【作者】 史健芳； 汤洪彪； 杨盘洪； 朱庆双； 龚海燕；

【Author】 Shi Jianfang Tang Hongbiao Zhu Qingshuang Gong Haiyan (College of Information Engineering of Taiyuan University of Technology Taiyuan 030024,China)

【机构】 太原理工大学信息工程学院；

【摘要】 小波神经网络(WNN)是用小波基函数取代神经网络中的非线性激励函数而构成的一种新的神经网络,它把小波变换与神经网络有机地结合了起来。本文将小波神经网络用于机动飞行目标跟踪中,构建了适于机动飞行目标跟踪的多分辨率小波神经网络的结构和基于该结构的自适应目标跟踪模型及跟踪算法。该跟踪方法不需建立多个模型,运算量不随所跟踪机动目标运动方式的复杂性而增加。仿真结果表明,该算法较单纯的神经网络具有更好的跟踪性能和实际应用价值。更多还原

【Abstract】 Wavelet Neural Network (WNN) is such a network that combines wavelet transform with conventional neural network.It takes nonlinear wavelet bases as hidden nodes activation to replace nonlinear activation function in neural network.A new maneuvering target tracking method based on WNN is proposed. The adaptive structure of multi-resolution WNN fitting for maneuvering target tracking is constructed.The method needless to establish multi-models and the computing quantity doesn’t increase with the maneuvering state of target.Simulation results show that the algorithm can track maneuvering target exactly,meanwhile,it has better tracking performance than usual neural network.更多还原