论文部分内容阅读
针对电厂烟气污染物SO2监测中出现的安装费用高、维护困难等问题,提出了一种烟气SO2排放量预测方法.首先应用BP神经网络建立模型,然后应用遗传算法对BP神经网络的连接权值和阈值进行优化配置,建立新的软测量模型,并对改进后的软测量模型的预测结果与实际运行数据进行了比较,结果表明:所建立的新模型具有较高的准确性和稳定性,可以较好地预测电厂烟气SO2排放质量浓度.
Aiming at the problems such as high installation cost and difficulty of maintenance in SO2 monitoring of flue gas pollutant in power plant, a method of SO2 emission prediction is proposed.Firstly, BP neural network is used to establish the model, and then genetic algorithm is used to connect BP neural network Weights and thresholds are optimized, and a new soft-sensing model is established. The prediction results of the improved soft-sensing model are compared with the actual operating data. The results show that the new model has higher accuracy and stability Which can better predict the SO2 emission concentration of power plant flue gas.