Numerical Evaluation of Applicability of Artificial Neural Networks (MLP) in Active Control of Structures

Numerical Evaluation of Applicability of Artificial Neural Networks (MLP) in Active Control of Structures

Vahid Shamsi1 Seyed Reza Elyasi2 Parviz Ahadi3

1) Graduate Student of Structural Engineering, Mohaghegh Ardabili University, Ardabil,Iran -
2) M.Sc Student of Earthquake Engineering, Mohaghegh Ardabili University, Ardabil,Iran -
3) Faculty Member of Islamic Azad University, Germi Branch,iran -

IOI Code : XYEA-BZADK
Publication : The 4th International Congress On Civil Engineering, Architecture & Urban Development(4icsau.ir)
Abstract :
    High-performance structures requires an environment free from vibration and excitation; therefore, such structure required new design approach. this design requirements has been created interest for using active control systems. active control system, controls the current situation of structure, decides According to the structural situation and take appropriate action to apply the structures. in this paper, the numerical assessment of active control of structures with multilayer perceptron artificial neural network (MLP) is studied. in this study active control conducted as e closed loop with regarding the structural responses. for this purpose, a structure considered with five degrees of freedom and network training by using Manjil, Tabas and Bam earthquake and eventually, efficiency of active control system have been compared with non-control and passive control systems. which represents the efficiency of active control system with artificial neural networks.

Keywords : mlp artificial neural networks active control optimized network structure equation of state time history analysis