Abstract: Aiming at identifying the fault types of plant fans rapidly and correctly, a new immune population network algorithm was proposed in this paper, which combined clone selection algorithm and immune network algorithm. The immune population network algorithm can execute both local and global search to multiantigen population, and form two-layer searching mechanism. It improves the capability of global and local search, effectively eliminates the premature convergence and increases the diversity of antigen population. The simulation results indicats that the immune population network algorithm can recognize the plant fans' faults efficiently.