Density Sequence Estimation Based on BP Neural Network
Taking into account the impact of the weak correlation error caused by the firing impact load on the dispersion,and the small sample size of the firing test,it is impossible to obtain a full pic-ture of the overall dispersion characteristics.The credibility of the test conclusion would be affected if continuing to use the traditional method,so an improved density estimation method of rapid-fire gun was proposed.On the basis of the consistency test of multiple sets of test data,the timing information of impact points were analyzed,and impact points were simulated and predicted by the BP neural net-work method to solve the small sample problem.Finally,the probability density function with coordi-nates of impact points obeying statistical laws was obtained based on the maximum entropy method,and the density was calculated.Example analysis shows that this method can more accurately describe the real dispersion situation of rapid-fire gun,and analysis results are accurate and reliable.This method can provide an effective reference for the analysis and evaluation density test data of rapid-fire gun.
NETL
NSTL
万方数据
