Optimizing Complex Bioeconomic Simulations Using an Efficient Search Heuristic

Abstract

For simulation to be truly useful for investigating many problems in agricultural economics, nonsimplifying optimization techniques need to be employed. General methods for simulation optimization that do not inhibit system characterization or analysis are available, and they would appear to provide much of the mathematical and optimizing rigor demanded by economists. This paper describes the theory and algorithm of a robust and efficient simulation optimization approach, the Complex Method. An example of implementing the algorithm is illustrated using a pest management problem.

Hybrid PSO-Complex Algorithm Based Parameter Identification for a Composite Load Model

“…The complex method is another efficient optimization algorithm searching from the initial simplexes for the optimization of either physical processes or mathematical functions [17]. Even though the complex method is a direct method, it does not require the gradients of the objective function for searching.…”

“…The need to improve the simplex method and become the complex method is presented in [17]. With the concavity of nonlinear functions, the simplex method may be directed to a less feasible solution space during the reflection and expansion operations.…”

“…In [11], reducing the number of parameters for the composite model was discussed using the concept of trajectory sensitivity. This paper presents a parameter estimation method for a composite load model that applies hybrid optimization with particle swarm optimization (PSO) [14][15][16] and a complex method [17]. PSO algorithms have the advantages of simplicity in implementation and quick convergence to a reasonably good solution [18].…”

Effects of different objective functions on optimal decision variables: a study using modified complex method to optimize hamburger cooking

“…Umeda, Shindo and Ichikawa (1972) used this method to solve variational problems with state-variable inequality constraints and demonstrated its applicability. Kazmierczak (1996) gave an example of a pest management problem to illustrate that the Complex Method was a very efficient optimization approach and mathematically very simple 10 compared to some other methods. Erdogdu and Balaban (2002) modified this method and showed that it allowed the incorporation of different constraints on microbial sterility and final temperature at the coldest point as implicit constraints in thermal processing problems.…”

“…These problems are especially difficult to solve because of the non-linear and distributed nature of the system dynamics and the existence of explicit and implicit constraints on both control variable and objective function (Banga et al, 2001). There have been numerous methods suggested for this kind of problem and accumulated knowledge in the literature (Gallardo & Casares, 1991;Noronha, Hendrickx, Suys, & Tobback, 1993;Banga & Seider, 1996;Kazmierczak, 1996;Noronha, Loey, Hendrickx, & Tobback, 1996a,b;Terajima & Nonaka, 1996;Banga, Perez-Martin, Banga, Alonso & Singh, 1997;Chalabi, Van Willigenburg, & Van Straten, 1999;Banga et al, 2001;Balsa-Canto, Alonso & Banga, 2002;Erdogdu & Balaban, 2002). As seen in the literature, the maximum principle of Pontryagin has been used by many authors to solve the thermal processing optimization problems.…”