基于混合智能的锅炉飞灰含碳量实时目标值模型  

Real-time Optimal Value of Unburned Carbon in Fly Ash Based on Hybrid Intelligence Technique

作　　者：洪军[1] 司风琪[1] 徐治皋[1] 

机构地区：[1]东南大学能源与环境学院,江苏南京210096

出　　处：《锅炉技术》2007年第2期15-19,共5页Boiler Technology

摘　　要：为了降低飞灰含碳量,提高锅炉运行水平,运用混合智能技术建立了飞灰含碳量目标值模型。从运行优化角度提出了飞灰含碳量目标值的定义和技术可行方案。通过对锅炉的历史运行工况数据库进行数据挖掘,建立了锅炉历史最优工况数据库,以此作为训练样本建立飞灰含碳量目标值的神经网络模型,在进行了实例验证后对模型进行了分析讨论。实际应用表明该模型具备自调节能力,能够向运行人员实时提供当前工况下的飞灰含碳量目标值,为飞灰含碳量的实时优化指明了调整的方向。To improve boiler operation level, a new conceptualization of optimal value of unburned carbon content in the fly ash was expatiated. A viable technical solution of optimal value of unburned carbon content in the fly ash was developed by hybrid intelligence technique. Boiler historical optimal operation database were established by data mining based on boiler original operation database. An artificial neural network model for optimal value of unburned carbon content in the fly ash was set up and verified. An example was presented to demonstrate the effectiveness of the model. The application result shows that the model could provide real-time optimal value of unburned carbon content in the fly ash for operational staff with better predicting performance, higher calculation speed, outstanding self-tuning ability and generalization ability comparing with other modeling methods.

关 键 词：飞灰含碳量 混合智能 电站锅炉 运行优化 数据挖掘 神经网络 

分 类 号：TK223[动力工程及工程热物理—动力机械及工程]