一种新的制导炸弹智能控制系统

A New Intelligent Control System of Guided Bomb

设计了一种具有抗干扰能力的制导炸弹智能控制系统。该系统采用非单点模糊推理系统为核心,利用梯度下降算法和遗传算法构成的混合并行学习算法训练建立推理规则,在自学习建立规则库的过程中,能够自动滤除训练数据中的噪声,获取准确的控制信息。通过计算机仿真,并与基于ANFIS的制导炸弹智能控制系统进行比较,证明该系统抗干扰能力强、控制精度高。

This paper presents an intelligent control system of guided bomb with anti-jamming capability. Based on the non-singleton fuzzy inference system, this system establishes its inference rules by using a new combined parallel learning algorithm of Gradient descent algorithm and genetic algorithm. In the course of building-up of the rule base through self-study, the system can automatically filter out the noise of training data, and acquire exact control information. A comparison is made through computer simulation between this system and the intelligent control system of guided bomb based on adaptive neuro-fuzzy inference system （ANFIS）, which shows this system has good anti-jamming capability and high control precision.