Theoretical and practical aspects of the convergence properties of the dynamically defined reaction path method

In spite of much work on path‐following methods, a solid mathematical foundation (especially convergence conditions and their practical measures) are lacking in most cases. In our previous articles the general theory of a new global searching procedure, the dynamically defined reaction path (DDRP) method, its rigorous mathematical formulation, the algorithm, a practical computation program, and some applications to abstract mathematical functions and simple chemical examples have been presented. In this article we give a short theoretical description and some practical criteria and measures for the convergence of the method and illustrate the principles and uses by numerical mathematical and chemical examples. © 1996 by John Wiley & Sons, Inc.