Surrogate-model-based dynamic-sensing optimization method for high-speed elevator car horizontal vibration reduction

Abstract

The horizontal vibration of high-speed elevators has a significant impact on ride comfort and safety stability. Plenty of intelligent methods have been implemented in horizontal vibration reduction. The conventional optimization process for vibration reduction design involves conducting numerous complex dynamics simulations to establish the correlation between design parameters and vibration response. However, this method is inefficient and costly due to the demanding computational power requirements and time-consuming nature of the simulations. This study proposes a surrogate-model-based dynamic-sensing optimization method. A Kriging model is developed based on dynamics simulations, providing an efficient approach for obtaining the vibration response with specified design parameters. The DSPSO algorithm was proposed by incorporating perceptual particles and environmental adaptation mechanisms into the PSO algorithm. The DSPSO algorithm can be nested and interact with the Kriging model to optimize design parameters. During evolution, feasible and uncertain solutions are extracted into the sample space to enhance the accuracy of the Kriging model. The effectiveness of the proposed method is validated using both the ZDT series test functions and elevators from Canny Elevator Co., Ltd.