A Linear Programming, Tabu Search method for solving forest-level bucking optimization problems

Abstract

The forest-level bucking optimization problem consists of determining the bucking program that maximizes the global profit subject to demand constraints, merchandising restrictions, and forest-estate, A Linear Programming-Tabu Search (LP/TS) method was developed for obtaining near-optimal, rule-based solutions to this problem. TS is used for generating an efficient bucking rule for every stand and merchandising restrictions considered. The LP model is used for determining the area of each stand that should be allocated to each merchandising alternative land by doing so selecting the bucking rule defined for the corresponding specifications). The goal is to capture the synergy that can be obtained by combining logs from different sources that are produced applying different production guidelines. The results obtained for several estate, price, and demand conditions indicate that the LP/TS method consistently achieves an efficiency level of approximately 97% compared to the optimal solution for the same scenarios. Considering that the simplicity of the bucking rules offers an important operational advantage (Laroze and Greber 1997), the method developed for this study represents an interesting approach for solving forest-level bucking problems.