基于GRNN的船舶柴油机NOx排放预测-《集美大学学报（自然科学版）》

文章信息/Info

Title:: Prediction of NOxEmissions from Four-stroke Marine Diesel Engine Based on General Regression Neural Network

作者:: 孙思远1; 黄朝霞2; 徐虎1; 李孟杰3; 黄加亮1; 4; （1.集美大学轮机工程学院，福建厦门 361021；2.集美大学理学院，福建厦门 361021；3.中华人民共和国河北海事局，河北曹妃甸063200；4.福建省船舶与海洋工程重点实验室，福建厦门 361021）

Author(s):: SUN Siyuan1; HUANG Zhaoxia2; XU Hu1; LI Mengjie3; HUANG Jialiang1; 4; (1.School of Marine Engineering,Jimei University,Xiamen 361021,China；2.School of Science,Jimei University,Xiamen 361021,China;3.Hebei Maritime Safety Administration of People＇s Republic of China,Caofeidian 063200,China；4.Fujian Provincial Key Laboratory of Naval Architecture and Ocean Engineering,Xiamen 361021,China)

Keywords:: four-stroke diesel engine; GRNN; propulsion characteristics; load characteristics; nitrogen oxide emission

摘要:: 基于4190ZLC-2船用四冲程增压柴油机实验测得的数据，运用广义回归神经网络（GRNN）相关理论，以转速、功率、喷油提前角作为样本的输入量，以排放气体氮氧化物（NOx）的体积分数作为样本的输出量，对输入数据进行归一化处理，对输出数据进行反归一化处理，建立船舶柴油机广义回归神经网络排放预测模型，在推进特性与负荷特性工况下利用该模型进行柴油机NOx的排放预测。仿真结果表明，所提出的模型具有较高的预测精度，可为柴油机减少NOx排放提供依据。

Abstract:: Based on the experimental data measured from the 4190ZLC-2 marine four stroke turbocharged diesel engine,the related theory general regression neural network(GRNN) was employed to establish the model for predicting the emission of nitrogen oxides(NOx) from marine diesel engine.The prediction was carried out on the model by setting the rotation speed,power and fuel injection advance angle as the input,and the volume fraction of nitrogen oxides (NOx) within the exhausted gases as the output. The inputs of were normalized and the outputs were anti normalized.The model was finally used to predict nitrogen oxides emissions under the condition of propulsive characteristic and load characteristic.Results show that the proposed model has higher prediction accuracy and is suitable for providing reference for the reduction of NOx emissions.