优化动力配煤数学模型寻优算法的研究  

Algorithms for Optimizing the Model of Steam Coal Blending

作　　者：董平[1] 陈彦杰[1] 

机构地区：[1]黑龙江科技学院资源与环境工程学院,黑龙江哈尔滨150027

出　　处：《洁净煤技术》2005年第4期60-64,共5页Clean Coal Technology

基　　金：黑龙江省科技厅攻关项目(GC03J102)

摘　　要：根据用户对配煤煤质特性指标和燃烧特性指标的要求,笔者确立了非线性动力配煤优化模型,并详细研究了几种常用的寻优算法的基本原理、特点及其在优化动力配煤数学模型中的应用效果。研究结果表明,穷举法准确但效率太低,而混合离散变量优化设计(MDOD)法只能局部优化并易陷入局优解;遗传算法鲁棒性强,具有较好的全局搜索性能;模拟退火算法简单、通用,具有很好的局部搜索能力。在此基础上,笔者总结出一种遗传算法与模拟退火法相结合的混合算法。This paper introduced the process of nanoflltration for treating mining water containing suspended matters. The influence of nanofiltration membrane operating condition to the filtration function of membrane and the effect of the combined process on coal mining water treatment were investigated. The experimental results showed that when operated at the cycle of T = 30min and adapted to back - aeration, the fouling of the membrane decreased effectively. The effluent quality of the system is stable and the tested items of the effluent met the national drinking criterion. The nanofiltration membrane is a high - efficiency and economical process in coal mining water treatment.

关 键 词：动力配煤 穷举法 MDOD法 遗传算法 模拟退火算法 

分 类 号：TD849[矿业工程—煤矿开采] O242[矿业工程—矿山开采]