is an open access repository that collects the work of Arts et Métiers ParisTech researchers and makes it freely available over the web where possible. This is an author-deposited version published in: https://sam.ensam.eu Handle ID

Abstract-This paper presents an original method, based on artificial neural networks, to reduce the torque ripple in a permanent-magnet non-sinusoidal synchronous motor. Solutions for calculating optimal currents are deduced from geometrical considerations and without a calculation step which is generally based on the Lagrange optimization. These optimal currents are obtained from two hyperplanes. The study takes into account the presence of harmonics in the back-EMF and the cogging torque. New control schemes are thus proposed to derive the optimal stator currents giving exactly the desired electromagnetic torque (or speed) and minimizing the ohmic losses. Either the torque or the speed control scheme, both integrate two neural blocks, one dedicated for optimal currents calculation and the other to ensure the generation of these currents via a voltage source inverter. Simulation and experimental results from a laboratory prototype are shown to confirm the validity of the proposed neural approach.

A novel centroid-based diagnostic method of the power switches in five-leg Voltage Source Inverter (VSI) is proposed in this paper. Using a vectorial multi-machine description, a five-phase drive presenting an opened switch or an opened phase faults has typical operating characteristics in comparison to classical three-phase drives. Based on such characteristics, this work aims to provide a simple and robust diagnostic process for switches fault regardless of the shape of the back-EMFs (harmonic components) and the transient states due to the load variation. Original theoretical developments are presented. Experimental results are shown to validate the proposed strategy.

In the context of future Permanent Magnet Synchronous Machines (PMSMs) with a high number of phases (>7) in integrated drives, this paper proposes several control strategies when multiphase PMSMs with non-sinusoidal back electromotive forces (back-EMFs) operate in healthy and open-circuit faults. In all operation modes, the considered constraint on current is related to the maximum root mean square (RMS) current allowable in one phase of the machine. The constraint on voltage limits the maximum peak value of the phase voltage determined by the DC-bus voltage of the converter. When one or two phases are opencircuited, to maximize torque and respect the constraints, new current references obtained by several proposed methods in rotating and natural frames are imposed to the machine. Due to the non-sinusoidal waveform of back-EMFs and the considered constraints, numerical computations based on analytical formulations are required to obtain maximal torque-speed characteristics, including the flux-weakening operation. The usefulness of the proposed strategies is verified by numerical and experimental results.

is an open access repository that collects the work of Arts et Métiers ParisTech researchers and makes it freely available over the web where possible. Abstract -This paper analyzes two fault-tolerant dualmultiphase motor drives, a series connected topology and a standard H-bridge topology. Previous studies have shown that the series connected topology is appropriate to an aerospace application and has lower peak current in degraded mode in comparison with the H-bridge topology, which may consequently diminish the system's weight and cost. This paper extends the study to compare different control strategies of these structures under two fault conditions: short-circuit of an inverter's switch and an open-phase of the machine. The control strategies analyzed in this paper do not impact the fundamental current or the torque generation, but the amplitudes of some harmonics in degraded mode are expected to be narrowed down in order to reduce the inverter's size. Some analyses of maximum voltage and peak current in degraded mode have been used for inverter dimensioning. Experimental results are shown and compared to the simulated ones to confirm the validity of this study.

For non-sinusoidal electromotive force (NS-EMF) multiphase machines, this paper proposes a new strategy and control scheme to guarantee smooth torque under an open-phase fault. Notably, the conventional proportional-integral (PI) controllers implemented for vector control in healthy mode can be used in the faulty mode. The strategy is based on reduced-order transformations while the control scheme applies a simple artificial intelligence algorithm using a specific online-trained Adaptive Linear Neuron (ADALINE). Indeed, the inputs of ADALINE require the knowledge of rotor position and NS-EMF harmonic rank to optimize the learning time. The proposed strategy and control scheme are tested on a seven-phase machine with a strong Total Harmonic Distortion (THD) of NS-EMFs, containing numerous harmonics Hk (THD=38% with 100% H1, 32.3% H3, 9.4% H7, 12.5% H9, 10.3% H11). Numerical and experimental results are presented in this paper. This paper is accompanied by a video demonstrating the experimental results. 1

This paper proposes a procedure that is suitable for experimental investigation of real-time open-switch and open-phase faults diagnosis of a five-leg Voltage Source Inverter (VSI) feeding a five-phase bi-harmonic Permanent Magnet Synchronous Machine (5-Φ B-PMSM). The algorithm is based on the specific characteristics of multiphase machines, which allows inverter fault detection with sufficient robustness of the algorithm in the presence of fundamental and third harmonic components. Firstly, the inverter fault effects analysis is achieved in the characteristic subspaces of the five-phase PMSM. Specificities that are interesting for the elaboration of a real-time Fault Detection and Identification (FDI) process are highlighted. Original and particular algorithms are used for an accurate two-dimensional normalized fault vector extraction in a defined fault reference frame. This frame is dedicated only for fault detection and identification. To ensure the high immunity of the FDI process against transient states, a particular normalization procedure is applied. The normalized diagnostic signals are formulated from the defined frame and others variables derived from the reference and measured currents. Simulation and experimental results of open-switch and open-phase faults are provided to validate the proposed algorithm. Index TermsSpace vector theory, multiphase systems, fault diagnosis, five-phase permanent magnet synchronous machine, inverter open-circuit fault.

Please cite this article in press as: O. Fall, et al., Variable speed control of a 5-phase permanent magnet synchronous generator including voltage and current limits in healthy and open-circuited modes, Electr. Power Syst. Res. (2016), http://dx. a b s t r a c tThis paper proposes a novel variable speed control strategy of a particular 5-phase Permanent Magnet Synchronous Generator (PMSG) in healthy and faulty modes by taking into account the constraints on voltages and currents. These constraints are related to the converter and machine design. The considered faults are open-circuited phases (one phase, two adjacent phases and two non-adjacent phases). A variable speed control strategy is presented, including flux weakening operations. Based on analytical formulations, a numerical computation is proposed to bring out the torque−speed characteristics. This method allows the determination of the current references which ensure the functioning of a 5-phase PMSG at variable speed while keeping phase voltages and currents below their limits. Theoretical, numerical and experimental results are presented. These results are compared in order to validate the proposed approach.