摘要
为了提高雷达的抗干扰(ECCM)能力,采用混沌调频(FM)信号进行传输,但其回波经传统匹配滤波处理后,存在较高的随机分布的距离旁瓣,严重影响了雷达的性能。提出了一种以量子粒子群优化算法(QPSO)训练径向基函数(RBF)神经网络进行混沌FM信号旁瓣抑制的新方法。将RBF神经网络的参数组成一个多维向量,作为算法中的粒子进行进化,由此可在全局空间中搜索最优解。仿真结果表明,该方法计算简单,收敛速度快,并具有较好的数值稳定性,能很好的抑制距离旁瓣。
The chaotic frequency modulation （FM） signal is used to improve the electronic counter-countermeasure（ECCM） capabilities of radar. However, the sidelobe level of the signal after matching pro- cessing is very high, which greatly debases the radar＇ s performance. Based on the radial basis function （RBF） network, a novel range sidelobe processing technique was proposed, in which the quantum-behaved particle swarm optimization （QPSO） algorithm is applied to realize the optimization computing. A multidimensional vector composed of RBF network parameters is regarded as a particle to evolve, then, the feasible sampling space is searched for the global optima. The simulated results show that the algorithm has easier computation, more rapid convergence compared with traditional algorithms, and can also successfully suppress the sidelobe with good numerical stability.
出处
《兵工学报》
EI
CAS
CSCD
北大核心
2010年第2期177-183,共7页
Acta Armamentarii
基金
国防预研基金资助项目(9140A0120106BQ02)
关键词
雷达工程
混沌调频
神经网络
旁瓣抑制
radar engineering
chaotic frequency modulation
neural network
sidelobe suppression
