Besides , considering the high ratio between quality and price , high scalability , the convenience and flexibility in development and utilization , selecting workstations cluster as parallelization environment can reduce the development and application cost of image reconstruction and has great significance upon increasing the ratio between quality and price of high power x - ict 另外基于工作站機(jī)群的高性價(jià)比、高擴(kuò)展性、開發(fā)和使用上的方便靈活性等種種優(yōu)點(diǎn),選擇工作站機(jī)群作為并行實(shí)現(xiàn)環(huán)境,還能降低圖像重建的開發(fā)和應(yīng)用成本,對(duì)提高高能x射線工業(yè)ct機(jī)的性價(jià)比具有重要意義。
This algorithm can identify and choose in - core or out - of core algorithm based on different scale of computation and cluster during each computing phase . and parallelization during each phase is implemented . this algorithm can solve the residual equations parallelly and the current data distribution of residual equations does not change ( 4 )給出了一種可以在計(jì)算的每一個(gè)階段根據(jù)不同的計(jì)算規(guī)模和機(jī)群規(guī)模,自動(dòng)識(shí)別選用內(nèi)存或外存算法的邊界元子域并行算法;實(shí)現(xiàn)了各主要計(jì)算步驟的并行化;對(duì)于剩余方程組的并行求解,算法可以在不改變當(dāng)前數(shù)據(jù)分布狀態(tài)下,實(shí)行并行求解。
The idea of combining the practical technique with parallel technique is presented and the parallelization of practical gmres is implemented . the corresponding serial and parallel programs are developed and different numerical tests are carried out . ( 4 ) a be sub - region parallel algorithm is presented 分析了經(jīng)典gmres算法的復(fù)雜度和并行性,提出了實(shí)用化技術(shù)和并行化技術(shù)結(jié)合使用的思想,通過(guò)對(duì)矩陣向量運(yùn)算的分布并行處理實(shí)現(xiàn)了實(shí)用化gmres算法的并行計(jì)算;編制了相應(yīng)的串并行程序,進(jìn)行了數(shù)值試驗(yàn)驗(yàn)證。
We analysed the traditional automatic parallelization technology , including dependency analysis theory , program transformation technology , parallel scheme and the optimization of related synchronization and communication etc , which are the theoretical basis of the whole article . cfd computing features , especially the features of explicit difference computing , have also been further ananlysed . we also summarized drawbacks of traditional automatic parallelization technology used in cfd : small parallel granularity , difficulty in attaining global identical data partition , and difficulty in attaining high parallel efficiency on distributed memory system 本文討論、分析、總結(jié)了通用的自動(dòng)并行化技術(shù):相關(guān)性分析理論、程序變換技術(shù)、并行模式以及同步通信與優(yōu)化問(wèn)題等等,它們是本文研究工作的理論基礎(chǔ);針對(duì)研究對(duì)象,深入分析了cfd計(jì)算的特點(diǎn),特別是顯式差分計(jì)算的特點(diǎn);并歸納出傳統(tǒng)的自動(dòng)并行化技術(shù)在cfd應(yīng)用中存在的問(wèn)題:并行粒度小、難以獲得全局統(tǒng)一的數(shù)據(jù)劃分方式,對(duì)于分布存儲(chǔ)結(jié)構(gòu)的并行機(jī)難以獲得高效率。
The optimization time and the optimization quality of evolutionary computation ca n ' t keep up with the actual demand . in order to solve the massive complicated optimization problems , the author analyzes the parallelization principle and the application environment of parallel evolutionary computation , and presents internet - based parallel evolutionary computation ( ipec ) 為解決大規(guī)模復(fù)雜優(yōu)化問(wèn)題,本文就并行進(jìn)化計(jì)算的并行化原理和應(yīng)用平臺(tái)進(jìn)行分析,提出了基于internet環(huán)境的并行進(jìn)化計(jì)算( internet - basedparallelevolutionarycomputation ,簡(jiǎn)稱ipec ) 。
Some crucial design principles , methods and techniques in petsc are highlighted . propose a general scheme for the parallelization of an unstructured mesh , including graph partitioning , data division and mesh managing , based on message passing programming style . propose a series of comprehensive considerations about how to achieve a well - designed , object - oriented and data - distributed parallel software ?從軟件外在的功能組織、使用模式與內(nèi)在的設(shè)計(jì)思想、實(shí)現(xiàn)技術(shù)及其因果關(guān)聯(lián)性等多重角度出發(fā),對(duì)petsc做了深入的分析探討,尤其從源代碼的層次,重點(diǎn)剖析了petsc的面向?qū)ο?、中性?shù)據(jù)結(jié)構(gòu)、上下文環(huán)境、并行設(shè)計(jì)與通信等重要核心技術(shù)的實(shí)現(xiàn)方法及其對(duì)petsc使用模式和性能所產(chǎn)生的意義。
In order to solve the massive complicated optimization problems , the author analyzes the parallelization principle and the application environment of parallel genetic algorithm , and presents a kind of extended network - based distributed genetic algorithm ( endga ) which combines the respective ideal of sega and parallel genetic algorithm 為解決大規(guī)模復(fù)雜優(yōu)化問(wèn)題,本文就并行遺傳算法的并行化原理和應(yīng)用平臺(tái)進(jìn)行分析,并結(jié)合sega算法和分布式遺傳算法兩種思想提出了一種基于網(wǎng)絡(luò)環(huán)境的分布式遺傳算法( extendednetwork - baseddistributedgeneticalgorithm ,簡(jiǎn)稱endga ) 。
The narrow fan - beams are reset into parallel beam then the image reconstruction is conducted by parallel beam " convolution back projection algorithm , which combines the high scan efficiency with the convenience of image reconstruction . this paper analyses the parallelism in narrow fan - beam " convolution back projection algorithm , divides the task of image reconstruction into several subtasks , and discusses the parallelization of narrow fan - beam " decomposition and reset , parallel beam " convolution back projection , and image accumulation 對(duì)于窄角扇束掃描方式,把窄角扇束重排成平行束,再由平行束卷積反投影重建算法來(lái)重建圖像,是把掃描的高效率和重建方式的簡(jiǎn)便易行很好的結(jié)合起來(lái)。本論文對(duì)窄角扇束卷積反投影算法進(jìn)行了并行性分析,指出把圖像重建任務(wù)分解為多個(gè)子任務(wù)并行工作,并在工作站機(jī)群上討論了窄角扇束的分解、重排、平行束卷積反投影圖像重建、圖像合成的并行實(shí)現(xiàn)。
Aiming at the implicit parallelism of ga and the characteristic of dps system , we study the parallelization of the former two algorithms . the basic idea is to put forward an agents - based model of parallel coalition formation algorithm on the basis of coarse - grained parallel genetic algorithm 由于遺傳算法的隱并行性以及dps系統(tǒng)的特性,我們?cè)谖闹袑?duì)上述兩種算法的并行化做了研究,基本的思想是在粗粒度并行遺傳算法模型的基礎(chǔ)上提出一種基于agents的并行聯(lián)盟形成算法模型。
The difference between these two algorithms is that the former uses sub - domain as the basic unit of task to be allocated and the latter uses the node - super - row as the basic unit of task . ( 6 ) the original problem is transformed into transformed domain by using laplace transform method . by the parallelization of the bem in the transformed domain , the parallelization of the elasto - dynamic be analysis is implemented by introducing the time related fimdamental solution , the time dependency is released from the formation of time - domain be equations ( 6 )通過(guò)拉氏積分變換法將彈性動(dòng)力問(wèn)題轉(zhuǎn)換至變換域,通過(guò)變換域上邊界元的分布并行處理實(shí)現(xiàn)了彈性動(dòng)力邊界元分析的并行化;引入與時(shí)間有關(guān)的基本解,解除了時(shí)域邊界元系統(tǒng)方程組形成階段的時(shí)間順序依賴性,通過(guò)矩陣向量運(yùn)算的分布并行處理實(shí)現(xiàn)方程組時(shí)間步進(jìn)求解方法的并行化,這種方法是一種部分時(shí)間并行算法。