基于LSSVM-MODE的水煤浆优化配煤研究  

Research on Optimization Coal Blending of Coal Water Mixture Based on LSSVM-MODE

作　　者：刘定平[1] 叶向荣[1] 邓华裕[1] 

机构地区：[1]华南理工大学电力学院,广东广州510640

出　　处：《选煤技术》2008年第3期16-20,共5页Coal Preparation Technology

摘　　要：根据茂名洁能水煤浆厂常用的三种配煤方式,利用最小二乘支持向量机(LSSVM)对球磨机电流和水煤浆浓度进行多目标建模,并采用基于Pareto最优概念的多目标微分进化(MODE)算法,对运行工况进行寻优,然后根据模糊集理论在Pareto解集中求得满意解,获得了水煤浆浓度高、生产成本低的三种配煤优化调整策略,并得出了最佳配煤方案,从而确定了水煤浆生产优化配煤策略。according to the three kinds coal blending methods frequently used in Maoming CWM plant,the LSSVM(Least Square Support Vector Machines) was proposed to construct multi-objective optimization model for CWM concentration and ball mill electric current.MODE(multi-objective differential evolution) based on Pareto optimal concept is used to perform a search for determining the optimum solutions,from which three kinds optimum adjustments that could increase concentration and lower production cost of CWM is obtained based on fuzzy theory.Then thebest coal blending project is acquired that could optimize preparation of coal water mixture.

关 键 词：水煤浆 优化配煤 最小二乘支持向量机 多目标微分进化算法 

分 类 号：TQ536[化学工程—煤化学工程]