A two-phase optimization procedure for integer programming problems

This paper presents a simple two-phase method for optimizing integer progra mming problems with a linear or nonlinear objective function subject to mul tiple linear or nonlinear constraints. The primary phase is based on a vari ation of the method of steepest descent in the feasible region, and a hem-s titching approach when a constraint is violated. The secondary phase zeros on the optimum solution by exploring the neighborhood of the suboptimum fou nd in the first phase of the optimization process. The effectiveness of thi s method is illustrated through the optimization of several examples. The r esults from the proposed optimization approach are compared to those from m ethods developed specially for dealing with integer problems. The proposed method is simple, easy to implement yet very effective in dealing with a wi de class of integer problems such as spare allocation, reliability optimiza tion, and transportation problems. (C) 2001 Elsevier Science Ltd. All right s reserved.