基于KPCA-IPOA-BiGRU的联合循环余热锅炉主蒸汽参数预测

Prediction of main steam parameters of combined cycle recovery steam generator based on KPCA-IPOA-BiGRU

余热锅炉主蒸汽参数对于联合循环机组的健康运行判断至关重要。针对余热锅炉运行参数非线性和时延性导致主蒸汽参数预测精度低的问题,提出了一种联合循环余热锅炉主蒸汽参数预测模型。首先,采集燃机电厂的SIS运行数据,通过灰色相关性分析法确定输入变量;其次,通过核主成分分析(kernel principal component analysis,KPCA)提取输入参数的特征信息,并根据主成分贡献率选取输入维度;最后,利用改进的鹈鹕优化算法(improved pelican optimization algorithm,IPOA)优化双向门控循环神经网络(bi-directional gated recirculation neural network,BiGRU),并构建KPCA-IPOA-BiGRU进行三压余热锅炉主蒸汽参数的预测测验。结果表明,现场采集的10000组数据点,其中8000组用作训练集,2000组用作测试集,提出的模型将28维输入参数降低至8维,可准确预测3个压力级的蒸汽参数,且R2均大于98%,为预测时延性的联合循环余热锅炉主蒸汽参数提供了技术支持。

Main steam parameters of heat recovery steam generator(HRSG)are crucial for the healthy operation of the combined cycle unit.Aiming at the problem of low prediction accuracy of main steam parameters due to non-linearity and time-delay of operating parameters of HRSG,a model for predicting main steam parameters of the combined cycle HRSG is proposed.Firstly,operation data of a gas turbine combined cycle were collected to determine the input variables by grey correlation analysis.Secondly,the feature information of the input parameters was extracted by KPCA and the input dimensions were selected according to the principal component contribution ratio.Finally,BiGRU was optimised by IPOA and KPCA-IPOA-BiGRU was constructed to conduct the prediction test of the three-pressure-HRSG main steam parameters.The results show that the proposed model reduces the 28-dimensional input parameters to 8 dimensions and can predict the three-pressure steam parametersusing 8000 data points as the training set,and 2000 data points as the test set.The R2 is greater than 98%for all of the tests,which provides technical support for monitoring time-delayed main steam parameters.