Abstract: The cluster optimization for traditional multi-level energy wireless sensor networks is generally only for cluster head election process optimization without the cluster generation process optimization. Corresponding to two stages including the formation and operation of the clusters, we optimize the formation and election of the clusters with two levels. The first level optimization is organizing the nodes in the network into clusters using Fuzzy c-Means algorithm, then we make the second optimization with the construction of mathematical model for the election of cluster heads considering the residual energy of the nodes, the distance between the nodes in the clusters and the distance between the nodes and the base station. Simulation result shows that the energy consumption of the nodes in the network is balanced and the lifetime of the network is prolonged with the algorithm. The protocol with the algorithm is more suitable than DEEC and TDEEC protocols for multi-level energy heterogeneous wireless sensor networks.