摘要
硅压阻式压力传感器因对温度具有敏感性,工作时受环境温度的影响会产生温度漂移现象,降低了测量精度,为提升压力传感器的检测精度,提出了一种基于深度信念网络(Deep Belief Network,DBN)的高精度温度补偿模型。研究了压阻式压力传感器的工作原理和温度补偿的数学模型,利用深度学习强大的数据表征能力,设计了区间定位的温度补偿模型构建算法,建立并优化DBN模型的网络结构,将DBN温度补偿模型对实验数据进行训练拟合,结果表明:温度补偿后的满量程相对误差由原来的7.013×10-3提升至8.240×10-5,验证了所提出的方法能具有较好的稳定性和温度补偿效果,较大幅度地提升了传感器的检测精度。
The silicon piezoresistive pressure sensor is sensitive to temperature,and the temperature drift phenomenon will occur when it works under the influence of ambient temperature,which reduces the measurement accuracy. In order to improve the detection accuracy of the pressure sensor,a high-precision temperature compensation model based on DBN( Deep Belief Network) was proposed. The working principle and the mathematical model of temperature compensation of the piezoresistive pressure sensor were studied,using the powerful data representation capabilities of deep learning,a temperature compensation model construction algorithm for interval positioning was designed,the network structure of the DBN model was established and optimized,and the DBN temperature compensation model was trained and fitted to the experimental data,the results show that: after temperature compensation,the full-scale relative error is increased from 7.013×10-3 to 8.240×10-5. It is verified that the proposed method has good stability and temperature compensation effect,and greatly improves the detection accuracy of the sensor.
作者
石文兵
葛斌
苏树智
SHI Wenbing;GE Bin;SU Shuzhi(School of Computer Science and Engineering,Anhui University of Science and Technology,Huainan Anhui 232001,China)
出处
《传感技术学报》
CAS
CSCD
北大核心
2021年第7期995-1000,共6页
Chinese Journal of Sensors and Actuators
基金
国家自然科学基金项目(51874003,61806006)
中国博士后科学基金项目(2019M660149)
安徽省自然科学基金项目(1808085MG221)
安徽省高等学校自然科学研究项目(KJ2018A0083)
安徽省质量工程基金项目(2020jyxm0454)。
关键词
温度补偿模型
深度信念网络
压阻式压力传感器
温度漂移
temperature compensation model
deep belief network
piezoresistive pressure sensor
temperature drift
