Virtual sensing-based monitoring and maintenance in elevators

Abstract

This paper presents a virtual sensing approach for monitoring the elevator installations. The aim of the proposed virtual sensing approach is to monitor the condition of elevators without installing additional sensors. The adopted virtual sensing approach comprises a coupled-electromechanical elevator model presented in state space domain and a generalized output equation which includes not only the available measurements but also a set of virtual sensors. This virtual sensing approach employs an observer, and this observer estimates other physical quantities that cannot be directly measured with the already installed sensors. The estimated physical quantities are then employed for the calculation of four easy-to-interpret Key Performance Indicators (KPIs). These four indicators will enable to report the condition and triggers maintenance actions of elevator installations. Two of the indicators are related to the elevator ride quality and the energy efficiency which are currently employed by elevator manufacturers for evaluating their elevators. The rest two indicators are related to guide system diagnosis and electrical traction machine diagnosis. Finally, the proposed procedure is experimentally evaluated, by comparing the KPIs calculated based on the estimated virtual sensors and actual measurements, with the values obtained from the additional set of sensors installed in an elevator test tower.