<p>Ensuring k-coverage in Wireless Sensor Networks (WSNs), where each target is monitored by k sensors, is a critical issue that continues to impede the improvement of network fault tolerance. Current methodologies fail to address dynamic sensor malfunctions or appropriate deployment strategies. To address this deficiency, we offer an optimization framework that integrates Ant Colony Optimization (ACO) for the static placement of sensor nodes and Proximal Policy Optimization (PPO) for the dynamic repositioning of existing sensor nodes to achieve k-coverage (k = 1,3). The proposed hybrid ACO–PPO model introduces a novel adaptive framework that unites heuristic and reinforcement learning strategies for fault-tolerant k-coverage optimization. In this study, we developed a system that ensures optimal sensor placement in a grid capable of recovering from sensor failures. The coverage evaluation is conducted by simulation across multiple iterations, examining failures and adaptive recovery effects. The experimental findings indicate that the hybrid ACO-PPO model outperforms ACO and comparable techniques, maintaining a high coverage rate despite the absence of some sensor components. The results demonstrate that integrating reinforcement learning with heuristic algorithms significantly improves the efficiency and robustness of wireless sensor network installations, offering a powerful foundation for adaptive sensor system coverage.</p>