A Parallel Supervision System for Vehicle CO2 Emissions Based on OBD-Independent Information

A parallel supervision system is built in this paper in order to accurately estimate vehicleCO(2) emissions. Only on-board diagnostics (OBD)-independent information is used, making the model capable of making predictions based on future road gradients and planned speed trajectories. Based on the parallel theory, the actual traffic environment is considered the physical world, while the combined CO2 model (which consists of physical and data-driven models) is the core part of the artificial world. The physical model uses a cascaded structure with engine speeds and torques as intermediate variables, and the data-drivenmodel relies on a modified long short-term memory (LSTM) neural network. When the historical data is sufficient in size and diversity, the data-driven model is appropriate and achieves more accurate estimations; otherwise, the physical model is preferable because of its greater robustness. Based on this combined model, the supervision system can leverage both the learning ability and physics-based knowledge. Two real-world experimental case studies have been performed to validate this system. According to the research analysis, both the physical and data-driven models achieve sufficient accuracy. The physical model indicatesmore robustness even when some primary parameters (gear ratios) are unknown, which can be used as a supplement to the data-driven model. Moreover, the deterioration factor (DF) of vehicleCO(2) emissions is considered to simulate aged vehicles. This parallel supervision system can effectively address the gap between regulatory test cycles and real-world carbon emissions.

Automated summary

A parallel supervision system is developed to accurately estimate vehicle CO2 emissions by using only onboard diagnostics (OBD)-independent information. The system can predict future road gradients and planned speed trajectories. The combined CO2 model, consisting of physical and data-driven models, is considered the core part of the artificial world, while the actual traffic environment is regarded as the physical world. Two real-world experimental case studies validate the accuracy of both the physical and data-driven models, with the physical model showing more robustness. The system effectively bridges the gap between regulatory test cycles and real-world carbon emissions.