The Research And Application Of Bionic Intelligent Algorithm Based On Artificial Bee Colony And Cloud Model

Bionic intelligent algorithm, as an important branch of the field of artificial intelligence, has a strong interdisciplinary, involving bionics, mathematics, physics, physiology, psychology, neuroscience, computer science, sociology and other disciplines. It shows strong vitality and development potential in solving complex optimization problems and practical application. Here aimed at bionic intelligent algorithm based on artificial bee colony and cloud model, artificial bee colony algorithm(ABC algorithm), which has the characteristics of the global exploration and local exploitation, combined with the cloud model, particle swarm optimization and cultural algorithm, is used so that a number of algorithms based on artificial bee colony for numerical function optimization problems was proposed. In addition, an improved artificial bee colony algorithm with particle velocity was designed to solve minimum attribute reduction problem.First of all, an improved artificial bee colony algorithm, which was applied to the numerical function optimization problems, based on cloud model was proposed to solve the slow convergence speed and the drawback of fall into local optimum easily of the classical ABC algorithm. The improved algorithm makes use of the normal cloud particle operator to generate the new location of nectar and dynamically adjusts the local search range by the nonlinear decreasing strategy. In addition, a new selection strategy was designed so that the probability, which the poor individuals of the employed foragers was chose to follow, is larger. The experimental results show that the improved algorithm not only improves the solution quality, but also enhances the ability to escape from the local optimal solution.Secondly, combining the characteristics of the particle swarm optimization and cultural algorithm, we propose two bionic intelligent algorithms based on artificial bee colony. First, because the GABC algorithm inspired by particle swarm optimization still has the drawback of the solution quality, IGABC algorithm adjust weights adaptively in order to improve the local exploitation capacity. What’s more, it uses the disruptive selection strategy to avoid prematurely fall into local optimal solution. Second, using the dual evolution structure of the cultural algorithm for reference, this paper proposes a cultural algorithm based on artificial bee colony, named NKABC. It uses the normative knowledge of the belief space and historical optimal solution information to guide the scouts searching, and adopt the similarity measurement criteria based on Euclidean Distance, as the diversity of the population, to improve the group’s ability of global exploration. The experimental results validate the effectiveness of the IGABC algorithm and NKABC algorithm.Finally, we design a new algorithm for solving the minimum attribute reduction based on IGABC algorithm combined with the concept of particle velocity, named IGABCAR. It denotes individuals as binary encoding, combined with the particle velocity to calculate the new nectar position of bees. The experimental results on several UCI datasets validate that IGABCAR algorithm can effectively improve the quality of solution.