隨機搜尋法用於有限元素動態模型之調整

A Random Search Algorithm for Tuning Finite Element Dynamic Models with Measured Data

郭信川(H. C. Kuo) ； 官佳慶(C. C. Kuan)

《中國造船暨輪機工程學刊》 22卷4期 (2003/11) Pp. 213-219

https://doi.org/10.29982/JTSNAME.200311.0006

摘要

本文以一實驗模型所測得數據與有限元素法所建構之分析模型分析比較動態特性，調整分析模型的動態參數，以最佳化方法期其動態特性與實驗樑測數據盡可能的吻合。文中利用Borland C++ Builder 軟體，撰寫一介面程式，將隨機搜尋法、有限元素分析軟體ANSYS 與前處理及後處理所須之應用軟體，建構一結構分析與最佳化設計之整合系統。

關鍵字

單純形搜尋法 ； 隨機搜尋法 ； 有限元素動態模型 ； 整合系統

並列摘要

In this study, a random search algorithm based on the conventional simplex method is used to develop aprocedure for tuning an analytical finite element dynamic model with the experimental data measured from the testing model. In this procedure, several tuning strategies are developed to minimize deviation between measured data andthe tuned analytical modal frequencies and displacements at the measured points equal to their corresponding exciting
frequencies. The results demonstrate that it has a better performance to meet the tuned analytical finite element model
with measured data. Meanwhile, the integration system for the tuning finite element model, which is written in C++ language and connects structural analysis software ANSYS with the random search algorithm based an optimization program, is also developed by using software BC++Builder.

並列關鍵字

Nelder And Mead Simplex Algorithm ； Random Search Algorithm