Research On Swarm Intelligence Algorithm Based On Single Objective And Multi-dimension Space

In recent decades,swarm intelligence algorithm has become a research hotspot in solving optimization problems all over the world.Traditional mathematical models,such as gradient descent method,Newton method and Lagrange multiplier method,have been in short supply in the face of increasingly complex nonlinear and complex optimization problems.By imitating the social behavior of creatures in the nature and according to natural or physical phenomena,researchers put forward a variety of swarm intelligence algorithms,which are widely used in design,planning,management and control.Although the existing swarm intelligence algorithm is simple,efficient and universal,its mechanism still leads to the problem of local convergence.This paper studies the improvement of fireworks algorithm,and proposes a new swarm intelligence algorithm named solar system optimization algorithm.The research content and innovation achievements of this paper are summarized as follows:1)An enhanced fireworks algorithm with dynamic explosion radius(EFWA-DER)is proposed.EFWA-DER divides the search space equally in order to avoid the problem that the initial solutions are too concentrated or scattered.It designs a double elite-tournament strategy.Two elite individuals are retained for the next generation selection in each iteration so as to enhance the exploration ability of the algorithm.The dynamic explosion radius operator is introduced to generate adaptive changes.As a result,the convergence rate and solution precision are improved due to enhanced exploitation ability by dynamically controlling the explosion radius.2)A new swarm intelligence algorithm,called solar system optimization algorithm(SSOA),is proposed.Inspired by the fact that the sun as the only star in the solar system is surrounded by eight planets,and the planets are also surrounded by their own satellites,a novel swarm intelligence algorithm named solar system algorithm is proposed.In the solar system algorithm,the simulated solar system will undergo a process of continuous shrinking and searching.The sun is surrounded by eight planets,and each planet generates a certain number of satellites.By simulating the replacement of planets and generation of satellites,the population diversity will be ensured and the exploration and exploitation are completed.