【摘要】 基于混沌退火算法和BPNN模型的末敏弹系统效能参数优化,引入贝叶斯正则化方法的BPNN模型,使神经网络具有自适应性和推广能力。交替使用贝叶斯正则化和混沌退火算法,对网络参数进行优化,并对训练后的网络用系统效能参数。结果表明该模型不仅能拟合原系统,而且末敏弹系统效能参数更优,命中率更高。更多还原

【Abstract】 Efficiency parameters optimization about terminal-sensitive projectiles was based on algorithms of chaos-Annealing strategy and BPNN model. The BPNN model of Bayesian regularization method was adopted to create the adaptivity and generalization of BPNN. The Bayesian regularization method and algorithms of chaos-annealing strategy were used by turns to optimize the network parameters; and the system efficiency parameters were used in trained network. The results shows that the model could fit the original system, and the efficiency parameters of terminal-sensitive projectile system are better and hit rate is higher.更多还原