Painleve analysis of nonlinear evolution equations - an algorithmic method

This paper refines existing techniques into an algorithmic method for deriving the generalization of a Lax Pair directly from a general integrable nonlinear evolution equation via the use of truncated Painleve expansions. The resulting algorithm is also applicable to multicomponent integrable systems, and is thus expected to be of great value for complicated variants of such systems in various applications areas. Although a related method has existed for simple scalar integrable evolution equations for many years now, nevertheless no systematic procedure has been given that would work in general for scalar as well as for multicomponent systems. The method presented here largely systematizes the necessary operations in applying the Painleve method to a general integrable evolution equation or system of equations. We demonstrate that by following the concept of enforcing integrability at each step (referred to here as the Principle of Integrability), one is led to an appropriate generalization of a Lax Pair, although perhaps in nonlinear form, called a "Lax Complex". One new feature of this procedure is that it utilizes, as needed, a technique from the well-known Estabrook-Wahlquist method for determining necessary integrating factors. The end result of this procedure is to obtain a Lax Complex, whose integrability condition will contain the original evolution equation as a necessary condition. This in itself is sufficient to ensure that the Lax Complex may be used to construct Backlund solutions of the evolution equation, to obtain Darboux Transformations, and also to obtain Hirota's tau functions, in a manner analogous to the procedure for single component systems. The additional problem of finding a general procedure for the linearization of any Lax Complex is not treated in this paper. However, we do demonstrate that a particular technique, which can be derived self-consistently from the Pain leve-Backlund equations, has proven to be sufficient so far. The Nonlinear Schrodinger equation is used to illustrate the method, and then the method is applied to obtain, for the first time via the Painleve method, a Lax Complex for the vector Manakov system. Limitations in the algorithm remain, especially for cases with more than one principal branch, and these are briefly mentioned as directions for future work. (c) 2005 Elsevier Ltd. All rights reserved.