非常感谢审稿人的意见，本文采用的是离线监测。离线监测和在线区别是在线监

Thank you very much for the comments of the reviewers. This article uses offline monitoring. The difference between offline monitoring and online monitoring is that online monitoring refers to real-time monitoring of the status of the OLTC, while offline monitoring refers to sampling the OLTC to the laboratory for analysis. This paper sets up faults on OLTC through the laboratory, and then collects 50 groups of vibration signals of 4 different faults respectively, and then performs feature extraction and feature selection accuracy calculation. Online monitoring is the next step. Through real-time observation of OLTC vibration conditions, feature extraction is compared with typical signal features, and the results are given online. The entire diagnosis process is sampled at about 100ms to 120ms, including the entire contact switching process of the M-type OLTC.

Thank you very much for the comments of the reviewers, this article is using offline monitoring. Offline monitoring and online differentiation are online monitoring refers to real-time monitoring of the status of OLTC, and offline monitoring is the sampling of OLTC to laboratory analysis. In this paper, the fault is set up in the laboratory on OLTC, and then 50 sets of vibration signals for 4 different faults are collected, and then the feature extraction is carried out, and the accuracy of feature selection is calculated. Online monitoring is the next step, by real-time observation of the vibration of OLTC vibration to compare feature extraction with typical signal characteristics, and give online results. The entire diagnostic process is sampled from around 100ms to 120ms and includes the entire M-type OLTC contact switching process.

Thank you very much for the comments of the reviewers. This paper uses offline monitoring. The difference between off-line monitoring and on-line monitoring is that online monitoring refers to real-time monitoring of OLTC, while offline monitoring is to sample OLTC to the laboratory for analysis. In this paper, through the laboratory in the OLTC set fault, and then four kinds of different fault vibration signals were collected 50 groups, and then feature extraction, feature selection accuracy calculation. Online monitoring is the next step of the plan, through the real-time observation of OLTC vibration conditions to extract features, compared with typical signal features, and online results are given. The whole diagnosis process is sampled from 100ms to 120ms, including the whole contact switching process of the whole M-type OLTC.<br>