Fulltext

Concurrent optimization of harvesting and road network layouts under steep terrain

Bont, Leo ; Heinimann, Hans ; Church, Richard

In: Annals of Operations Research, 2015, vol. 232, no. 1, p. 41-64

Aggiungi alla tua lista
    Titre
    • Concurrent optimization of harvesting and road network layouts under steep terrain
    Auteur
    • Bont, Leo. Institute of Terrestrial Ecosystems, ETH Zürich, Universitaetstrasse 16, CHN K 75.1, 8092, Zurich, Switzerland
    • Heinimann, Hans. Institute of Terrestrial Ecosystems, ETH Zürich, Universitaetstrasse 16, CHN K 72.2, 8092, Zurich, Switzerland
    • Church, Richard. Department of Geography, University of California, Santa Barbara, 93106-4060, Santa Barbara, CA, USA
    Type de document
    • Postprint
    Langue
    • Inglese
    Publié dans
    • Annals of Operations Research, 2015, vol. 232, no. 1, p. 41-64. Springer US; http://www.springer-ny.com
    Autre version électronique
    Identifiant OAI-PMH
    • oai:doc.rero.ch:331271
    Summary
    Timber extraction is based on two transportation modes—off-road and on-road—that are connected by a set of landing nodes. Forest operations planning that is oriented toward harvesting consists of concurrently locating a truck road network, designating landing/loading sites along that network, and allocating stands to specific harvest technologies (e.g., cable roads). In central Europe, this problem has, until recently, been solved primarily with rules of thumb. By contrast, only a few attempts, designed for plantation systems, have been made to find the mathematical optima. Here, we present a modeling approach to identify a minimal-cost solution for this problem of laying out truck roads and cable roads when the terrain is steep. This technique is based on a Mixed Integer Linear Programming formulation. Our approach is as good as or better than state-of-the-art methods. Here, the overall costs of harvesting and roading were decreased by about 7% compared with techniques that called for a heuristic solution only. Depending upon parameter choices, we also determined that a computing time ranging from 4 min to 8 hrs was required when assessing a logging area of 4.3km2.