Neural Network Method for Ballistic Parameters Design Taking Uncertainty into Account

To investigate the effect of design parameters with uncertainty characteristics on performance in ballistic design, neural network method was adopted to determine optimized standard deviations of corresponding parameters. A trajectory model was established first and Monte Carlo simulation was done to analyze statistical performance of the flight range, which is used as the objective. By controlling distribution variances of parameters using a neural network approach, Circular Error Probability (CEP) is limited within a satisfactory range and therefore precision is improved.