A Bilevel Genetic Algorithm for Global Optimization Problems.

Genetic algorithm is an important intelligent optimization algorithm that operates on specific population by simulating the natural evolution process and using artificial evolution to continuously optimize the population so as to search for the optimal solution. At present, there are a large number of methods focus on improving genetic algorithms, but the current stage of genetic algorithm tends to have the problems of falling into local optimal premature and slow convergence. In this paper, we try to design a bilevel evolutionary particle swarm optimization algorithm based on the idea of genetic algorithm within the framework but without increasing the complexity, using a data-driven idea, and verify it by the genetic algorithm in the commercial software MATLAB. Numerical experiments show that the data-driven a bilevel genetic algorithm-based algorithm significantly improves the algorithm performance.