查詢結果分析
來源資料
相關文獻
- 一般化適應性M估測式應用於動態系統參數估測問題之探討
- 一般化適應性M估測式之新適應性法則及其於動態系統上之應用
- 適應性輸入估算法則於線上逆向熱傳問題之研究
- Stability of Linear Constrained Dynamical Systems with Modeling Uncertainties
- 混合式隔振系統在頻率域之隔振特性分析
- A Forward Iterative Dynamic Programming Technique for Optimal Control of Nonlinear Dynamic Systems
- A Haar Wavelet Aproach to the Analysis of Distributed-Parameter as Well as Lumped-Parameter Systems
- 動態系統模型建構與模擬--強化淨評估的工具
- 修正田口動態系統之參數設計與多重品質特性最佳化方法
- 兒童踢球動作型式之轉移:工作限制與年齡效應
頁籤選單縮合
題名 | 一般化適應性M估測式應用於動態系統參數估測問題之探討= |
---|---|
作者 | 劉正瑜; |
期刊 | 中正嶺學報 |
出版日期 | 19930100 |
卷期 | 21:2 1993.01[民82.01] |
頁次 | 頁33-40 |
分類號 | 319.5 |
語文 | chi |
關鍵詞 | 一般化適應性M估測式; 動態系統; 參數估測; 最大相似估測式; Dynamic system; sParameters estimation; Generalized adaptive M estimator; Maximum likelihood estimator; |
中文摘要 | 動態系統參數之估測,在應用科學中為一重要之課題。而日前常用之最 大相似估測式,對密度函數之靈敏度甚高。不利於廣泛之應用。一般化適應性M 估測式為一可隨量測值特性,調整其性能指標之強韌性估測式,可改進最大相似 估測式之缺點。但在應用於動態系統參數估測時,由於在調整性能指標過程中, 須使用常態化量測值,而有應用上之困難。本篇論文,將探討此一問題,並提出 以一般化適應性M估測式之剩餘值,代替常態化量測值的方法。由實例之說明, 此一方法確能解決應用之問題,並重現雜訊,此對日後雜訊之分析有很大助益。 而應用於動態系統之估測結果,亦遠較最大相似估測式為優。 |
英文摘要 | Estimation of dynamic system parameters is an important topic for applied sciences.The maximum likelihood estimator is the most popular estimation method recently. However, its higher sensitivity due to variation of density function would cause bad results inwide range of applications. The generalized adaptive M estimator, whose performance index is chosen by the normalized measurements, had been proved as a well-known robustestimator. The normalization of the measurements is the critical problem as being appliedin a dynamic system. In this paper, we concentrate our attention on this problem andpropose a method using the residuals induced from the generalized adaptive M estimatorto replace the normalized measurements. From the results of an example, we obtain theproof of this method and robustness of this estimator which is superior to the maximumlikelihood estimator. In addition to, the proposed method would be good for noise analysissince the generated residuals are look like noises. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。