Title :
Constrained Global Optimization by Constraint Partitioning and Simulated Annealing
Author :
Wah, Benjamin W. ; Chen, Yixin ; Wan, Andrew
Author_Institution :
Dept. of Electr. & Comput. Eng., Illinois Univ., Urbana-Champaign, IL
Abstract :
In this paper, we present constraint-partitioned simulated annealing (CPSA), an algorithm that extends our previous constrained simulated annealing (CSA) for constrained optimization. The algorithm is based on the theory of extended saddle points (ESPs). By decomposing the ESP condition into multiple necessary conditions, CPSA partitions a problem by its constraints into subproblems, solves each independently using CSA, and resolves those violated global constraints across the subproblems. Because each subproblem is exponentially simpler and the number of global constraints is very small, the complexity of solving the original problem is significantly reduced. We state without proof the asymptotic convergence of CPSA with probability one to a constrained global minimum in discrete space. Last, we evaluate CPSA on some continuous constrained benchmarks
Keywords :
computational complexity; constraint theory; simulated annealing; asymptotic convergence; constrained optimization; constrained simulated annealing; constraint partitioning; constraint-partitioned simulated annealing; continuous constrained benchmarks; discrete space; extended saddle points; global constraints; global optimization; Constraint optimization; Convergence; Electrostatic precipitators; Engineering profession; Neodymium; Partitioning algorithms; Programming profession; Simulated annealing; Stochastic processes;
Conference_Titel :
Tools with Artificial Intelligence, 2006. ICTAI '06. 18th IEEE International Conference on
Conference_Location :
Arlington, VA
Print_ISBN :
0-7695-2728-0
DOI :
10.1109/ICTAI.2006.47