基于流程模拟与Elman神经网络的燃煤电厂CO_2吸收系统动态数据检验方法

Dynamic Data Validation Method of CO_2 Absorption System in Coal-fired Power Plant Based on Process Simulation and Elman Network

提出了结合流程模拟与Elman神经网络的燃煤电厂联胺吸收CO2系统动态数据检验方法。通过流程模拟软件Aspen HYSYS建立动态模型,获得动态数据,采用一步前预测方法构造样本训练Elman网络,通过训练后的网络对测量参数进行预测估计。通过Aspen HYSYS与Matlab的数据通信实现了Elman网络对HYSYS动态过程数据的在线检验。仿真试验表明,该方法能有效地对CO2吸收系统动态过程的数据进行在线检验,提高脱碳系统监测的可靠性。

A dynamic data validation method of CO2 absorption system with diamide in coal-fired power plant based on a combination of process simulation and Elman neural network was proposed.The dynamic model was constructed by the process simulation soft Aspen HYSYS,and the dynamic data was obtained.The one-step-ahead prediction method was adopted to create sample training set for Elman network,and the prediction of the measurement parameters were obtained through the trained network.The online validation of Elman network to dynamic process data of HYSYS was implemented by data communication between Aspen HYSYS and Matlab.The simulation tests showed that this method was effective in data online validation of CO2 absorption system dynamic process and the monitoring reliability of CO2 absorption system could be improved.