Real-Time Path Planning for Gas Leak Monitoring Based on PPO Algorithm

As industrial production continues to expand, gas leakage accidents have become more frequent, posing serious threats to both operational safety and environmental health. Unmanned Aerial Vehicles (UAVs), due to their rapid response and adaptability, play an increasingly important role in gas leak monitoring. However, gas diffusion is influenced by factors such as leakage rate and wind speed, making the monitoring environment highly dynamic. Ignoring such variations in path planning may lead to inefficient monitoring or even result in serious accidents. To address this problem, we propose a real-time UAV path planning strategy based on the Proximal Policy Optimization (PPO) algorithm. This approach leverages a multi-objective reward function to guide the UAV in avoiding high-risk areas while prioritizing information-rich regions, thereby maximizing monitoring efficiency. Simulation results demonstrate that the proposed method enables adaptive and efficient path planning in dynamic and uncertain environments.