Evaluation of Capability for Gun-missile System Based on BP Neural Network

Gun-missile system is an anti-aircraft system, whose aim is low-altitude air target. The capability of gun-missile system influences the air defence capability. This paper made the structure and the function analysis of the system first, then decided twelve characters, such as the maximal detection range, the number of target that can be detected and tracked synchronously, the maximal steagy tracking range, firing rate, initial velocity, dispersion of fire, hit probability, mass of projectile, mass of missile, reliability, supportability, and the weight of system, that attributed system capability. Built the prediction model using BP neural network with feedback and predicted the capability of gun-missile system. It is convenient to achieve BP network utilizing those functions relevant with neural net design, training and simulating, which are supplied with the neural network tool kit of MATLAB. The correctness of the model has been verified by calculating a certain example of gun-missile system barrel.