论文部分内容阅读
针对导弹武器系统生存能力评估问题,建立了评估的指标体系,给出了指标的隶属函数.利用BP神经网络的自学习、自适应、强容错性,并通过遗传算法(GA)优化BP神经网络的连接权重和阀值,弱化了评价中的人为因素,提高了评价结果的准确性和权威性,解决了BP神经网络存在落入局部最小点和收敛速度慢的问题.实例研究表明,遗传神经网络的评价模型具有有效性和可行性.
Aiming at the assessment of the survivability of missile weapon system, an index system of assessment is established and the membership function of the indicator is given.Using self-learning, self-adaptation and strong fault tolerance of BP neural network, BP neural network is optimized by genetic algorithm (GA) , Weakened the human factors in the evaluation and raised the veracity and authority of the evaluation results, and solved the problem of BP neural network falling into the local minimum and slow convergence rate.Examples show that genetic neural The evaluation model of the network is effective and feasible.