Intelligent vibration control for ultra-high-speed elevators: A type 2 variable universe fuzzy neural network method with input saturation

Abstract

When elevators accelerate to ultra-high speed, the horizontal vibration suppression is significantly challenged by the complex coupled excitation and uncertain time-varying parameters. Therefore, this paper proposes an intelligent control method with type 2 variable universe fuzzy neural network considering input saturation. Firstly, considering the multilevel relationship of the guide rails, guide shoes, frame and car, a horizontal vibration dynamic model is constructed. To improve the tolerance and adaptability of the control system to the uncertain parameters, the type 2 variable universe fuzzy control is used as the main controller. The discourse of universe is adjusted by the T-S fuzzy neural network, and the anti-saturation controller is designed to compensate for the over-saturation problem aggravated by the networks. The non-parametric inverse model of intelligent adjustable dampers represented by the magnetorheological damper is constructed to accurately calculate the ideal output damping force. The effectiveness of the proposed control method is verified by the elevator experimental data and the bidirectional gas-solid coupling aerodynamic excitation. The results show that the proposed intelligent control method can significantly improve the mean values, peak values and comfort indexes of the horizontal vibration acceleration and tilt angle acceleration of the elevator car. The research provides a new intelligent control scheme for high-speed/ultra-high-speed elevators and other similar vibration suppression systems.