Abstract: A new method of on-line calculation of heat efficiency of circulating fluidized bed boiler (CFB Boiler) was proposed. Based on some measurable operation parameters, including pressure, temperature and the flow rates of main steam, water and primary and secondary air, which were collected online, input energy and the utilization of energy in effect could be deduced respectively, and then heat efficiency could be calculated. Using the plant-site data from a 135 t/h CFB boiler unit in Fujian, the online calculation of heat efficiency was simulated and the result of industrial simulation indicated that this method was accurate and effective. Furthermore, orthogonal experiment was performed to analyze seven factors which contributed to heat efficiency of CFB boiler. Because it is hard to get corresponding data on the orthogonal layout, a Levenberg-Marquardt (LM)_back propagation (BP) algorithm-based predictive model was established by using the data collected from plant-site and on-line calculation model. The result of orthogonal experiment indicated that air/coal ratio, main steam pressure and boiler firebox temperature were key factors. The result was validated by the industrial optimization experiment. This method will help on-line optimal operation of the CFB boiler.

摘要： 提出一种基于正平衡方法的热效率在线计算方法：通过在线采集锅炉运行中主蒸汽、给水、一二次风等锅炉运行参数，计算锅炉有效利用能量和输入能量，实现热效率的在线计算；采用福建某135 t/h工业CFB锅炉发电机组现场数据进行仿真计算，仿真结果验证了算法的有效性。在此基础上采用正交试验设计方法分析影响CFB锅炉热效率的影响因素；由于工业现场采样的操作数据难以得到各个因素不同水平下的热效率的对应值，故首先采用工业数据和热效率计算数据建立基于LM-BP算法的人工神经网络热效率预测模型。根据模型预测数据设计正交试验，结果表明对于该锅炉风煤比、主汽压力和炉床温度对该锅炉热效率的影响最大。最后进行工业调优试验，试验结果验证了分析结果并初步实现了锅炉热效率的在线优化。