應用數位信號處理晶片於模糊邏輯辨識系統之研究

Applied the Digital Signal Processing Chip on Fuzzy-Logic Recognition System

古德順(Te-Sheng Ku)

指導教授 ： 杜筑奎

中原大學/工學院/資訊工程研究所/碩士(2005年)

https://doi.org/10.6840/cycu200500332

摘要

摘　要

本研究首先對聲源訊號作特徵的擷取設計，再以模糊邏輯作識別處理，最後整合完成聲源辨識之應用。在接收聲源訊號過程中，受到外在噪音影響，必須對訊號作適確的處理，才能將遠距離傳送，而受到雜訊滲入之聲源，正確的辨識出來。處理過程可分為兩個步驟：(1)聲源訊號作小波轉換去雜訊後作特徵擷取;(2)利用模糊邏輯識別技術辨認聲源訊號。
模糊識別系統建立時，利用各樣本類別之特徵參數訂定模糊邏輯規則，以作為模糊邏輯推論之用，並以模糊識別原則獲得辨識結果，並根據辨識之結果與實際類別作誤差修正，調整歸屬函數，以增進辨識率的效果。
為達即時辨識目的，使用TI TMS320C6711 DSK為開發平台，利用其快速的數位信號處理速度，減少聲源訊號的訓練和辨識時間，建構快速且有效之辨識系統。

關鍵字

雜訊消除 ； 模糊邏輯 ； 聲源訊號 ； 小波轉換

並列摘要

Abstract

This research start with the feature extraction of the acoustic signals and then analyze the acoustic signal with Fuzzy-Logic. Finally, integrate them as a acoustic signal recognition system. In the process of various acoustic signals that interfered by the environment and all kinds of noises which permeate through the propagation channel, it is necessary to have an appropriate signal processing technology to recognize signals and eliminate the noises. The procedure is divided into two stages. : (1)using wavelet transform of the acoustic signals to reduce noise and extract features (2) Using Fuzzy-logic technology to recognize acoustic signals.

When constructing the fuzzy recognizing system, various classified pattern features have been considered to build the Fuzzy-Logic rule, for good performance, the effective fuzzy inference and membership function are established.

In order to reach the purpose of real-time recognition, the system is developed with TI TMS320C6711 DSK. The C6711 chip reveals high performance of digital signal processing, it can reduce the processing time for the acoustic signal recognition.

並列關鍵字

Fuzzy logic ； Acoustic signal ； de-noising

參考文獻

[1]B. William ,Underwater acoustic system analysis,2nd ,Prentice Hall, 1991.

[2]W.C. Knight, “Digital Signal Processing for Sonar”, Proc. of IEEE, pp. 1451-1506,Nov. 1981.

[8]I. Daubechies, Ten Lecture on Wavelets,SIAM Press,1992.

[11]R.M. Rao&A.S. Bopardikar,Wavelet transforms: introduction to theory and applications,Addison Wesley, 1998.

“Wavelet De-noising for Highly Noisy Source Separation,” IEEE International Symposium on Circutis and Systems, Vol.1,pp. I-201 -I-204, 26-29 May.2002.