论文部分内容阅读
循环流化床锅炉具有污染物排放少、燃料适应性广、负荷调节能力强等优点,近年来在电力、供热等行业中得到广泛应用。然而目前大部分循环流化床锅炉均存在自动投入率低,操作依赖人工经验的特点,造成这一状况的一个重要原因是缺乏合理的数学模型。首先对工艺流程进行分析,选取对锅炉效率影响最大的7个参数作为建模对象:过量空气系数、床温、排烟温差、飞灰含碳量、一次风机电流、二次风机电流、引风机电流。每个参数有各自不同的特点,对不同的统计模型的适用性也不尽相同。为了达到最佳建模效果,分别应用多元线性回归、多元逐步回归、偏最小二乘回归及BP神经网络对这些参数进行建模。实例研究表明,过量空气系数和二次风机电流适合采用偏最小二乘回归法建模;床温、排烟温差、一次风机电流和引风机电流适合采用多元线性回归法建模;飞灰含碳量采用BP网络模型对其预测效果相对较好。本文所建的模型对循环流化床锅炉的节能分析和进一步的操作优化研究具有一定的实际意义。
Circulating fluidized bed boilers have the advantages of less emission of pollutants, wide adaptability of fuel and strong ability of load regulation. In recent years, they have been widely used in electric power, heating and other industries. However, at present, most circulating fluidized bed boilers have the characteristics of low automatic input rate and operation dependent on human experience. An important reason for this situation is the lack of a reasonable mathematical model. Firstly, the process flow was analyzed, and the seven parameters with the greatest influence on the boiler efficiency were selected as modeling objects: excess air ratio, bed temperature, exhaust temperature difference, carbon content in fly ash, primary fan current, secondary fan current, Current. Each parameter has its own different characteristics, the applicability of different statistical models are not the same. In order to achieve the best modeling results, these parameters were modeled using multiple linear regression, multiple stepwise regression, partial least-squares regression and BP neural network respectively. The case study shows that excess air coefficient and secondary fan current are suitable for modeling by PLS regression. Bed temperature, exhaust temperature difference, primary fan current and induced draft fan current are suitable to be modeled by multiple linear regression. The BP neural network model has a better forecasting result. The model built in this paper has some practical significance for energy-saving analysis and further optimization of operation of CFB boiler.