摘要
以非规则物体尺寸测量为研究目标,以超声波测距和线性回归理论为研究基础,进行了基于超声波的多元线性回归模型分析。系统硬件设计以STM32为主控平台,利用超声波模块采集5路测距值作为线性方程的自变量;再通过非规则物体周长的人工测量,得到方程的因变量,展开模型的多元线性回归分析。通过回归方程的建立,定量展示了基于机器学习的测量方案具备显著优势,为同类测量系统提供了一种成本的设计参考。
In this paper, we analyze the multiple linear regression model based on ultrasonic for irregular object size measurement. The hardware design of the system is based on the STM32 control platform, and the 'independent variable of linear equation is provided by using the ultrasonic module to collect the range value of 5 channels; then through artificial measurement of irregular objects of the perimeter, we get the equation dependent variables, and then the multiple linear regression analysis of the model is carried out. The establishment of regression equation offers the measurement scheme based on machine learning significant advantage, which provides the design reference for the similar measurement system.
出处
《电子科技》
2017年第7期58-61,共4页
Electronic Science and Technology
基金
新疆农业大学校前期基金(XJAU201516)
关键词
超声波测距
机器学习
多元线性回归
STM32处理器
ultrasonic distance measurement
machine learning
multiple linear regression
STM32 processor