A basis decomposition linear programming approach to limit analysis

A solution technique for limit analysis linear programming (LP) problems is presented. The method rests on the ordinary simplex method based on Winkler's decomposition scheme. It takes full advantage of a special data structure often encountered in LP models used in storage requirements and in computation time. The main features of the proposed method are outlined. Then, an analytical comparison of the computational efficiency of this technique vs. the ordinary simplex method (and the Dantzig-Wolfe algorithm to a limited extent) is given. Numerical results are also presented and discussed.