基于多目标进化算法的锅炉热损失优化研究

Research on Optimization of Boiler Heat Loss Based on Multi-objective Evolutionary Algorithm

针对锅炉热损失模型的特点,提出基于Pareto最优概念的多目标进化算法实现运行工况寻优,然后根据模糊集理论在Pareto解集中求得满意解,获得最佳的锅炉燃烧调整方式。通过某600MW锅炉热损失的优化研究,并与基于神经网络的寻优结果比较,数值计算表明支持向量机模型寻优结果在Pareto前沿具有更好的多样性,结果更优,可指导运行人员进行参数优化调整,提高燃烧经济性。

In accordance with the features of the model of boiler heat loss, the scheme of optimal boiler combustion adjustment is proposed. In this scheme, using multi-objectlve evolutionary algorithm based on Pareto optimal concept the optimal operational status is searched, then sarisfactory solution is obtained in Pareto optimal sets based on fuzzy sets theory. Through the research on optimization of boiler heat loss in 600MW power plant, and comparing the results with optimal search results by neural network method, the calculation shows the results of optimization from support vector machine model offer more diversified in Pareto frontier, more optimal results. It is a good operator guide for optimal parameter adjustment to enhance economics of combustion.