Spotted Hyena Optimizer And Its Applications

Spotted hyena optimizer(SHO)is an emerging intelligent method for group intelligence that mimics the predation behavior of spotted hyenas in nature.SHO has a simple structure,clear concept,easy implementation,and good overall performance.However,the research and application of SHO is still in its infancy,and there are still some shortcomings,such as slow convergence in the later period and weak local search ability.In this paper,the shortcoming existed in SHO algorithm are analyzed and improved.The purpose is to improve and broaden the theoretical basis and application scope of the spotted hyena optimizer,in order to solve the large-scale optimization problem of complex systems.An effective method will be provided in this paper.The main work of this paper is as follows:(1)In order to enhance the global search ability of the SHO,to better balance its global and local search ability,improve its convergence precision,and introduce the neighborhood centriod reverse learning strategy,spotted hyena individual based on neighborhood centriod reverse learning is proposed.The optimization algorithm(NCOSHO)enhances the global search ability of the SHO algorithm by introducing a neighborhood c centriod reverse learning strategy.NCOSHO was used for 23 classical test functions and infinite impulse response(IIR)engineering problems and compared with different algorithms.The experimental results show that NCOSHO performs better on function optimization and IIR problems.(2)A method based on the SHO to determine the optimal proportional integral derivative(PID)controller parameters of AVR system is proposed.The method has the advantages of simple implementation,good convergence stability and high calculation efficiency.Quickly adjust the optimal PID controller parameters to produce a high quality solution.In order to evaluate the superiority of theperformance of the SHO-PID controller,this algorithm is compared with the comparison algorithms such as SCA,FPA,PSOGSA,WWO and GWO algorithms.The experimental results show that the proposed method has more steps to improve the step response of the system.It is high efficiency and robustness.(3)A small target detection method based on side suppression is proposed.The principle of side suppression in vision is applied to image preprocessing,and the background suppression and target enhancement are performed on the image,so as to achieve the purpose of adaptive preprocessing.The search task is performed on the feasible domain space of the preprocessed image by using the SHO algorithm.An optimization algorithm LI-SHO for suppressing the spotted hyena is proposed,and the image matching problem is used.The experimental results show that LI-SHO solves the image matching problem better than other algorithms to solve the image matching problem,and has better robustness.(4)The training goal of artificial neural networks is to find a set of optimal weights and deviations to achieve the minimum error of the network.Spotted hyena optimizer for training neural network is proposed.Compared with other meta-heuristic algorithms,the detection range is wide and the mining ability is strong.