In wireless sensor networks (WSN), efficient data distribution technology is critical to meet the requirements of group communication and data fusion. D2MD (Device-to-Multidevice) multicast technology can directly transmit data streams from a sensor node to multiple terminal devices or processing centers, significantly improving data transmission efficiency and enhancing network robustness. However, after multicast clustering, cluster heads face the challenge of high energy consumption due to massive content distribution, which rapidly depletes battery capacity, shortens the survival time of multicast clusters, and ultimately reduces overall network lifespan. To address this issue, a network lifespan model under a dual-cluster-head collaboration mechanism is established by introducing a secondary cluster head as a forwarding node within the cluster. An improved Evolution-Fruit Fly Optimization Algorithm (EFOA)-based cluster head search algorithm is proposed to rapidly and efficiently select the optimal cluster head pair. Simultaneously, aiming at maximizing network lifespan, an energy efficiency model is constructed, and cluster head transmission power optimization is achieved based on the Dinkelbach algorithm. Simulation results demonstrate that the proposed EFOA algorithm outperforms Neighborhood Search (NS) and Exhaustive Search Method (ESM) in convergence speed, while its achieved network lifespan closely approximates the results of the ESM algorithm.