| 摘要: |
| 提出了一种基于多Agent的遗传算法（MAGA）。基于三峡机组非线性全数字仿真模型，利用JADE中间件建立了实施MAGA的分布式移动计算平台。在此平台上，对机组不同运行工况下自适应PID调速器进行了优化。仿真结果表明，MAGA能够获得遗传算法的寻优效果且优化时间显著缩短。 |
| 关键词: 多Agent 遗传算法 自适应变参数PID调速器 JADE 优化 |
| DOI： |
| 分类号: |
| 基金项目:广东省产学研结合项目（100003）；广东省节能与新能源重点实验室项目（IDSYS200701） |
|
| Multi-Agent genetic algorithm for PID governor |
|
MENG Anbo, YIN Hao
|
|
Faculty of Automation，Guangdong University of Technology，Guangzhou 510006，China
|
| Abstract: |
| A MAGA（Multi-Agent Genetic Algorithm） is proposed and a distributed computing platform applying JADE midwares is built to implement MAGA，which is based on the nonlinear full-digital simulation model of governing system for Three-Gorge units. The adaptive PID governor of unit is optimized on the platform for different operating conditions. Simulative results show that，MAGA reserves the optimization effect of genetic algorithm while shortens the optimization time significantly. |
| Key words: multi-Agent genetic algorithms adaptive governor with variable PID parameters JADE optimization |
