基于改进圆域分割方法的混沌微弱信号检测

针对Melnikov方法和Lyapunov指数法在确定Duffing振子混沌阈值时计算繁琐、耗时较长的问题,提出基于改进圆域分割的相图识别方法。以圆域内相点比例为指标,利用其在系统相变时急剧变化这一特性进行微弱信号的检测。改进圆域分割方法的圆域半径取值更加灵活,使混沌阈值的确定更加方便。仿真实验表明,信噪比为-60 d B时检测正确率达89%,验证了该方法的有效性。

Chaotic Weak Signal Detection Based on Improved Circle Zone Partition Method

To solve the complex computation and time consuming problem when determining the threshold value of chaosstate of Duffing oscillator using Melnikov method and Lyapunov exponent method, improved circle zone partition (ICZP)method was presented to identify the phase state. ICZP method uses phase point percentage within the circle area to deter.mine the phase state, which will change dramatically when the state changes from chaos to periodic state. This characteris.tic was used to perform weak signal detection. The radius of the circle zone may take different values, which makes it moreconvenient to determine the chaos threshold. The simulation showed that the detection accuracy could reach up to 89%even when the signal to noise ratio (SNR) got as low as -60dB, which proved the validity of the proposed method.