循环流化床锅炉床温动态模型  

Dynamic Modeling on Bed Temperature of Circulating Fluidized Bed Boiler

作　　者：刘吉臻 洪烽 吕游 杨婷婷 高明明 

机构地区：[1]新能源电力系统国家重点实验室(华北电力大学),北京市昌平区102206

出　　处：《中国电机工程学报》2016年第8期2168-2174,共7页Proceedings of the CSEE

基　　金：中央高校基本科研业务费专项资金项目(2015ZZD15;2016XS40;2016MS28);国家重点基础研究发展计划项目(973计划)(2012CBC215203)~~

摘　　要：循环流化床(CFB)锅炉的结构和控制系统与煤粉炉差异较大,最明显的区别在于床温控制。通过分析床温动态过程,定义了输入量的动态阶,利用最小二乘支持向量机(LSSVM)算法实时预测动态床温,采用遗传算法(GA)对不同负荷段下模型最佳输入动态阶组进行离线寻优,同时,更新LSSVM输入阶,保证准确性和有效性。由此建立GA-LSSVM床温动态模型。基于机组运行的实际数据,对所建模型进行仿真验证,结果表明模型具有较高精度,实时性强,能达到预测动态床温的目的。此外,通过模型仿真验证了不同负荷段下CFB锅炉床温动态性的差异,揭示了床温响应与锅炉蓄热、惯性的联系。The structure and control system between circulating fluidized bed（CFB） boiler and pulverized coal are of great difference, of which the most obvious is the control of the bed temperature. On the analysis of bed temperature dynamic process in the mechanism of bed temperature, the dynamic order for the inputs was defined. The GA-LSSVM dynamic bed temperature model of CFB boiler established by the least squares support vector machine（LSSVM） algorithmfor real-time prediction of dynamic bed temperature, which obtains the best dynamic order of inputs among each working span through genetic algorithm（GA） off-line optimization. Based on the running data of actual unit, the simulation results indicate that the dynamic bed temperature model owns high accuracy, real-time, which can be applied to predict the dynamic bed temperature. In addition, the change of the boiler dynamic performance is verified under different load sections,which reveals the relationship between the bed temperature response and the heat storage as well as inertia of the boiler.

关 键 词：循环流化床锅炉 数据建模 床温动态模型 动态阶 

分 类 号：TK39[动力工程及工程热物理—热能工程]