收斂加速 的英文怎麼說

中文拼音 [shōuliǎnjiāsù]
收斂加速 英文
acceleration of convergence
  • 收 : Ⅰ動詞1 （把攤開的或分散的事物聚集、合攏） put away; take in 2 （收取） collect 3 （收割） harvest...
  • 斂 : Ⅰ動詞1 （收起; 收住） hold back; keep back 2 （約束） restrain 3 （收集; 徵收） gather; collect; ...
  • 速 : Ⅰ形容詞（迅速; 快） fast; rapid; quick; speedy Ⅱ名詞1 （速度） speed; velocity 2 （姓氏） a surna...
  • 收斂 : 1 （減弱或消失） weaken or disappear 2 （約束言行） restrain oneself 3 [數學] convergence; constr...
  1. It is shown in the case that super matrix is easy to constringe the selection sequence assisted by the software, which is efficient and convenient for manager to evaluate and choose the project

    案例顯示，加權超級矩陣在軟體求解的基礎上可以快速收斂，得出各種方案的優先排序，從而幫助管理者簡便、有效地進行方案決策。
  2. A kind of accelerating convergence factors for limit periodic continued fraction

    一類極限循環連分式的加速收斂因子
  3. Different algorithms were compared in the numerical simulation and model experiment of concrete velocity inversion the results showed that, compared with the least - square solutions and the damped least - square solutions, the damped and natural weighted least - square solutions reflected the inner defections of the tested object more reliably and exactly due to the usage of correct priori information, which benefited the suppression of noises and made the iterations of inverse stable and convergent

    結果表明，相對于最小二乘法和阻尼最小二乘法，基於自然權的加權阻尼最小二乘演算法，由於利用了正確的先驗信息，不但使反演過程收斂，而且具有數值穩定、抗噪能力強的優點，其成像結果能真實有效地反映對象內部缺陷，因此更適用於混凝土的超聲波速度反演。
  4. Compared with straight ray tracing, inversion of bending ray tracing is better when data of test are accurate. on the other hand, natural weight is introduced and numerical simulations and experiments of wlm on inversion of velocity and attenuation are also made, which show that wlm, with resistance of noises and convergence of iteration, may get good re - construction images

    本文引入了物理意義明確的自然權，對基於自然權的加權阻尼最小二乘反演演算法（ wlm ）在速度反演、衰減反演中進行了數值模擬和試驗研究，結果表明wlm均能獲得良好的重建圖像，迭代收斂，抗噪能力強。
  5. The preconditioned accelerate the convergence of 2ppj method

    型方法收斂性的加速
  6. This article puts forward a solution named divide - assemble by deducing the size of bp neural network to overcome entering the local best point, the dividing process is that a big bp neural network is divided into several small bp neural networks, every small bp neural network can study alone, after all small bp neural networks finish their study, we can assemble all these small bp neural networks into the quondam big bp neural networks ; on the basis of divide - assemble solution, this article discusses the preprocessing of input species and how to deduce the size of bp neural network further to make it easy to overcome entering the local best point ; for the study of every small bp neural network, this article adopts a solution named gdr - ga algorithm, which includes two algorithms. gdr ? a algorithm makes the merits of the two algorithms makeup each other to increase searching speed. finally, this article discusses the processing of atm band - width distribution dynamically

    本文從bp網的結構出發，以減小bp神經網路的規模為手段來克服陷入局部極小點，提出了bp神經網路的拆分組裝方法，即將一個大的bp網有機地拆分為幾個小的子bp網，每個子網的權值單獨訓練，訓練好以後，再將每個子網的單元和權值有機地組裝成原先的bp網，從理論和實驗上證明了該方法在解決局部極小值這一問題時是有效的；在拆分組裝方法基礎上，本文詳細闡述了輸入樣本的預處理過程，更進一步地減小了bp網路的規模，使子網的學習更加容易了；對于子網的學習，本文採用了最速梯度? ?遺傳混合演算法（即gdr ? ? ga演算法） ，使gdr演算法和ga演算法的優點互為補充，提高了收斂速度；最後本文闡述了用以上方法進行atm帶寬動態分配的過程。
  7. And raga can steadily adjust and compress the variable changed space. the times of accelerating circle is not many, commonly, the times is under 10 times. even though the problem is very complicated, the times of accelerating circle is less than 50 times

    該方法具有處理復雜優化問題的能力，並且，它調整、壓縮搜索區間的性能也是穩健的，不易早熟收斂，加速循環的次數也不會很高，一般就在10次以下，即便是特別復雜的問題其加速循環次數也在50次以內，所以說raga的計算量很少。
  8. In order to quicken styptic v elocity, under - relaxation a nd source term linearization is appilied

    為了加快收斂的速度，採用了欠鬆弛技術和源項線性化。
  9. The viscid flux is discretized by second - order central difference scheme. baldwin - lomax turbulence model is implemented in navier - stokes flow solver. for steady - state calculations, a four - stage runge - kutta scheme with convergence acceleration techniques such as local - time stepping and implicit residual smoothing is used

    其中，定常計算中的時間推進採用四步runge ? kutta方法，並應用了當地時間步長、隱式殘值光順等加速收斂措施；非定常計算中的時間推進採用jameson的隱式雙時間方法。
  10. In this dissertation, we firstly prove that any dirichlet problem is indeed equal to a voltages problem of networks. we give five solutions to dirichlet problem in two dimensions ; among these five solutions, we prove that the iteration solution and the solution of relaxations are exponential convergence, then we estimate their respective convergence rates ; secondly, we discuss random walks on general networks, prove that there is an one to one correspondence between networks and reversible ergodic markov chains ; thirdly, we give probabilistic interpretation of voltages for general networks : when a unit voltage is applied between a and b, making va = 1 and vb = 0, the voltage vx at any point x represents the probability that a walker starting from x will return to a before reaching b ; furthermore, we study the relationship between effective resistance and escape probability : starting at a, the probability that the walk reaches b before returning to a is the ratio of the effective conductance and the total conductance

    本文證明了任何邊值的dirichlet問題都可轉化為求解電路電壓的問題：給出了計算平面格點上dirichlet問題的5種方法：證明了迭代法和松馳法都是指數收斂的，並分別給出收斂速度的估計；討論了一般電路上的隨機徘徊，驗證了電路與可逆的遍歷markov鏈是一一對應的；給出了電路電壓的概率解釋：當把1伏電壓加於a ， b兩端，使得v _ a = 1 ， v _ b = 0時，則x點的電壓v _ x表示對應的markov鏈中，從x出發，到達b之前到達a的概率；進一步地，給出了逃離概率與有效電阻之間的關系：從a出發，在到達b之前到達a的概率為有效傳導率與通過a的總傳導率之比。
  11. First, we consider an additive schwarz algorithm for the solution of ax 4 - f ( x ) 0, x when coefficient a is an m - matrix and f ' ( x ) 0. by applying the theory of weak regular splitting of matrices to the above considered algorithm, we obtain the weighted max - norrn bound for iterations. moreover, under the assumption that f ( x ) is concave, we establish monotone convergence of the considered algorithm

    本文內容如下：首先，應用加性schwarz演算法求解非線性互補問題，其中a是m陣，應用弱分解理論，我們獲得了在加權范數意義下誤差的幾何收斂速度，在f （ x ）是凹函數的假設下我們還獲得了此演算法的單調收斂性，同時我們給出此演算法的一種修改演算法，無需前面的假設，該演算法具有單調收斂性。
  12. 4. a 2 - d and 3 - d euler equations and n - s equations are solved using the cell - centered finite volume method and four - step runge - kutta scheme on the cartesian grids with standard convergence acceleration techniques such as local time stepping, enthalpy and implicit residual smoothing

    使用jameson中心有限體積法和runge - kutta時間推進方法，求解了關於二維、三維復雜流場的euler 、 navier - stokes方程，採用了當地時間步長、隱式殘值光順等多種加速收斂方法。
  13. Through ( multi - level ) bordered block partition of the power system topology matrix, a bbdf coefficient matrix which is suitable for above decomposition is formed. when the equation is solved with parallel iterative method, convergence acceleration is achieved through damping newton method

    通過採用（多重）對角加邊的方法進行電力網路拓撲分割，以形成適合以上分解法求解的對角加邊形式的系數矩陣，并行求解后並對迭代過程進行阻尼牛頓法修正，以達到加速收斂的作用。
  14. In this paper, the upwind scheme and the central scheme are presented for solving 3 - d n - s equations using the cell - center finite volume spatial discretization and four - stage runge - kutta time stepping scheme, with standard convergence acceleration techniques such as local time stepping and implicit residual smoothing

    在n - s方程的數值計算上，採用了中心差分格式和迎風格式，用格心格式的有限體積法進行了空間離散，用四步龍格?庫塔法作顯式時間推進，並採用了當地時間步長和隱式殘差光順等加速收斂措施。
  15. A normal transform is introduced, and there are enough much grids in the region between the critical layer and the wall, where the variation of the disturbance is the quickest. the finite - difference of governing equations with fourth - order accuracy in the normal direction is utilized in full regions including points close the wall, and is very important for accurately discreting equations. the combination of global and local methods is implemented, and a new iterative formula is derived

    對于曲線坐標系下的拋物化穩定性方程，文中發展了求解的高效數值方法:引進法向變換，使得在臨界層與壁面之間的擾動量變化最快的區域有更多的法向網格點;導出包含邊界鄰域在內的完全四階精度的法向有限差分格式，這對方程精確離散至關重要;採用全局法和局部法相結合的方法及其新的迭代公式，大大加速收斂並得到更精確的特徵值。
  16. A novel dynamic evolutionary clustering algorithm ( deca ) is proposed in this paper to overcome the shortcomings of fuzzy modeling method based on general clustering algorithms that fuzzy rule number should be determined beforehand. deca searches for the optimal cluster number by using the improved genetic techniques to optimize string lengths of chromosomes ; at the same time, the convergence of clustering center parameters is expedited with the help of fuzzy c - means ( fcm ) algorithm. moreover, by introducing memory function and vaccine inoculation mechanism of immune system, at the same time, deca can converge to the optimal solution rapidly and stably. the proper fuzzy rule number and exact premise parameters are obtained simultaneously when using this efficient deca to identify fuzzy models. the effectiveness of the proposed fuzzy modeling method based on deca is demonstrated by simulation examples, and the accurate non - linear fuzzy models can be obtained when the method is applied to the thermal processes

    針對模糊聚類演算法不適應復雜環境的問題，提出了一種新的動態進化聚類演算法，克服了傳統模糊聚類建模演算法須事先確定規則數的缺陷.通過改進的遺傳策略來優化染色體長度，實現對聚類個數進行全局尋優;利用fcm演算法加快聚類中心參數的收斂;並引入免疫系統的記憶功能和疫苗接種機理，使演算法能快速穩定地收斂到最優解.利用這種高效的動態聚類演算法辨識模糊模型，可同時得到合適的模糊規則數和準確的前提參數，將其應用於控制過程可獲得高精度的非線性模糊模型
  17. The main characteristics and advantages are : on one hand, we adopted the low - complexity bussgang algorithm, and did blind estimation to ofdm sub - channel according to the mean square error criterion ( mse ) and peak distortion criterion ; on the other hand, we used the differential detection, which accelerates the convergence speed and avoid the error transmission problem resulted from the bussgang algorithm

    這兩種演算法一方面採用了低復雜度的bussgang自適應演算法，分別依據均方誤差準則和峰值失真準則對ofdm系統子通道進行盲估計；另一方面通過引入差分檢測技術，加快了演算法收斂速度，克服了bussgang演算法帶來的誤差傳播問題。
  18. In the last of this paper we apply our algorithms to the learning of feed - forward neural network, and get some new learning algorithms. we also give some numerical experiments to compare our algorithms with others

    最後，將得到的這些優化加速收斂方法應用到了多層前饋神經網路的學習過程，給出了加速收斂的bp演算法，通過實際神經網路學習問題驗證了工作的成效。
  19. Abstract : based on the iterative bit - filling procedure, a computationally efficient bit and power allocation algorithm is presented. the algorithm improves the conventional bit - filling algorithms by maintaining only a subset of subcarriers for computation in each iteration, which reduces the complexity without any performance degradation. moreover, a modified algorithm with even lower complexity is developed, and equal power allocation is introduced as an initial allocation to accelerate its convergence. simulation results show that the modified algorithm achieves a considerable complexity reduction while causing only a minor drop in performance

    文摘:基於迭代的比特和功率分配機制，提出了一種低復雜度的比特和功率分配演算法.與傳統的迭代分配演算法不同，該演算法在每次迭代中只需要比較幾個特定的子載波.該方法在保持傳統迭代演算法性能的前提下極大地減小了迭代分配演算法的復雜度.此外，通過選擇等功率分配方案作為初始方案加快了演算法的收斂速度，進一步降低了演算法復雜度.模擬結果表明，提出的改進演算法在基本不犧牲系統性能的前提下有效地降低了演算法復雜度
  20. For the variables with greater gradient, a high - precision interpolating function, such as quasi - consistence hexahedral element method, is naturally necessary to be adopt

    開邊界給定方式宜採用速度dirichlet條件和流動壓力neumann條件，以利用壓力無反射條件，加快計算收斂速度。
分享友人
分享到 Line

分享到臉書