Genetic algorithm for constrained global optimization in continuous variables (Q814754)

A genetic algorithm for a constrained optimization in continuous variables over a compact domain, which is a stochastic global optimization algorithm, is presented. It is demonstrated both theoretically and numerically that the real-coded genetic algorithm can locate the optimal solution to any prescribed accuracy and reliability for the constrained global optimization problem. Convergence in probability is established under presence of constraints and specific procedures for treating linear inequality constraints are provided along with their theoretical convergence properties.