基于神经动力学的目标跟踪算法  

The algorithms of target tracking based on neural dynamic

作　　者：朱安民[1] 明仲[1] 

机构地区：[1]深圳大学计算机与软件学院,深圳518060

出　　处：《深圳大学学报（理工版）》2009年第3期283-288,共6页Journal of Shenzhen University(Science and Engineering)

基　　金：国家自然科学基金资助项目(60673122);深圳大学科研启动基金资助项目(200744)

摘　　要：通过对神经网络模型的研究,将神经动力学的Shunting模型应用到目标跟踪系统中.该算法具有模型简单、参数不敏感、网络运算不易饱和、算法速度快,且不需要预设工作环境等优点,适用于未知动态环境下的优化目标跟踪.Matlab仿真试验表明,该算法可解决动态环境下的目标跟踪.A neural network based neuron dynamics model was studied and applied to solve the target tracking problem in unknown dynamic environments. The shunting-model-based algorithm has advantages sensitive to algorithm parameters, quick, and suitable to various dynamic environments. The that it is simple, not effectiveness and effi- ciency of the proposed algorithm are demonstrated by simulation studies with several cases in Matlab.

关 键 词：神经动力学 目标跟踪 路径选择 最短路径 动态环境 

分 类 号：O232[理学—运筹学与控制论]