A PKGV-ANN model for vehicle high emitters identification based on remote sensing data

Vehicle emission remote sensing system is an effective real-time method of monitoring vehicle emissions on road. This paper presents an artificial neural network model for identifying high emitters by combing the remote sensing data and the idle test data. On the base, an improved model called PKGV-ANN is proposed. The model combines several advanced methods and useful achievements, including principle components analysis, k-nearest neighbor algorithm, genetic algorithm and the results obtained from the studies about vehicle specific power. Experiments results show that the model is very valid. The percentage of hits reaches 89.40%.