A Modified Sine-Cosine Algorithm For Solving Optimization Problems

Continuous unconstrained optimization problem is a basic problem in the optimiza-tion.Its research and development have major impact on the process of the whole opti-mization field.Furthermore,efficient solutions for solving continuous unconstrained op-timization problems is the cornerstone of solving complex optimization problems.From general sequential optimization methods(such as simple method,Newton method etc)to modern optimization methods(such as Genetic Algorithm,Particle Swarm Optimiza-tion),the methods and algorithms for solving continuous unconstrained optimization problems emerge in endlessly.The aplenty research results and optimization algorithms of the unconstrained optimization problems have also prompted us to transform the complex optimization problems frequently encountered in scientific research,economic management and engineering design into unconstrained optimization problems.This in turn enhances the demand for higher performance algorithms.In this thesis,the basic theory of continuous unconstrained optimization are re-viewed,along with the main ideal and detailed steps of some modern intelligent op-timization algorithms.Then the Sine-Cosine algorithm and the related improvements under the framework are dicussed in detail,together with their advantages and disad-vantages.Considerfing the shortages of the sine-cosine algorithm,this thesis designs promising strategies from two aspects:the sine-cosine parameter and candidates' inertia weight.Finally,this paper proposes the modified sine-cosine algorithm.In the end,the modified sine-cosine algorithm is verified by the test function set,and the effectiveness and superiority of the modified sine-cosine algorithm is illustrated through the comparison with other algorithms and statistics test.