Silva Fennica : quarterly issues : 33(3) : sa333207.htm
Multiple solution harvest scheduling
Silva Fennica 33(3): 207–216
Application of the Metropolis algorithm for forest harvest scheduling is extended by automating the relative weighting of objective function components. Previous applications of the Metropolis algorithm require the user to specify these weights, which demands substantial trial and error in practice. This modification allows for general incorporation of objective function components that are either periodic or spatial in nature. A generic set of objective function components is developed to facilitate harvest scheduling for a wide range of problems. The resulting algorithm generates multiple feasible solutions rather than a single optimal solution.
Keywords
Metropolis algorithm, simulated annealing, Gibb’s sampler
Addresses
Principal Research Scientist, NCASI, Northeast Regional Center, Tufts University, 1 Anderson Hall, Medford, Massachusetts 02155, USA. Fax +1 617 627 3831, E-mail pvandeus@tufts.edu
Received 11 March 1999 Accepted 27 July 1999
ISSN 0037-5330