物种生灭算法的改进策略

Improved Strategies of Species Explode and Deracinate Algorithm

物种生灭算法(Species Explode and Deracinate Algorithm,SEDA)是一种简单、高效的群智能优化算法。为了进一步提高SEDA算法的寻优速度、解的质量,首先,通过一种无排序筛选幸存物种的递归算法,提出了基于递归筛选的SEDA算法,减少了SEDA算法的时间复杂度,提高了算法的寻优速度;其次,通过引入衍生趋势的方法,提出了基于衍生趋势的SEDA算法,提高了SEDA算法对复杂、难以寻优的优化问题解的质量。三个测试函数的仿真结果表明,改进的方法具有更小的时间复杂度,能够有效改善SEDA算法解的质量。

To further improve the convergence speed and solution quality of Species Explode and Deracinate Algorithm(SEDA)which is one of simple and effective swarm intelligence algorithm,some improved SEDAs are proposed.Firstly,by no sorting all of species,the survival spices is selected,and a new SEDA based on recursive screening is proposed.The new algorithm reduces the time complexity of SEDA.Then,by introducing derive tendency,a new SEDA algorithm is proposed,called SEDA based on derive tendency.The new algorithm improves the solution quality of complex and hard to seek the optimal solution questions.The simulation results indicate that these improved strategies of SEDA have smaller time complexity than SEDA,and can improve the solution quality effectively.