ACTIVE CONTROL SYSTEM FOR LOCAL NONLINEAR STRUCTURE USING MODAL MODEL AND ARTIFICIAL NEURAL NETWORK

• Main Author: MUHAMMAD ABDAT (NIM : 35005002); Tim Pembimbing : Dr. Ir. Herlien D. Setio; Prof. Ir. Adang, DJAMAL
• Format: Dissertations
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/15422
• Institution: Institut Teknologi Bandung

The problems that occur regarding the dynamic loading on civil structure systems have been a matter of great concern for a very long time. Conventional structure system can cause very large intertory drifts in flexible structures and high floor accelerations in stiff structures. These two factors cause difficulties in insuring the security of the building components and contents. <br /> <br /> Passive base isolation systems are one of the most successful and widely implemented technologies for seismic hazard mitigation. The fundamental concept is to isolate the structure from ground motion, especially in the frequency range where the building is most affected. The goal is to limit or avoid damage to the structure and its contents by reducing the value of interstory drifts and floor accelerations. <br /> <br /> The effectiveness of the passive base isolation system can be further improved by combining it with an active vibration control system. Active hybrid control system <br /> <br /> and hydraulic actuators.This "hybrid" base isolation system can actively adapt to seismic excitation to minimize the induced damage to the structure. With these hybrid systems, the benefits of basic isolation system capable of reducing the horizontal movement of the building, will be preserved, while the safety and integrity protected by the active control. <br /> <br /> The purpose of this study is to investigate and develop solutions in the formulation of a mathematical model of non-linear state space that was applied to the new active hybrid control systems using the proposed modal model and <br /> <br /> optimal control algorithms. Data input and output results of this study further used to study neural network control algorithms. <br /> <br /> The study result showed that the solution of mathematical models can be used in general on active control system for local nonlinear structures, and has shown the results of its application to the active hybrid control systems more effective and superior compared to conventional systems and passive or active control only in reducing the structural response to all the seismic excitation is considered in this research study.