Constant energy supply for sensor nodes is essential for the development of the green Internet of Things (IoT). Recently, WRSNs have been proposed to resolve the energy limitations of nodes, aiming to realize continuous functioning. In this article, a coverage-aware hierarchical charging algorithm in WRSNs is proposed, considering energy consumption and the degree of node coverage. The algorithm first performs network clustering using the K-means algorithm. In addition, nodes are classified into multiple levels in each cluster to calculate respective anchor points based on the energy consumption rate and coverage degree of nodes. Then, the anchor points converge to an optimized anchor point in each cluster. To reduce charging latency, the optimized anchor points form two disjoint internal and external polygons. Next, mobile chargers travel along the internal and external polygons, respectively. Experimental results indicate that the proposed algorithm can improve charging efficiency and reduce charging latency substantially.