Applying ant colony optimization metaheuristic to solve forest transportation planning problems with side constraints

摘要

Forest transportation planning problems (FTPP) have evolved from considering only the financial aspects of timber management to more holistic problems that also consider the environmental impacts of roads. These additional requirements have introduced side constraints, making FTPP larger and more complex. Mixed-integer programming (MIP) has been used to solve FTPP, but its application has been limited by the difficulty of solving large, real-world problems within a reasonable time. To overcome thislimitation of MIP, we applied the ant colony optimization (ACO) metaheuristic to develop an ACO-based heuristic algorithm that efficiently solves large and complex forest transportation problems with side constraints. Three hypothetical FTPP were created to test the performance of the ACO algorithm. The environmental impact of forest roads represented by sediment yields was incorporated into the economic analysis of roads as a side constraint. Four different levels of sediment constraints were analyzedfor each problem. The solutions from the ACO algorithm were compared with those obtained from a commercially available MIP solver. The ACO solutions were equal to or slightly worse than the MIP solution, but the ACO algorithm took only a fraction of thecomputation time that was required by the MIP solver.