An Arduino-Based Wireless Sensor Network for Soil Moisture Monitoring Using Decagon EC-5 Sensors

It is undeniable that wireless communication technology has become a very important component of modern society. One aspect of modern society in which application of wireless communication technologies has tremendous potential is in agricultural production. This is especially true in the area of sensing and transmission of relevant farming information such as weather, crop development, water quantity and quality, among others, which would allow farmers to make more accurate and timely farming decisions. A good example would be the application of wireless communication technology to transmit soil moisture data in real time to help farmers make irrigation scheduling decisions. Although many systems are commercially available for soil moisture monitoring, there are still many important factors, such as cost, limiting widespread adoption of this technology among growers. Our objective in this study was, therefore, to develop and test an affordable wireless communication system for monitoring soil moisture using Decagon EC-5 sensors. The new system uses Arduino-compatible microcontrollers and communication systems to sample and transmit values from four Decagon EC-5 soil moisture sensors. Developing the system required conducting lab calibrations for the EC-5 sensors for the microcontroller operating in either 10-bit or 12-bit analog-to-digital converter (ADC) resolution. The system was successfully tested in the field and reliably collected and transmitted data from a wheat field for more than two months.

干旱荒漠区EC-5土壤水分传感器的校准和应用

应用EC-5土壤水分传感器在沿黄干旱荒漠绿洲地区的风沙土和粘土两种土壤类型中对不同土层深度的土壤含水率进行测定,同时以烘干称重法测定数据为标准值,采用回归分析方法建立EC-5在两种土壤中测定含水率的校准模型。结果表明,风沙土的校准模型为线性函数,且与埋设土壤深度有关,QS=0.71393S+0.01543D+1.71354(R2=0.920,P<0.05),D为土层深度(cm),S为传感器输出值(%);粘土的校准模型则为非线性的,lnQN=1.112S0.316(R2=0.949,P<0.05)。通过模拟回代结果显示,模拟值与标准值间均具有显著的相关性。在EC-5的应用验证阶段,校准值能够反映该地区风沙土和粘土土壤含水率的年变化特征,校准值与传感器输出值之间呈极显著相关关系(P<0.01),说明EC-5土壤水分传感器的校准模型稳定可靠,应用效果良好。

基于高精度传感器和卷积神经网络的智能灌溉系统研究

为了实现智能精准灌溉,以STM32为主控板,将基于卷积神经网络图像识别技术应用在灌溉系统中.结合EC-5土壤水分和温度传感器来检测土壤的湿度和实时温度,利用图像识别技术对土壤不同的干旱情况下植物叶片状态的识别.通过结合农作物叶片状态和土壤的湿度情况进行精准灌溉.在Keil5中完成各个功能模块的软件设计,通过主控板将采集到的数据与设定值做对比,实现对植物的精准灌溉.同时,用户可以在手机端上实时观察土壤湿度和植物的状态,也可以通过手机端进行远程操作.经测试,本系统可以实现对植物的精准灌溉.

冻融与静荷载双重作用下土体中水分场变化规律模型试验研究

在季节冻土区,反复冻融作用能够导致土体中水分迁移和重分布;对土体施加静荷载时,土体内部应力场发生变化,水分场也发生变化。通过对EC-5土壤水分传感器进行标定发现,相同密度条件下,含水量低于26.2%时,测试结果误差较小,高于26.2%时,则误差较大;不同密度时,密度越大,所测得的电压值越大。通过模型试验,研究冻融循环作用和冻融与静荷载双重作用下土体中水分场的变化规律,试验结果表明：两种情况下,土体内部水分含量都随冻融循环次数的增加而逐渐增大;无荷载时,在距表层0~20 cm范围内有含水量较高区域;施加静荷载时,在荷载下方和距离荷载最远的土体表层有三个水分集中区域,而在荷载两侧则有两个低含水量区域。