Conventional approaches to the estimation problem add initial/boundary conditions to imbed the ODE in a system which can be solved explicitly in order to generate comparison solutions for insertion in an optimization procedure. However, this requires a priori information about system stability to guarantee that the imbedding choices are appropriate. This problem can be avoided by Lagrangian methods, and orthogonal cyclic reduction has a fascinating role in reducing the resulting constrained optimization problems. Augmented Lagrangian methods lend themselves to the solution of the reduced problem

MINISIMPOSIUM SESSION: 1. Optimization problems in differential equations