| 论文标题 |
Simulation Research on Exploiting Electrostatic Potential Difference Sensing for Shaft Centerline Orbit Reconstruction |
| 作者 |
Kaihao Tang , Hongli Hu*, Lin Li, Yong Qin, and Xiaoxin Wang |
| 发表/完成日期 |
2019-05-29 |
| 期刊名称 |
IEEE Access |
| 期卷 |
Volume: 7, Issue:1;DOI:10.1109/ACCESS.2017 |
| 相关文章 |
08725598.pdf
|
| 论文简介 |
To realize non-contact measurement of rotational shaft centerline orbit, a shaft centerline
orbit reconstruction method using a specifically designed electrostatic sensor in conjunction with neural
network approximator is proposed in this paper. The sensing principle of the designed electrostatic sensor is
based on partial capacitance theory. The proposed sensor consists of one electrostatic field excitation
electrode, two sensing electrodes, and a cylindrical metal shield covering the three electrodes. It directly
establishes the map from sensor output (voltages between sensing electrodes and shield) to shaft center
position through neural network approximator, so that the shaft centerline orbit can be reconstructed. The
CAE software COMSOL Multiphysics is used for sensor modeling and simulation analysis. The simulation
reconstruction experiment is carried out to test the performance of the proposed sensor and to obtain
training data for the neural network. The experiment result shows that the RMS error of the proposed sensor
is within 0.4 μm, which has proved the feasibility of the novel sensor combined with neural network fitting
algorithm for industrial applications. |
(创新港)
