A multi-objective optimization strategy for timber forwarding in cut-to-length harvesting operations

Abstract

Optimization of forwarding work on the trail network created by the harvester in cut-to-length operations carries much potential for increasing the efficiency and reducing the environmental impact of timber removal. While several approaches to this optimization task have been proposed, existing methods focus on optimizing a single objective at a time. In this work, a heuristic multi-objective optimization strategy for the problem is presented, where the route length, operation duration, and soil damage can be optimized simultaneously. The strategy is based on a tailored version of ant colony optimization and linear scalarization of the full multi-objective problem. The relative importance of the objectives can be set via tunable weights, and the approach is of a modular nature. Results on a realistically sized clearcut harvest site indicate that the strategy produces promising results and gives the expected qualitative behavior when optimizing either a single objective or two objectives at a time.