I have recently written a paper on applying Q-Learning – a reinforcement learning algorithm – to the acceleration problem brought up in my Going Fast on Your Own Power series. The paper’s abstract, the paper itself, and a corresponding presentation are below.
In this report, Q-Learning is used to maximize velocity and minimize driver fatigue in a human powered vehicle attempting a land speed record. Q-Learning results in a set of Q-values for state-action pairs, called a policy. That policy is utilized here to provide a function of power input over time for a given driver, vehicle, and environment combination that maximizes velocity for the driver’s fatigue limit.This function of power input over time results in higher achievable speeds than other methods of developing a power over time function.