【作者】 翟宝峰；

【导师】 许会；

【作者基本信息】 沈阳工业大学 ， 测试计量技术及仪器， 2002， 硕士

【摘要】 水分检测技术是一门涉及多学科的综合性技术。几十年来，通过科技人员的不懈努力，已取得了令人可喜的成果。但由于水分检测影响因素较多，现有的水分仪在处理数据时，普遍存在考虑因素单一，忽视一些次要因素影响等缺点，所以，在检测精度及重复性上一直不能取得令人满意的效果。本次毕业设计，针对以往水分检测技术的不足，引进了先进的数据处理方法，进行了一些探索性的尝试，为水分检测技术的数据处理提供了新的依据。本文的主要内容可归纳如下： 1．分析了现代各种水分检测技术的基本原理，指出了各自的优缺点。 2．进行了电容传感器结构设计优化，并对各种电容检测方法分析比较，找出了适合于粮食水分检测的方法。设计了具体的检测电路，对电路进行了仿真、制版、调试。 3．水分检测数据所用样本的制作，数据采样方案的设计及具体实施。 4．通过对影响水分检测各因素之间的数据分析，指出了水分与电压、温度、紧实度之间的关系，并指出了水分检测中的主要误差来源。 5．利用多维回归拟合算法及人工神经网络—BP网对多传感器数据进行融合处理，提出了具体的操作步骤及方法，对处理结果做了误差分析，指出了两个方法的优缺点。 6．分析了采用基于参数估计方法进行数据融合的基本原理，验证了此方法对于减小不确定因素影响提高检测精度的作用。更多还原

【Abstract】 The technology of moisture measurement is a synthetic technology relating to many subjects. During the past several decades, by the researcher’s hard work, some prominent successes have been achieved. Whereas, due to the influences of several other factors in moisture measurement, existing moisture instruments have some defects such as ignoring influences by subordinate factors. Hence, the measurement precision and the repeatability are not satisfied. In the thesis, more advanced method of data processing is introduced and exploring study is carried out. These provide new ideas for data processing in the moisture measurement technology. The major points of the thesis include:1.The principles of current moisture instruments are analyzed and their merits and defects are pointed out.2.The capacitive sensor structure is optimized. By comparing various capacitance measurement circuits, a measuring circuit is designed and simulated. The performance of the circuit is practically tested.3.Samples of grain for moisture measurement are made. Data acquisition procedures are planed and executed.4.The relationships of moisture and voltage, temperature and density are indicated by data analysis. The main error sources in the moisture measurement are discussed.5.The data from several type sensors are merged using data fitting method by multi-dimension regressions and artificial neural networks-HP networks method respectively. The particular operating processes are presented and the errors of each methods are analyzed in detail.6.The principle of data merging method based on parameter estimation is used. The validation is proved by increasing the measurement precision and reducing the uncertain factor.更多还原