In this paper, we introduce the Improved Chaotic Lévy Particle Swarm Optimization (ICLPSO), a cluster routing algorithm based on PSO optimization, which effectively extends network lifetime of Wireless Sensor Networks (WSNs) and enhances communication quality. Targeting the large number of IoT nodes, the single aggregation node, and environmental interference in large shipbuilding bases, we first incorporate Tent chaotic mapping in the initialization phase of the PSO population. We then combine the Lévy flight strategy in the particle update phase to prevent the algorithm from getting stuck in local optima, enhance stability, and improve global search capability and optimization accuracy. In the cluster head (CH) election process, a new fitness function is designed based on residual energy, the distance between nodes within the cluster and the CH, and the transmission signal strength. ICLPSO is used to optimize this process and to comprehensively select the optimal CH. Simulation results show that ICLPSO effectively achieves node load balancing, reduces network energy consumption, and extends network lifetime in this environment.