Timothée Gerber

PhD student at Gipsa-lab, France

Portrait of Timothée Gerber

About me

I graduated in 2012 from the french engineering school ENSE3 in the signal processing department. Since then, I am a PhD student at Gipsa-lab, involved in the SAIGA team. Perhaps you want to know what I am working on? Or maybe you came here to have a look to my publications. You can also check the tools I am releasing under free licence.

signal processing, spectral analysis, system monitoring, time-frequency representation
timothee.gerber [at] gipsa-lab.grenoble-inp.fr
+33 (0)4 76 82 63 13
PhD supervisor
Nadine Martin & Corinne Mailhes
Last webpage update
October 15, 2015


Dynamic tracking of modulated components under non-stationary random excitation
Application to automatic condition monitoring of failures in wind farms

The energy produced by wind farm is becoming more and more important worldwide. Its impact on the electrical grid grows every day. The economical and ecological interests reside in the availability of the equipment. Therefore, the capacity to plan the maintenance based on a reliable condition monitoring of the equipment is a crucial challenge. The KAStrion project takes up this challenge.

My thesis is part of KAStrion project and has the objective of designing and developing a part of the condition monitoring system with signal processing methods. More precisely, I work on the non-stationarity issue due to the wind variations and on the surveillance of the system along time. In the end, the main idea is to track relevant features in order to monitor the turbine components.


IEEE Trans. on Industrial Electronics Time-Frequency Tracking of Spectral Structures Estimated by a Data-Driven Method

Timothée Gerber, Nadine Martin & Corinne Mailhes

The installation of a condition monitoring system aims to reduce the operating costs of the monitored system by applying a predictive maintenance strategy. However, a system-driven configuration of the condition monitoring system requires the knowledge of the system kinematics and could induce lots a false alarms because of predefined thresholds. The purpose of this paper is to propose a complete data-driven method to automatically generate system health indicators without any a priori on the monitored system or the acquired signals. This method is composed of two steps. First, every acquired signal is analysed: the spectral peaks are detected and then grouped in more complex structure as harmonic series or modulation sidebands. Then, a time-frequency tracking operation is applied on all available signals: the spectral peaks and the spectral structures are tracked over time and grouped in trajectories, which will be used to generate the system health indicators. The proposed method is tested on real-world signals coming from a wind turbine test rig. The detection of a harmonic series and a modulation sideband reports the birth of a fault on the main bearing inner ring. The evolution of the fault severity is characterised by three automatically generated health indicators and is confirmed by experts.

Insight AStrion strategy: from acquisition to diagnosis. Application to wind turbine monitoring

Zhong-Yang Li, Timothée Gerber, Marcin Firla, Pascal Bellemain, Nadine Martin & Corinne Mailhes

This paper proposes an automatic procedure for condition monitoring. It presents a valuable tool for the maintenance of expensive and spread systems, such as wind turbine farms. Thanks to data-driven signal processing algorithms, the proposed solution is fully automatic for the user. The paper briefly describes all the steps of the processing, from preprocessing of the acquired signal to interpretation of the generated results. It starts with an angular resampling method with speed measurement correction. Then comes a data validation step, in both the time/angular and frequency/order domains. After the preprocessing, the spectral components of the analysed signal are identified and classified into several classes, from sine wave to narrowband components. This spectral peak detection and classification allows the harmonic and side-band series to be extracted, which may be part of the signal spectral content. Moreover, the detected spectral patterns are associated with the characteristic frequencies of the investigated system. Based on the detected side-band series, the full-band demodulation is performed. At each step, the diagnosis features are computed and dynamically tracked, signal by signal. Finally, system health indicators are proposed to conclude the condition of the investigated system. All the steps mentioned create a self-sufficient tool for a robust diagnosis of mechanical faults. The paper presents the performance of the proposed method on real-world signals from a wind turbine drive train.

CMMNO Monitoring based on time-frequency tracking of estimated harmonic series and modulation sidebands

Timothée Gerber, Nadine Martin & Corinne Mailhes

A condition monitoring system is a key element in a predictive maintenance strategy allowing to reduce the operating costs of the monitored system. However, the system-driven generation of health indicators requires the knowledge of the system kinematics and the configuration of thresholds which may induce lots of false alarms. In this paper, we propose a generic and data-driven method to automatically generate system health indicators without any a priori knowledge on the monitored system or the acquired signals. The proposed method is based on the automatic detection of spectral content characterising every acquired signal. Within these successive spectral contents, peaks, harmonics series and modulation sidebands are then tracked over time and grouped in time trajectories which will be used to generate the system health indicators

CM & MFPT Identification of harmonics and sidebands in a finite set of spectral components

Timothée Gerber, Nadine Martin & Corinne Mailhes

Spectral analysis along with the detection of harmonics and modulation sidebands are key elements in condition monitoring systems. Several spectral analysis tools are already able to detect spectral components present in a signal. The challenge is therefore to complete this spectral analysis with a method able to identify harmonic series and modulation sidebands. Compared to the state of the art, the method proposed takes the uncertainty of the frequency estimation into account. The identification is automatically done without any a priori, the search of harmonics is exhaustive and moreover the identification of all the modulation sidebands of each harmonic is done regardless of their energy level. The identified series are characterized by criteria which reflect their relevance and which allow the association of series in families, characteristic of a same physical process. This method is applied on real-world current and vibration data, more or less rich in their spectral content. The identification of sidebands is a strong indicator of failures in mechanical systems. The detection and tracking of these modulations from a very low energy level is an asset for earlier detection of the failure. The proposed method is validated by comparison with expert diagnosis in the concerned fields.

CM & MFPT Consequences of non-respect of the Bedrosian theorem when demodulating

Christian Pachaud, Timothée Gerber, Marcin Firla, Nadine Martin & Corinne Mailhes

Vibration data acquired during system monitoring periods are rich in harmonics characterizing the presence of several mechanical parts in the system. Periodic variations of the torque or of the load create modulation sidebands around those harmonics. Even if the energy impact of the sidebands is small compared to the total energy of the signal, they are strong indicators of failures in mechanical systems. Unfortunately, these effects are of little concern in most condition monitoring systems. When considering the problem from a signal processing point of view, the demodulation of those sidebands allows for a time visualization of the modulating functions which are a precise image of the torque or the load variations. This demodulation can be done on the analytical signal directly derived from the original data. But to do that, data and specifically its spectrum should respect some constraints. The purpose of this paper is to underline those often neglected constraints. In particular, the respect of the non-overlapping condition in the Bedrosian theorem is discussed for signals and modulation rates that can be encountered on rotating machines. The respect of the constraints depends on the monitored phenomenon (e.g., gear mesh, rotating shaft), the modulation phenomenon (e.g., belt frequency, rotor current) and the type of medium (e.g., vibrations, electrical current). In the case where the constraints are not satisfied, we explain the consequences in terms of signal processing. These results are illustrated by an industrial case study.

ISMIR Professionally-Produced Music Separation Guided by Covers

Timothée Gerber, Martin Dutasta, Laurent Girin & Cédric Févotte

This paper addresses the problem of demixing professionally produced music, i.e., recovering the musical source signals that compose a (2-channel stereo) commercial mix signal. Inspired by previous studies using MIDI synthesized or hummed signals as external references, we propose to use the multitrack signals of a cover interpretation to guide the separation process with a relevant initialization. This process is carried out within the framework of the multichannel convolutive NMF model and associated EM/MU estimation algorithms. Although subject to the limitations of the convolutive assumption, our experiments confirm the potential of using multitrack cover signals for source separation of commercial music.


LaTeX Makefile

Timothée Gerber

A cool Makefile creating awesome PDF files from amazing LaTeX documents with wonderfull SVG images in an extraordinary clean directory -- So swag!


Info-info IT architecture and services

Timothée Gerber

Description of the IT architecture of Gipsa-lab and the associated services you can access to.

Info-info Computation server usage

Timothée Gerber

Local and remote connections to the computation servers, file transfers and some good practices.

Info-info Versioning with Git

Raphaël Bacher

Save the history of your codes and text documents.

Info-info Don't be afraid of the console anymore!

Timothée Gerber

Basic knowledge and tips to use efficiently the console.