Energy Consumption Prediction for Electric Vehicles Based on Real-World Data

Photo by Oxana Melis on Unsplash

Why does your EV sometimes run out of charge faster, even on the same route?

That’s what a group of researchers from Vrije Universiteit Brussel tried to understand. They studied how different real-world factors affect how much energy an electric vehicle (EV) uses.

Problem

Most EV energy tests are done in labs under fixed conditions. But in real life, things are different, roads have slopes, traffic slows you down, temperatures change, and everyone drives differently. These factors make it hard to predict how much battery power an EV will actually use.

What researchers did

The researchers collected real driving data from a Nissan Leaf for two years, over 23,000 kilometers of city, rural, and highway trips around Brussels. They recorded the car’s speed, acceleration, road slope, temperature, and battery use every second.

Then, instead of just using physics equations, they combined physics with real data using statistical modeling to build formulas that can predict how much energy an EV will need for any trip.

Formula

They started from the physics of driving:

• A car needs energy to overcome rolling resistance (friction with the road).
• It uses more energy when going uphill.
• Air drag increases quickly as speed goes up.
• Acceleration burns extra energy, while braking can recover some through regeneration.
• Air conditioning and heating also drain the battery.

By analyzing thousands of trips, they built a formula that links these effects to energy use:

E = B0 
  + B1 * distance 
  + B2 * v^2 
  + B3 * slope 
  + B4 * acceleration_metric 
  + B5 * temperature_term

Each BBB is a coefficient that tells how strongly that factor affects energy. In simple terms: speed, slope, acceleration, and temperature all matter.

What They Found

• Acceleration and stop-and-go traffic increase energy use by about 20%.
• Uphill roads add a big energy penalty, while downhill roads can help recover up to 40% through regenerative braking.
• Heating or cooling systems can consume up to 40% of total energy on short trips.
• Their model predicted actual trip energy with about ±25% accuracy, which is quite good for messy real-world data.

Why It Matters

This research shows that real-world data gives a much clearer picture of EV performance than lab tests. Accurate energy formulas like this can help:

• EV owners plan trips better.
• Delivery companies predict range and charging needs.
• Researchers and engineers design smarter routing and battery management systems.

Conclusion

Physics explains the forces that drain an EV battery, but real-world data reveals how those forces behave in daily driving. Combining both gives us a more accurate way to understand — and improve — electric mobility.

References

De Cauwer et al.