Evolutionary-Based Hybrid Optimizer Applicable to Large-Scale Design Problems

Abstract

Design Informatics has three points of view. First point is the efficient exploration in design space using evolutionary computation. Second point is the structuring and visualizing of design space using data mining. Third point is the application to practical problems. The investigation of efficient evolutionary-based optimizer for the above first point is essential in order to generate hypothetical database for data mining. In the present study, the performance regarding diversity and convergence has been compared among pure and their hybrid methods using three standard mathematical test problems with/without noise. The result indicates that the hybrid method between the genetic algorithm based on the elitist non-dominated sorting genetic algorithm and the differential evolution is better performance for efficient exploration in the design space under the condition for large-scale engineering design problems within 10 2 order evolution at most. Moreover, the comparison among eight crossover indicates that the principal component analysis blended crossover is good performance on the hybrid method between the genetic algorithm and the differential evolution.

Influence of Difference among Evolutionary Computations for Design Information

“…When the practical evolution is imitated for the evolutionary computation, the different evolutionary algorithms might ultimately be applied to each individual in population. In the present study, three optimizers, as GA, DE, and PSO, will be coupled [9]. First, multiple individuals are generated randomly as an initial population.…”