AI Based Drive Cycle Prediction for HEV

HEVs (Hybrid Electrical Vehicles) is regarded as one of the most effective solutions in the context of environmental pollution and fuel consumption. Nevertheless, formulating the sustainable Energy Management Strategies (EMSs) for HEVs is not an easy task while at the same time one has to meet different design objectives and achieve maximum fuel economy. Among numerous EMSs, the present paper highlights the new system of AI-powered Driving Cycle Prediction. Using the Machine Learning Algorithms and Time Series Forecasting, we apply Dynamometer Drive Schedules data for EVs, meanwhile, Mean Squared Error (MSE) helps to evaluate the model quality. XGBoost and NBeats algorithms comparison shows that the last one has much higher forecasting accuracy; however, it is computationally heavier. Such study is focused on the development of HEV technology and hence the availability of sustainable transport solutions through the process of providing detailed information on the algorithmic outputs.