【作者】 张明社；

【导师】 何东健； 李小昱；

【作者基本信息】 西北农林科技大学 ， 机械设计及理论， 2003， 硕士

【摘要】 泥沙测量是泥沙问题研究中的重要内容之一，而含沙量则是泥沙测量的必测参数。目前世界各国大多仍采用悬沙采样器取水样的方法通过“烘干称重法”来检测泥沙含量。这种方法既费时又费力，现已成为泥沙测量与研究中迫切需要解决的问题。本研究尝试着采用电容法测量水流泥沙含量，并用自行研制的电容式传感器进行试验，研究准确、快速、在线测量水流中泥沙含量的方法和技术。主要进行了如下研究： (1)通过试验，论文确定了电容式传感器的输出特性为线性输出，并对电容传感器从0到800kg／m~3的泥沙含量进行了标定。 (2)通常的传感器多存在着交叉灵敏度，本文研究对象电容式传感器也存在着这种情况。为了确定传感器在输出时的干扰因素，分别做了温度对传感器输出的影响、土壤种类对传感器输出的影响、土壤含盐量对传感器输出的影响等试验。通过试验确定出温度是影响传感器输出的主要影响因素。同时测出了温度从0到45℃变化对传感器输出的影响。 (3)为了消除温度对传感器输出的影响，使得测量精度和稳定性得到进一步的提高，运用曲面回归法对所测得的数据进行了二传感器信息融合处理。融合结果表明，融合后的数据比未融合前相比其准确度有所提高。 (4)用人工神经网络法对所测得的数据进行了处理，结果表明，该方法能够有效的消除温度对传感器输出所产生的影响，在相同的温度变化下，传感器的输出稳定性提高了96倍。通过对比两种融合方法，选择人工神经网络法运用到了测试系统中。 (5)在网络训练过程中，对神经网络结构进行了分析，建立了计算输出和理想输出关系非线性方程组，依据非线性方程理论阐述设计变量、样本数量和输出层单元数量的关系，确定了隐层神经元的数量，经比较该方法很有效。 (6)利用Visual Basic编制了一套泥沙测量软件，可用于在线和离线测量。该软件界面友好，功能完备能够完成数据的采集、存储、融合处理等功能。更多还原

【Abstract】 Measuring the sediment is one of the important questions in the sediment question research. In which the sediment concentration is a parameter that must be measured . Presently, the means of drying and weighing the sediment liquor are usually used to measure the sediment concentration in many countries. But this method is not only taking time but also strenuosity. This is becoming a question which must be solved urgently in sediment measuring and studying. This paper tries to use the capacitive method to measure sediment concentration in flow-water and use the self-regulating capacitance sensor to carry experimentation in order to study a way and technology used to measure the sediment concentration in a nicety - speediness and on-line way. The contents and results of this paper as follow:1. The paper ascertained the response characteristic of the capacitive senser is linear. The sediment concentration from 0 to 800kg/m3 measured by capacitive sensor is demarcated.2. Usually,there is cross responses in common sensors. The sensor in the paper has also with the character. In order to ensure the interferefering factor, the examination of the temperature effect, the earth kind, the saline earth are done in the paper. The conclusion that the temperature is the main effect factor is gotten from the examination. The effect of temperature from 0 to 45 C to the sensor is also measured in the paper.3. In order to eliminate the effect of the temperature and to advance the precision of the sensor, the method based on curl suface is used to fusion the information. The fusing result indicate that the precision of the sensor is improved.4. The method based on artificial neural network has also been used to fusion the information. The fusing result indicate that the precision of the sensor is improved 96 times. Comparing the two fusion ways, the latter is used in the measuring system.5. In course of training the network, the structure of the network has been analysed. Using the nonlinear theory, the node of the hidden layer is ascertained. By comparing this method is very efficient.6. Software for measuring the sediment was developed by using of Visual Basic. It has a friendly interface and can be used in on-line measuring.更多还原