Learning internal representations by error propagation
Parallel distributed processing: explorations in the microstructure of cognition, vol. 1
Hi-index | 0.00 |
Artificial neural network (ANN) isemployed to present a fiber-optic ice sensor (FOIS) with wide measurement range. Comparing with existing FOIS signal processing methods, this approach is not limited by the double-valued problem of output curve. Instead, it performs a measurement range from front-slope areas to back-slope areas. Moreover, this approach also handles the nonlinear problem of the sensor. As an application of the ANN, a calibration experiment platform is set up. The training samples are employed to train the ANN, and the testing samples are applied to surveil the predict ability of the ANN. The results obtained demonstrate the applicability of the proposed approach.