【作者】 李志浩；

【导师】 王艳霞；

【作者基本信息】 青岛科技大学 ， 检测技术与自动化装置， 2010， 硕士

【摘要】 传统超声波流量测量中,温度对声速的影响较大,管道内外径、管道压力等流体相关环境因素也会影响超声波的测量精度,为了保证超声波流量计的高准确度和高灵敏度测量,本文研究基于神经网络的超声波流量测量及温度补偿系统。根据被测流体的流场分布,通过分析流动性对测量的影响因素,确定利用超声波测量流量的时差法工作原理。针对温度变化引起声速变化的特点,通过对神经网络特性、结构、算法等内容的研究,确立利用神经网络技术来进行温度补偿,并建立超声波流量测量温度补偿模型。完成了具有温度补偿的超声波流量测量系统硬件设计,硬件电路设计在保证测量精度的前提下,对各个相关电路进行合理的设计。主要硬件电路包括超声波发射电路、接收电路、温度测量电路、系统控制电路、高频振荡和计数电路等。根据相应的系统硬件结构,进行了软件程序设计。并利用MATLAB仿真软件使用测量样本对网络进行训练,以此对超声流量测量系统进行温度补偿。利用神经网络技术进行温度补偿,达到了减小误差、提高测量准确度的目的。神经网络技术的应用极大提高了超声波流量测量精度,实现了超声波流量测量温度补偿智能化,为高精度超声波流量计的实际应用奠定了基础。更多还原

【Abstract】 In traditional ultrasonic flow measurement, temperature has greater infection on the speed of sound, furthermore, the accuracy of ultrasonic will be affected by the related environmental factors of fluid, which include internal diameters and external diameter of the pipeline and pipeline pressure etc. In order to ensure the high accuracy and high sensitivity measurements of ultrasonic flowmeter, this text researches the ultrasonic flow measurement and temperature compensation based on neural network.According to the flow field of measured fluid, through analyzing the impact of mobility on the measurement of factors, time difference works of ultrasonic flow measurement is determined to use.According to the characteristics of sound velocity variation caused by temperature variation, through the research of the characteristics, the structure, algorithms and other contents of neural network, determining to use neural network technology to carry out temperature compensation, and temperature compensation model of ultrasonic flow measurement is established.Ultrasonic flow measurement system hardware design with temperature compensation is completed. Related circuits are designed rationally by the premise of ensuring accuracy of hardware design. Major hardware circuits include ultrasonic transmitter circuit, receiver circuit, temperature measurement circuit, system control circuit, high frequency oscillation and counting circuit.The software programs are designed according to the corresponding hardware structure. Using MATLAB simulation software to train the network by using measurement samples, thus temperature compensation of the ultrasonic flow measurement system is established. By using neural network technology to temperature compensation, the purpose for reducing error and increasing measurement accuracy are attained. The precision of the ultrasonic flow measurement is improved greatly by the application of neural network technique, and the intelligent temperature compensation of the ultrasonic flow measurement is realized, which lays the foundation for the practical application of high precision ultrasonic flowmeter.更多还原