传感技术学报2012，Vol.25Issue(10)：1354-1360,7.DOI:10.3969/j.issn.1004-1699.2012.010.006
改进型BP神经网络对电容称重传感器的非线性校正
Nonlinear Calibration of Capacitance Weighing Sensor with Improved BP Neural Network Model
摘要
Abstract
Considering characteristics of the nonlineanty of the capacitance weighing sensor, i. e. the nonlinear relationship between the output voltage of the sensor and the loading, an improved BP neural network based on the Levenberg-Marquardt algorithm of Bayesian-Regularization was established to improve the nonlinear calibration capabilities. Simulation results show that the improved BP neural network achieved faster rate of convergence, higher accuracy and stronger generalization capability in comparison with the traditional gradient descent algorithm, which can effectively upgrade the nonlinear calibration of the capacitance weighing sensor.关键词
电容称重传感器/非线性校正/贝叶斯正则化/Levenberg-Marquardt算法/梯度下降算法Key words
capacitance weighing sensor/ nonlinear calibration/ Bayesian regularization/ Levenberg-Marquardt algorithm/ Gradient descent algorithm分类
信息技术与安全科学引用本文复制引用
郭伟,张栋,李巨韬,王磊..改进型BP神经网络对电容称重传感器的非线性校正[J].传感技术学报,2012,25(10):1354-1360,7.基金项目
青年科学基金项目(51005162) （51005162）
国家863计划项目(2011AA040601) （2011AA040601）
