Improved Sine Cosine Algorithms And Their Applications

As an important method to upgrade algorithm,intelligent optimization algorithm has been widely used and made remarkable achievements in many fields,such as transportation,military affairs,medical treatment and engineering,etc.The sine cosine algorithm(SCA)has attracted great attention since it was first proposed,because it has a simple formula and is easy to work out as a new group intelligent optimization algorithm which finds the optimal solution by using the mathematical model of trigonometric function.The traditional sinusoidal cosine algorithm performs poorly in optimization when it comes to different conditions.In order to perfect the optimal performance of the sine cosine algorithm and put the sine cosine algorithm in more application in engineering,the main research contents of this paper are as follows:(1)A new backbone sine cosine algorithm based on neighborhood structure is proposed.On the basis of backbone idea,the new algorithm introduces neighborhood structure and gaussian sampling learning.What's more,it controls the weight of gaussian sampling learning in the process of updating to ensure the algorithm has strong exploration ability at the beginning and behaves well in development in later period.The new algorithm also has obvious improvement in robustness and calculation accuracy,according to the results of the simulation experiment which optimizes benchmark functions with the new algorithm and other intelligent optimization algorithms.The simulation experiment of the test function proves that the improved strategy significantly improves the performance of the sine and cosine algorithm.(2)A hierarchical multi-subgroup cooperative sine cosine algorithm is proposed.The main idea of the new algorithm is to divide the update structure of the population into two layers: in the bottom layer,the population is divided into several sub-groups with equal number of individuals,and the individuals of each group can only evolve in parallel in the corresponding population;in the top layer,the population is composed of the optimal individuals in each sub-group in the bottom layer,and the individuals in the top layer are updated according to the Gaussian sampling learning strategy.In order to maintain the diversity of the population,the random recombination strategy is used to maintain the diversity of the population in the renewal process.(3)A discrete multi-objective sine cosine algorithm is proposed.According to the characteristics of community detection problems in complex networks,the new algorithm discretizes the sine cosine algorithm which dealt with the continuous optimization problem by redefining the operation rules of the sinusoidal algorithm.By analyzing the result of simulation experiment of the real data set,the new algorithm is approved to have obvious advantages in handling problems concerning community detection in complex networks.