In the last ten years the number of electric vehicles on the roads has increased significantly. One reason for this popularity growth is that the range of electric vehicles has reached a level that makes it usable for more people in everyday life. However, the range is also the biggest disadvantage compared to combustion engines. The average range of an electric car is currently 200-300 km (some models can reach 500 km), while a conventional car has 700-1000 km (or even more) , a further disadvantage is the charging time of the accumulators takes much more time for electric cars.
In my thesis I made a plan for a program that helps to optimize the energy usage of electric vehicles by predicting the different kind of energy losses on a planned route. The core of the program is a vehicle model, that estimates vehicle energy usage.
The model can be used to calculate energy needed for a defined route. The model uses three energy type to estimate the total energy usage. These are the kinetic, potential energy, and work due to air resistance (drag) force. As the real energy usage is differ from the estimated energy requirement, I have developed two correction methods to correct the model's estimation. The first method corrects the estimated values during the calculation of the energy demand. The second method corrects the energy demand by using measured data realtime.
I implemented the model and the related features. These are the estimation of energy demand for the route, the implementation of correction methods, and an algorithm for estimating the range of the vehicle.
Testing is a necessary part of all models and takes a significant amount of time in the development. To make testing easier, I wrote two additional programs. The first program simplifies the generation of test data, while the second supports the test run. The audit of the program was divided into two major parts: testing the correctness of the implemented model and the correction during vehicle progression. In the first part, I ran six tests to verify the correct functioning of the model, that confirmed my preliminary theoretical expectations. In the second audit section I conducted four tests to check and compare the correction methods. By evaluating the results I could choose the correction that could provide the most accurate result, while the error remains within an acceptable level.