## Fuel Consumption Forecasting

Fuel consumption forecasting is based on mileage prediction (D), which is calculated using a machine learning method with an additive model and the MCMC (Markov Chain Monte Carlo) algorithm. The model is trained considering trends, seasonality factors, and a series of regressive components.

E – Efficiency

If there is more than 12 months of full history in the app, the current value is used for each predicted month. July is copied to July, and December to December. This allows for the most accurate consideration of seasonal changes in your car’s fuel consumption.

If the history is less than 12 months, the app calculates the average value and applies it to the entire predicted period. In this case, the seasonal factor is not taken into account.

P – Price

The ARBA Auto app does not attempt to predict changes in fuel and electricity prices. The last (current) non-zero price at which you purchased fuel or electricity is used for all calculations.

T – Total Fuel (Charged)

The amount of fuel needed to travel the predicted distance is calculated by the formula: T = D / E; Where D is the distance, and E is the fuel efficiency.

FC – Fuel (Energy) consumption

The fuel consumption forecast is calculated by the formula: FC = T / D * 100

Where T is the amount of fuel, and D is the distance.

A – Total Spent (Fuel /energy expenses)

Calculated by the formula: A = T * P

Where T is the amount of fuel, and P is the price.

C – Trip cost (Cost per km/mi)

The trip cost is calculated by the formula: C = A / D

Where A is the fuel expenses, and D is the distance.

US MPG – How many miles a car travels using 1 gallon of fuel

For the US, this is calculated by the standard formula: US MPG = D / T,

Where D is the distance, and T is the total fuel input. For all other countries, a conversion from the metric system is additionally applied.