In this paper a new line search for a Newton-Raphson learning control algor
ithm is presented. Theorems and rigorous proofs of its increased robustness
over existing line searches are provided, and numerical examples are used
to further validate the theorems.