摘要
构造了一种基于最小二乘支持向量机和多目标进化算法的锅炉燃烧优化控制系统,通过从电厂分散控制系统上采集数据,利用最小二乘支持向量机对锅炉燃烧特性建模并通过样本的机器学习,提出了以锅炉效率与NOX排放为组合的锅炉燃烧多目标优化模型,采用基于Pareto最优概念的多目标进化算法实现运行工况寻优,根据模糊集理论在Pareto解集中求得满意解,获得锅炉燃烧优化调整方式。
An optimized control system for high efficiency and low emission combustion of power plant boiler is constructed based on LS-SVM（Least Square Support Vector Machines） and MOEA （Multi-Objective Evolutionary Algorithms）. A LS-SVM model of boiler combustion response property is set up based on data acquisition from power plant distributed control system. Through data sam- ples machine learning,a multi-objective optimization model for high efficiency and low NOx emission combustion is established. MOEA based on Pareto optimal concept is used to perform a search for determining the optimum solutions,from which the optimum combustion adjustment mode of boiler is obtained based on fuzzy theory.
出处
《电力自动化设备》
EI
CSCD
北大核心
2006年第5期46-49,共4页
Electric Power Automation Equipment
基金
国家自然科学基金项目(50276019)
关键词
燃烧优化
NOX排放
支持向量机
多目标进化算法
combustion optimization
NOx emission
support vector machine
multi-objective evolutionary algorithm
