Improvements And Applications Of Marine Predator Algorithm

Optimization problems widely exist in real life,such as function optimization,engineering design,etc.These problems generally have the characteristics of high dimension,many constraints and nonlinearity,but the solution accuracy and solution efficiency of traditional optimization methods are difficult to meet the actual requirements,the intelligent optimization algorithm came into being in this situation.Due to the characteristics of intelligence and universality,intelligent optimization algorithms are favored by researchers and are widely used to solve complex optimization problems.Compared with other intelligent optimization algorithms,the marine predator algorithm has the advantages of simple principle,few parameters,easy implementation and strong global exploration ability.However,the algorithm has problems such as slow convergence speed and low optimization accuracy,and its research and application is still in its infancy.Therefore,the marine predator algorithm has the significance of further research and exploration.In order to improve the optimization performance of the algorithm and expand its application range,first,the advantages and disadvantages of the marine predator algorithm are analyzed and summarized;then,the algorithm is improved around three different types of optimization problems:single-objective continuous optimization,multi-objective continuous optimization and discrete optimization.Related research is divided into the following three aspects:(1)Aiming at the single-objective continuous optimization problem,a marine predator algorithm combined with students’ psychological optimization is proposed.Firstly,the Tent chaotic sequence is used to improve the diversity of the initial population;secondly,the nonlinear convergence factor is introduced to improve the convergence speed and the ability to explore the optimal solution of the algorithm;finally,the strategies of learning and self-learning in the optimization algorithm based on students’ psychology are integrated into the marine predator algorithm to strengthen the information exchange between populations,and in the iterative process of the algorithm,the adaptive mutation strategy is used to improve the algorithm’s ability to jump out of the local optimum.The experimental results show that the improved algorithm has advantages in both convergence speed and optimization accuracy.(2)Aiming at the multi-objective continuous optimization problem,a multi-objective marine predator algorithm combined with a two-stage optimization strategy is proposed.Firstly,the Henon chaotic sequence is used to generate a two-dimensional initial population to improve its ergodicity;secondly,a non-dominated sorting method is introduced to improve the exploration ability of the algorithm;finally,a two-stage optimization strategy that takes into account both exploration and development is integrated into the update iteration stage of the algorithm to improve its optimization performance.Experimental results show that the improved strategy improves the diversity and convergence of Pareto solutions.(3)For the discrete optimization problem represented by the electric vehicle routing,an adaptive large-neighborhood marine predator algorithm is proposed.First,the initial solution is generated by the mileage saving method;secondly,the deletion operator and the repair operator in the large neighborhood search strategy are used to further optimize the performance of the solution;finally,the performance of the improved algorithm is verified by solving a discrete optimization problem represented by electric vehicle routing.The experimental results show that the improved algorithm is more competitive than other algorithms for solving the electric vehicle routing problem.This thesis analyzes the shortcomings of the marine predator algorithm,and improves the algorithm from three different types of optimization problems: single-objective continuous optimization,multi-objective continuous optimization and discrete optimization.The experimental results show that the improved algorithm has achieved good results in solving single-objective continuous optimization problems and multi-objective continuous optimization problems represented by function optimization and engineering design,and can well solve discrete optimization problems represented by electric vehicle routing.