双人工神经网络建模及约束条件下的遗传算法优化.pdfVIP

27 3 动 力 工 程 Vol. 27 No. 3 2007 6 Journal of Power Engineering June 2007 : 2007) 0335705 刘福国 ( , 250002) : 采用双人工神经网络模型, 能使人工神经网络从高维度的有限样本中提取 靠的信息 此外, 还给出了基于双网络模型的有约束遗传算法的优化方法在锅炉燃烧区段风量和粉量分配 建模及优化中的应用表明: 双网络模型有较高的 靠性; 风量和粉量的合理分配能够改善锅炉运行 的经济和环保效益; 双网络模型对该过程的推断与现有的知识和经验相符 : 自动控制技术; 人工神经网络; 双网络建模; 遗传算法; 数据挖掘; 锅炉; 燃烧优化 : TK223; N 4 : B Dual Aritificial Neural Network Modeling and Optimization by Genetic Algorit hm wit h Constraints LI U Fug uo ( Shandong Research Institute of Electric Power, Jinan 250002, China) Abstract: Reliable information may be drawn from multidimensional finite numbered samples. Moreover, a way of optimization by genetic algorithm with constraints, based on dual neural network models, is being presented. Application in modeling and ooptimizing the allotment, of air and pulverized coal, in combustion zones of boilers, shows that dual network models are quite reliable; rational allotment of air and pulverized coal improves operationeconomical and environmental benefits. Ascertions concerning boiler combustion processes conform with present day knowledge and experiences Keywords: automatic control technique; artificial neural network; dual network modeling; genetic algorithm; data mining; boiler; combustion optimization 8~ 15 m , , , , , NO , x , , , ; , NO ; , [ 1] [ 2] [ 3] [ 4] , x , 1 : : : ( 1 6 ) , , , , 1. 1