【正文】 ng coef?cients variations around the average value) and the weighting vector lag (difference between the average and the optimal value), [3]. It can be expressed as: ? ?? ? ? ?? ?*kkkkk WWEWEWV ???? , (2) According to (2), the ith element of kV is: (3) where ? ?? ?kWbias i is the weighting coef?cient bias and ??ki? is a zeromean random variable with the variance 2? .The ? ? ? ?? ? ? ?? ? ? ? ? ?? ?? ?? ?? ? ? ?kkWbi as kWEkWkWkWEkViiiiiii ??? ???? * variance depends on the type of LMSbased algorithm, as well as on the external noise variance 2n? .Thus, if the noise variance is constant or slowlyvarying, 2? is time invariant for a particular LMSbased algorithm. In that sense, in the analysis that follows we will assume that 2? depends only on the algorithm type, . on its parameters. An important performance measure for an adaptive ?lter is its mean square deviation (MSD) of weighting coef?cients. For the adaptive ?lters, it is given by, [3]: ? ?kTkk VVEM S D ??? lim. 3. Combined adaptive ?lter The basic idea of the bined adaptive ?lter lies in parallel implementation of two or more adaptive LMSbased algorithms, with the choice of the best among them in each iteration [9]. Choice of the most appropriate algorithm, in each iteration, reduces to the choice of the best value for the weighting coef?cients. The best weighting coef?cient is the one that is, at a given instant, the closest to the corresponding value of the Wiener vector. Let ? ?qkWi , be the i ?th weighting coef?cient for LMSbased algorithm with the chosen parameter q at an instant k. Note that one may now treat all the algorithms in a uni?ed way (LMS: q ≡ 181。最常用的自適應(yīng)系統(tǒng)對(duì)那些基于最小均方（ LMS）自適應(yīng)算法及其 改進(jìn) （ 基于 LMS 的 算法）。 這種方法可以適用于所有的 LMS 的算法，雖然我們?cè)谶@里只考慮其中幾個(gè)。在非平穩(wěn)情況下，未知 系統(tǒng)參數(shù) （即 *kW 最佳載體） 是隨 時(shí)間 變化的 。自適應(yīng)濾波器 的 一個(gè)重要性能衡量標(biāo)準(zhǔn)是其均方差（ MSD）的加權(quán)系數(shù)。 基于LMS 算法的 行為 主要依賴于 q， 在每個(gè)迭代 中 有一個(gè)最佳值 optq ，生產(chǎn)的最佳表現(xiàn)的自適應(yīng)算法。 提出的聯(lián)合算法 (CA)現(xiàn)在可以被總結(jié)為下面的步驟 : 第 1 步 ： 從不同預(yù)定義設(shè)置 ? ??, 2qqQ i? 中 為算法 計(jì)算 ? ?qkWi , 。 第 4 步 ： 轉(zhuǎn)到下一 個(gè)瞬間。 CA 的復(fù)雜性取決于組成算法（第 1 步），并在決策算法（步驟 3）。結(jié)果，獲得了平均超過 100（蒙特卡羅方法） 個(gè) 獨(dú)立 的 運(yùn)行， 其中 μ = 。 仿真結(jié)果 提出的基于 LMS 的算 法不同類型的自適應(yīng) 組合 濾波器是實(shí)行固定和非平穩(wěn) 情況 ，合并后的過濾系統(tǒng)識(shí)別 。每個(gè)算法 AMSD 考慮是： AMSD = （ SA1， μ）， AMSD = （ SA2， μ/ 2）， AMSD = （ SA3， μ/ 8）和 AMSD = 。這表明了各自增長(zhǎng) 了LMS 算法。t help but sing the folk songs, Nasun says. The vastness of Inner Mongolia and the lack of entertainment options for people living there, made their lives lonely. The nomadic people were very excited about our visits, Nasun recalls. We didn39。s president, who is also a renowned tenor, tells China Daily. During a tour in 1985, he went to a village and met an elderly local man, who told him a story about his friendship with a solider from Shenyang, capital of Northeast China39。s Zhangye city during their journey to Kazakhstan, May 5, 2021. The caravan, consisting of more than 100 camels, three horsedrawn carriages and four support vehicles, started the trip from Jingyang county in Shaanxi on Sept 19, 2021. It will pass through Gansu province and Xinjiang Uygur autonomous region, and finally arrive in Almaty, formerly known as AlmaAta, the largest city in Kazakhstan, and Dungan in Zhambyl province. The trip will cover about 15,000 kilometers and take the caravan more than one year to plete. The caravan is expected to return to Jingyang in March 2021. Then they will e back, carrying specialty products from Kazakhstan A small art troupe founded six decades ago has grown into a household name in the Inner Mongolia autonomous region. In the 1950s, Ulan Muqir Art Troupe was created by nine young musicians, who toured remote villages on horses and performed traditional Mongolian music and dances for nomadic families. The 54yearold was born in Tongliao, in eastern Inner Mongolia and joined the troupe in says there are 74 branch troupes across Inner Mongolia and actors give around 100 shows every year to local nomadic people. I can still recall the days when I toured with the troupe in the early 39。這對(duì) VS AMSD 是 AMSD = ，而 在CA（ CoLMS） 中 AMSD = 。 圖 2（ a）顯示了每個(gè)算法的 AMSD 特點(diǎn)。也就是說(shuō)，如果 CA 選擇， 那么 在 k 次迭代中，加權(quán)系數(shù)向量 PW ，然后 根據(jù)每一個(gè)獨(dú)立的 算法計(jì) 算出 加權(quán)系數(shù) 在（ k +1） 次迭代： 0)(,1 1 22 ?? ? ? ixf oreT Ti in? ? ?kkpk XeEWW ?21 ??? （ 9） 圖 1 快速平均算法 圖 2 快速平均算法 在前面的示例應(yīng)用 中，圖 1（ b）顯示了這種改進(jìn)。 未知的系統(tǒng)有四個(gè)時(shí)間不變系數(shù)， 而且 FIR 濾波器的 N = 4。對(duì)于標(biāo)準(zhǔn)的 LMS 算法在穩(wěn)定狀態(tài)， 2n? 和 2q? 是相關(guān)的。因此，檢查了一 對(duì) 新的 加權(quán) 系數(shù)，或者，如果 ??kDi 是最后一對(duì)，只選擇具有最小方差 的 算法 。 由于我們 對(duì) 有關(guān)信息 ? ?? ?qkWbias i , 沒有先驗(yàn) 知識(shí)，我們將使用一種特定的統(tǒng)計(jì)學(xué)方法得到的標(biāo)準(zhǔn) ,即自適應(yīng)算法選擇的 q 值問題。,GLMS: q ≡ a,SA:q ≡ 181。 因此，如果噪聲方差為常數(shù)或 是 緩慢變 化的 ，2? 為某一特定的基于 LMS 時(shí)間不變的算法。 正在研究中的 自適應(yīng)濾波問題在于 嘗試 調(diào)整權(quán)重系數(shù)，使系統(tǒng)的輸出 kTkk XWy ?跟蹤參考信號(hào)， kkTkk nXWd ?? * 中 n 是一個(gè)零均值與方差 2n? 的高斯噪聲， *kW 是最佳權(quán)向量（維納向量）。我們提出了一個(gè)自適應(yīng)濾波器的性能改善的方法 。 仿真 結(jié)果證實(shí)了提出的自適應(yīng)濾波器的優(yōu)點(diǎn) 。 is the algorithm step, E{ various parameters affecting the step for VS LMS). These parameters crucially in?uence the ?lter output during two adaptation phases:transient and steady state. Choice of these parameters is mostly based on some kind of tradeoff between the quality of algorithm performance in the mentioned adaptation phases. We propose a possible approach for the LMSbased adaptive ?lter performance improvement. Namely, we make a bination of several LMSbased FIR ?lters with different parameters, and provide the criterion for choosing the most suitable algorithm for different adaptation phases. This method may be applied to all the LMSbased algorithms, although we here consider only several of them. The paper is anized as follows. An overview of the considered LMSbased algorithms is given in Section 3 proposes the criterion for evaluation and bination of adaptive algorithms. Simulation results are presented in Section 4. 2. LMS based algorithms Let us de?ne the input signal vector Tk NkxkxkxX )]1()1()([ ???? ?and vector of weighting coef?cients as