摘要：A particular numerical analytical technique for solving systems of simultaneous equations which offers several advantages to the user over other numerical techniques is discussed. The Gauss-Seidel algorithm is simply an iterative technique which requires no derivatives, matrix inversion, eigenvalue computation, or any other sophisticated numerical methodology. While the technique has been used successfully by a few large scale model builders and by the authors for several commodity models, the experience gained in the use of the technique has not been generally disseminated. Obtaining convergence with such an iterative technique is critically dependent on a number of factors which are taken up in some detail by the authors.
