The invention relates to the technical field of fault detection of speed reducers for mine hoists, in particular to a model training method, a speed reducer fault detection method and a device for a mine hoist. The method comprises the following steps: carrying out ITD decomposition on a fault vibration signal to obtain an inherent rotation component of the fault vibration signal; carrying out phase-space reconstruction on a target inherent rotation component with the maximum correlation degree with the fault vibration signals in the inherent rotation components, obtaining fault feature information of each type of fault vibration signals, and constructing a training set of the fault vibration signals; constructing an initial fault detection network model; and iteratively training the initial fault detection network model according to the training set, and obtaining a target fault detection network model for fault detection of the speed reducer for the mine hoist. According to the method, the relevancy between the training set of the fault vibration signals and the real fault is improved, so that the target fault detection network model obtained according to the training set can detect various faults more accurately and finely, and accurate and efficient fault detection of the speed reducer is realized.
本发明涉及矿井提升机用减速器故障检测技术领域,具体涉及一种模型训练方法、矿井提升机用减速器故障检测方法和装置。该方法包括:对故障振动信号进行ITD分解,获取故障振动信号的固有旋转分量;对固有旋转分量中与故障振动信号相关度最大的目标固有旋转分量进行相空间重构,获得每一类故障振动信号的故障特征信息,构建故障振动信号的训练集;构建初始的故障检测网络模型;根据训练集,迭代训练初始的故障检测网络模型,获取用于矿井提升机用减速器故障检测的目标故障检测网络模型。本发明提高了故障振动信号的训练集与真实故障之间的相关度,使根据该训练集获得的目标故障检测网络模型能够更准确、精细地检测各类故障,实现了准确高效的减速器故障检测。
Model training method, speed reducer fault detection method and device for mine hoist
模型训练方法、矿井提升机用减速器故障检测方法和装置
2021-11-09
Patent
Elektronische Ressource
Chinesisch
Accidental falling prevention speed reducer of mine hoist
Europäisches Patentamt | 2023
|Connecting device for height limiter and hoist speed reducer
Europäisches Patentamt | 2021
|