Explainable Reinforcement Learning for Formula One Race Strategy