中文摘要：

针对电厂烟气含氧量难以进行有效预测的问题,提出一种烟气含氧量的智能混合预测方法。首先采用RBF神经网络、主元分析方法对输入变量进行降维处理;其次利用上述分析结果运用案例推理方法进行烟气含氧量的预测;然后,为反映烟气含氧量数据中的时间累积效应,采用过程神经网络方法对当前时刻烟气含氧量进行预测;最后基于方差-协方差方法的权值组合预测方法,获得最终的烟气含氧量。基于实际运行数据的分析和工业试运行表明,所提出的智能混合预测模型具有较高的精度和鲁棒性,可以较好地解决电厂烟气含氧量的预测问题。

英文摘要：

Aiming at the fact that the power plant flue gas oxygen content is hard to predict effectively,an intelligent hybrid prediction method is proposed.Firstly,RBF neural network and principal component analysis method are utilized to reduce the dimension of the input variables;secondly,the case-based reasoning method is adopted to predict the flue gas oxygen content from the above analysis results.Then in order to reflect the time cumulative effect of the flue gas oxygen content,the process neural network method is used to predict the current flue gas oxygen content value.Finally the final flue gas oxygen content is obtained using variance-covariance method of weight combination forecasting.The results of actual operation data analysis and industrial trial operation show that the proposed intelligent hybrid prediction model has high accuracy and robustness,and can effectively solve the problem of predicting flue gas oxygen content.