预估节能型粮食水分测定仪

Grain is the treasure of a nation.Food is the first necessity of the people.The grain moisture content has tremendous influence on the agricultural product quality rating, therefore the measurement which is precise or not of the moisture content is extremely important to the producers and the consumers. The loss on drying is the standard method to detecting the grain moisture content for its high accuracy. The detection process needs 1-2 houses each time, so it time and energy consuming. This project has made significant improvements on the traditional loss on drying. By using electromagnetic force balance sensor and processing circuit to obtain the information of weight of samples, the aim of this project is to build up the drying kinetics model of samples and put forward an information fusion estimation method of loss on drying for rapidly moisture detection. At the early stage of drying, this method can accurately estimate the weight of the dry matter under completely drying condition of samples so that moisture content can be determined accurately and rapidly. Base on this information fusion estimation method, a new type of rapid moisture analyzer with DSP as information processing core has be made. This analyzer can perform rapid determination of moisture content during the process of loss on drying. It shortens 40% of moisture detection time of the traditional loss on drying, and saves more than 60% energy consumption. With function of network, this analyzer can be managed by PC. The project will present the fast testing theory of fast moisture detection, which can be applied into the fields of biometrics recognition and engineering and material quality testing, etc.

“民以食为天，粮乃国之宝”。粮食是人类赖以生存的基本物质。粮食质量安全与社会的和谐、政治的稳定、经济的持续发展息息相关。水分含量是影响粮食质量评定的关键因素，其测定精确与否对生产者和消费者都十分重要。烘干失重法是粮食水分测定标准方法，精确高、应用广，但测定过程耗电费时，测定效率较低。本项目对烘干失重法进行重大改进，利用电磁力平衡传感器与后续处理电路获取被测试样的质量信息，建立烘干失水过程中被测试样的质量变化模型和烘干失重法水分快速测定信息融合预估算法，在烘干初期准确估计出被测试样完全干燥失水时的质量；在此基础上研制一种以DSP为信息处理核心的预估节能型粮食水分快速测定仪，实现水分快速测定，测定时间缩短为传统烘干失重法的40%，节约能耗60％以上。本项目研究可形成物质水分含量快速检测理论，关键技术与方法可应用于生物特征识别及工程与材料质量检测等领域。

烘干失重法是粮油食品、医药化工等诸多行业采用的水分测定标准方法，应用广泛。这种方法依据试样初始质量与其完全干燥状态下的质量差计算水分百分比含量，耗电费时，一次测量需要1-2小时。本项目对烘干失重法进行重大改进，利用电阻应变片式传感器与后续处理电路获取被测试样的质量信息，建立烘干失水过程中被测试样的质量变化模型和烘干失重法水分快速测定信息融合预估算法，在被测试样的烘干初期准确估计出被测试样完全干燥失水时的质量，实现水分的准确、快速测定。.烘干失重法水分测定作为一种“无筛选”水分测定方法，其被测对象的广泛性是该方法的最大优势，也是建立预估融合方法的最大难点。本项目以干燥动力学、热力学理论为基础分析含水试样干燥特性曲线的阶段性特征，借鉴干燥动力学理论对物质干燥特性的分类方法，筛选典型试样进行全面的烘干失重法水分测定试验研究，将烘干失水曲线、失水速度曲线作为考察指标，分析试样品类、烘干温度、试样粒径、初始水分含量及初始质量对试样干燥特性的影响，确定降速干燥阶段的稳定性和升速干燥阶段的差异性是试样烘干失重过程可预估性的主要特征。.以不同品类粮食试样的干燥动力学实验研究为理论基础和支撑，建立了一种基于Levenberg-Marquardt（LM）算法的烘干失重法水分含量预估融合方法。针对传统LM算法阻尼因子定值更新导致的初值依赖性强、收敛性弱的问题，将信赖域搜索方法应用于阻尼因子的自适应更新，完成基于阻尼因子自适应LM算法的串行迭代结构预估算法设计，在试样未达到完全烘干的状态下估算试样的水分含量。.最终在理论研究的基础上研制一种以DSP TMS320C6748为信息处理核心的预估节能型烘干失重法水分快速测定仪，实现烘干失重法的快速测定，节省测定时间40~60%，有明显的节能效益，避免了完全干燥对试样的损坏。本项目研究的预估方法对于其它缓变量的快速、准确测量具有重要的参考与借鉴价值。