本文關(guān)鍵詞： 漁船模擬器 漁具系統(tǒng)水動力模型 數(shù)值模擬 GPU加速　出處：《大連海事大學(xué)》2016年博士論文　論文類型：學(xué)位論文

【摘要】：利用漁船模擬器進(jìn)行漁船船員的培訓(xùn),可降低風(fēng)險、節(jié)省成本、縮短培訓(xùn)周期、提高受訓(xùn)人員的綜合操控能力和實際分析能力。為進(jìn)行漁船模擬器中雙船拖網(wǎng)和單船圍網(wǎng)作業(yè)過程的動態(tài)仿真,本文基于集中質(zhì)量法分別建立了雙船拖網(wǎng)和單船圍網(wǎng)漁具系統(tǒng)的高可信度水動力數(shù)學(xué)模型,通過采用高效的精細(xì)積分算法求解龐大的非線性微分方程組和基于CUDA的GPU(Graphic Process Unit)并行計算加速技術(shù)保證了所建立數(shù)學(xué)模型解算的實時性,考慮漁船與漁具系統(tǒng)的耦合,搭建了雙船拖網(wǎng)和單船圍網(wǎng)作業(yè)的仿真系統(tǒng),從而為漁船模擬器的開發(fā)奠定了堅實的基礎(chǔ)。高可信度水動力數(shù)學(xué)模型是雙船拖網(wǎng)和單船圍網(wǎng)作業(yè)過程動態(tài)仿真的基礎(chǔ)。本文基于集中質(zhì)量法提出一種雙船拖網(wǎng)和單船圍網(wǎng)漁具系統(tǒng)數(shù)學(xué)模型的建模方法。在建模過程中,將漁具系統(tǒng)分為曳綱(括綱)和網(wǎng)具系統(tǒng)兩部分,分別建立其水動力模型,并通過邊界條件實現(xiàn)了二者的耦合；將漁具系統(tǒng)視為一系列無質(zhì)量彈簧相連的質(zhì)量點集合,將結(jié)節(jié)視為球體,目腳視為圓柱體,所有外力均作用于質(zhì)量點上,根據(jù)牛頓第二定律分別建立各質(zhì)量點的水動力數(shù)學(xué)模型；在此基礎(chǔ)上,考慮漁船與漁具系統(tǒng)間的耦合,從而可進(jìn)行雙船拖網(wǎng)和單船圍網(wǎng)作業(yè)的動態(tài)仿真。將仿真結(jié)果與模型試驗結(jié)果進(jìn)行比較和分析,驗證了模型的有效性。漁具系統(tǒng)由大量有彈性的柔性體組成,使用集中質(zhì)量法對其進(jìn)行數(shù)字離散后,構(gòu)成了數(shù)目巨大的非線性微分方程組,且該非線性微分方程組穩(wěn)定求解的時間步長受到柔性體剛度的限制。高效的數(shù)值解法對保障其水動力模型求解的穩(wěn)定性及提高模型解算的效率是十分必要的。本文提出了一種基于精細(xì)積分法的漁具系統(tǒng)水動力模型的數(shù)值解法,運用指數(shù)矩陣的2N類算法及函數(shù)矩陣的分部積分法對網(wǎng)具系統(tǒng)的非線性方程進(jìn)行數(shù)值求解。與傳統(tǒng)漁具系統(tǒng)數(shù)值解法相比,該數(shù)值方法精度較高,計算效率高,對采用集中質(zhì)量法建模的方程求解具有普適性。大型拖網(wǎng)及圍網(wǎng)系統(tǒng)的物理節(jié)點數(shù)量通常達(dá)百萬級別,即便利用網(wǎng)目合并技術(shù)對網(wǎng)衣進(jìn)行簡化處理,使用上述數(shù)值解法仍不能保證水動力模型解算的實時性。本文基于CUDA的GPU并行加速技術(shù)進(jìn)行漁具系統(tǒng)水動力數(shù)學(xué)模型的解算,通過在全局內(nèi)存中開辟數(shù)組,存儲上一幀所有質(zhì)量點的位置,通過線程索引查找相連節(jié)點位置計算張力,網(wǎng)具系統(tǒng)中所有結(jié)節(jié)和目腳的模型解算采用并行執(zhí)行。實驗結(jié)果表明：利用GPU并行解算在保證計算精度的前提下能有效提高計算效率,保證了大型拖網(wǎng)及圍網(wǎng)系統(tǒng)動態(tài)仿真的實時性。最后,利用場景管理軟件OSG(OpenSceneGraph)實現(xiàn)了雙船拖網(wǎng)和單船圍網(wǎng)作業(yè)過程的三維可視化,并與航海模擬器相結(jié)合,初步實現(xiàn)了漁船模擬器中雙船拖網(wǎng)和單船圍網(wǎng)操作模塊的功能。本文研究對提高水下柔性體的仿真效果、推進(jìn)漁船模擬器產(chǎn)品研發(fā)具有現(xiàn)實意義,可為漁船模擬器的進(jìn)一步研究提供理論依據(jù)和技術(shù)支撐。
[Abstract]:By using the fishing crew fishing simulator training, can reduce the risk, save the cost, shorten the training cycle, improve the comprehensive ability to control the trainees and the actual analysis. Dynamic simulation of double boat trawl fishing simulator and ship seining process, this paper is based on the lumped mass method were established and two boat trawler single boat Seine fishing gear system with high reliability of hydrodynamic mathematical model, through the use of highly precise integration algorithm for solving large nonlinear differential equations based on CUDA and GPU (Graphic Process Unit) parallel real-time calculation speed technology to ensure the mathematical model calculation, considering the coupling of fishing boats and fishing gear system, set up simulation system of double and single boat trawl fishing, thus laying a solid foundation for the development of fishing simulator. High reliability hydrodynamic mathematical model is a double boat The basic network and single ship seining process dynamic simulation. The modeling method of lumped mass method and presents a double boat trawl and single boat Seine fishing gear system based on mathematical model. In the process of modeling, the gear system is divided into uarp (including class) two and the net system, and respectively establish its hydrodynamic model, and realized the coupling of the two through the boundary conditions; fishing gear system as a series of springs without mass connected to the quality of the collection, the nodules as sphere, mesh feet as a cylinder, all external forces are applied at the quality point, the hydrodynamic mathematical model of the quality points are established according to the second law of Newton; on this basis, considering the coupling between fishing boats and fishing gear system, dynamic simulation can be used for two boat trawler and single boat seiner. The simulation results and the experimental results were compared and analyzed to verify the validity of the model. Fishing gear system consists of a large number of flexible elastic components, the use of digital discrete lumped mass method, a set of nonlinear differential equations in large numbers, and the nonlinear differential equations for solving the stable time step by the flexible body stiffness limit. Efficient solution to guarantee the stability of numerical hydrodynamic model and to improve the efficiency of model calculation is very necessary. This paper presents a numerical solution of the hydrodynamic model of gear system based on precise integration method, the numerical solution of nonlinear equation integral method by using 2N algorithm and function matrix exponential matrix of the gear system. Compared with the traditional numerical method of fishing gear system. The numerical methods with high accuracy, high calculation efficiency and universality of the equations of lumped mass method. A large number of physical nodes and fence system trawl Usually one million level, even on the net is simplified by using mesh merging technique, using the numerical method still can not guarantee real-time dynamic model of water solution. The CUDA GPU parallel accelerating technology of fishing gear hydrodynamic mathematical model solution based on, by opening up in the global memory array, a frame of all position the quality of the storage, index lookup nodes connected by thread tension position calculation system, net all nodules and eye foot model are solved by parallel execution. The experimental results show that parallel computation in ensuring the computational accuracy can effectively improve the computing efficiency by GPU, real-time and dynamic system of large trawl net the simulation. Finally, using the scene management software OSG (OpenSceneGraph) to realize 3D visualization of two boats trawl and single boat seiner process, and with the navigation simulator. Combined, initially realized the two ships in fishing simulator trawl and single boat Seine operation modules. The simulation results in this paper to improve the water body flexible, has practical significance to promote fishing simulator product development, further study can be used as a fishing simulator to provide theoretical basis and technical support.

【學(xué)位授予單位】：大連海事大學(xué)
【學(xué)位級別】：博士
【學(xué)位授予年份】：2016
【分類號】：U661.74

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 張錚錚;李勝忠;;我國遠(yuǎn)洋漁業(yè)裝備發(fā)展戰(zhàn)略與對策[J];船舶工程;2015年06期

2 ;2014年全國漁業(yè)經(jīng)濟統(tǒng)計公報[J];中國水產(chǎn);2015年06期

3 ZHOU Cheng;XU Liuxiong;ZHANG Xinfeng;YE Xuchang;;Application of Numerical Simulation for Analysis of Sinking Characteristics of Purse Seine[J];Journal of Ocean University of China;2015年01期

4 姚杰;任玉清;于曉利;鄧長輝;;高素質(zhì)遠(yuǎn)洋漁業(yè)航海技術(shù)人才培養(yǎng)的實踐探討[J];航海教育研究;2014年03期

5 ;2013年全國漁業(yè)經(jīng)濟統(tǒng)計公報[J];中國水產(chǎn);2014年07期

6 陳英龍;趙勇剛;周華;黃洪亮;;大型中層拖網(wǎng)網(wǎng)具系統(tǒng)的仿真研究[J];浙江大學(xué)學(xué)報(工學(xué)版);2014年04期

7 徐義軍;魏少峰;;漁政執(zhí)法培訓(xùn)和漁船船員培訓(xùn)管理存在的問題分析與探討[J];河南農(nóng)業(yè);2014年07期

8 彭志英;;提高漁業(yè)船員培訓(xùn)質(zhì)量的研究[J];價值工程;2014年07期

9 林秀玲;;淺議素質(zhì)培養(yǎng)對船員培訓(xùn)的重要性[J];武漢船舶職業(yè)技術(shù)學(xué)院學(xué)報;2014年01期

10 楊智棟;潘光;杜曉旭;;變纜長拖纜AUV縱向運動建模與仿真[J];魚雷技術(shù);2014年01期

，

本文編號：1483660