This paper introduces a centrality-guided modularity optimization algorithm for overlapping and nested community detection (CG-MONCD), aimed at addressing community structure identification in opportunistic networks. Characterized by data transmission through sporadic encounters between nodes without the need for stable communication links, the CG-MONCD algorithm constructs a weighted graph model by quantifying the frequency and duration of encounters, as well as energy consumption. Incorporating a centrality-based initial community construction strategy and a method for identifying overlapping nodes, it effectively detects overlapping and nested communities within the network. Experimental results demonstrate that this algorithm outperforms existing methods in community detection, modularity optimization, and intra-community edge density, accurately revealing the network's nested communities and hierarchical features, thereby enhancing communication efficiency and network stability.