A Fiber-Optic Aircraft Icing Detection System Using Double Optical Paths Based on the Theory of Information Fusion

A method to resolve the single-valued problem of the traditional fiber-optic icing sensor is presented.Based on the theory of information fusion,an improved measuring configuration using double optical paths is employed to overcome the optical path disturbances.An inverse model of the sensor is obtained based on neural network trained by the sensor's output and the actual sensed ice thickness.The experiment results show that the method can largely widen the measuring range and enhance the measuring accuracy.