PHÁT HIỆN HƯ HỎNG TRONG DẦM THÉP THÔNG QUA TẦN SỐ TỰ NHIÊN BẰNG MÔ HÌNH RỪNG NGẪU NHIÊN=DAMAGE DETECTION IN STEEL BEAMS THROUGH NATURAL FREQUENCY USING A RANDOM FOREST MODEL

0 Người đánh giá. Xếp hạng trung bình 0

Tác giả: Anh Dũng Đặng, Hải Đăng Lâm, Đức Thịnh Nguyễn, Trung Đức Trần, Văn Tuấn Vũ

Ngôn ngữ: vie

Ký hiệu phân loại:

Thông tin xuất bản: Journal of Science and Technique: Section on Special Construction Engineering, 2024

Mô tả vật lý: tr.150

Bộ sưu tập: Metadata

ID: 251756

Thêm vào giỏ

Gần đây, các thuật toán học máy (ML) đã được chứng minh là công cụ có hiệu quả cao để dự đoán hư hỏng kết cấu. Tuy nhiên, dữ liệu được sử dụng trong quan trắc tình trạng kết cấu thường trong các điều kiện vận hành bình thường hoặc những sai lệch nhỏ so với trạng thái ban đầu và rất ít dữ liệu thu thập được trong điều kiện nguy hiểm. Các dữ liệu này nếu có thì cũng sẽ liên quan đến các tham số như phạm vi hệ số cường độ ứng suất và tỉ lệ ứng suất, đây là các tham số rất khó đo lường trong kết cấu thực. Hạn chế này khiến việc tạo bộ dữ liệu thực tế để huấn luyện các mô hình ML nhằm phát hiện hư hỏng kết cấu trở nên khó khăn. Trong nghiên cứu này, mô hình Rừng ngẫu nhiên (RF) đã được phát triển để dự đoán vị trí, chiều rộng và độ sâu của vết cắt trên dầm thép dựa trên sự thay đổi tần số tự nhiên. Các tần số tự nhiên này trong các kịch bản thiệt hại khác nhau được xác định bằng phương pháp phần tử hữu hạn (FEM). Tần số tự nhiên của kết cấu ở trạng thái không bị hư hại sẽ được so sánh giữa hai phương pháp (phương pháp phần tử hữu hạn và phân rã miền tần số) để xác định tính chính xác của mô hình FEM. Sau khi huấn luyện, mô hình RF cho kết quả R-squared (trên tập kiểm tra) của vị trí là 0,996
0,876 với chiều rộng và 0,88 với chiều sâu vết nứt. Tham số đo độ chính xác MSE lần lượt là 0,0003 đối với vị trí, 0,0313 với chiều rộng và 0,042 với chiều sâu. Kết quả chỉ ra rằng việc kết hợp FEM và phân rã miền tần số (FDD) với mô hình RF là rất có tiềm năng trong việc giám sát sức khỏe công trình.Recently, machine learning (ML) algorithms have proven to be highly effective tools for predicting structural damage. However, the data used in structural health monitoring often consists primarily of normal operational conditions or slight deviations from the original state, with a scarcity of data representing potentially dangerous conditions. This limitation makes it challenging to create realistic datasets for training ML models to detect structural damage. If such data were available, it would likely involve parameters like the stress intensity factor range and stress ratio, which are difficult to measure in real-world structures. In this study, a random forest (RF) model was developed to predict the locations, widths, and depths of saw-cuts in steel beams based on variations in natural frequencies. These natural frequencies under various damage scenarios were determined using the Finite Element Method (FEM). To ensure accuracy, the natural frequencies in the undamaged state were compared between the FEM and Frequency Domain Decomposition (FDD). After training, the RF model showed an R-squared value of 0.996 for location, 0.876 for width, and 0.880 for depth. The mean squared error (MSE) was found to be 0.0003 for location, 0.0313 for width, and 0.0420 for depth. The results indicate that combining the FEM and FDD with the RF model holds significant potential for applications in structural health monitoring.

1. Đọc trực tuyến

1. Đọc trực tuyến

Tạo bộ sưu tập với mã QR