Reinforcement learning (RL) has significant potential across various fields, with its application in sports decision optimization emerging as a prominent research focus amid rapid advances in AI technologies. This paper develops tennisDecisionRL, an RL-driven decision optimization framework for tennis competitions that captures critical match dynamics through the integration of RL techniques such as Deep Q-Networks (DQN) and Proximal Policy Optimization (PPO). The proposed system enhances the sophistication and adaptability of competitive decision-making processes. Empirical evaluations using two datasets demonstrate its superior performance compared to traditional approaches. The paper concludes by identifying current limitations and outlining future research directions, providing novel insights for decision optimization in tennis and other athletic competitions.