Improved Electromagnetism-Like Algorithm for Nonlinear Bilevel Programming Problem

This paper presents an efficient hybrid heuristic search method based on electromagnetism-like algorithm and chaos searching technique to solve the nonlinear bilevel programming problem. In order to increase the population diversity, chaotic opposition-based population initialization is first applied to generate initial population. Considering the structure and characteristic of the bilevel programming problem, we introduce a simple fitness function only related to the upper level objective function. Instead of the random neighbourhood local search, chaos searching technique is used as local search. Based on these techniques, the hybrid algorithm improves the global search capability, avoids convergence to local minima and accelerates the convergence. Simulation results on test problems show that the proposed algorithm has a better performance than the compared algorithms.