Emperor Penguin Optimizer For Some Kinds Of Optimization Problems

Emperor Penguin Optimization(EPO)is a new swarm intelligence optimization algorithm proposed in 2018,which simulates the behavior of emperor penguin colonies huddling together to keep warm in winter.In this algorithm,the population are guided by the optimal individual,which accelerate the convergence speed.Therefore,it has been extensively studied since it was proposed,and has been applied to engineering optimization,wireless communication,neural network,image processing and other fields.EPO is simple and easy to understand,which has a fast convergence speed.But it also has drawbacks,such as easy to trapped into a local optimum and low convergence accuracy.For solving unconstrained optimization problems,constrained optimization problems,and multi-objective optimization problems,the improved EPO algorithms are proposed to overcome these shortcomings.The main research contents are as follows:1.A hybrid emperor penguin optimization(HEPO)is proposed for unconstrained optimization problems.In order to improve the exploration ability of the algorithm,chaotic initialization method is adopted to make the initial population uniformly distributed in the search space.Meanwhile,the variation operator is used to increase the population diversity and prevent the algorithm from trapping into local optimum.Finally,learning strategies are integrated to improve the global search ability of the algorithm.The simulation results show that the HEPO is effective.In addition,the HEPO is used to solve the nonnegative linear least squares problem.2.An improved emperor penguin optimization(IEPO)is proposed for constraint optimization problems.The dynamic linear adjustment of particle number strategy combined with two kinds of mutation operation modes is used to balance the exploration ability and exploitation ability of the algorithm.Furthermore,an archive replacement operation mechanism is introduced to improved the performance of the feasibility criterion and accelerate the convergence speed of the algorithm.The simulation experiments results of benchmark test function and two engineering optimization problems show that the IEPO can effectively solve the constrained optimization problems.3.A multi-objective emperor penguin optimization based on decomposition method(d MOEPO)is proposed for multi-objective optimization problems.Firstly,the method based on penalty boundary intersection is used to replace the parent individuals by child individuals,which increases the searching ability of the algorithm.Secondly,non-dominant individuals are stored in an external archive,and the binary tournament method is used for optimal individual selection.Finally,the simulation results show that the d MOEPO is effective for solving multi-objective optimization problems.