Objectives  Warships will be attacked by missiles or shells when carrying out combat tasks, and armor piercing damage identification has certain characteristics, so it is of great significance to study this specific area.

  Methods  Taking the stiffened plate structure which commonly exists in ships as the research object, a probability neural network (PNN) with the natural frequencies of the first three modes of stiffened plate as its characteristic parameters is used to study the damage identification of a stiffened plate with round hole-shaped armor piercing; the symmetry identification of the natural frequencies is then solved by combining the structural damage identification method with the modal characteristics and acceleration variance.

  Results  The results show that this method has a good effect on the damage location and damage degree identification of round hole-shaped armor piercing damage.

  Conclusions  The PNN has such advantages as fast convergence, ease of implementation in hardware and high consistency with the problem of identifying round hole-shaped armor piercing damage in stiffened plate structures.