Abstract

This investigation focuses on predicting race times and results of Formula One races, using historic data as well as the application of some innovative models. Prior to producing a prediction engine, an in depth study of historic success of different approaches to pit stops was undertaken to gain an important understanding of strategy. Throughout the investigation MatLab was used to program solutions due to its data analysis capabilities.

 

Firstly a simple mathematical model was constructed to predict the drivers lap times, and hence race result, based upon their optimal lap time set pre race and the phenomenon of fuel effect. The investigation then concentrates on modelling four of the most important variables in Formula One, consisting of time lost through driver mistakes, lapping slower cars, being in a battle for position with another driver and tyre degradation. These models were then collated to evaluate the success of their application in predicting race times and results for the 2008 Malaysian Grand Prix, with an analysis of their respective accuracy. Finally several applications of this investigation are discussed with initial work being performed on a strategy package which could be used at race weekends for a range of motorsport events.

Project in support of the degree of Master of Engineering

Prediction of Formula One Results

Onto the initial model.