现代电子技术2016，Vol.39Issue(16)：6-9,4.DOI:10.16652/j.issn.1004-373x.2016.16.002
基于BP神经网络的超声波流量计的设计
Design of ultrasonic flowmeter based on BP neural network
摘要
Abstract
A ultrasonic flowmeter taking time⁃difference method for pipe flow measurement was designed based on the con⁃troller STM32. In order to improve the environmental suitability and flow measurement accuracy,the generalized cross⁃correla⁃tion time⁃delay estimation algorithm based on the BP neural network filtering is applied to time⁃difference detection. The algo⁃rithm filters the mixed noise by combining and optimizing the multiple filters with specific statistical characteristics,has self⁃learning and self⁃adjusting ability,and is able to adapt to the dynamic changing environment. The basic principles of the flow measurement,hardware structure and software design are mainly introduced in this paper. The strong adaptability and high ac⁃curacy of the flowmeter were proved through test in a variety of environment and the result analysis.关键词
STM32/神经网络/时差法/广义互相关算法Key words
STM32/neural network/time-difference method/generalized cross-correlation algorithm分类
信息技术与安全科学引用本文复制引用
高正中,张晓燕,龚群英,宋森森,张政..基于BP神经网络的超声波流量计的设计[J].现代电子技术,2016,39(16):6-9,4.基金项目
中国博士后科学基金（2015T80729）；研究生科技创新基金 （）
