提示
基于empso指数加权粒子优化的研究
首发时间：2024-07-18
摘要：针对粒子群优化算法无法计算复杂行为的函数导数并且无法达到最优迭代次数方面的问题，提出一种新的指数加权粒子优化器（emchanism particle swarm optimization，empso）。本文提供指数加权粒子优化器梯度近似的数学证明，测试结果表明任何函数的梯度，无论是在可微还是不可微的测试函数或者标准单目标测试函数中，都可以用指数加权粒子优化器的参数来近似。
关键词：
for information in english, please click here
research on exponential weighted particle optimization based on empso
abstract：aiming at the problem that the particle swarm optimization algorithm cannot calculate the functional derivative of complex behavior and cannot reach the optimal number of iterations, a new exponentially weighted particle optimizer (empso) and a new layerless optimizer are proposed. the ability of the new layerless optimizer to solve optimization problems is attributed to its ability of good layer approximation. this paper provides a mathematical proof of the layer approximation of the exponentially weighted particle optimizer. the test results show that the layer of any function, whether differentiable or not, can be approximated by the parameters of the exponentially weighted particle optimizer. this is a new technique for simulating the layer descent algorithm (gradient descent，gd), which is at the boundary between numerical methods and swarm intelligence.
keywords：
基金：
论文图表：
引用
导出参考文献
no.****
动态公开评议
共计0人参与
勘误表
基于empso指数加权粒子优化的研究
评论
全部评论