Real-World Range Estimation Of Electric Vehicles Using Linear Regression Model

  • Mr. Deepak S, Dr. Nitin Khedkar


Even though electric vehicles have been introduced in the market, they are still seen as an incomplete solution to mobility due to their limited capacity range per single charge. The real-world range estimation is a problem to solve for gaining customer trust and warmth. The estimations currently available for estimations use a standard test cycle approach rather than the real world situation of vehicle running on the road. To add on, the driver-specific inputs on the average driving speed, loading condition of the vehicle also need to be captured and subsequent corrections need to be made in the pre-estimated range. The data generated over various conditions are used to train a mathematical model. The model can then be used to efficiently estimate the real world range of vehicles. The model used for prediction returns the estimated real-world range with a minimum permissible error. The predictions are then visualized through the use of a mobile app for ease of use.