Luc Loron

Torque ripple reduction in Permanent Magnet Synchronous Machines using angle-based iterative learning control

Permanent Magnet Synchronous Machines (PMSM) are used in many applications, particularly in high-performance drive systems. However, one inherent problem of PMSM is its parasitic undesired torque ripple, which reduces the PMSM performances. Several kinds of parasitic torque ripples are periodic functions of the rotor position. To reduce them, a method called Iterative Learning Control (ILC) seems well suited, but is efficient for only one constant speed chosen by the designer (and all the integer multiples of this speed). To design a solution for a varying speed, an ILC variant named angle-based ILC is proposed, which no longer uses time as the reference, but the angular position. This angle-based ILC can simply be implemented and does not require a heavy computational load. Simulations of a PMSM variable speed drive system and experiments have been done to assess the performances of the proposed technique.

Design of a lying sensor for permanent magnet synchronous machine torque ripple reduction using the iterative learning control technique

Permanent magnet synchronous machines (PMSM) are widely used for high-performance drive systems. However, an important problem for PMSMs is that parasitic torques may degrade the performances of the drive system. These torque ripples generally vary periodically with the rotor position and lead to speed ripple. To suppress these speed ripples, an iterative learning control (ILC) is used, because it is a good candidate for dealing with periodical errors. In this paper, an original approach called “lying sensor technique” is proposed and analyzed. Compared to the torque ripple reduction approaches which realize the current compensation calculation in the controller, this technique consists in modifying the feedback speed information of the sensor. ILC is integrated into this technique for computing the lying speed information. Simulation is used to check the effectiveness of this approach. Simulation results prove that the ILC lying sensor technique has a good performance.

Parameters and States Estimation with Linear Quadratic Regulator Applied to Uninterruptible Power Supplies (UPS)

This paper presents a control strategy applied for UPS with a low commutation frequency. An adaptive linear quadratic controller (LQR) for a DC/AC converter is presented. The controller gains are determined by minimizing a cost function. A recursive least square (RLS) estimator is used to identify the parameters model at different load conditions. The inductor current is estimated by a Kalman filter. The capacitor voltage and the inductor current are used as state variables. Simulation results show that the proposed strategy offers good performances for either linear and non-linear loads with low total harmonic distortions even at low frequencies making it very useful for high power applications

High Frequency Modelling of Stator Windings Dedicated to Machine Insulation Diagnosis by Parametric Identification

This paper presents a novel state-space modelling dedicated to stator insulation diagnosis of form wound electrical machines by parametric identification. The underlying idea of the proposed diagnosis approach is that there exists a wide frequency domain (from several kHz to hundreds of MHz) which has not yet been exploited by conventional electrical diagnosis tools. The originality of our work, but also its difficulty, consists in searching for the best tradeoff between simplicity for model identiflability and complexity for its diagnosis ability. This is important for diagnosis and maintenance procedures which require to interpret the parameter estimated values and theirs drifts. It is possible only if the model parameters are related closely enough to the system physical parameters. A high voltage Mosfet pulse generator with ultra fast rise time allows theoretical and experimental system identification with respect to main industrial constraints. The first experiments deal with coils taken from a 5 kV induction machine. This allows to isolate and understand the complex phenomena which occur in stator insulation during identification procedures.

A new Approach for Electrical Machine Winding Insulation Monitoring by Means of High Frequency Parametric modelling

This paper investigates a new approach for stator winding insulation diagnosis. Unlike conventional methods such as insulation resistance tests or partial discharge monitoring, this approach is based on the effective use of parametric modelling and identification. The guiding principle consists in modelling the high-frequency input-output behavior of the winding ground-wall insulation system. The originality and the difficulty of our work are to find the best tradeoff between simplicity for system identifiability and complexity for diagnosis abilities. More specifically, we try to distinguish between winding pollution (moisture, coal or metallic dust) and insulation material ageing because the maintenance procedures and their induced costs are very different. Experiments have been carried out in order to appreciate the feasibility of this approach. A high voltage MOSFET pulse generator with ultra fast rise time has been specially designed. It allows to apply an input excitation covering the wide frequency range required for the ground-wall insulation system identification. In order to isolate numerous phenomena which could be mixed together in the whole machine, the experiments first deal with one bar and one coil taken from a 5 kV induction machine. Diagnosis model parameters are identified by means of advanced identification algorithms with different moisture contents. The first results are promising

Comparative Analysis of Some Parametric Model Structures Dedicated to EDLC Diagnosis

Electrical double-layer capacitors (EDLCs) are now widely used as power buffers in energy storage systems, especially for electrical transport applications. Because of their high acquisition costs, their predictive maintenance represents a key issue for the development of electrical traction systems. This paper investigates the diagnosis of EDLCs by a parametric estimation of a dynamic continuous-time model. Indeed, the physical interpretation of their parameters simplifies the supercapacitor diagnosis and prognosis. Due to the complex electrochemical phenomena occurring in EDLC cells, modeling errors are far more important than measurement noise. Therefore, we propose to use the output error minimization techniques and the relative sensitivity functions to evaluate and compare the diagnosis potential of several model structures. This allows the use of some comprehensive comparison criteria without any assumption about the statistical characteristics of the system disturbances. An experimental accelerated aging process of 200 commercial super capacitors (Nichicon, 1F, 20 weeks long aging at 65 °C, 0 and 2.5 V dc polarization) is also performed to make statistics and illustrate the methodology.

Online monitoring of marine turbine insulation condition based on high frequency models: Methodology for finding the "best" identification protocol

This paper investigates the online monitoring of electrical machine winding insulation systems based on the parametric modeling and identification. The proposed method consists in monitoring the drift of diagnostic indicators built from in-situ estimation of high-frequency electrical model parameters. The involved model structures are derived from the RLC network modeling of the winding insulation. Because they often present an important modeling noise, we propose to use the output error method not only to estimate the model parameter values but also to evaluate their uncertainty. This approach is based on the numerical integration of the model sensitivity functions. The so-called global identification scheme is coupled with an optimization algorithm that brings the best combination of any diagnostic model structure and its excitation protocol usable in operating conditions. Experimental data recorded from an industrial wound machines are used to illustrate the methodology.

A new Approach for Form Wound Machine Goundwall Insulation Diagnosis by Means of High Frequency Model Parameter Monitoring

This paper investigates a new approach for stator winding insulation diagnosis. The guiding principle consists in modelling the input-output behavior of the groundwall insulation system in a high-frequency (HF) domain, ranging from several kHz to hundreds of MHz. The first motivation of this work is that this frequency area has been seldom used by conventional insulation diagnostic techniques such as partial discharge monitoring or time domain measurements. Another originality consists in modelling the groundwall insulation system with a diagnostic dedicated modelling approach, i.e. searching for the best tradeoff between simplicity for system identifiability and complexity for diagnosis abilities. More specifically, we try to distinguish between winding contamination (moisture, coal or metallic dust) and insulation material ageing because the maintenance procedures and their induced costs are very different for high voltage (HV)) machines ranging from a few to several kilovolts or more. At last, the theoretical and industrial constraints impose to use very efficient identification algorithms. Experiments have been carried out in order to appreciate the feasibility of this approach. A high voltage MOSFET pulse generator with ultra fast rise time has been specially designed. It allows to apply a high energy input excitation covering the wide frequency range required for the ground-wall insulation system identification. In order to isolate numerous phenomena which could be mixed together in an entire winding, the experiments first deal with stator bars and coils taken from high voltage induction machines. The first results are promising.

Easy and Efficient Tuning of PI Controllers for Electrical Drives

PI controllers are very often used because of their simplicity, but there are frequently poorly tuned. This article discusses and compares the efficiency of three analytical tuning methods for PI or IP controllers. The comparison criteria are the reference following bandwidth and the disturbance rejection versus the open-loop phase margin or the closed-loop peak resonant. The simplicity and the efficiency of the considered approaches are illustrated by the tuning of typical controllers (current and speed loops) for electrical drives

A Multirate Simulation Method for Large Timescale Systems Applied for Lifetime Simulations

This paper introduces an original approach of the modeling and simulation of multiphysics systems that exhibit a wide range of time scales. This approach will be illustrated by the simulation of the energy storage unit (ESU) of an all-electric ferry. The implementation of the models was performed using two different modeling approaches (causal and acausal) highlighting the main difficulties of modeling multiphysics systems. In order to optimize the sizing of the ESU considering its ageing, the system should be simulated for a lifetime of 20 years. For this purpose, a multirate method was developed to speed-up the simulation and the optimization by a factor of 100 or more.