基于Android的车载式土壤电导率和光谱反射率检测系统(英文)

This paper aims to develop a vehicular integrated system collecting soil electrical conductivity(EC)and spectral reflectance.The system could collect soil EC and spectral reflectance automatically when the vehicle moves and save these measurement results with GPS.The information could reflect the characteristics of soil parameters such as soil salinity and water content.Hence,an Android-based vehicular detection system for soil conductivity and spectral reflectance information was developed.Soil electrical conductivity measurement was performed based on improved"current-voltage"four-terminal method.A STS-NIR spectrometer was used for collecting near-infrared spectral reflectance.The system collected the information of soil electrical conductivity and soil spectral reflectance while collecting GPS,which could be used for precision agriculture.The system was tested in a farm of Beijing on March 25,2014.Soil electrical conductivity of the farmland was measured and soil samples were collected.Water content and soil electrical conductivity were measured in the laboratory.The result of these experiments showed that the system could work stably in farmland and had good prospects.

车载式土壤光-电特性参数采集系统研究

土壤电导率是衡量土壤传导电流能力的一种固然属性,除了能够反映土壤质地外,还能够反映土壤的含水率、含盐量、有机物含量等特性。利用土壤光谱数据可以分析出土壤含水率、养分等,且测量过程中无需采样、搅动土壤。通过土壤电导率数据和光谱数据的综合、校正,能够提高系统的精度。基于嵌入式技术开发了土壤电导率和光谱反射率综合检测系统。土壤电导率测量系统采用基于改进的电流-电压四端法原理,使用深松犁的尖端作为电极传感器,能够在测量的同时完成松土的作用。光谱测量系统使用微型光谱传感器采集光谱数据,并进行实时处理。在采集土壤电导率、光谱数据的同时,系统同步采集GPS信息,并和土壤电导率、光谱数据一起保存,供进一步绘制土壤特性分布图使用。系统能够实时综合处理多种数据,并进行显示、保存等操作,具有良好的应用前景。

三阶非线性泛函微分方程的振动性

文章主要研究了三阶非线性泛函微分方程1/p(t)1r(t)x′(t)′′=Q(t)f(x)+R(t)h(x) when∫∞t0r(t)∫st0p(u)duds=∞分别在满足条件∫∞t0r(t)∫st0p(u)duds=∞和∫∞t0r(t)∫st0p(u)duds<∞时的振动准则,因此我们的结果不同于已有文献的一些结果.

二阶非线性中立型微分方程的振动准则

文章主要研究了如下方程的振动性(r(t)|(x(t)+p(t)x(τ(t)))′|α-1(x(t)+p(t)x(τ(t)))′)′+q0(t)|x(τ0(t))|α-1x(τ0(t))+∑ni=1qi(t)|x(τi(t))|βi-1x(τi(t))=e(t),t≥T.其结果推广和改进了已有结论.

基于深度自编码器的PRI调制方式识别的一种新方法

脉冲重复间隔(PRI)是雷达信号调制的重要参数,也是雷达信号分选的重要组成部分,对分选出的每个雷达脉冲序列的PRI调制特性进行分析,一方面有助于解决参数模糊问题,提高信号分选的可靠性;另一方面,PRI调制形式反映了雷达信号的某些特性,对PRI调制类型的正确识别将有助于推定雷达的用途与性能,实现辐射源识别。针对此问题,提出了一种基于数据降维和深度学习的PRI调制方式识别新方法。该方法将包含在信号中的PRI重频变化规律用一种滤波器思想提取出来,达到了凸显PRI变化规律的目的,与此同时,通过采样思想实现了对长序列的数据降维,然后采用深度自编码器(DAE)模型训练样本数据,最后采用Softmax分类器完成分类识别任务。通过仿真实验表明,对于不同信噪比下的信号,该方法识别准确率较高,且相对于传统识别方法识别速度较快。