Optimization of a fuzzy logic controller for MR dampers using ANFIS

• Main Author: Braz-César, Manuel
• Publication Date: 2015
• Other Authors: Barros, Rui
• Language: eng
• Source: Repositórios Científicos de Acesso Aberto de Portugal (RCAAP)
• Download full: http://hdl.handle.net/10198/17547
• topic: Fuzzy control; Semi-active control; MR dampers; Structural dynamics; ANFIS

Fuzzy controllers have been successfully applied to a wide range of engineering problems due its robustness and the ability to deal with non-linear plants. Despite the inherent advantages of these controllers, there is no systematic technique for converting human knowledge into the rule base of a fuzzy inference system. Adaptive neuro-fuzzy inference system (ANFIS) is an artificial intelligence technique that has been successfully used for mapping input-output relationships based on available data sets, i.e., to automatically adjust a fuzzy inference system with a backpropagation algorithm based on training data. This paper presents the application of a ANFIS model to optimize the parameters of a fuzzy controller for structural control of a building structure using a MR damper. The results obtained with the neurofuzzy controller are compared with those of a passive control modes to assess the performance of the proposed control system in reducing the seismic response of the structure.