توجه: محتویات این صفحه به صورت خودکار پردازش شده و مقاله‌های نویسندگانی با تشابه اسمی، همگی در بخش یکسان نمایش داده می‌شوند.
۱Prediction of performance and emissions of diesel engine fueled with various biodiesel using by neural networks
اطلاعات انتشار: هشتمین همایش موتورهای درونسوز، سال
تعداد صفحات: ۱۱
The purpose of this paper is to develop two artificial neural network(ANN) models, back propagation andgeneralized regression neural networks, for estimating exhaust emissions and engine performance of a dieselengine operated with various blends of three different biodiesels with diesel fuel under a variety of operationconditions. Experimental data, which obtained from a semi‐heavy duty, turbocharged, four cylinder, directinjection diesel engine, has been used for designing both generalized regression (GRNN) and back propagation(BPNN) neural networks. Predictive abilities of these two neural networks are compared. The predicted resultsshow that the coefficient of determination (R2) values of developed BPNN model are 0.9456, 0.9961, 0.9960,0.9912, 0.9838, 0.8952 and 0.9901 for, CO, CO2,O2, NOx, PM, power , and exhaust temperature respectively.However, these values for developed GRNN model are equal to 0.9812, 0.9935, 0.9861, 0.9878, 0.9879, 0.9096and 0.9880, respectively. Also, the relative root mean square error (R‐rmse) values for the BPNN and GRNN are0.0435 and 0.0496, respectively. The comparison of predicted results indicate that while generalized regressionneural networks are better than the traditional back propagation neural networks in terms of speed andsimplicity, back propagation neural network can predict more accurately than generalized regression network ata well‐trained condition. Thus, BPNN is a robust virtual sensing tool for prediction and modeling of performanceand emissions of diesel engine fueled with diverse biodiesels and their diesel blends.<\div>
نمایش نتایج ۱ تا ۱ از میان ۱ نتیجه