摘要
针对评价问题的实际情况,利用神经网络良好的自适应性和遗传算法强大的全局搜索能力,采用遗传算法和BP神经网络相结合的遗传神经网络算法解决装备研制作战需求方案评价问题。通过样本数据训练,提高了评价的准确性和适应性。仿真实验表明,该算法能够较好地克服人为因素和随机性的影响,其稳定性和精确性也有较高。
Aimming at the actual conditions of assessment problem, this paper makes use of the fair self-adapting capacity of the neural network and the strong global search capacity of the genetic algorithm, and it adopts genetic neural network to solve the assessment on the battle demand plan of weapon development by combining genetic algorithm with neural network. By sample data training, the algorithm raises the accuracy and applicability of estimate. It is proved that the algorithm can overcome the human factor and random influence, and its steadiness and nrecision are also higher.
出处
《指挥控制与仿真》
2008年第6期93-97,共5页
Command Control & Simulation
关键词
遗传算法
神经网络
装备研制
评价
genetic algorithm
neural network
weapon development
assessment
作者简介
傅勉（1978-），女，安徽霍邱人，讲师，研究方向为决策支持系统。
乔冠军（1979-），男，硕士研究生。
张杰（1980-），男，讲师。
金波（1979-），男，讲师。
