英文翻譯伍杰杰.doc_第1頁
英文翻譯伍杰杰.doc_第2頁
英文翻譯伍杰杰.doc_第3頁
英文翻譯伍杰杰.doc_第4頁
英文翻譯伍杰杰.doc_第5頁
已閱讀5頁,還剩2頁未讀, 繼續(xù)免費閱讀

下載本文檔

版權(quán)說明:本文檔由用戶提供并上傳,收益歸屬內(nèi)容提供方,若內(nèi)容存在侵權(quán),請進(jìn)行舉報或認(rèn)領(lǐng)

文檔簡介

中南民族大學(xué)畢業(yè)論文(設(shè)計)英文翻譯材料學(xué)院: 計算機科學(xué)學(xué)院 專業(yè): 自動化 年級: 2007 學(xué)生姓名:伍杰杰 學(xué)號: 07064098 指導(dǎo)教師: 程立 職稱:講 師 2011年 05 月 05 日譯文:數(shù)字IIR濾波器設(shè)計使用蟻群算法算法文獻(xiàn)專業(yè)詞匯DSP數(shù)字信號處理; recursive 遞歸; non-recursive非遞歸; FIR有限脈沖響應(yīng)響應(yīng); IIR無限脈沖響應(yīng); multi-modal error surface多模態(tài)誤差表面; SA現(xiàn)代算法; GA遺傳算法; TS禁忌搜索; ACO蟻群算法; TACO蟻群優(yōu)化算法; optimization優(yōu)化; digital filter數(shù)字濾波器; adaptation process適應(yīng)過程; the parameter space參數(shù)空間; the global minima全局最小值; probabilistic transition rules概率轉(zhuǎn)換規(guī)則; a flexible memory可變內(nèi)存; the global optimum solution全局優(yōu)化方案; continuous optimization problems連續(xù)優(yōu)化問題; objective function目標(biāo)函數(shù); a string of binary bits二進(jìn)制位串; sub-path子路徑; the intelligent problem人工智能問題; peaks峰值; memory內(nèi)存 摘要: 改造和分析從模擬信號源采樣而來的信號,應(yīng)用信號處理(DSP)算法 。DSP的優(yōu)點是基于這樣的事實,即應(yīng)用算法的性能一直是可預(yù)見的。沒有依賴公差的電器元件作為模擬系統(tǒng)。DSP算法能合理地描述為一個數(shù)字濾波器。數(shù)字濾波器大體上分為兩類:有限和無限的脈沖響應(yīng)濾波器(檢索)過濾器。因為錯誤的表面過濾器通常是信息檢索的多式聯(lián)運方式的要求,為了避免局部最小值和設(shè)計有效的數(shù)字信息檢索過濾器。在這部作品中,提出了一種基于螞蟻算法和能力提出了一種用于全球最佳化的數(shù)字信息檢索濾波器的設(shè)計。仿真結(jié)果表明,本文提出的方法是準(zhǔn)確和具有收斂速度快,計算結(jié)果表明,提出的方法可以有效地用于數(shù)字信息檢索濾波器的設(shè)計。關(guān)鍵詞: 蟻群;禁忌搜索;持續(xù)最優(yōu)化改造和分析從模擬信號源采樣而來的信號,應(yīng)用信號處理(DSP)算法 。價格便宜,功能強大的通用計算機和定制設(shè)計的DSP芯片已經(jīng)開發(fā),DSP已經(jīng)在通訊,生物醫(yī)學(xué),氣象學(xué)和控制工程等幾個非常重要的領(lǐng)域得以廣泛應(yīng)用。例如,DSP已經(jīng)在儲存和復(fù)制的音頻和視頻信號的領(lǐng)域廣泛應(yīng)用。 DSP的優(yōu)勢在于該算法的性能應(yīng)用總是可以預(yù)測的,且不依賴于電子元件在模擬系統(tǒng)中的公差。任何DSP算法或處理器可以說是合理的數(shù)字濾波器。數(shù)字濾波器可大致分為兩類:遞歸和非遞歸濾波器。非遞歸的,或者有限脈沖響應(yīng)響應(yīng)濾波器(FIR),只取決于現(xiàn)在的輸入信號及其前任值。遞歸,或無限脈沖響應(yīng)過濾器(IIR),不僅依賴于輸入數(shù)據(jù),而且還取決于一個或多個先前的輸出值。IIR數(shù)字濾波器的主要優(yōu)點是與擁有相同數(shù)量系數(shù)的FIR濾波器相比,它可以提供更好的性能。然而,關(guān)于IIR濾波器設(shè)計還存在一些問題。最根本的問題是,它們可能有一個多模態(tài)錯誤表面。另一個問題是過濾器在適應(yīng)的過程可能變得不穩(wěn)定。雖然這第二個問題可以很容易地通過限制參數(shù)空間處理,以避免第一個問題,一個可以實現(xiàn)在多模態(tài)誤差全局最小的表面的設(shè)計方法是必需的。然而,傳統(tǒng)的設(shè)計梯度搜索的方法可以很容易停留在表面的局部極小的誤差。因此,一些研究人員試圖發(fā)展諸如模擬全局優(yōu)化設(shè)計算法的現(xiàn)代方法的算法(SA)和遺傳算法(GA)?,F(xiàn)代算法和遺傳算法采用概率轉(zhuǎn)換規(guī)則來搜尋錯誤表面的全球最低值。 遺傳算法是一個人口為基礎(chǔ)的算法(荷蘭,1975年)和演進(jìn)的解決方案,人口問題作為遺傳算法試圖改善單一的解決方案,使用一個鄰域搜索機制(柯克帕特里克等,1983)。雖然遺傳算法是很容易進(jìn)行編程和地方銜接 ,這取決于最初的解決方案,它可能需要花費太多功能評估,以收斂到全局極小。通常發(fā)現(xiàn)遺傳算法搜索到希望的地區(qū)十分迅速,但它往往需要太多的計算,從而達(dá)到局部極小概率轉(zhuǎn)換規(guī)則,因為工作和鄰域搜索機制不使用。這兩種算法的缺點均不希望出現(xiàn)在IIR數(shù)字濾波器設(shè)計。另外兩個具有全局優(yōu)化能力的禁忌搜索(TS)和蟻群優(yōu)化算法已被廣泛采用的組合型問題啟發(fā)式流行。格洛弗開發(fā)的TS算法是一種迭代搜索的形式,并根據(jù)智能解決問題的原則。TS有一個靈活的內(nèi)存來保存有關(guān)搜尋的資料,過去的步驟,并使用它來創(chuàng)建和利用在搜索空間的新的解決方案。一個可能與傳統(tǒng)的TS達(dá)到一個合理的計算時間時,最初的解決辦法是遠(yuǎn)離所在地區(qū)存在的最優(yōu)解全局最優(yōu)解的問題。蟻群算法(ACO)模擬了真實蟻群的行為。該算法的主要特點是分布式計算,正反饋和反對構(gòu)造性貪婪搜索。因此,蟻群算法的表現(xiàn)是由于本地搜索的積極反饋,由于分布計算全局搜索功能良好。在文獻(xiàn)中,目前只有一對蟻群的不斷優(yōu)化和工程應(yīng)用中提出的模型的幾部作品。提出的旅游蟻群優(yōu)化(南京塔塔汽車零部件)算法。在這項工作中,我們首先描述一個簡單的策略對記憶功能的TS算法的基礎(chǔ),以提高南京塔塔汽車零部件非凸連續(xù)優(yōu)化問題的一種表現(xiàn),其次提出了一種新方法,南京塔塔汽車零部件為基礎(chǔ)的數(shù)字IIR濾波器的設(shè)計。第2節(jié)描述了一個基本的蟻群算法,南京塔塔汽車零部件算法和提出的戰(zhàn)略。第3節(jié)的測試功能和由南京塔塔汽車零部件的模擬結(jié)果和修改后的南京塔塔汽車零部件算法獲得。第4節(jié)介紹了如何南京塔塔汽車零部件可應(yīng)用于數(shù)字IIR濾波器的設(shè)計。并比較了TS,南京塔塔汽車零部件和修改后的南京塔塔汽車零部件的設(shè)計IIR濾波器算法的性能。蟻群算法是優(yōu)化進(jìn)行自然真實蟻群的過程人為的版本。在這種情況下,一個解決方案的目標(biāo)函數(shù)值對應(yīng)于一個真實的螞蟻遵循的方式的長度。因此,由于信息素的金額存入一個自然的方式取決于它的長度,目標(biāo)函數(shù)值可以被用來確定問題的解空間的人工信息素量的方法。因此,一個簡單的原理圖算法建模的主要步驟的真實蟻群的行為,可以概括如下:開始初始化重復(fù)為所有人工螞蟻人工方式計算長度的所有人工方式更新信息素量就如何保持連接發(fā)現(xiàn)最近的路到現(xiàn)在直到結(jié)束。在南京塔塔汽車零部件算法描述Hiroyasu等。每個解決方案都代表了一個設(shè)計參數(shù)向量,其中每一個編碼的二進(jìn)制位串,即解決方案是一個二進(jìn)制位向量。因此,人工螞蟻為每個字符串中位值搜索,換句話說,他們試圖決定是否位值是0或1。在為一個位的值決策階段,螞蟻只使用信息素的信息。一旦螞蟻完成對字符串中所有值得決策過程,這意味著它已經(jīng)產(chǎn)生了解決問題的方法。該解決方案進(jìn)行評估的問題和一個數(shù)字值顯示解決方案的質(zhì)量是通過使用一個函數(shù)調(diào)用的評價函數(shù)來計算的。人工信息素是一種附著在子路徑形成的解決方案是使用此值計算。畢竟在殖民地螞蟻已經(jīng)產(chǎn)生了解決方案和屬于每個解決方案的信息素的金額已經(jīng)被計算出來,分位之間的路徑的信息素更新。這是進(jìn)行信息素通過降低以前存入的金額和新的路徑信息素量。假設(shè)被首選的子路徑介于0和1(0-1)在一個階段的概率計算。TS演算法是Glover為解決組合優(yōu)化的困難問題在1986年提出來的 。TS是在傳統(tǒng)計算機優(yōu)化上,通過避免搜索空間中已經(jīng)訪問了點來增強局部搜索。事實上,它模擬人工智能的問題來解決在使用過程中的問題。TS算法的主要特點是它有一個明確的記憶,內(nèi)存存儲一個關(guān)于搜索過去步驟的信息,新舉措會根據(jù)這個內(nèi)存在一定的區(qū)域內(nèi)產(chǎn)生。換句話說,搜索方向是受內(nèi)存控制的。通過這種記憶方法,最近的舉動在不在生產(chǎn)的這段時間內(nèi)復(fù)制,因此這個搜索可以擺脫局部最小值,并找到帶有多個峰的搜索空間中的一個。一個簡單的TS采用兩種存儲器:頻率和新近的回憶。頻率的內(nèi)存中存儲的是如何經(jīng)常的舉動是在一個時間間隔,而新近產(chǎn)生的內(nèi)存寄存器信息有關(guān)的時間(迭代)特定此舉已最后一次嘗試的信息。如果一個特定的移動頻率超過預(yù)定期限,那么這將被列為禁忌。一動近因值等于現(xiàn)在之間的迭代,在哪個此舉已嘗試最后一次迭代的差異。正如頻率一樣,如果一動近因值超過預(yù)定限額,那么這一舉動也列為禁忌。歸類為禁忌的舉動并沒有再次嘗試,直到他們得到的禁忌分類出來。使用這些記憶時,TS可以克服的單車周圍局部極小的問題,并找到發(fā)生在一個多維,簡稱AE,下同搜索空間的全球最低。為了避免過早收斂問題遇到塔科的頻率為基礎(chǔ)的記憶體為基礎(chǔ)的戰(zhàn)略已經(jīng)提出。在塔科的頻率為基礎(chǔ)的記憶存儲有關(guān)的頻率子路徑是由蟻后的信息。雖然子路徑的頻率似乎是與子路徑連接的信息素量一樣,但實際上并不是這樣的。通過檢查的信息素量,很難斷定是否大多數(shù)螞蟻遵循一個子路徑總價。但這是很容易通過評估頻率信息的方法來實現(xiàn)。 原文:Designing digital IIR filters using ant colonyoptimisation algorithmAbstract: In order to transform and analyse signals that have been sampled from analogue sources, digital signal processing (DSP) algorithms are employed. The advantages of DSP are based on the fact that the performance of the applied algorithm is always predictable. There is no dependence on the tolerances of electrical components as in analogue systems. DSP algorithms can be reasonably described as a digital filter. Digital filters can be broadly divided into two-sub classes: finite impulse-response filters and infinite impulse-response (IIR) filters. Because the error surface of IIR filters is generally multi-modal, global optimisation techniques are required in order to avoid local minima and design efficient digital IIR filters. In this work, a new method based on the ant colonyoptimisation algorithm with global optimisation ability is proposed for digital IIR filter design. Simulation results show that the proposed approach is accurate and has a fast convergence rate, and the results obtained demonstrate that the proposed method can be efficiently used for digital IIR filter design. Keywords: Ant colony; Tabu search; Continuous optimisation In order to transform and analyse signals that have been sampled from analogue sources, digital signal processing (DSP) algorithms are employed. After the cheap and powerful general-purpose computers and custom-designed DSP chips have been developed, DSP has found very significant applications in several engineering areas from communication, biomedical, and control to meteorology. For example, DSP has obtained wide application in the storage and reproduction of audio and video signals. The advantages of DSP are based on the fact that the performance of the applied algorithm is always predictable. There is no dependence on the tolerances of electrical components as in analogue systems. Any DSP algorithm or processor can be reasonably described as a digital filter. Digital filters can be broadly classified into two groups: recursive and non-recursive filters. The response of non-recursive, or finite impulse-response (FIR) filters is dependent only upon present and previous values of the input signal. Recursive, or infinite impulse-response (IIR) filters, however, depend not only upon the input data but also upon one or more previous output values. The main advantage of a digital IIR filter is that it can provide a much better performance than the FIR filters having the same number of coefficients. However, there are some problems with the design of IIR filters. The fundamental problem is that they might have a multi-modal error surface. A further problem is the possibility of the filter becoming unstable during the adaptation process. Although this second problem can be easily handled by limiting the parameter space, in order to avoid the first problem, a design method which can achieve the global minima in a multi-modal error surface is required. However, the conventional design methods based on gradient search can easily be stuck at local minima of error surface. Therefore, some researchers have attempted to develop the design methods based on modern global optimisation algorithms such as the simulated annealing (SA) algorithm and genetic algorithm (GA) . SA and GA employ probabilistic transition rules to search the global minima in a error surface. GA is a population based algorithm (Holland, 1975) and evolves a population of solutions to the problem as SA attempts to improve a single solution using a neighbourhood search mechanism (Kirkpatrick et al., 1983). Although SA algorithm is quite easy to be programmed and good at local convergence, depending on the initial solution it might often require too many cost function evaluations to converge to the global minima. GA usually discovers the promising regions of search space very quickly, however it often needs too many computations to reach a local minima since the probabilistic transition rules are employed and a neighbourhood search mechanism is not used. These disadvantages of both algorithms are not desired in the design of digital IIR filters. Other two popular heuristics which have global optimisation ability are tabu search (TS) and ant colony optimisation algorithms which have been widely employed for combinatorial type problems. TS algorithm developed by Glover is a form of iterative search and based on intelligent problem solving principles. TS has a flexible memory to keep the information about the past steps of the search and uses it to create and exploit the new solutions in the search space. A conventional TS might have problem with reaching the global optimum solution in a reasonable computation time when the initial solution is far away from the region where optimum solution exists. Ant colony optimisation (ACO) algorithm simulates the behaviour of real ant colonies. The main features of the algorithm are distributed computation, positive feedback and con structive greedy search. Therefore, the performance of ACO algorithm is good for local search due to the positive feedback and for global search because of the distribution computation features. In the literature, there are just a few works on the models of ACO proposed for continuous optimization and their engineering applications .Hiroyasu have presented the touring ant colony optimisation (TACO) algorithm. In this work, we firstly describe a simple strategy based on the memory feature of TS algorithm to improve the performance of TACO algorithm for non-convex continuous optimization problems and secondly propose a new method based on TACO for digital IIR filter design. Section 2 describes a basic ACO algorithm, TACO algorithm and the proposed strategy. Section 3 presents the test functions and the simulation results obtained by TACO and the modified TACO algorithms. Section 4 describes how TACO can be applied to digital IIR filter design. It also compares the performance of TS, TACO and the modified TACO algorithms on IIR filter design. ACO algorithm is the artificial version of the natural optimisation process carried out by real ant colonies. If an optimisation problem can be expressed in the form of a minimisation problem a possible solution to this problem can be considered as a possible way between the nest and food in real ants world. In this case, the value of objective function for a solution corresponds to the length of the way followed by a real ant. Therefore, since the pheromone amount deposited on a natural way depends on its length, the objective function value can be used to determine the pheromone amount of artificial ways in the solution space of the problem. Hence, the main steps of a simple schematic algorithm modeling the behaviour of real ant colonies can be summarised as below: BEGIN Initialise REPEAT Generate artificial ways for all artificial ants Compute the length of all artificial ways Update the amount of pheromone attached on the ways Keep the shortest way found up to now UNTIL (maxiteration or a criteria is satisfied) END. In TACO algorithm described by Hiroyasu et al. (2000), each solution is represented by a vector of design parameters of which each is coded with a string of binary bits, i.e. a solution is a vector of binary bits. Therefore, artificial ants search for the value of each bit in the string, in other words, they try to decide whether the value of a bit is 0 or 1. At the decision stage for the value of a bit, ants use only the pheromone information. Once an ant completes the decision process for the values of all bits in the string, it means that it has produced a solution to the problem. This solution is evaluated for the problem and a numeric value showing the quality of the solution is calculated by using a function called the evaluation function. An artificial pheromone to be attached to the sub-paths forming the solution is computed using this value. After all ants in the colony have produced their solutions and the pheromone amount belonging to each solution has been calculated, the pheromones of sub- paths between the bits are updated. This is carried out by lowering the previous pheromone amounts and depositing the new pheromone amount on the paths. Assume that the probability of being preferred of the sub-path between 0 and 1 (0-1) at a stage is calculated. TS algorithm has been proposed by Glover in 1986 to solve difficult combinatorial optimisation problems. TS is a general heuristic for optimisation in conventional computers that enhance local search by attempting to avoid already visited points in the search space. In fact, it simulates the intelligent problem solving process used by human being. The main feature of TS algorithm is that it has an explicit memory. The memory stores an information about the past steps of search and new moves are produced in a certain neighbourhood according to this memory. In other words, the direction of search is controlled by the memory. By means of this memory, the moves produced recently are not reproduced within a period of time and hence the search can get out of a local minimum and find the global one of the search space with several peaks. A simple TS employs two kinds of memory: frequency and recency- based memories. The frequency-based memory stores an information about how often a move was produced during a time interval while the recency-based memory registers an information regarding the time (iteration) a specific move has been last time

溫馨提示

  • 1. 本站所有資源如無特殊說明,都需要本地電腦安裝OFFICE2007和PDF閱讀器。圖紙軟件為CAD,CAXA,PROE,UG,SolidWorks等.壓縮文件請下載最新的WinRAR軟件解壓。
  • 2. 本站的文檔不包含任何第三方提供的附件圖紙等,如果需要附件,請聯(lián)系上傳者。文件的所有權(quán)益歸上傳用戶所有。
  • 3. 本站RAR壓縮包中若帶圖紙,網(wǎng)頁內(nèi)容里面會有圖紙預(yù)覽,若沒有圖紙預(yù)覽就沒有圖紙。
  • 4. 未經(jīng)權(quán)益所有人同意不得將文件中的內(nèi)容挪作商業(yè)或盈利用途。
  • 5. 人人文庫網(wǎng)僅提供信息存儲空間,僅對用戶上傳內(nèi)容的表現(xiàn)方式做保護處理,對用戶上傳分享的文檔內(nèi)容本身不做任何修改或編輯,并不能對任何下載內(nèi)容負(fù)責(zé)。
  • 6. 下載文件中如有侵權(quán)或不適當(dāng)內(nèi)容,請與我們聯(lián)系,我們立即糾正。
  • 7. 本站不保證下載資源的準(zhǔn)確性、安全性和完整性, 同時也不承擔(dān)用戶因使用這些下載資源對自己和他人造成任何形式的傷害或損失。

最新文檔

評論

0/150

提交評論