Eight heuristic planning techniques applied to three increasingly difficult wildlife planning problems

Abstract

Chung, W. 2002. Eight heuristic planning techniques applied to three increasingly diffi cult wildlife planning problems. Silva Fennica 36(2): 561-584.As both spatial and temporal characteristics of desired future conditions are becoming important measures of forest plan success, forest plans and forest planning goals are becoming complex. Heuristic techniques are becoming popular for developing alternative forest plans that include spatial constraints. Eight types of heuristic planning techniques were applied to three increasingly diffi cult forest planning problems where the objective function sought to maximize the amount of land in certain types of wildlife habitat. The goal of this research was to understand the relative challenges and opportunities each technique presents when more complex diffi cult goals are desired. The eight heuristic techniques were random search, simulated annealing, great deluge, threshold accepting, tabu search with 1-opt moves, tabu search with 1-opt and 2-opt moves, genetic algorithm, and a hybrid tabu search / genetic algorithm search process. While our results should not be viewed as universal truths, we determined that for the problems we examined, there were three classes of techniques: very good (simulated annealing, threshold accepting, great deluge, tabu search with 1-opt and 2-opt moves, and tabu search / genetic algorithm), adequate (tabu search with 1-opt moves, genetic algorithm), and less than adequate (random search). The relative advantages in terms of solution time and complexity of programming code are discussed and should provide planners and researchers a guide to help match the appropriate technique to their planning problem. The hypothetical landscape model used to evaluate the techniques can also be used by others to further compare their techniques to the ones described here.

Tree-Level Harvest Optimization for Structure-Based Forest Management Based on the Species Mingling Index

“…Given that the harvested trees are the main decision, formulation of this problem through mixed integer non-linear programming was deemed quite difficult. Therefore, threshold accepting was chosen due to its speed and reputation for producing high-quality results in comparison with other heuristics [29]. Further, the results from the heuristic were generated from random initial solutions in order to develop statistically independent samples.…”

“…Threshold accepting was developed initially by Dueck and Scheuer [28] and has been applied to forest-level optimization problems in [29][30][31][32], among others. Threshold accepting is an iterative search process for developing near-optimal solutions to problems.…”

Possibilities to aggregate raster cells through spatial optimization in forest planning

“…Threshold accepting (TA) is a deterministic version of another local search method, simulated annealing. Threshold accepting has been shown to be able to produce as good, or even better solutions, than simulated annealing (Dueck and Scheuer 1990, Bettinger et al 2002, Pukkala and Heinonen 2006. In Monsu software, TA uses the best of a set of random combinations of calculation units' treatment schedules as the initial solution.…”

Stand-Level Forest Management Planning Approaches

“…The advantage of heuristics is mainly due to the quickness of generating very good and feasible solutions to complex problems, once the heuristic is developed appropriately. Previous research including Bettinger et al [48] and Pukkala and Miina [31] have shown in forest management that quality solutions can be obtained using heuristics. If the quality of the solution is deemed high compared to exact solution techniques, then heuristics are the preferred solution technique.…”