非线性最优化问题的一种混合解法（A hybrid algorithm for solving nonlinear optimization problems）.docVIP

非线性最优化问题的一种混合解法（A hybrid algorithm for solving nonlinear optimization problems） Abstract: combining BFGS method with chaotic optimization method, a hybrid optimization method for solving nonlinear optimization problems with variable boundary constraints is proposed based on chaotic variables. The hybrid algorithm takes into account the advantages of chaotic optimization, global searching ability and the fast convergence of BFGS method. It has become an effective method to solve the global optimization problem of Nonconvex Optimization problems. The example shows that when the number of chaotic search reaches a certain amount, the hybrid optimization method can guarantee the algorithm to converge to the global optimal solution, and the computational efficiency is greatly improved than the chaotic optimization method. Keywords: hybrid method, BFGS method, chaotic optimization method, global optimum 1 Introduction In the field of systems engineering, control engineering, statistics, inverse problems and optimization, many problems are non convex. Only this common optimization technique for local optimal solution, because the deterministic algorithm always solution is an extreme point of the nearest [1], only to be able to give a good initial point can be obtained by the global optimal solution needs. Therefore, in the practical application by using traditional numerical method in many initial points to get the global solution is still being used by people, but the probability of this method for the global optima is not high, low reliability, set up as large as possible to solve the global probability solution algorithm is still an important problem. In recent years, global optimization method based on gradient method has been studied [2], genetic algorithm, random search and simulated annealing algorithm in the application of global optimization problems have been more and more attention based on [3-4]. In this paper, a hybrid algorithm for solving optimization problem