Solving the integrated forest harvest scheduling model using metaheuristic algorithms

Author(s):  
Nader Naderializadeh ◽  
Kevin A. Crowe ◽  
Melika Rouhafza
FLORESTA ◽  
2010 ◽  
Vol 40 (3) ◽  
Author(s):  
Lucas Rezende Gomide ◽  
Julio Eduardo Arce ◽  
Arinei Lindbeck da Silva

O objetivo do estudo foi aplicar restrições espaciais de adjacência no agendamento da colheita florestal em um modelo tradicional de planejamento florestal. Foi analisado e comparado o impacto no VPL (valor presente líquido) e na produção volumétrica de madeira. A área de estudo foi composta por 52 talhões de eucalipto. Foram simuladas 254 alternativas de manejo, sendo então criados 4 cenários de agendamento da colheita florestal envolvendo o uso da programação linear inteira, seguindo a formulação pelo modelo tipo I com maximização do VPL e um horizonte de planejamento de 7 anos. O cenário 1 não considerou a adjacência, enquanto que os cenários 2 (URM), 3 (ARM50) e 4 (ARM70) continham restrições de adjacência. Os resultados demonstraram que as restrições de adjacência reduzem o VPL em 3,74%, 2,24% e 2,10%, e a produção volumétrica em 2,92%, 1,79% e 1,73%, nos cenários 2, 3 e 4 respectivamente. Porém, os cenários 2, 3 e 4 obtiveram sucesso no controle de corte de talhões adjacentes, segundo suas restrições, e impedindo a formação de extensas áreas contiguas como encontrado no cenário 1 (153,25 ha).Palavras-chave: Programação linear inteira; restrição URM; restrição ARM. AbstractSpatial adjacency constraints effect in optimized forest planning. The objective of the study was to apply the spatial adjacency constraints in the forest harvest scheduling when using the traditional forest planning model. It was analyzed and compared the impact on NPV (Net Present Value) and the volumetric production of wood. The area of study was formed by 52 stands of eucalyptus. A total of 254 forest management alternatives were simulated, where 4 forest harvest scheduling scenarios contained the integer linear programming were created, following the model type I and the maximization of the NPV for 7 years of the horizon planning. Spatial adjacency constraints were applied to scenarios 2 (URM), 3 (ARM50) and 4 (ARM70) but not to scenario 1. The results showed that the spatial adjacency constraints reduced NPV in 3,74%, 2,24% and 2,10%, and the volumetric production in 2,92%, 1,79% e 1,73% through the scenarios 2, 3, and 4, respectively. Therefore, the scenarios 2, 3, and 4 obtained success in controlling the adjacent harvested stands, according to their restrictions, and avoiding the creation of large and continuous areas such as observed in the scenario 1 (153,25ha).Keywords: Integer linear programming; URM constraint; ARM constraint.


1994 ◽  
Vol 24 (6) ◽  
pp. 1260-1265 ◽  
Author(s):  
David N. Holland ◽  
Robert J. Lilieholm ◽  
David W. Roberts ◽  
J. Keith Gilless

Three indices of forest stand structural and compositional diversity were incorporated into a linear programming timber harvest scheduling model to examine the trade-offs between managing stands for timber production and biodiversity objectives. The indices, based on Shannon's diversity index, characterized stand species diversity, basal area diversity, and vertical crown diversity. While harvest-level objectives were often compatible with the maintenance of vegetative diversity, the maximization of present net value was accompanied by substantial reductions in all three measures of diversity.


2016 ◽  
Vol 359 ◽  
pp. 11-18 ◽  
Author(s):  
Andrew P. Robinson ◽  
Michael McLarin ◽  
Ian Moss

1979 ◽  
Vol 9 (4) ◽  
pp. 525-531 ◽  
Author(s):  
Chiang Kao ◽  
J. Douglas Brodie

To resolve the traditional quantifiable but incommensurate objectives of perfect regulation, maximization of present net worth, and even-flow harvest, goal programming (GP) was applied to a sample forest, providing optimal solutions for each goal and a compromise solution that jointly considered all three as weighted goals. Goal programming overcame problems of infeasible specification and satisfied alternate criteria in cases with multiple optima. The GP approach provided a means of considering each of the three goals and minimizing the appropriately weighted deviations.


2009 ◽  
Vol 24 (2) ◽  
pp. 61-66 ◽  
Author(s):  
Kevin Boston ◽  
John Sessions ◽  
Robin Rose ◽  
Will Hoskins

Abstract A variable green-up period is incorporated into a tactical harvest scheduling model to allow for the regeneration policy to be included as a decision variable. The benefit of this formulation was demonstrated by solving a 91-logging unit forest plan under four different green-up policies.The first three policies used fixed green-up periods of 2, 3, and 4 years. The fourth policy uses a variable green-up period where the model selects the regeneration effort that determines the green-up period stand by stand. This policy allows each harvested stand to choose among a 2-, 3-,or 4-year green-up period. The variable green-up period resulted in a slight improvement in the net present value because its total was approximately $3,000 or $1.11/ac higher for the total planning area, not just the harvested acres when compared with the best solutions when 4-, 3-, and 2-year fixed green-up periods were used. The variable green-up constraint adds a level of complexity to the spatial harvest scheduling problem that is easily incorporated into a variety of heuristic procedures because they do not require that all combinations of harvest units be specified before solving the problem. The results cannot be generalized to other forests because the spatial arrangement of the stands is a major component in the value determination. However, using the methodology presented, forest managers can evaluate their own alternatives that may improve the returns from managing their resources.


Sign in / Sign up

Export Citation Format

Share Document