基于HPSO的钢坯加热过程炉温优化设定  

Hybrid particle swarm optimization of temperature settings of billet-reheating process

作　　者：廖迎新 吴敏[1] 佘锦华[2] 曹卫华[1] 

机构地区：[1]中南大学信息科学与工程学院 [2]东京工科大学仿生学学部

出　　处：《控制理论与应用》2007年第6期1010-1014,共5页Control Theory & Applications

基　　金：国家杰出青年科学基金资助项目(60025310);国家863计划项目(2006AA04Z172);湖南省科技计划项目(04FJ3029).

摘　　要：针对一种蓄热推钢式加热炉三个加热区的炉温稳态优化问题,本文提出了一种混合粒子群优化(HPSO)方法.首先,基于钢坯导热偏微分方程和边界条件,建立钢坯温度预报模型.然后,采用HPSO算法确定最佳稳态炉温,即炉温控制的参考输入.该方法利用混沌机制产生初始种群,通过免疫和克隆来提高粒子群优化(PSO)算法的全局搜索能力和搜索精度.A hybrid particle swarm optimization （HPSO） approach is presented to solve the problem of optimizing the steady-state temperatures of the three zones of a billet reheating furnace. Firstly, a model for predicting billet temperature is constructed based on a heat transfer equation and its boundary conditions. Then, HPSO is used to determine the optimal steady-state temperatures, which are reference inputs for the control of furnace temperature. This approach uses a chaos mechanism to create the initial population, and adopts immunity and cloning to improve the global search capability and search precision of narticle swarm ootimization.

关 键 词：加热炉 稳态优化 混合粒子群优化 免疫 克隆 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]