Applying Transformation to Reduce Model Sizes for Constrained Optimization Problems

Many practical problems in the areas of manufacturing, distribution, supply chain logistics, transportation, urban logistics, etc. are formulated in the form of COPs (Constrained Optimization Problems). These problems are most often discrete in nature and are characterized by large size, i.e. a large number of constraints and decision variables. This results in their usually high computational complexity which significantly hinders the use of exact methods to solve them. This paper presents the author's transformation of the modeled problem using the structure of the modeled problem and data instances to reduce its size, which consequently results in reduced computation time. Computational experiments were also conducted to test the effectiveness of the proposed approach against mathematical programming methods for the selected COP-variant of the vehicle routing problem.