應用多尺度熵分析法於 三維結構健康診斷系統之數值分析

Abstract

本研究以前期研究所提出之單軸向立面多尺度交叉取樣熵分析為基礎，進一步研發應用於三維結構之健康診斷系統。藉由SAP2000程式建立數值模型，將結構雙軸向微振訊號視作結構物的生理訊號，先以多尺度熵(Multi-scale Entropy, MSE)分析結構物破壞與否，再利用結構物每個樓層之速度訊號，進行雙軸向立面之多尺度交叉取樣熵分析法(Multi-scale Cross Entropy, MSCE)搭配立面破壞指數，進而識別出破壞樓層。在得知破壞樓層後，量測破壞樓層四個方位雙軸向訊號，對其進行平面多尺度交叉取樣熵分析，最後以平面破壞指數顯示出破壞方位。本研究藉由立面及平面漸進式分析，將兩分析結果以破壞指數呈現，可有效且快速地指出破壞位置，不同於其他健康診斷系統需倚靠大量的資料庫，本系統僅需於破壞發生前量測一組初始值，且不需要長時間的量測，達到省時又降低整體成本之效果。理論分析結果顯示，以結構物的微振訊號進行多尺度熵分析，可初步了解結構是否發生破壞，若結構有破壞出現，則可依序進行雙軸向立面及平面多尺度交叉取樣熵分析，最後有效的分辨出破壞樓層及該樓層破壞方位，搭配破壞指數的結果，提高此系統之診斷速度與準確率。本研究結構健康診斷系統的驗證，首先建立模擬資料庫，利用數值分析驗證所提出之診斷系統，藉由結構物速度訊號判斷結構物立平面之破壞位置，討論其結果並評估現地實驗施做可行性。

A Three-dimensional structural health monitoring (SHM) system based on multi-scale Entropy (MSE) and Multi-scale cross-Sample entropy (MSCE) is proposed in this paper. According to the early stage reseach, by measuring the ambient vibration signal from the structure, which is treated as the biological signal of the structure, the damage condition can be rapidly evaluated by the MSE analysis. The vertical damage location can then be detected by analyzing the biaxial signals of different floors under the same damage condition with the vertical MSCE analysis and display the result by the vertical damage index. Moreover, the research use the planar MSCE analysis by measuring four directions biaxial signals of the story roof and backplane to detect the damage directions of the damage floor and display the result by the planar damage index. With the progressive analysis of vertical and plane, we can diagnose with the damage floor and the damage directions correctly and fast. Unlike some existing SHM methods, no experimental database or numerical simulation is required. Simply a reference measurement of the current stage can initiate and launch the SHM system. Numerical simulation of a seven-story steel structure has verified that the damage floors and directions can be detected by the proposed SHM algorithm, and the results can be quantified by the damage index efficiently. As only the ambient vibration signal is required with a set of initial reference measured, the proposed SHM system can be implemented practically with low cost.