Research on emission prediction for series hybrid vehicles using AI technology

Vehicle exhaust emission standards have been tightened step by step. However, there are still problems such as localized pollution, and further improvement of the atmospheric environment is required. Against this background, OBM(On-Board Monitoring) system is planned to be introduced to monitor exhaust gas emissions performance when users use vehicles. The objective of this study is to establish virtual exhaust gas emissions sensor technologies applicable to OBM system. Based on data from chassis dynamometer tests for a series hybrid vehicle, the models to predict exhaust emissions were developed using multiple AI techniques and evaluated in terms of prediction accuracy and computation time.