# HybridPSODevelopment
## Update and Notification :
<br> All the work in this rep is finished as a research assistant of Advanced Manufacturing Group of Department of Mechanical and Energy Engineering, Southern University of Science and Technology,cooperating with PhD candidate Zhanying CHEN , and under the supervised of Prof. Xuekun LI. Both of my supervisor and cooperator are from Institude of Manufacturing Engineering, Tsinghua University.</br>
<br> Reference literature will be supplied with review later. (when I'm out of endless exams and projects)</br>
<br> *ALL COMPYRIGHTS RESERVED*
---
## Depiction
<br>This project aimed to develop extended PSO algorithm that is of praticle value for optimization problems in real engineering.
## Features of the algorithm:
### 1. Multple objectives
Two sets of codes, uses relatively *density measure* and *superposition method* when select leader for the searm is provided.
### 2. Multiple constraints
In all codes provided, multiple constraints are handled with *feasible area* method (which is proved to be a more efficient method compared to basic *projection method* ).
### 3. Setable weights on different objectives
Although codes selecting leader with density measure is provided, all codes select a final global optima relys on *Grid Index*, which is essentially superposition of objective functions. The weights attached to different objectives are setable.
### 4. Produces a unique global optima together with Pareto Front
The codes in the pack are set to graph the Pareto Front, as well as using a blue square to suggest position of the global optima. Codes which can be used to graph the process of all particles together with Pareto Front is also reserved (and commented). However there may be particles whose values of objective functions might include imaginary numbers, so it's not suggested to be used.
## Codes in the pack
### MOPSO(density meausre method)
Matlab codes of MOPSO algorithm uses density measure.
#### mopsoDensityMeasure
Basic MOPSO using density measure.
#### mopsoVariantParameter
MOPSO using density measure, and improved with vairnt weight and learning factor (c1 and c2) duirng runtime.
(I've examined this two algorithms in several different ways, but end up failling to decide which one performs better).
### MOPSO(Superposition Method)
MOPSO using superposition method and basic, non-variant coeeficient.
# Reference
Very original version of codes comes from : https://www.mathworks.com/matlabcentral/fileexchange/52870-multi-objective-particle-swarm-optimization--mopso-
More detaied description will be presented in technical document (which by now, may still reamianed unfinished).
You are welcome to contact me for any problems at:
derizsy@foxmail.com
开发扩展的MOPSO算法,在实际工程中具有实用价值的优化问题.zip
版权申诉
89 浏览量
2023-05-10
11:04:44
上传
评论
收藏 42KB ZIP 举报
JGiser
- 粉丝: 7567
- 资源: 5112
最新资源
- 三层别墅图纸编号D099-三层-13.80&10.40米-效果图.dwg
- 三层别墅图纸编号D098-三层-16.20&13.20米-效果图.dwg
- 三层别墅图纸编号D097-三层-15.00&13.20米-施工图.DWG
- 三层别墅图纸编号D096-三层-10.10&09.00米-施工图.dwg
- 三层别墅图纸编号D095-三层-13.30&14.60米-施工图.dwg
- 三层别墅图纸编号D094-三层-11.40&12.90米- 施工图.dwg
- 3层独栋别墅图纸编号D093-三层-13.90&20.70米-施工图.dwg
- IIS多站点管理工具,可支持同时运行多个IIS服务
- Navicat15与keygen下载
- vue+pdf.js预览本地pdf文件(可以复制文本,滚动页码展示)
资源上传下载、课程学习等过程中有任何疑问或建议,欢迎提出宝贵意见哦~我们会及时处理!
点击此处反馈