in the diagnosis of rotor faults in closed-loop drives by providing information about the condition of the machine in a way that is independent of the working conditions of the drive such as the load level, reference speed, and bandwidth of the control loops. Although the VCT was primarily developed for traction drives used in railway applications, it can be incorporated in any DRFOC drive at almost no additional cost. Several simulation results, obtained with different types of DRFOC drives, as well as experimental results obtained in the laboratory, demonstrate the effectiveness of this new diagnostic approach.

Magnets faults characterization for Permanent Magnet Synchronous Motors

Nowadays Permanent Magnet Synchronous Motor (PMSM) are an attractive alternative to induction machines for a variety of applications due to their higher efficiency, power density and wide constant power speed range. In this context the condition monitoring of magnets status is receiving more and more attention since is critical for industrial applications. This paper presents a characterization of rotor faults for such a motor due to local and uniform demagnetization by means of two dimensional (2-D) Finite Element Analysis (FEA) and proposes a new non-invasive method for their detection by means of a Fourier transform of the back-EMF. The proposed approach is then validated for three Permanent Magnet Synchronous Motors with different winding configurations.

Diagnostic Technique based on Rotor Modulating Signals Signature Analysis for Doubly Fed Induction Machines in Wind Generator Systems

The paper introduces an advanced monitoring and diagnosis system for the detection of incipient electrical faults in doubly fed induction generators (DFIGs) used in wind generator systems. In this application the rotor power is supplied by a converter for the control of active and reactive power flow from the generator to the mains. The paper deals with a new diagnostic method based on the analysis of the rotor modulating signals. Simulation and experimental results confirm that the analysis of the spectra of rotor input modulating signals leads to an effective diagnostic procedure. The system is suitable to be easily embedded in the drive control system

Diagnosis of induction machines in time-varying conditions

A method is here presented that allows the diagnosis of broken rotor bars in time varying operations with simple post processing of input currents. Extensive simulations are presented to validate the proposed approach for open loop and closed loop induction machines. A diagnostic index is presented also, that is quite robust versus load and inertia variations.

Advanced diagnosis of broken bar fault in induction machines by using Discrete Wavelet Transform under time-varying condition

The diagnosis of induction machine faults is commonly carried out by means of Motor Current Signature Analysis (MCSA), i.e., by classical spectrum analysis of the input currents. Specifically in case of broken bars, the amplitude of the left sideband component of a phase current is monitored in order to sense its signature. However MCSA has some drawbacks that are still under investigation. The main concern is that an efficient frequency transformation cannot be made under speed-varying condition, since slip and speed vary and so does the left sideband frequency. In this paper, an advanced use of the Discrete Wavelet Transform (DWT) is proposed to overcome the limitation of the classical approaches based on Fourier Analysis (FA). Experimental and simulation results show the validity of the developed approach, leading to an effective diagnosis method for broken bars in induction machines.

Experimental fault characterization of doubly fed induction machines for wind power generation

For modern large wind farms, it is interesting to design an efficient diagnosis system oriented to wind turbine generators based on doubly-fed induction machine (DFIM). Intensive research effort has been focused on the signature analysis to predict or detect electrical and mechanical faults in induction machines. Different signals can be used, voltage, current, stray flux. In the case of wind generators, considering that both the machine signals and the control signals are accessible, for diagnostic purposes, there are interesting additional possibilities, for example the use of the converter modulating voltages. In this paper, a complete system is analyzed by suitable simulations and experimentations to deeply study fault influence and to identify the best diagnostic procedure to perform predictive maintenance

Closed loop bandwidth impact on doubly fed induction machine asymmetries detection based on rotor voltage signature analysis

This paper investigates the impact of the control system bandwidth on the detection of incipient faults in doubly fed induction generators used in wind turbines. In this application, the rotor is supplied by a static converter for the control of active and reactive power flows from the generator to the electrical grid. The Impact of the closed loop control system cannot be neglected when the detection of asymmetries in the machine are based on the signature analysis of electrical variables. An investigation of the behavior of faulty spectral components for different bandwidths of closed loop regulators is therefore necessary in order to evaluate the effectiveness of diagnostic indexes based on those variables. Simulation and experimental results are here presented to show the limits of two different approaches for fault detection. The first one based on current signature analysis and the second one based on rotor voltage signature analysis.

Doubly Fed Induction Machine stator fault diagnosis under time-varying conditions based on frequency sliding and wavelet analysis

The paper introduces a monitoring and diagnostic technique for the detection of incipient stator electrical faults in Doubly Fed Induction Machine (DFIM) for wind power systems. Operating in aggressive environments, the detection of anomalies at an incipient stage is crucial to decide about the operating continuity of the machines. Discrete Wavelet Transform (DWT) is used to detect stator faults under time varying-condition in two mainly different contexts: Transient-Speed conditions and Fault-Varying conditions. A frequency sliding (FS) with High Multiresolution Analysis (HMRA) approach is proposed for improving the ability of DWT in extracting the most relevant stator fault frequency component dynamically over time thereby. A dynamic mean power calculation at different resolution levels was introduced as a diagnostic index to quantify the fault extent. Simulation and experimental results show the effectiveness of the proposed approach in discriminating stator fault severities leading to an effective diagnostic procedure for stator faults in DFIM.