• 한국공간구조학회논문집
• /
• 제17권4호
• /
• pp.123-132
• /
• 2017

A smart tuned mass damper (TMD) was developed to provide better control performance than a passive TMD for reduction of earthquake induced-responses. Because a passive TMD was developed decades ago, optimal design methods for structural parameters of a TMD, such as damping constant and stiffness, have been developed already. However, studies of optimal design method for structural parameters of a smart TMD were little performed to date. Therefore, parameter studies of structural properties of a smart TMD were conducted in this paper to develop optimal design method of a smart TMD under seismic excitation. A retractable-roof spatial structure was used as an example structure. Because dynamic characteristics of a retractable-roof spatial structure is changed based on opened or closed roof condition, control performance of smart TMD under off-tuning was investigated. Because mass ratio of TMD and smart TMD mainly affect control performance, variation of control performance due to mass ratio was investigated. Parameter studies of structural properties of a smart TMD was performed to find optimal damping constant and stiffness and it was compared with the results of optimal passive TMD design method. The design process developed in this study is expected to be used for preliminary design of a smart TMD for a retractable-roof spatial structure.

• 한국공간구조학회논문집
• /
• 제17권3호
• /
• pp.107-115
• /
• 2017

In this paper, a structural design method of a smart tuned mass damper (TMD) for a retractable-roof spatial structure under earthquake excitation was proposed. For this purpose, a retractable-roof spatial structure was simplified to a single degree of freedom (SDOF) model. Dynamic characteristics of a retractable-roof spatial structure is changed based on opened or closed roof condition. This condition was considered in the numerical simulation. A magnetorheological (MR) damper was used to compose a smart TMD and a displacement based ground-hook control algorithm was used to control the smart TMD. The control effectiveness of a smart TMD under harmonic and earthquake excitation were evaluated in comparison with a conventional passive TMD. The vibration control robustness of a smart TMD and a passive TMD were compared along with the variation of natural period of a simplified structure. Dynamic responses of a smart TMD and passive TMD under resonant harmonic excitation and earthquake load were compared by varying mass ratio of TMD to total mass of the simplified structure. The design procedure proposed in this study is expected to be used for preliminary design of a smart TMD for a retractable-roof spatial structure.

• 한국공간구조학회논문집
• /
• 제21권2호
• /
• pp.41-48
• /
• 2021

A smart tuned mass damper (TMD) is widely studied for seismic response reduction of various structures. Control algorithm is the most important factor for control performance of a smart TMD. This study used a Deep Deterministic Policy Gradient (DDPG) among reinforcement learning techniques to develop a control algorithm for a smart TMD. A magnetorheological (MR) damper was used to make the smart TMD. A single mass model with the smart TMD was employed to make a reinforcement learning environment. Time history analysis simulations of the example structure subject to artificial seismic load were performed in the reinforcement learning process. Critic of policy network and actor of value network for DDPG agent were constructed. The action of DDPG agent was selected as the command voltage sent to the MR damper. Reward for the DDPG action was calculated by using displacement and velocity responses of the main mass. Groundhook control algorithm was used as a comparative control algorithm. After 10,000 episode training of the DDPG agent model with proper hyper-parameters, the semi-active control algorithm for control of seismic responses of the example structure with the smart TMD was developed. The simulation results presented that the developed DDPG model can provide effective control algorithms for smart TMD for reduction of seismic responses.

• 한국전산구조공학회논문집
• /
• 제24권1호
• /
• pp.69-78
• /
• 2011

본 연구에서는 스마트 TMD를 효과적으로 제어할 수 있는 퍼지제어알고리즘을 개발하기 위하여 다목적 유전자알고리즘을 이용한 최적화기법을 제안하였다. 예제구조물로는 풍하중을 받는 76층 벤치마크건물을 선택하였다. 스마트 TMD를 구성하기 위하여 100kN 용량의 MR 감쇠기를 사용하였고, 스마트 TMD의 진동주기는 예제구조물의 1차모드 고유진동주기에 맞추어 조율되었다. MR 감쇠기의 감쇠력은 예제구조물의 풍응답을 최소화할 수 있도록 퍼지제어기를 통해서 조절된다. 퍼지제어기의 입력변수는 75층의 가속도 응답과 스마트 TMD의 변위응답으로 하였고, 출력변수는 MR 감쇠기로 전달되는 명령전압으로 하였다. 퍼지제어기의 최적화를 위하여 다목적 유전자알고리즘인 NSGA-II 기법이 사용되었고, 이때 75층의 가속도 응답과 스마트 TMD의 변위응답을 목적함수로 사용하였다. 최적화 결과, 구조물의 풍응답과 STMD의 변위응답을 동시에 적절히 제어할 수 있는 다수의 퍼지제어기를 얻을 수 있었다. 수치해석을 통해서 스마트 TMD의 성능이 수동 TMD에 비하여 월등히 뛰어남을 알 수 있었고 경우에 따라서는 샘플 능동 TMD보다 더 우수한 제어성능을 발휘하였다.

• 한국공간구조학회논문집
• /
• 제21권1호
• /
• pp.79-86
• /
• 2021

Control performance of a smart tuned mass damper (TMD) mainly depends on control algorithms. A lot of control strategies have been proposed for semi-active control devices. Recently, machine learning begins to be applied to development of vibration control algorithm. In this study, a reinforcement learning among machine learning techniques was employed to develop a semi-active control algorithm for a smart TMD. The smart TMD was composed of magnetorheological damper in this study. For this purpose, an 11-story building structure with a smart TMD was selected to construct a reinforcement learning environment. A time history analysis of the example structure subject to earthquake excitation was conducted in the reinforcement learning procedure. Deep Q-network (DQN) among various reinforcement learning algorithms was used to make a learning agent. The command voltage sent to the MR damper is determined by the action produced by the DQN. Parametric studies on hyper-parameters of DQN were performed by numerical simulations. After appropriate training iteration of the DQN model with proper hyper-parameters, the DQN model for control of seismic responses of the example structure with smart TMD was developed. The developed DQN model can effectively control smart TMD to reduce seismic responses of the example structure.

• 한국공간구조학회논문집
• /
• 제11권4호
• /
• pp.79-88
• /
• 2011

근래에 들어와서 3T (Twisted, Tapered, Tilted)로 대별되는 비정형 초고층 건축물이 다수 계획되고 있다. 이러한 비정형 초고층 건물을 위해서 구조적인 효율성 및 조형성 때문에 다이어그리드 구조시스템이 현재까지 가장 널리 사용되고 있는 구조시스템 중의 하나이다. 건축적인 조형미 등의 이유로 경사진 비정형 초고층 건물에 대한 계획안이 다수 발표되고 있으며 다수의 구조물들이 다이어그리드 구조시스템을 활용하고 있다. 경사진 비정형 초고층 건물은 횡하중뿐만 아니라 자중에 의해서도 횡방향 변위가 발생한다. 따라서 정형적인 초고층 건물보다 횡방향 응답을 저감시카는 젓이 더 중요한 문제로 대두된다. 본 연구에서는 경사진 다이어그리드 비정형 초고층 건물의 지진응답을 저감시키기 위하여 스마트 TMD를 적용하였고 그 제어성능을 평가하였다. 스마트 TMD를 구성하기 위하여 MR 감쇠기를 사용하였으며 스마트 TMD는 그라운드훅 제어알고리즘을 사용하여 제어하였다. 100 층의 예제구조물에 대하여 제어를 하지 않은 경우와, 일반적인 TMD를 사용한 경우, 그리고 스마트 TMD를 사용하여 제어한 경우를 비교 검토하였다. 수지해석결과 스마트 TMD가 변위 응답 제어에는 우수한 성능을 나타냈지만 가속도응답제어에는 효과적이지 못했다.

• 한국산학기술학회논문지
• /
• 제18권7호
• /
• pp.720-725
• /
• 2017

본 연구에서는 멀티 해저드를 고려한 빌딩 구조물의 안전성 및 사용성에 대한 평가를 수행하였고 지진 하중 및 풍하중에 대한 안전성과 사용성이 관련된 구조 성능을 개선하기 위하여 TMD 기반 적응형 스마트 구조 제어 시스템을 제안하였다. TMD 기반 적응형 스마트 구조 제어 시스템은 MR 감쇠기를 이용하여 구성하였다. 멀티 해저드 하중을 작성하기 위하여 미국의 대표 강진 지역 및 강풍 지역을 선택 하여 해당 지역의 특성을 고려한 인공 지진 하중 및 인공 풍 하중을 작성 하였다. 작성된 하중을 사용하여 20층 예제 구조물의 안전성 및 사용성을 검토하였다. 대상 예제 구조물의 안전성 및 사용성을 개선하기 위하여 스마트 TMD를 적용 하였고 성능 개선 정도를 평가하였다. 스마트 TMD는 MR 감쇠기를 이용하여 구성하였다. 수치 해석 결과 예제 구조물은 멀티 해저드에 대하여 안전성 및 사용성 측면에서 모두 설계 기준 값을 벗어났다. 스마트 TMD가 안전성과 연관되는 지진 응답과 사용성과 연관되는 풍 응답을 모두 효과적으로 저감시키는 것을 확인하였다.

• 한국공간구조학회논문집
• /
• 제16권4호
• /
• pp.91-100
• /
• 2016

A retractable-roof spatial structure is frequently used for a stadium and sports hall. A retractable-roof spatial structure allows natural lighting, ventilation, optimal conditions for grass growth with opened roof. It can also protects users against various weather conditions and give optimal circumstances for different activities. Dynamic characteristics of a retractable-roof spatial structure is changed based on opened or closed roof condition. A tuned mass damper (TMD) is widely used to reduce seismic responses of a structure. When a TMD is properly tuned, its control performance is excellent. Opened or closed roof condition causes dynamic characteristics variation of a retractable-roof spatial structure resulting in off-tuning. This dynamic characteristics variation was investigated. Control performance of a passive TMD and a smart TMD were evaluated under off-tuning condition.

• Smart Structures and Systems
• /
• 제33권6호
• /
• pp.415-430
• /
• 2024

To control vertical and lateral compound vibration simultaneously using an integrated smart controller, passive tuned mass damper (TMD) and tuned liquid damper (TLD) are updated and combined to an adaptive-passive TMD-TLD (AP-TMD-TLD) system. As for the vertical AP-TMD part on top of the vertical spring, it can retune itself through varying the level of liquid in the tank to adjust its mass, while the lateral AP-TLD part at the bottom of the vertical spring can retune itself by changing the level of liquid. Further, for multimodal response control, the multiple AP-TMD-TLD (MAP-TMD-TLD) system is proposed as well. Each AP-TMD-TLD in the system can identify the structural vertical and lateral modal frequencies through the wavelet-transform (WT) based algorithm and retune its vertical and lateral natural frequencies both through adjusting the level of liquid in the AP-TMD and AP-TLD parts respectively. A cantilever cable-stayed landscape bridge which is sensitive to both human-induced and wind-induced vibrations is presented as a case study. For comparison, initial parameters of MAP-TMD-TLD are mistuned. Results show that the presented system can retune its vertical and lateral frequencies precisely, while the retuned system has a better bi-directional compound control effect than the mistuned system before the retuning operation and can improve the serviceability significantly.

• 한국공간구조학회논문집
• /
• 제22권2호
• /
• pp.39-46
• /
• 2022

Recently, machine learning is widely used to solve optimization problems in various engineering fields. In this study, machine learning is applied to development of a control algorithm for a smart control device for reduction of seismic responses. For this purpose, Deep Q-network (DQN) out of reinforcement learning algorithms was employed to develop control algorithm. A single degree of freedom (SDOF) structure with a smart tuned mass damper (TMD) was used as an example structure. A smart TMD system was composed of MR (magnetorheological) damper instead of passive damper. Reward design of reinforcement learning mainly affects the control performance of the smart TMD. Various hyper-parameters were investigated to optimize the control performance of DQN-based control algorithm. Usually, decrease of the time step for numerical simulation is desirable to increase the accuracy of simulation results. However, the numerical simulation results presented that decrease of the time step for reward calculation might decrease the control performance of DQN-based control algorithm. Therefore, a proper time step for reward calculation should be selected in a DQN training process.