Nicolas Heraud

Diagnosis of the Doubly-Fed Induction Generator of a Wind Turbine

This paper presents a method of identifying and estimating gross errors for linear dynamic systems. The method is applied to wind power, in particular the doubly-fed induction generator. Measurements have errors, but it is possible to reduce the effect of such errors on control by exploiting relationships between the different variables of the system. Such analysis is called ‘data validation’. Data validation uses a mathematical model, based on equations, to simulate the real dynamic system. An analysis of systematic errors is made using a measurement test. The method has the potential to support the on-line multiparameter data analysis, and hence maintenance, of complex systems, such as wind turbines.

Data reconciliation and gross error detection applied to wind power

Abstract The current paper presents a method to identify and estimate gross errors for linear dynamic systems using polynomial approximation. The method presented in a previous paper which uses linear dynamic reconciliation, is extended to allow the dynamic estimation of errors. This method is applied to the doubly fed induction generator (DFIG) of a wind turbine. The technique is validated on an experimental system used to emulate the working of the wind turbine.

Hourly Solar Irradiations Estimation : From Horizontal Measurements to Inclined Data

We quantify the accuracy of various well-known models of the literature: the seven first type of models calculate diffuse horizontal irradiations from global ones and the fifteen second type models compute global irradiation on inclined planes from diffuse and global components on a horizontal surface. Then, the two classes of models are combined and the adequacy of each association is calculated for experimental hourly data collected on a French Mediterranean site. We compared these combinations with the Olmo model transforming directly global irradiations on tilted planes from horizontal ones.

Condition monitoring & fault diagnosis system for Offshore Wind Turbines

Due to the technological development, the electronic power progress and economic stake, through the use of Wound Rotor Induction Motor (WRIM) has taken more and more places in different domains (transport, energy production, electric drive..,) thanks to their robustness, efficiency and lower costs. Despite the performed work researches and the improvement that has been brought, these machines still remain the potential seats of failures both in stator and rotor levels. Consequently, WRIM faults detection is currently one of the centers of interest of several researches of both academic and industrial laboratories. In fact, this article addresses this problem by the use of Principal Components Analysis (PCA) for faults detection in Offshore Wind Turbine Generator (OWTG). An accurate analytic modeling of healthy and faulted OWTG is suggested to perform the data matrix needed for PCA method. Tests were achieved using a numeric simulator on Matlab/Simulink software. Analysis of OWTG simulation proves the efficiency of PCA method. Several simulation results will be presented and discussed.

Coupling hydro and wind electricity production by water - pumping storage

We quantify the capability of coupling wind farm and hydraulic plant, in view to produce energy without fluctuations, like wind farm alone. In fact, we suppose the hydraulic plant is like a buffer, and the wind farm fed this buffer. The aim is to satisfy a certain quantity of energy, more particularly, a percentage of the peak load. We can so optimize the system : power of the wind farm with the power of the hydraulic plant.

Observability and sensor positioning for processes described by linear-bilinear equations

The control of a process requires a minimum of representative information on its operating state. This information can be obtained by both direct sensors or indirectly using mathematical models. It frequently occurs that the number of direct measures is not sufficient to allow unmeasured variables to be deduced and additional sensors have to be installed to render the required variables observable. This paper presents a new technique to place sensors in systems described simultaneously by linear and bilinear equations. It takes into account placement constraints and a complete observability is obtained. Moreover, it is pointed out that from an observability point of view a bilinear system may be studied by considering a succession of linear systems.

Modeling and faults detection of small power wound rotor induction machine

In this work, we focus specifically in winding faults that may occur to the small power wound rotor induction machine. Inter-turn short-circuit faults are often critical and must be detected earlier with a good precision in order to inform a supervisor of the existence of a system failure to limit the risks incurred by the material and human located around the machine in fault. Using winding function approach for modeling the machine, two models of inter-turn short-circuit are selected: the first model isolates the turns short-circuited to form a new coil and the faulty coil has a new number of turns minus the number of turns short-circuited. In this faulty coil appears a current called circulating current. The second model has a same aspect of the first model but in this time, the point of contact of the inter-turn short-circuited is modelised by a resistance. In order to ensure the performance of the used methods, a validation using another model, electric circuits magnetically coupled followed by experimental validation is effectuated. The obtained results validate the proposed techniques and demonstrate the effectiveness of the developed approach.

Performances of the PCA Method in Electrical Machines Diagnosis Using Matlab

Nowadays, faults diagnosis is almost an inevitable step to be maintained in the optimal safety operating of every physical system. Electrical machines, main elements of every electromechanical system, are among the research topics of many academic and industrial laboratories because of the importance of their roles in the industrial process. Lots of technologies of these machines are old and well controlled. However, they remain the seat of several electrical and mechanical faults [1-4]. Thus, this article deals with faults detection of a wound rotor synchronous machine (WRIM) by the principal component analysis (PCA) method.

Fault diagnosis of variable speed wind turbine

This paper addresses the fault detection and isolation issue of wind turbine systems. The underlying problem is very critical to enhance the reliability, the availability and reduce the cost of maintenance of wind turbines. In this work, we use a doubly fed induction generator and a nonlinear observer is utilized to generate the residual for fault detection purpose.

Cost-optimal design of reliable sensor networks extended to multi-linear system

In this paper, we develop a single procedure that can be used both linear and multi-linear systems to place sensors with low cost and with respect to observability and reliability properties. A reference to similar works limited to linear cases is made as an example. Simulation results demonstrate the efficiency and consistency of the proposed methodology.