植基於退火粒子群最佳化演算法之奇異質分解於二維濾波器之設計

2-d Filters Designed by Usingsingular-Value Decomposition Based on an Annealed Particleswarm Optimization

林灶生(Jzau-Sheng Lin) ； 張蓺英(Yi-Ying Chang) ； 黃世演(Shi-Yuang Huang)

《技術學刊》 24卷1期 (2009/03) Pp. 69-75

https://doi.org/10.29507/JT.200903.0007

摘要

本論文旨在提出以退火粒子群最佳化(Annealed particle swarm optimization，APSO)演算法求得不等距取樣點之空間頻率取代傳統等距取樣點空間頻率，以產生平面響應矩陣。粒子群最佳化演算法是一種具有群體智慧概念、屬於演化計算領域的一種計算方法，在決策過程中利用個體所擁有的經驗以及他人的經驗，經由考慮這二項資訊，來做最佳的決策。當不等距取樣點之空間頻率取得後，再以奇異值分解法(Singular value decomposition, SVD)求得多級可分離之一維濾波器；最後以這些一維濾波器組合成二維濾波器。由實驗數據可看出，以粒子群演算法求得不等取樣點之空間頻率，由奇異值分解法所設計出來之二維濾波器，比傳統等距取樣點空間頻率可得到較佳之增益響應。

關鍵字

二維濾波器 ； 奇異值分解 ； 粒子群最佳化

並列摘要

In this paper, an annealed particle swarm optimization (APSO) is proposed to randomly select non-equal scaling spatial frequency points in order to generate the spatial-response matrix for the singular-value decomposition (SVD) algorithm to a 2-D filter design. PSO is a swarm intelligent strategy which also is a new evolutionary computation technique. In the decision process, each particle adjusts its position to get a promising position in accordance with its own flying experience and its companion's flying experience. The SVD algorithm is used to get multiple separated 1-D filters after generating non-equal scaling spatial frequency points. Finally, we can combine these 1-D filters to construct a 2-D filter. From the experimental results, the non-equal scaling spatial frequency points randomly selected by the PSO for the SVD algorithm to design a 2-D filter can get a better frequency response than those gotten by the traditional method, which generates equal scaling spatial frequency points.

並列關鍵字

2-D filter design ； SVD ； PSO