基于智能算法的燃煤电站锅炉燃烧优化  

Combustion Optimization of a Coal-fired Boiler Based on Intelligent Algorithm

作　　者：余廷芳[1] 耿平[1] 霍二光 曹孟冰 

机构地区：[1]南昌大学机电工程学院,南昌330031

出　　处：《动力工程学报》2016年第8期594-599,607,共7页Journal of Chinese Society of Power Engineering

基　　金：国家自然科学基金资助项目(61262048)

摘　　要：基于Matlab人工智能工具包对某300MW燃煤电站锅炉进行了燃烧优化混合建模:利用BP神经网络建立了锅炉燃烧特性的BP神经网络模型,用以预测锅炉热效率和NO_x排放质量浓度.基于该模型,以锅炉热效率和NO_x排放质量浓度为目标,结合Matlab遗传算法工具包对锅炉进行燃烧优化,并采用权重系数法将多目标优化问题转化为单目标优化问题.结果表明:锅炉热效率和NO_x排放质量浓度校验样本的相对误差平均绝对值分别为0.142%和1.790%,该模型具有良好的准确性和泛化能力;权重系数法可根据实际情况,以锅炉热效率或NO_x排放质量浓度为优化重点选取相应的权重系数,对燃烧优化具有一定的指导意义.A hybrid model was set up using Matlab artificial intelligence toolkit to optimize the combustion in a 300 MW coal fired boiler. The specific way is to establish a BP （back propagation） neural network model for boiler combustion properties to predict the thermal efficiency and NO, emission concentration of the boiler, and then to optimize the boiler combustion with Matlab artificial intelligence toolkit based on the model by taking the thermal efficiency and NO, emission concentration as the target variables, during which the multi-objective optimization problems were transformed into single-objective optimization problems by weight coefficient method. Results show that the average relative errors of boiler thermal efficien cy and NOz emission are 0. 142% and 1.790% respectively, indicating good accuracy and strong generaliza tion ability of the model. By weight coefficient method, boiler thermal efficiency and NOx emission concentration can be chosen as the key optimization objectives by selecting corresponding weight coefficients, which therefore may serve as a reference for combustion optimization of similar coal-fired boilers.

关 键 词：电站锅炉 锅炉热效率 NOX排放 遗传算法 多目标优化 

分 类 号：TK227.1[动力工程及工程热物理—动力机械及工程]