Hybrid Taguchi-chaos of Artificial Bee Colony Algorithm for Global Numerical Optimization

In this paper, a new evolutionary learning algorithm is proposed by hybridizing the Taguchi method and chaos artificial bee colony (CABC). The algorithm is thus called HTCABC. First, the chaos search algorithm and adaptive bound method is adopted to improve the ABC performance and convergence rate. Then, the Taguchi method and crossover operation are incorporated into the CABC to produce good food sources, thus accelerating the search capacity. The Taguchi method has also been utilized to establish a proper balance between the exploration and exploitation by incorporating the information from the best global solution into the solution search equation. Third, the natural phenomenon of the elite strategy is adopted and the recruitment of new scout bees is used for HTCABC, which can have a rapid convergence rate maintain the diversity of the population, and escape from local optima. Additionally, there is no complex parameter setting in the algorithm design. Therefore, the HTCABC can be a more robust, quickly convergent and more accurate optimal solution. Finally, the algorithm is examined by using a set of benchmarks and the proposed approach is effectively applied to solve the parameter identification of a chaotic system. Simulation results show that the proposed algorithm is more efficient than the existing algorithm reported in the literature.