The invention provides an elevator car system fault early warning method, which comprises the following steps: firstly, carrying out offline model training, carrying out deviation standardization processing on collected vibration acceleration data during normal operation of an elevator car system, and carrying out training through a deep auto-encoder (DAE) model; then calculating an upper limit value K of the reconstruction error set E in the box plot as an early warning threshold value th; and then online fault early warning is conducted, a reconstruction error e is calculated through a trained DAE model according to collected vibration acceleration data of real-time operation of the elevator car system, and the health condition of the elevator is analyzed by comparing an early warning threshold value th with the obtained reconstruction error e. And finally, mapping the image to a two-dimensional space and a three-dimensional space by using a t-distributed random neighborhood embedding (t-SNE) algorithm for visual display. According to the invention, elevator management personnel can be helped to master the health condition of the elevator car system in real time, and early warning can be carried out on the car system with degraded performance.
本发明提出了一种电梯轿厢系统故障预警方法,首先进行线下模型训练,将采集到的电梯轿厢系统正常运行时的振动加速度数据进行离差标准化处理,并通过深度自编码器(Deep Auto‑Encoder,DAE)模型进行训练,然后计算重构误差集E在箱线图中的上限值K作为预警阈值eth;接着进行在线故障预警,将采集到的电梯轿厢系统实时运行的振动加速度数据通过训练好的DAE模型来计算重构误差e,将预警阈值eth与得到的重构误差e进行比较来分析电梯健康状况,最后利用t‑分布随机邻域嵌入(t‑SNE)算法将其映射到二维以及三维空间中进行可视化展示。本发明可以帮助电梯管理人员实时掌握电梯轿厢系统的健康状况,对性能退化的轿厢系统进行预警。
Fault early warning method for elevator car system
一种电梯轿厢系统故障预警方法
2022-08-30
Patent
Electronic Resource
Chinese
Elevator fault detection and early warning system and method
European Patent Office | 2023
|