Design of Semi-Active Seismic Vibration Controllers Using Fuzzy Logic and Evolutionary Optimization

Design of Semi-Active Seismic Vibration Controllers Using Fuzzy Logic and Evolutionary Optimization

Monica Patrascu (University Politehnica of Bucharest, Romania)
Copyright: © 2018 |Pages: 26
DOI: 10.4018/978-1-5225-3531-7.ch002
OnDemand PDF Download:
List Price: $37.50
10% Discount:-$3.75


In the light of recent technological advances, semi-active structural damping systems for seismic vibration mitigation are considered part of the civil engineering design process. Various actuating devices have been integrated into structures along with specifically designed control strategies. Semi-active dampers are nonlinear switching systems that require enhanced controllers. In order to minimize instability risk, a cascaded fuzzy control system that integrates the switch behaviour is designed. The inner loop uses a PI (proportional-integral) controller that is tuned with evolutionary optimization. The case study uses an electrohydraulic damper and a three-story building. To anticipate robustness to real-world disturbances and equipment failure, an uncertainty effect analysis is included in which three control systems are compared. First, structural and actuator disturbances are considered. Then, the case of switch failure brings forth the high reliability of the fuzzy control system, in what concerns using semi-active dampers in the development of civil structures.
Chapter Preview


During the lifetime of a structure located in seismically active geographical areas, high levels of ground acceleration can be observed. In recent years, special attention has been paid to the development of processes and structural vibration damping mechanisms to assess and mitigate the response of buildings and bridges to the action of winds and earthquakes. Notable results have been achieved in what concerns the damping of structural vibrations, with a wide variety of mechanisms, usually part of one in four categories: passive, active, hybrid, and semi-active. By adding damping devices, the natural frequencies of a structure, its vibration modes, and its corresponding damping factors change during seismic motion. Some of these devices require the design of specialized control algorithms, supported by developments in the field of automation and control systems, incorporating both intelligent control strategies, and the necessary hardware and software support for their implementation.

This chapter discusses the design of control systems for semi-active seismic dampers, which are inherently nonlinear. These actuators have behaviours that cannot be linearized, approximated or reduced to linear models, so formal design of control systems, for those strategies that require the use of well identified models, is difficult. Although simulation validation is often successful, implementation requires further tuning or functioning modes that could not have been taken into account at an earlier design phase, as is the case in control engineering. With the development of computing power, the real world usage of more computationally expensive techniques has become practicable. Thus, controllers based on fuzzy systems or neural networks have started to enter the interdisciplinary field of control for civil engineering, while automated tuning or learning techniques have started to include meta-heuristic approaches like evolutionary computing or particle swarm optimization.

There are questions that arise, however, due to the highly uncertain nature of intelligent control systems design. Can these techniques be applied safely to systems as critical as seismic vibration mitigation? Does the control system retain robustness? Is the end result reliable? Or would it be better to consider combinations between formal and intelligent design, so that the advantages of both are used efficiently.

The main objective of the chapter is to analyse the robustness of fuzzy controllers for seismic vibration mitigation, when applied to semi-active dampers and tuned with genetic algorithms. Moreover, the discussion includes how conventional control systems (like the standard PID proportional-integral-derivative) can be enhanced for structural vibration and nonlinear actuators, with case-based robustness tests.

Ultimately, intelligent techniques can significantly improve the performance of conventional control systems, if used wisely. Implementation is already in progress for biological systems (such as drug regulation and heart rate control), and around the corner for seismic protection systems.

Complete Chapter List

Search this Book: