Improved Cuckoo Search Algorithm And Its Application In Antenna Array Beam Pattern Optimizations

Swarm intelligence algorithm,as an emerging evolutionary computing method in optimization problems,has become a hot topic in the field of optimization problems.The Cuckoo Search algorithm(CS)is a popular bionic optimization algorithm.It has the advantages of easy implementation,fast convergence speed,and strong concurrency.Based on the research of many scholars,this paper proposes two improved cuckoo search algorithms,and selects a practical array antenna pattern synthesis problem to verify the effectiveness of the improved algorithms.Array antenna pattern is an important part of array antenna performance research.Array antennas with strong radiation directivity will provide people with better communication quality.The main work in this article is as follows:(1)An improved cuckoo search with reverse learning and invasive weed operators is proposed.The algorithm starts from population initialization,uses the reverse learning to generate the reverse population,and selects the elite individuals in the mixed set of the original population and the reverse population to form a high-quality initial population;the algorithm then introduces weeds operator,first it uses the weed growth and reproduction mechanism to generate offspring nests;and next,it uses the diffusion mechanism to spread the offspring nests around the parent nests int the normal distribution to increase the search range and increase the diversity of the population.At last,the algorithm eliminates disadvantage individuals by competition.In this paper,five swarm intelligence algorithms and the standard cuckoo algorithm are selected as comparison algorithms.The comparison experiment of solving the CEC2014 test set function and significance test verify that the algorithm has excellent optimization performance.(2)A tolerance cuckoo search based on the beetle antennae search is proposed.This algorithm introduces the beetle operator into the standard cuckoo search algorithm,and uses the beetle to guide the cuckoo nest update.This operator enhances the optimization ability of the cuckoo nest individuals;at the same time,in order to avoid the nest falling into the local optimum,an adaptive strategy is introduced into the cuckoo search algorithm.First,a tolerance counter is defined.The tolerance adjustment probability is used to compare the global optimal guidance ability of the candidate nest generated by self-learning and the bird nest trapped in a local optimum.Based on this,the algorithm can adaptively adjusts search direction.This adaptive learning strategy can effectively improve the global Optimizing ability.This paper uses PSO,BBO,CSO,FA,GA,CS and improved algorithms to perform simulation experiments.By solving the experimental results of 30 functions,it can be concluded that the comprehensive performance of the improved algorithm is better than the comparative algorithms.(3)The paper selects linear array antennas and circular array antennas as the representative models of array antenna pattern optimization problem.With a certain number of array elements,in order to obtain an array antenna pattern with a minimum maximum sidelobe level,two improved algorithms are used to optimize the distance of array elements and excitation amplitude in the model.Experiments show that,compared with the comparison algorithm,the improved algorithm proposed in this paper effectively suppresses the maximum sidelobe level and is also effective in the field of practical engineering.