摘要
针对目前清洗机器人倾角传感器测量精度低、控制算法复杂等缺陷,提出了一种新的倾角传感器信号处理方法,通过Savitzky-Golay平滑滤波技术对传感器采集数据进行滤波预处理,建立了以温度、滤波后准确的传感器数字信号为输入变量的BP神经网络模型,将倾角传感器的测量精度提高到0.91%,实现了对机器人倾角传感器信号的高精度处理。研究结果表明,该方法能有效提高倾角传感器的测量精度,可为清洗机器人准确完成清洗作业提供技术借鉴。
Aiming at the defects of low measurement accuracy and complicated control algorithm of the inclination sensor of cleaning robots,this paper presented a new method of signal processing for inclination sensor to resolve these problems,which used the Savitzky-Golay smoothing filtering technique to pretreat the data collected by inclination sensor.Then a BP neural network model was established,which used temperature and filtered digital signals as the input variables.This model raised the measurement accuracy of the inclination sensor to 0.91%,and achieved the high precision processing of the signal datathat collected by inclination sensor.The results showed that this method could improve the measurement accuracy of inclination sensor effectively,and provide cleaning robots with technical reference for completing the cleaning work accurately.
作者
谭爽
李丽宏
崔国强
TAN Shuang;LI Lihong;CUI Guoqiang(Information Engineering College,Taiyuan University of Technology,Taiyuan 030024,China;Shanxi Wanli Technology Limited Company,Taiyuan 030032,China)
出处
《传感技术学报》
CAS
CSCD
北大核心
2018年第7期1067-1071,共5页
Chinese Journal of Sensors and Actuators
