A robust line search for underdetermined learning control

In this paper a new line search for a Newton Rhaphson learning control algorithm 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.