Mitigation techniques
Lors de l'utilisation des capteurs à modulation d'intensité, il est attendu que les variations de la puissance optique proviennent uniquement de la quantité à mesurer. Cependant ceci n'est généralement pas le cas, et de nombreuses perturbations indésirables viennent s'additionner au signal. On peut recenser trois sources importantes d'erreur :
-
Fluctuations of temperature which affect the emitting devices, so that the optical output power is not constant.
-
Leakages due to curves, connectors, modal interferences, etc.
-
Aging of the electronic components, notably for the sources, so that the emitted power decreases.
So it is necessary to resort to mitigation techniques for all sensors whose performances must remain stable. Mitigation techniques are not unique because they depend on the structure of the transducers. They are all based on the principle of adding extra information in the system in order to extract the fluctuations due to measurands. Then we will present some of the most used types of mitigation techniques, even if it is difficult to draw up an exhaustive classification.
Spatial division
The signal is divided into at least two and injected into identical fibers. The fibers are introduced at the same place, the light intensity transmitted in each of them will be modulated by disturbances and by the measurand, the signal can be inferred by combining all the measured intensities. Here for example two of the most frequently used equations:
where I1 and I2 are the intensity coming from each fiber. This technique is often used for systems with coupling sensors [18]. This simple technique depends a lot on the kind of sensors and it is difficult to apply it with intrinsic sensors (i.e. sensors which modulate the light intensity without making it coming out of the fiber).
Reference fiber
This technique can be considered as an exceptional case of spatial division. An extra fiber which transmits a light signal follows the same path as the fiber with the sensor. This signal is exposed to the same leakages and external disturbances, so that you can use it as a signal of reference [19]. The asset of this technique is to make it possible to follow the drift of the light power emitted by the source, which is critical for the intensity modulation sensor.
Wavelength of reference
In its simplest configuration, this technique uses two optical sources with thin spectral widths and with two different wavelengths. We can use a source with a large spectral width instead but you will need to use filters in order to assess the information coming from each wavelength. If we assume that the leakages induced by the system are independent of the wavelength but that the amplitude of the measurand we want to detect does depend on it, then it is possible to use the results of the spatial division. Provided that we choose the correct wavelengths, it is sometimes possible to take one of the two wavelengths as a reference.
Intrinsic characteristics of the measurand:
If the quantity you have to measure is alternating (for example vibrations), you simply have to make a spectrum analysis on the electric signal coming from the sensor, so that the ratio between the alternating and continued components will directly give the amplitude of the measurand. In other cases, the measurand will be proportional to the continued component. It is also possible to use modulated sources [20].