节点文献

边防哨所重要路口无人值守自动监控系统的研究

Research of Automatic Supervising System of Nobody on Duty Applying to Important Crossing of Frontier Sentry

【作者】 杨兴国

【导师】 郭勇;

【作者基本信息】 成都理工大学 , 测试计量技术及仪器, 2007, 硕士

【副题名】基于DSP的小型便携式地震动目标探测识别系统

【摘要】 本文主要针对三种典型战场地面运动目标(人员、轮式车、履带式车),设计了基于DSP的小型便携式地面运动目标探测识别系统,主要内容分为三部分:目标地震动信号特征提取及目标识别算法的DSP实现,目标识别系统的硬件电路设计和系统软件设计。硬件主要由三部分组成:信号调理电路对检测到的微弱地震动信号进行放大:基于DSP的信号分析与目标识别电路将放大后的模拟信号进行A/D转换,进一步将得出的数字信号分析、处理,实现目标识别。并对系统的小型化、低功耗、抗干扰技术进行了研究。软件上采用小波包分析对典型地面运动目标(人员、轮式车、履带式车)的地震动信号进行特征提取,采用BP神经网络分类识别算法进行目标识别。结合定点DSP算法的实现原理,完成了基于小波包分析的特征提取算法和BP神经网络分类识别算法的定点DSP编程。利用DMA结合McBSP来实现采集数据,保证了系统的实时性要求。最后,该系统体积小、重量轻,可以在100米之内接收到识别信号。结果表明,系统达到了预期的技术要求,并且应用广泛,适用于多种场合下的目标识别,可以起到预警作用。

【Abstract】 The seismic target identification micro-system based on DSP is mainly discussedin the thesis. It consists of three parts, which are feature extracting of seismic signaland target identification realizing in DSP hardware, designing system circuit,programming software of system.The hardware mainly consists of three parts: signal pre-amplifying andconditioning circuit is used to amplifying the micro-seismic signals which havedetected, signal processing and target identifying circuit based on DSP is used toconvert the analog signals which have been amplified to digital signals and go throughsignal analysis, processing to identify the moving target. Accordingly, we studyhardware miniaturization, lower-power, and anti-jamming technology.As to the software, It realized feature extraction from the seismic signals ofrepresentative moving targets using wavelet package analyses method and targetidentification using BP method. Based on the arithmetic theory of fix DSP, weaccomplish the feature extracting arithmetic using wavelet package analyses andtarget identification using BP method on DSP.In order to insure the real time demandof system, we use DMA and McBSP technology to realize real time collection andprocess of seismic signal.Finally, the micro-system is small and light, which can receive the identifyingresult in 100 meters. The result shows software and hardware designing satisfy theanticipative demand and can be used widely. It can be used in many kinds of situationto distinguish the targets, and have the ability of pre-warning.

【关键词】 DSP目标识别特征提取小波包
【Key words】 DSPtarget identificationfeature extractingwavelet package
  • 【分类号】TP277
  • 【被引频次】6
  • 【下载频次】299
节点文献中: