论文部分内容阅读
基于复杂非线性系统相空间重构理论,提出了一种混沌背景中微弱信号检测的选择性支持向量机集成的方法,为了提高支持向量机集成的泛化能力,采用K均值聚类算法选择每簇中精度最高的子支持向量机进行集成,建立了混沌背景噪声的一步预测模型,从预测误差中检测湮没在混沌背景噪声中的微弱目标信号(包括周期信号和瞬态信号),最后分别以Lorenz系统和实测的IPIX雷达数据作为混沌背景噪声进行实验研究,结果表明该方法能够有效地将混沌背景噪声中极其微弱的信号检测出来,抑制噪声对混沌背景信号的影响,与神经网络和传统支持向量机方法相比,预测精度和检测门限方面的性能有显著提高.
Based on the theory of phase space reconstruction in complex nonlinear systems, a method of selective support vector machine (SVM) for weak signal detection in chaotic background is proposed. In order to improve the generalization ability of SVM, K-means clustering algorithm is used to select each Cluster with the highest degree of accuracy, a one-step prediction model of chaotic background noise is established, and the weak target signal (including periodic signal and transient signal) annihilated in chaotic background noise is detected from the prediction error. Finally, Lorenz system and measured IPIX radar data as chaotic background noise. The experimental results show that this method can effectively detect the extremely weak signals in chaotic background noise and suppress the influence of noise on the chaotic background signal. Compared with neural network and traditional support Compared with vector machine method, the performance of prediction accuracy and detection threshold has been significantly improved.