基于偏最小二乘回归的氨法烟气脱硫效率预测

摘要

Taking the main parameters affecting the ammonia flue gas desulfurization efficiency as input variables, and the desulfurization efficiency as output variables, a model for calculation of the desulfurization efficiency was built up using partial least squares regression, based on which efficiency predication was carried out for a certain power plant. Results show that the maximum relative error of desulfurization efficiency between predicted results and operation data is less than 4. 0%, and the average error of the model is 0. 53%, proving the calculation model to be accurate and applicable for efficiency prediction of flue gas desulfurization in actual engineering projects.%以影响氨法烟气脱硫效率的主要参数作为输入变量,以脱硫效率作为输出变量,采用偏最小二乘回归法建立了氨法烟气脱硫效率计算模型并对电厂脱硫效率进行了预测.结果表明:采用氨法烟气脱硫效率计算模型得出的预测值与实际运行数据的最大相对误差小于4.0％,模型的平均误差为0.53％,说明该烟气脱硫效率计算模型的预测精度较高,能够满足烟气脱硫效率预测的工程需要.