Abstract: A strategy optimization framework based on the proximal policy optimization (PPO) algorithm is proposed to address the problem of optimizing and controlling key parameters in the reflow soldering process. The goal is to meet key process indicators while minimizing the heating factor area. Firstly, the process constraints and optimization indicators in the reflow soldering parameter optimization process are analyzed, and the problem is transformed into a continuous control optimization problem under the framework of sequential decision-making. It is further formalized as a Markov decision process, clarifying key elements in the reinforcement learning process. Then, to enhance the stability and policy expression ability of the reinforcement learning algorithm, an Actor-Critic strategy optimization framework incorporating generalized advantage estimation (GAE) is adopted. Finally, relevant experiments for reflow soldering parameter optimization are designed, verifying that the intelligent optimization method based on the PPO algorithm exhibits better stability and generalization ability compared to traditional methods, thus providing effective technical support for intelligent parameter adjustment in actual production.