NOx Virtual Prediction Technology of Diesel Engine Based on HPO-LSTM
In the face of strict emission regulations,the diesel engine post-treatment system plays an immeasurable role,and the acquisition of NOx emissions is one of the prerequisites for the normal operation of SCR device in the post-treatment sys-tem.A virtual prediction model that used hunter-prey optimization(HPO)algorithm to optimize long short term memory(LSTM)network was established to accurately predict NOx emissions of diesel engine in place of existing physical sensors or as a parallel device to monitor their operation.The test was carried out on a dynamometer of diesel engine.During the highly transient operation cycle of diesel engine,several parameters that were easy to obtain and closely related to NOx formation were input into the model.The results show that,compared with the prediction results of non-optimized network,RMSE increases by 29.1％and 23.4％,and R2 is greater than and close to 0.95 respectively when the optimized network is applied to the test set or to a new unknown transient condition.The prediction results show a highly identical trend with the measured values of sen-sor,which meets the requirements of on-board application and accuracy and hence verifies the feasibility of this method.
diesel engineNOxpredictionhunter-prey optimization algorithmlong short term memory network
万方数据
