• 검색 결과가 없습니다.

Active Control for Seismic Response Reduction Using Probabilistic Neural Network

N/A
N/A
Protected

Academic year: 2021

Share "Active Control for Seismic Response Reduction Using Probabilistic Neural Network"

Copied!
10
0
0

로드 중.... (전체 텍스트 보기)

전체 글

(1)

한국구조물진단학회 제11권 제1호(2007. 1)

103

Abstract

Recently structures become longer and higher because of the developments of new materials and construction techniques. However, such modern structures are susceptible to excessive structural vibrations, which may induce problems of serviceability and structural damages. In this paper we attempt to control structural vibration using the probabilistic neural network(PNN) and the artificial neural network(ANN) based on the training pattern that consist of only the structural state vector and the control force. The state vectors of the structure and control forces made by linear quadratic regulator(LQR) algorithm are used for training pattern of PNN and ANN. The proposed algorithm is applied for the vibration control of the three story shear building under Northridge earthquake. Control results by the proposed PNN and ANN are compared with each other.

요 지

구조 재료와 시공기술의 발달로 구조물은 높고 길게 설계할 수 있게 되었으나, 그에 따른 진동문제와 사용 성에 관한 문제가 발생하였고, 구조물의 과다한 변위는 구조물에 심각한 손상을 발생시켰다. 이러한 구조물 의 진동 문제를 해결하기 위하여 본 논문에서는 구조물의 상태벡터와 제어력만으로 구성된 훈련패턴을 기본 으로 하여 인공신경망이론과 확률신경망이론을 사용하여 구조물의 진동을 능동제어하는 방법을 제안하였다.

구조물의 제어를 위해 LQR 제어알고리즘을 이용하여 구조물의 상태벡터와 제어력을 구한 후, 상태벡터를 입력으로 제어력을 출력으로 하는 인공신경망과 확률신경망의 훈련패턴을 구성하였다. 제안된 방법을 사용하 여 Northridge 지진하중을 받는 3층 빌딩구조물을 제어하였고, 제안된 인공신경망과 확률신경망의 제어 결 과를 비교하였다.

Keywords : Probabilistic nerual network(PNN), Training pattern, Active control, Control gain 핵심 용어 : 확률신경망, 훈련유형, 능동제어, 제어이득

지진하중을 받는 구조물의 능동제어를 위한 확률신경망 이론

Active Control for Seismic Response Reduction Using Probabilistic Neural Network

김 두 기* 이 종 재** 장 성 규*** 최 인 정****

Kim, Doo-Kie Lee, Jong-Jae Chang, Seong-Kyu Choi, In-Jung

1)

* 군산대학교 토목환경공학부 조교수

** 한국과학기술원 스마트사회기반시설연구센터 연구조교수 *** 군산대학교 토목환경공학부 박사과정

**** 군산대학교 토목환경공학부 박사과정

2)

E-mail : [email protected] 063-469-4770

•본 논문에 대한 토의를 2007년 2월 28일까지 학회로 보내 주시면 2007년 5월호에 토론결과를 게재하겠습니다.

(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)

한국구조물진단학회 제11권 제1호(2007. 1)

111

0 1 2 3 4 5 6 7 8

-0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5

time(sec)

Velocity(m/sec)

uncontrolled nn control

(a) First floor

0 1 2 3 4 5 6 7 8

-0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5

time(sec)

Velocity(m/sec)

uncontrolled pnn control

(a) First floor

0 1 2 3 4 5 6 7 8

-0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5

time(sec)

Velocity(m/sec)

uncontrolled nn control

(b) Second floor

0 1 2 3 4 5 6 7 8

-0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5

time(sec)

Velocity(m/sec)

uncontrolled pnn control

(b) Second floor

0 1 2 3 4 5 6 7 8

-0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5

time(sec)

Velocity(m/sec)

uncontrolled nn control

(c) Third floor

0 1 2 3 4 5 6 7 8

-0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5

time(sec)

Velocity(m/sec)

uncontrolled pnn control

(c) Third floor Fig. 11 Velocity time history of structure subjected to

Northridge earthquake(0.344g_ANN)

Fig. 12 Velocity time history of structure subjected to Northridge earthquake(0.344g, PNN)

(10)

수치

Fig.  12  Velocity  time  history  of  structure  subjected  to  Northridge  earthquake(0.344g,  PNN)

참조

관련 문서

Purpose: This study aimed at developing an Artificial Neural Network (ANN) model for predicting the amount of cooling energy consumption of the variable refrigerant flow

In this study, temperature-based artificial neural network (ANN) model for daily reference crop ET estimation was developed and compared with reference crop evapotranpiration (  

The estimation of the strength is based on the probabilistic neural network which is an effective tool for pattern classification problem and gives a

In this study, we estimated missing evapotranspiration (ET) data at a eddy-covariance flux tower in the Cheongmicheon farmland site using the Artificial Neural Network (ANN).. The

Artificial Neural Network inference results are used as the fuel flow control inputs to prevent compressor surge and flame-out for turbojet engine for UAV.. High Gain Observer

A classification map of natural forest and artificial forest were generated using KOMPSAT-3, 3A, 5 data by employing artificial neural network (ANN). And in order to

The prediction of final mass and optimized process conditions of injection molded products using Artificial Neural Network (ANN) were demonstrated.. The input parameters,

In the current study, the design methodology has been developed to produce shoe mold for a suitable walking shoes of the general Korean using ANN (Artificial Neural Network)..