Application of hedge algebra-based fuzzy controller to active control of a structure against earthquake

Abstract

Active control problems of seismic-excited civil structures have attracted considerable attention in recent years. In this study, a design procedure of the hedge algebra-based fuzzy controller (HAFC) is investigated for a class of high-rise building systems with active tendons, which are subjected to seismic activity. In the HAFC, semantics of linguistic terms is obtained through semantically quantifying mappings (SQMs) without using any fuzzy set. Hence, the design of a HAFC leads to determining parameters of SQMs, which are fuzziness measures of primary terms and linguistic hedges occurring in a fuzzy model. To prove the validity of the HAFC in the aforementioned problem, an earthquake-excited 10-floor building with two actuators is numerically simulated.

Nội Suy Dựa Trên Trọng Số Khoảng Cách Ngữ Nghĩa Kết Hợp Tối Ưu Bằng Giải Thuật Di Truyền Cho Bộ Điều Khiển Đại Số Gia Tử

“…Đường cong sẽ không còn mô tả đầy đủ ngữ nghĩa của LRBS ban đầu, thậm chí có thể gây nên sự mâu thuẫn giữa các luật, làm cho kết quả suy luận thiếu đúng đắn. Nội suy trực tiếp trên mặt ngữ nghĩa [16], [17], [18] hoặc sử dụng phương pháp nội suy tuyến tính 2 chiều (bi-linear interpolation) [19], [20] Từ QRBS trong Bảng 1, ta nhận thấy rằng thành sự biến thiên của phần đầu ra là "đồng biến" theo các thành phần đầu vào và . Vùng các ô không có giá trị trong Bảng 1 là thiếu tri thức từ hệ luật ( hoặc ).…”

Data‐driven optimal predictive control of seismic induced vibrations in frame structures

“…Many researchers have tackled this problem using different (model‐free) approaches (e.g. fuzzy sliding mode controller, fuzzy logic control, and fuzzy logic and neural networks). In Khalatbarisoltani et al, the authors propose a method based on an online control of active structural systems taking into account both seismic uncertainties and time delay combining reinforcement learning algorithms (which include a fuzzy gain‐scheduling controller) with a state predictor in order to face the input time delay.…”

Online control of an active seismic system via reinforcement learning

“…Fuzzy logic control (FLC), due to its inherent robustness, has also been extensively employed for active/semiactive control of structural systems with uncertainties. [18][19][20][21] This controller is a good match for control of the structural systems where mass of the structure is continuously varying according to the variable number of inhabitants or various types of the furniture. Moreover, stiffness of the structure may vary due to the partial damages occur in the structural elements by the external loads.…”