本文選題：盲均衡 + 正交小波變換��； 參考：《安徽理工大學(xué)》2014年碩士論文

【摘要】：目前,國內(nèi)外對(duì)果蠅優(yōu)化算法及其應(yīng)用的研究成果較少,也還未發(fā)現(xiàn)有關(guān)于小波變換理論、果蠅優(yōu)化算法和各類盲均衡算法相融合后應(yīng)用于水聲通信領(lǐng)域的研究報(bào)道。本文在深入地研究了果蠅優(yōu)化算法的基礎(chǔ)理論知識(shí)之后,將新穎的果蠅算法引入盲均衡技術(shù)之中來優(yōu)化均衡器的性能,并且在分析了果蠅優(yōu)化算法具有的優(yōu)勢(shì)和劣勢(shì)的基礎(chǔ)之上,結(jié)合當(dāng)前熱點(diǎn)新技術(shù)、新理論-模擬退火思想、小波變換理論和協(xié)同進(jìn)化策略,對(duì)果蠅優(yōu)化算法進(jìn)行了改進(jìn),將改進(jìn)后的新算法嘗試應(yīng)用于水聲通信盲均衡技術(shù)中以達(dá)到進(jìn)一步優(yōu)化算法的均衡性能來實(shí)現(xiàn)提高水聲信號(hào)傳輸效率的研究目的。本課題研究具體完成了以下工作： 1.果蠅優(yōu)化的小波盲均衡算法 傳統(tǒng)常數(shù)模盲均衡算法被廣泛應(yīng)用于水聲通信領(lǐng)域,它是通過利用隨機(jī)梯度下降的搜索方式來調(diào)整均衡器系數(shù)的,這種搜索方式不僅容易導(dǎo)致CMA陷入局部尋優(yōu),無法準(zhǔn)確找到全局極值點(diǎn)而且還對(duì)均衡器的代價(jià)函數(shù)有連續(xù)、可導(dǎo)的要求。果蠅優(yōu)化算法具有很強(qiáng)的全局搜索能力,將果蠅優(yōu)化算法引入CMA中提出了果蠅優(yōu)化小波盲均衡算法,可避免傳統(tǒng)CMA搜索方法存在的缺陷,正交小波變換可抑制信號(hào)之間自相關(guān)性。 2.果蠅優(yōu)化的小波自適應(yīng)軟約束常模盲均衡算法 均衡復(fù)雜的水聲信道時(shí)傳統(tǒng)常模盲均衡算法收斂速度很慢、穩(wěn)態(tài)誤差也比較大。自適應(yīng)軟約束常模盲均衡算法的均衡效果優(yōu)于CMA,處理水聲信號(hào)時(shí)收斂速度明顯快于CMA、穩(wěn)態(tài)誤差也相對(duì)較小,但SCS-CMA搜索最優(yōu)權(quán)向量的方式仍和常數(shù)盲均衡相同,都采用的隨機(jī)梯度下降法,常易陷入局部收斂。果蠅優(yōu)化小波自適應(yīng)軟約束常模盲均衡算法是在SCS-CMA中融入果蠅優(yōu)化算法和小波變換理論,運(yùn)用正交小波變換預(yù)處理均衡器的輸入信號(hào)來達(dá)到去除噪聲和降低輸入信號(hào)自相關(guān)性的作用,運(yùn)用果蠅優(yōu)化算法求解均衡器的代價(jià)函數(shù),并用SFOA迭代搜索所得的最優(yōu)權(quán)向量初始化均衡器,該算法的均衡效果明顯優(yōu)于CMA。 3.模擬退火-果蠅混合算法優(yōu)化小波廣義自適應(yīng)多模盲均衡算法 一般來說,我們?cè)诿ぞ饧夹g(shù)中運(yùn)用多模盲均衡算法來處理高階QAM信號(hào),而傳統(tǒng)多模盲均衡算法的均衡效果越來越不能滿足日益增長的實(shí)際應(yīng)用的需求。針對(duì)MMA和果蠅算法存在的缺點(diǎn),本文提出了模擬退火-果蠅混合算法優(yōu)化小波廣義自適應(yīng)多模盲均衡算法。這種新的算法結(jié)合模擬退火這一新技術(shù)與果蠅算法兩者的優(yōu)勢(shì),利用局部搜索能力強(qiáng)的模擬退火技術(shù)解決果蠅優(yōu)化算法搜索復(fù)雜的大規(guī)�？臻g時(shí)易陷入局部收斂的問題。模擬退火-果蠅混合優(yōu)化算法能夠精確快速地找到最優(yōu)權(quán)向量,加快算法的穩(wěn)定收斂速度,降低穩(wěn)態(tài)誤差。使用正交小波對(duì)均衡器的每路輸入信號(hào)進(jìn)行分解來除噪去信號(hào)的相關(guān)性,進(jìn)一步改善了廣義離散自適應(yīng)多模盲均衡器的性能,新算法更能有效地均衡高階QAM信號(hào)。 4.小波盲均衡多果蠅群協(xié)同優(yōu)化算法 (1)多果蠅群協(xié)同優(yōu)化算法 果蠅優(yōu)化算法的尋優(yōu)精度不高,當(dāng)尋優(yōu)復(fù)雜搜索區(qū)域時(shí),搜索性能較低,收斂速度較慢。針對(duì)果蠅優(yōu)化算法存在的不足,在SFOA中引入?yún)f(xié)同進(jìn)化思想,提出了多果蠅群協(xié)同優(yōu)化算法。新算法利用并行拓?fù)涞倪M(jìn)化結(jié)構(gòu)和正反反饋的信息共享方式來協(xié)同指導(dǎo)整個(gè)系統(tǒng)的進(jìn)化。搜索時(shí),將多個(gè)果蠅群作為獨(dú)立進(jìn)化的群體在同時(shí)進(jìn)行搜索中也相互跟蹤對(duì)方的全局最優(yōu)解。通過共享對(duì)各個(gè)果蠅群各自的尋優(yōu)結(jié)果進(jìn)行評(píng)價(jià)所得的群體當(dāng)前最優(yōu)解來指導(dǎo)各個(gè)種群在獨(dú)立進(jìn)化的同時(shí)協(xié)同進(jìn)化,直至獲得最優(yōu)解。 (2)多果蠅群協(xié)同優(yōu)化小波常模盲均衡算法 應(yīng)用多果蠅群協(xié)同優(yōu)化算法至盲均衡算法中,在CMA的基礎(chǔ)上融入多蠅協(xié)同的果蠅優(yōu)化算法尋找最優(yōu)權(quán)向量初始化均衡器,正交小波變換理論消噪、減小信號(hào)間存在的自相關(guān)性。該算法均衡信號(hào)的效果更好。 (3)多果蠅群協(xié)同優(yōu)化小波多模盲均衡算法 針對(duì)CMA均衡信號(hào)時(shí)相位模糊、誤差大、處理高階QAM信號(hào)均衡效果差等不足和果蠅優(yōu)化算法所存在的缺陷,分析了可有效糾正信號(hào)相位旋轉(zhuǎn)、適用于高階信號(hào)均衡的多模盲均衡算法的原理,將其與搜索能力強(qiáng)的多果蠅群協(xié)同優(yōu)化算法和抑制信號(hào)相關(guān)性強(qiáng)的小波變換相結(jié)合,提出了一種新算法-多果蠅群協(xié)同優(yōu)化小波多模盲均衡算法。
[Abstract]:At present , the research results of fruit fly optimization algorithm and its application are few , and it has not been found to be related to wavelet transform theory , fruit fly optimization algorithm and various blind equalization algorithms . After studying the basic theory knowledge of fruit fly optimization algorithm , the paper improves the algorithm of fruit fly optimization , and then attempts to apply the improved algorithm to underwater acoustic communication blind equalization technology to achieve the research purpose of improving the transmission efficiency of underwater acoustic signal .

1 . Small - wave blind equalization algorithm for fruit fly optimization

The traditional constant modulus blind equalization algorithm is widely used in the field of underwater acoustic communication .

2 . Optimal Blind Equalization Algorithm for Fruit - fly Optimization with Wavelet Adaptive Soft - constraint

The traditional common mode blind equalization algorithm has a slow convergence speed and a large steady state error when the complex underwater acoustic channel is balanced . The adaptive soft - constrained normal - mode blind equalization algorithm is better than CMA , and the convergence speed of the adaptive soft - constrained normal - mode blind equalization algorithm is obviously faster than CMA , and the steady - state error is relatively small , but the adaptive soft - constrained normal - mode blind equalization algorithm is used in the SCS - CMA to achieve the effect of removing noise and reducing the self - correlation of the input signal .

3 . Simulated annealing - fruit fly hybrid algorithm to optimize the wavelet generalized adaptive multi - mode blind equalization algorithm

In general , we use multi - mode blind equalization algorithm in blind equalization to deal with high - order QAM signals , and the traditional multi - mode blind equalization algorithm is more and more difficult to meet the need of increasing practical application .

4 . Cooperative optimization algorithm for small wave blind equalization multi - fruit fly group

( 1 ) Multi - fruit fly group cooperative optimization algorithm

In the search of complex search area , the search performance is low and the convergence speed is slow . In order to overcome the shortcomings of the optimization algorithm of fruit fly , a cooperative evolutionary algorithm is introduced in SFOA to guide the evolution of the whole system .

( 2 ) Multi - fruit fly group cooperative optimization wavelet normal - mode blind equalization algorithm

In this paper , a multi - fruit fly swarm optimization algorithm is applied to the blind equalization algorithm , and a multi - fly cooperative fruit fly optimization algorithm is integrated on the basis of CMA to find the optimal weight vector initialization equalizer , the quadrature wavelet transform theory denoising and the reduction of the self - correlation between the signals .

( 3 ) Multi - fruit fly group cooperative optimization wavelet multi - mode blind equalization algorithm

This paper analyzes the principle of multi - mode blind equalization algorithm which can effectively correct the phase rotation of the signal , applies to the high - order signal equalization , and combines the multi - fruit fly group cooperative optimization algorithm with the strong search capability and the wavelet transform with strong signal correlation , and proposes a new algorithm - multi - fruit fly group cooperative optimization wavelet multi - mode blind equalization algorithm .
【學(xué)位授予單位】：安徽理工大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2014
【分類號(hào)】：TN911.5

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 宋煒;劉強(qiáng);;基于模擬退火算法的過程挖掘研究[J];電子學(xué)報(bào);2009年S1期

2 郭業(yè)才;王麗華;;模糊神經(jīng)網(wǎng)絡(luò)控制的混合小波神經(jīng)網(wǎng)絡(luò)盲均衡算法[J];電子學(xué)報(bào);2011年04期

3 竇高奇;高俊;;適用于高階QAM系統(tǒng)的多模盲均衡新算法[J];電子與信息學(xué)報(bào);2008年02期

4 郭業(yè)才;何龍慶;韓迎鴿;趙俊渭;;盲均衡技術(shù)在水聲信道均衡中的應(yīng)用進(jìn)展[J];艦船科學(xué)技術(shù);2007年02期

5 姚祥光;周永權(quán);李詠梅;;二分微粒群協(xié)同進(jìn)化優(yōu)化算法[J];計(jì)算機(jī)工程與應(yīng)用;2009年35期

6 于同泉;韋周芳;程彬彬;;ARC-FSK中程多載波水聲通信新技術(shù)[J];計(jì)算機(jī)仿真;2006年11期

7 李金明;趙俊渭;陸晶;;支持向量機(jī)初始化的常數(shù)模盲均衡算法仿真[J];計(jì)算機(jī)仿真;2008年01期

8 譚勝蘭;;模擬退火遺傳算法在網(wǎng)絡(luò)負(fù)載均衡中應(yīng)用研究[J];計(jì)算機(jī)仿真;2011年12期

9 王欣;杜康;秦斌;徐海軍;;基于果蠅優(yōu)化算法的LSSVR干燥速率建模[J];控制工程;2012年04期

10 許肖梅;;水聲通信與水聲網(wǎng)絡(luò)的發(fā)展與應(yīng)用[J];聲學(xué)技術(shù);2009年06期

，

本文編號(hào)：1947946