A novel hybrid algorithm for function approximation
- Zne-Jung Lee
- Expert Syst. Appl.
This paper deals with a hybrid optimization method for solving the optimization problems with inequality constraints and equality constraints, in which an adaptive penalty strategy is firstly adopted to convert the optimization problem with both equality constraints and inequality constraints to one only with lower and upper bound constraint of decision variables, then an adaptive real-coded genetic algorithm combined with a local search algorithm is employed to optimize the problem. For improving the optimization effect of the real-coded genetic algorithm, the adaptive crossover probability and mutation probability are designed. Finally, to examine the validity of the hybrid optimization method, the method and other two algorithms are used to optimize the traffic signal timings optimization problem of an isolated intersection, and a large number of simulations show that the hybrid optimization method proposed in this paper can work well in the traffic signal timings optimization problem.