Viability of Artificial Neural Networks for Widen the Measurement Range of Interferometric Sensors
Ana Dinora Guzman–Chavez, Everardo Vargas-Rodríguez, Bertha Laura Vargas-Rodriguez, Mario Alberto García-Ramírez
Abstract
In this letter, the viability of Artificial Neural Networks (ANNs) for widen the overall measurement range of optical sensors, based on interferometric arrangements, is demonstrated. Moreover, it is proven that by using ensembles of ANN regressions the traditional <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$2\pi $ </tex-math></inline-formula> ambiguity of interferometric sensors can be overcome. Here, a simple temperature sensor was implemented and from measured spectra a dataset of features was formed. For this sensor, the measurement range was increased by a factor of 2 and the achieved mean absolute error was <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$0.17{^{\circ }}\text{C}$ </tex-math></inline-formula> . Finally, it is shown that based on numerical results it can be expected that by using these types of ensembles the measurement range and the precision of predictions can be further enhanced.