基于BP神经网络的PHEV油耗与排放研究

Research on Fuel Consumption and Emissions of PHEV based on BP Neural Network

文章以P1+P3结构PHEV为研究对象,设计了基于BP神经网络算法的动力匹配控制来提高PHEV的输出功率、降低排放及优化燃油经济性。结果表明:在山路、城市、高速和郊区四种路况下进行实车测试,嵌入算法后P1+P3结构PHEV的百公里油耗平均降低了0.61L,CO、CO_(2)、HC、NO_(X)排放分别降低了0.28g/km、0.198g/km、0.813g/km、0.021g/km,排放和燃油经济性均得到改善。

This article takes the P1+P3 structure PHEV as the research object,and designs a power matching control based on BP neural network algorithm to improve the output power of PHEV,reduce emissions,and optimize fuel economy.The results showed that during actual vehicle testing under four road conditions:mountain roads,cities,highways,and suburbs,the P1+P3 structure PHEV with embedded algorithm showed an average reduction of 0.61L in fuel consumption per 100 kilometers,and CO,CO_(2),HC,and NO_(X) emissions decreased by 0.28g/km,0.198g/km,0.813g/km,and 0.021g/km,respectively.The emissions and fuel economy were improved.