Application of composite fuzzy-PID algorithm to suspension system of maglev train

Author(s):  
Junyou Yang ◽  
Rongbin Sun ◽  
Jiefan Cui ◽  
Xinping Ding
Author(s):  
James Waldie ◽  
Brian Surgenor ◽  
Behrad Dehghan

In previous work, the performance of PID plus an adaptive neural network compensator (ANNC) was compared with the performance of a novel fuzzy adaptive PID algorithm, as applied to position control of one axis of a pneumatic gantry robot. The fuzzy PID controller was found to be superior. In this paper, a simplified non-adaptive fuzzy algorithm was applied to the control of both axes of the robot. Individual step results are first shown to confirm the validity of the simplified fuzzy PID controller. The fuzzy controller is then applied to a sinuosoidal tracking problem with and without a fuzzy PD tracking algorithm. Initial results are considered to be very promising. Future work requires developing an adaptive version of the controller in order to demonstrate robustness relative to changing tracking frequencies and changing supply pressures.


2016 ◽  
Vol 33 (6) ◽  
pp. 1659-1667 ◽  
Author(s):  
Chun-Tang Chao ◽  
Ming-Tang Liu ◽  
Juing-Shian Chiou ◽  
Yi-Jung Huang ◽  
Chi-Jo Wang

Purpose – The purpose of this paper is to propose a novel design for determining the optimal hybrid fuzzy PID-controller of an active automobile suspension system, employing the gravitational search algorithm (GSA). Design/methodology/approach – The hybrid fuzzy PID-controller structure is an improvement to fuzzy PID-controller by incorporating a fast learning PID-controller. Findings – The GSA can adjust the parameters of the PID-controller to achieve the optimal performance. Research limitations/implications – The GSA may have the advantage of quick convergence, but the required computation may be intensive. Practical implications – The simulation results demonstrate the effectiveness of the proposed approach on active automobile suspension system. Originality/value – In order to demonstrate the theoretical guarantee of the proposed method, comparisons with particle swarm optimization or other methods has also been carried out.


Sign in / Sign up

Export Citation Format

Share Document