Performance optimization of bundled fiber optic displacement sensors (original) (raw)

Smart Sensor Phenomena, Technology, Networks, and Systems Integration 2012, 2012

Abstract

ABSTRACT Bundled intensity-modulated fiber optic displacement sensors offer high-speed (kHz-MHz) performance with micrometer-level accuracy over a broad range of axial displacements, and they are particularly well-suited for applications where minimally invasive, non-contacting sensing is desired. Furthermore, differential versions of these sensors have the potential to contribute robustness to fluctuating environmental conditions. The performance limitations of these sensors are governed by the relationship between axial displacement and measured power at the locations of receiving fibers within a bundled probe. Since the propagating transmission's power level is spatially non-uniform, the relative locations of receiving fibers within a bundled probe are related to the sensor's output, and in this way fiber location is related to sensor performance. In this paper, measured power levels are simulated using a validated optical transmission model, and a genetic algorithm is employed for searching the intensity-modulated bundled displacement sensor's design space for bundle configurations that offer high-overall combinations of desired performance metrics (e.g., linearity, sensitivity, accuracy, axial displacement range, etc...). The genetic algorithm determines arrangements of fibers within the bundled probe that optimize a performance-based cost function and have the potential to offer high-performance operation. Multiple converged results of the genetic algorithm generated using different cost function structures are compared. Two optimized configurations are prototyped, and experimental sensor performance is related to simulated performance levels. The prototypes' linearity, sensitivity, accuracy, axial displacement range, and sensor robustness are described, and sensor bandwidth limitations are discussed. This paper has been approved by Los Alamos National Laboratory for unlimited public distribution (LA-UR 12-00642).

Erik Moro hasn't uploaded this paper.

Let Erik know you want this paper to be uploaded.

Ask for this paper to be uploaded.