摘要
针对目前测量高度方法的不足,以高性能AVR单片机Atmega16作为工作平台,选用新型气压传感器BMP180采集测量点的气压和温度,研究并实现了结合BP神经网络和Adaboost算法的高度测量系统。首先,对不同高度的样本集（压强、温度、海拔高度）进行线下训练,得到既稳定又精准的BP-Adaboost测量模型;然后,由新型气压传感器BMP180采集到的压强和温度作为输入,由线下训练得到的BP-Adaboost测量模型计算出该点的高度值,最后通过OLED显示屏将结果显示出来。多次实验表明,提出的方法能够较为准确地测量出不同高度值,相比传统的标准气压高度公式计算高度,具有更好的有效性和稳定性,为高度测量提供了一种实用可靠,且经济便携的解决方案。
In view of the shortcomings of the current measurement method, this paper uses a high performance AVR Atmega16 as the working platform, uses a new type of air pressure sensor BMP180 to collect the pressure and temperature of the measurement points, and has studied and realized a high measurement system combined with the BP neural network and the Adaboost algorithm. First, we train the sample sets of different height（pressure, temperature and altitude） under line training to get a stable and accurate BP-Adaboost measurement model. Then, the pressure and temperature collected by the new pressure sensor BMP180 are used as input, and the height of the point is calculated by the BP-Adaboost measurement model trained under the line. The result is displayed through the OLED display. Many experiments show that the proposed method can measure the different height values more accurately. Compared with the traditional standard pressure height formula, it has higher validity and stability. It provides a very practical, reliable and economical solution for height measurement.
作者
李伟
崔学林
高涛
匡昌武
Li Wei;Cui Xuelin;Gao Tao;Kuang Changwu(Hainan meteorological Survey Center,Haikou 570203,China)
出处
《电子测量技术》
2018年第19期28-33,共6页
Electronic Measurement Technology
基金
国家自然科学基金(41775011)
海南自然科学基金(2017CXTD014)
海南省气象局科研(HNQXQN201508)项目资助
