基于CSI的无线层析成像探测仓储粮食异常粮情的研究

WiFi environment perception technology is usually based on the received signal strength (RSS) information, its performance is affected by multipath interference. Compared with RSS, channel state information (CSI) can describe wireless channel state more precisely and distinguish the multipath components, so as to realize more robust and reliable environment awareness. This study put forward the WiFi Radio tomographic imaging(RTI) method based on the CSI for storage grain abnormal monitoring, the proposed method have the characteristics as no insertion, low cost, the common commercialized equipment, no effect on any operation of the granaries and online grain situation monitoring with higher precision, which can provide a reliable and scientific method for grain situation monitoring, analysis, forecast and management decision. The research contents include: Based on the capacity of MIMO system based on OFDM, establishing a wireless tomography model of storage grain on WiFi; Analyzing and integrating CSI amplitude information and phase information from multi-antenna and multi-channel, and establishing wireless tomography monitoring system based on CSI; explaining of the spatial location and type of spatial location from the reconstructed images, and correcting them by the interference factors of the grain temperature and density difference. This project is expected to solve the above problems, and the corresponding research results can provide theoretical support for the application of RTI technique based WiFi in granular online and real-time detection.

WiFi环境感知技术通常基于接收信号强度(RSS)信息，其性能受限于多径效应等因素影响。相比于RSS，信道状态信息(CSI)能够更细粒度地描述无线信道状态，区分多径成分，从而实现更为鲁棒、可靠的环境感知。本研究提出基于CSI的WiFi无线层析成像方法对仓储粮食异常粮情进行监测，所提方法具有非插入、成本低、设备已普通商用、不影响粮仓任何操作，具有真正意义上的在线、高精度粮情监测等特点，从而为仓储粮情的监测、分析、预报以及管理决策提供一种可靠的科学手段。研究内容包括：以基于OFDM的MIMO系统容量为依据，构建仓储粮食WiFi无线层析测量模型；融合分析多天线、多通道的CSI幅度信息和相位信息，建立基于CSI的无线层析成像粮情监测系统；研究异常粮情区域空间位置和类型的图像解释以及温度与密度等干扰因素的修正。上述问题的解决将为WiFi无线层析成像技术在仓储粮食实时在线体检中的应用提供理论支撑。

WiFi环境感知技术通常基于接收信号强度(RSS)信息，其性能受限于多径效应等因素影响。相比于RSS，信道状态信息(CSI)更细粒度地描述无线信道状态，区分多径成分，从而实现更为鲁棒、可靠的环境感知。本项目提出基于CSI的WiFi无线层析成像方法对仓储粮食异常粮情进行监测，所提方法具有非插入、成本低、设备已普通商用、不影响粮仓任何操作，具有真正意义上的在线、高精度粮情监测等特点，从而为仓储粮情的监测、分析、预报以及管理决策提供一种可靠的科学手段。本项目建立了单个频率下的小麦含水量/密度预测模型；优化了基于RSSI的无线层析成像模型和目标定位成像模型，构建了高分辨率无线层析成像监测系统；提出了基于WiFi信道状态信息的小麦水分/霉变/虫害异常状态检测方法，并开展了基于RFID标签识别粮食水分/温度异常的部分研究工作。从而解决了射频粮情探测中粮食介质射频传播机理与粮食水分/密度关系未探明以及WiFi无线粮情探测中异常粮情反演的数学模型建模的问题，为WiFi无线层析成像技术在仓储粮食实时在线体检中的应用提供理论支撑。