Elevator Abnormal State Detection Based on Vibration Analysis and IF Algorithm

Elevators play a crucial role in daily life, and their safety directly impacts the personal and property safety of users. To detect abnormal states of elevators and ensure people's personal safety, the acceleration signal of elevators is decomposed and Weiszfeld algorithm is used to estimate gravity acceleration. In addition, the study also introduces Kalman filtering to reduce error accumulation. To estimate the operating position of elevators, a method based on information fusion is studied and designed to construct a mapping relationship between elevator vibration energy and position, and to locate the height of elevator faults. Finally, an anomaly detection model combining vibration analysis and the Isolated Forest algorithm is developed. The results showed that the main distribution range of acceleration values in the horizontal direction was between 0.02 m2/s and -0.02 m2/s. The average estimation error and root mean square error of the research designed elevator position estimation method were 0.109 m and 0.113 m, respectively, which could solve the problem of accumulated position errors. The abnormal vibration energy and height corresponding to different operating conditions of elevators were different. The normal value ratios of the anomaly detection model under different sliding windows were 99.91% and 99.57%, respectively. The anomaly detection model designed for research has good performance and can provide technical support for the detection of elevator operation status.

@article{Wang2025,
  title     = {Elevator Abnormal State Detection Based on Vibration Analysis and IF Algorithm},
  journal   = {International Journal of Advanced Computer Science and Applications},
  volume    = {16},
  number    = {1},
  year      = {2025},
  publisher = {The Science and Information Organization},
  author    = {Zhaoxiu Wang},
  doi       = {10.14569/IJACSA.2025.0160177},
  url       = {https://doi.org/10.14569/IJACSA.2025.0160177}
}