Research Of Improved Marine Predators Optimization Algorithm And Its Application

Marine Predators Algorithm(MPA)is a metaheuristic algorithm based on the interaction between marine predators and prey,which can model the hunting strategies of marine organisms,such as Levy and Brownian motion,to better understand their behaviors and control them effectively.Due to the simplicity and efficiency of the algorithm,MPA is suitable for solving complex optimization problems in many engineering and scientific fields.However,MPA still have some shortcomings,such as a small range of feasible solutions,easily falling into local optimum solutions and low convergence rate when solving complex problems.To address the shortcomings of MPA,this paper proposes an improved Marine Predators Algorithm,ODMPA,and uses the CEC2014 benchmark functions set to demonstrate that the improved algorithm performs effectively in solving global optimization problems.ODMPA also successfully solves seven classical engineering design problems and PV model parameter identification problems.The main contributions of this paper are as follows:1.To overcome the shortcomings of the original MPA,which has low convergence accuracy,easily falls into local optimum,and too slow convergence in high dimensional problems,this paper proposes an improved Marine Predators Algorithm: ODMPA based on Tent chaotic mapping,outpost mechanism and Differential Evolution with Simulated Annealing(DE-SA)mechanism.The tent chaotic mapping is introduced in the population initialization stage to add randomness,while selecting the N individuals with the best fitness value from the chaotic population and the original randomly generated population,as the final initialized population.After that,the outpost mechanism is introduced to perform a Gaussian perturbation with a small step size near the optimal solution,which improves the convergence speed;the DE-SA mechanism introduced at the end is used as an important step to help the original MPA jump out of the local optima by generating a new solution with a larger step size.ODMPA can better achieve a balance between the exploration and exploitation phases in the process of finding the optimal solution.2.To verify the performance of ODMPA,ODMPA was compared with 20 other meta-heuristic algorithms on the CEC2014 benchmark functions set.ODMPA performed better than other algorithms in terms of convergence accuracy and speed.In addition,the Friedman assessment was also used and calculated by SPSS,and ODMPA obtained the minimum ARV value,ranking first in the algorithms.Practical engineering design problems like pressure vessel design are needed to be addressed,and they are also practical and effective methods for verifying the performance of single-objective optimization algorithms.ODMPA has obtained optimal results in seven engineering design problems,which proves the feasibility of ODMPA in solving engineering design problems.In addition,to solve the parameter identification problem of the photovoltaic model,ODMPA was used to optimize the key parameters of the model.Experimental results show that ODMPA exhibits the best results in terms of convergence speed and solution accuracy compared to other algorithms.