Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
7121790 | Measurement | 2018 | 17 Pages |
Abstract
In recent years, high-speed electric railway has become a major means of transport. Messenger wire is the main bearing part of electrified railway catenary, and once the damage occurs, it will easily cause a serious accident. In the harsh areas of climate, the messenger wire is covered with ice for a long time. To identify damages of electrical messenger wire in the icing environment, a synergetic damage recognition approach was put forward. On the basis of the synergetic principle, a dynamic decision-making mechanism was established. And the feature extraction and evaluation method for active acoustic emission signal was studied. Subsequently, a reconstruction method of damage status vector by using training sample fusion algorithm was introduced. Experiments were conducted in icing environment. Firstly, the impact of messenger wire galloping on identification performance was studied and found such influence can be ignored. In experiment of damage type recognition, the experimental results show that the proposed approach can identify different types of damages well. And compared with the cluster learning algorithm, the correct rate of recognition increased by 5% when training samples fusion method was adopted. This proposed approach demonstrates a strong anti-noise-interference ability, and it can provide a theoretical basis for the separation of stress wave modals in subsequent messenger wire structural damage monitoring imaging.
Keywords
Related Topics
Physical Sciences and Engineering
Engineering
Control and Systems Engineering
Authors
Xiaobin Hong, Jianxi Zhou, Guojian Huang, Lei Ni,