Overlooking Algorithm For Solving Global Optimization Problems

Global optimization problems are the hotspot in engineering,finance, aerospace, science and other areas of real life. At the same time, it is the basis for the researchs and application of other subjects. The great development of the theatrical and algorithm aspects of global optimization problems have been got by the wide application of global optimization problems in various fields, all kinds of experts and scholars continue to explore and study the in-depth method for global optimization. The development of global optimization methods got better and better. For example, the Tabu Search Algorithm,Genetic Algorithm, Neural Network Algorithm and Ant Colony Algorithm and others achieve the better algorithm performance continuously. However, as the scale of the global optimization problem increasing and the complexity of optimization problems is higher and higher, the requirements of optimization are also increasing, especially when the objective function with too many minim. The classical intelligent algorithms are not efficient enough. So it is an important research direction to find out a new intelligent optimization algorithm.This paper has been improved and perfected the Overlooking Algorithm for solving global optimization problems based on the analysis principle of human visual intelligence in literature 36.The algorithm simulates the vision intelligence for judgment, comparison and memory characteristics. The Overlooking algorithm gets supervision mechanism of overlook, the strategy of overlooking points, strategies of generating overlooking points, selection criterias, mechanism of constructing and solving local optimization problems through the three layers of the basis points, view points and local optimization process memory mechanism to improve the convergence speed of the algorithm, reduce the time-consuming. A large number of contrast tests with Genetic Algorithm show that Overlooking Algorithm has the higher convergence rate, the parameter selection is simpler and doesn’t depend on the initial points, avoid falling into local optimum and has a good algorithm performance. It brings a brand new way to solve global optimization problems.In order to improve the performance of Overlooking Algorithm and overcome the difficulties bring by the unideal initial points of Overlooking Algorithm, this paper tries to use intelligent genetic technological improve the Overlooking Algorithm, to form a new hybrid algorithm-IGOA(Intelligent Genetic Overlooking Algorithm). The algorithm convergence of Overlooking Algorithm, Tabu Search Algorithm and the Improved Simulated Annealing Algorithm are compared and analysized with IGOA. The results show that the convergence rate of Intelligent Genetic Overlooking Algorithm is better than Overlooking Algorithm and others.