In an attempt to facilitate the design and implementation of memory-less nonlinear sensors, the signal reconstruction schemes are analyzed and necessary modifications are proposed to improve the accuracy and minimize errors in sensor measurements. The problem of recovering chirp signal from the distorted nonlinear output is considered and an efficient reconstruction approach is developed. Model uncertainty is a serious issue with any model-based algorithms and a novel technique, which uses a norminal model instead of an accurate model and produces the results that are robust to model uncertainty, is proposed.

This content is only available via PDF.
You do not currently have access to this content.