本文選題：混輸泵 + 優(yōu)化設(shè)計(jì)�。� 參考：《蘭州理工大學(xué)》2012年碩士論文

【摘要】：多相混輸技術(shù)是近年來發(fā)展起來的一種高效經(jīng)濟(jì)的油田開發(fā)方式。多相混輸泵作為多相混輸技術(shù)的核心設(shè)備，它是一種集常規(guī)液相泵和氣體壓縮機(jī)性能于一體的多相輸送裝置。葉輪作為混輸泵的核心部件，直接影響著泵中流體的流動(dòng)穩(wěn)定性以及泵的效率。因此，對葉片進(jìn)行優(yōu)化設(shè)計(jì)，提高其工作性能對提高混輸泵的效率有著重要的意義。 本課題是基于CFD數(shù)值計(jì)算、神經(jīng)網(wǎng)絡(luò)及優(yōu)化算法相結(jié)合的混輸泵葉片優(yōu)化設(shè)計(jì)。該方法首先對葉輪葉片進(jìn)行了參數(shù)化設(shè)計(jì)，選取參數(shù)化后的12個(gè)控制點(diǎn)作為優(yōu)化變量，確定優(yōu)化變量的3個(gè)取值，通過改變優(yōu)化變量的取值，從而達(dá)到改變?nèi)~型的目的。以優(yōu)化變量為因素，3個(gè)取值為水平，使用L27_3_13正交實(shí)驗(yàn)設(shè)計(jì)表設(shè)計(jì)出27組實(shí)驗(yàn)方案。對每組方案進(jìn)行建模，通過fluent軟件預(yù)測每組方案的效率。27組方案做為神經(jīng)網(wǎng)絡(luò)的樣本數(shù)據(jù)，其中20組做為網(wǎng)絡(luò)訓(xùn)練樣本，7組作為預(yù)測樣本，優(yōu)化變量作為網(wǎng)絡(luò)輸入，計(jì)算效率作為網(wǎng)絡(luò)輸出。訓(xùn)練好的神經(jīng)網(wǎng)絡(luò)分別作為遺傳算法及粒子群算法的適應(yīng)度值函數(shù)，利用遺傳算法及粒子群算法的全局尋優(yōu)能力，對葉片進(jìn)行優(yōu)化。對兩種優(yōu)化算法輸出的最優(yōu)解進(jìn)行對比，結(jié)果顯示遺傳算法的輸出解更優(yōu)，最終以遺傳算法的輸出解作為混輸泵葉片的最優(yōu)模型。對優(yōu)化后及優(yōu)化前的混輸泵葉輪分別在含氣率為0.1、0.3、0.5、0.7、0.9五種工況下進(jìn)行數(shù)值模擬，對比其優(yōu)化效果。 從優(yōu)化結(jié)果來看，優(yōu)化后的混輸泵在各含氣率下效率及揚(yáng)程均有所提高，且在設(shè)計(jì)工況下優(yōu)化效果最為明顯，優(yōu)化后的計(jì)算效率最大提高了1.91%。這也在一定程度上證明了本文的優(yōu)化設(shè)計(jì)方法的可行性，該優(yōu)化設(shè)計(jì)方法適用范圍廣、魯棒性強(qiáng)，有較強(qiáng)的工程應(yīng)用價(jià)值。本文還從葉柵理論、速度三角形及泵內(nèi)多相流動(dòng)理論分析入手，，初步探討了混輸泵重要結(jié)構(gòu)參數(shù)與性能參數(shù)間的對應(yīng)關(guān)系，推導(dǎo)出混輸泵流量、揚(yáng)程、效率的表達(dá)式，為進(jìn)一步改善混輸泵性能提供了理論依據(jù)。
[Abstract]:Multiphase mixing technology is an efficient and economical oil field development method developed in recent years. As the core equipment of multiphase mixing technology, multiphase pump is a kind of multiphase transportation device which integrates the performance of conventional liquid phase pump and gas compressor. Impeller, as the core component of the mixed pump, directly affects the flow stability and efficiency of the pump. Therefore, it is of great significance to improve the efficiency of the mixed pump by optimizing the blade design and improving its working performance. This subject is based on CFD numerical calculation, neural network and optimization algorithm combined with the mixed pump blade optimization design. In this method, the impeller blade was first parameterized, 12 control points were selected as optimization variables, 3 values of optimization variables were determined, and the change of blade shape was achieved by changing the value of optimization variables. Taking the optimization variable as the factor and the three values as the level, 27 groups of experimental schemes were designed by using the L27313 orthogonal experimental design table. The efficiency of each group is predicted by fluent software. 27 groups of schemes are used as neural network sample data, 20 groups as network training samples and 7 groups as prediction samples, and the optimization variables as network input. The computational efficiency is the output of the network. The trained neural network is used as the fitness function of genetic algorithm and particle swarm optimization algorithm respectively. The global optimization ability of genetic algorithm and particle swarm optimization algorithm is used to optimize the blade. The results show that the output solution of genetic algorithm is better than that of genetic algorithm. Finally, the output solution of genetic algorithm is taken as the optimal model of mixed pump blade. After optimization and before optimization, the impellers of mixed transportation pump were numerically simulated under five conditions of 0.1g / 0.3g / 0.50.50.7g / 0.9, respectively, and the optimization results were compared. From the result of optimization, the efficiency and lift of the pump are improved under each gas content, and the optimization effect is the most obvious under the design condition, and the calculation efficiency of the optimized pump is improved by 1.91%. It also proves to some extent the feasibility of the optimal design method in this paper. This method has a wide range of application, strong robustness, and has a strong engineering application value. Based on the cascade theory, velocity triangle and multiphase flow theory, the corresponding relationship between the important structural parameters and the performance parameters of the pump is preliminarily discussed, and the expressions of the flow rate, lift and efficiency of the mixed pump are deduced. It provides a theoretical basis for further improving the performance of the mixed pump.
【學(xué)位授予單位】：蘭州理工大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2012
【分類號】：TH311

【引證文獻(xiàn)】

相關(guān)碩士學(xué)位論文 前3條

1 夏書強(qiáng);氣液兩相介質(zhì)下液力透平基本方程和換算關(guān)系曲線的研究[D];蘭州理工大學(xué);2013年

2 顧生琴;油氣混輸泵不同工況內(nèi)部流動(dòng)特性分析[D];蘭州理工大學(xué);2013年

3 趙文杰;導(dǎo)葉式螺旋離心泵內(nèi)固液兩相流動(dòng)的數(shù)值計(jì)算與 分析[D];蘭州理工大學(xué);2013年

本文編號：1775163