某柴油发动机缸内燃烧的数值模拟及优化分析
|
摘要
采用基于克里精(Kriging)代理模型的混合多目标梯度优化算法(HMGE),通过FLUENT软件平台,对柴油发动机燃烧过程进行了多目标优化,优化变量为缩口率、余隙高度、凸台高度、喷油锥角,优化目标为平均温度、平均压力、碳烟(Soot)排放、氮氧化合物(NOx)排放。完成多目标优化计算后,应用数据挖掘的总变差分析方法(ANOVA)和自组织映射分析方法(SOM)对优化变量和优化目标函数进行定性和定量分析,揭示出优化变量与优化目标之间的相互关系。分析结果表明:较好的动力性能需要较小的缩口率和较小的余隙高度;较少的碳烟排放需要较大的余隙高度,较低的氮氧化合物排放需要较小的凸台高度和较小的缩口率。
This research uses hybrid multi-objective Gradient exploration algorithm(HMGE) which is based on Kriging surrogate model to optimize and analysis combustion process and performance of diesel engine,and all of the numerical simulation are completed by FLUENT software.Caliber reducing rate,clearance height,convex platform height,injection cone angle are choose as optimization variables,average temperature,average pressure,soot and NOxemissions are theoptimization objectives.After achieving multi-objective optimization,two data mining methods,analysis of variance(ANOVA) and self-organizing feature map(SOM) are used to qualitatively and quantitatively analyze the design variables and objective functions,and reveals the influence of design variables to design objectives.The results show that a good dynamic performance demands smaller convex platform height and caliber reducing rate;less soot emissions need larger clearance,and lower emissions of nitrogen oxides needs smaller convex platform height and smaller caliber reducing rate.
This research uses hybrid multi-objective Gradient exploration algorithm(HMGE) which is based on Kriging surrogate model to optimize and analysis combustion process and performance of diesel engine,and all of the numerical simulation are completed by FLUENT software.Caliber reducing rate,clearance height,convex platform height,injection cone angle are choose as optimization variables,average temperature,average pressure,soot and NOxemissions are theoptimization objectives.After achieving multi-objective optimization,two data mining methods,analysis of variance(ANOVA) and self-organizing feature map(SOM) are used to qualitatively and quantitatively analyze the design variables and objective functions,and reveals the influence of design variables to design objectives.The results show that a good dynamic performance demands smaller convex platform height and caliber reducing rate;less soot emissions need larger clearance,and lower emissions of nitrogen oxides needs smaller convex platform height and smaller caliber reducing rate.
引文
[1]王欣.4D24柴油机燃烧过程的多维数值模拟[D].南昌:南昌大学,2013.
[2]胡林峰.柴油机对燃油喷射系统的要求和喷油系统的发展趋势[J].现代车用动力,2002(4):1-4.
[3]DUAN Menghua,LAI Chenguang.Design space exploration on Combustion Chamber international[J].Conference on Flow Dynamics,2015(11):701-702.
[4]袁方恩,林学东,田维,等.缩口燃烧室中气流特性与燃油喷雾匹配对柴油机燃烧及排放的影响[J].吉林大学学报(工学版),2011(5):629-634.
[5]邹林君.基于Kriging模型的全局优化方法研究[D].武汉:华中科技大学,2011:16-17.
[6]KAVEH R,NAGARAJAN G.Optimization of Diesel Engine Operating Parameters Using a Response Surface Method[J].SAE Technical Paper,2010(1):1262.
[7]赖晨光,陈小雄.基于遗传算法某汽车外形空气动力学优化[J].重庆理工大学学报(自然科学),2016,30(4):1-5.
[8]GUO Z,ZHOU Z,SONG L,et al.Aerodynamic Analysis and Multi-Objective Optimization Design of a High Pressure Ratio Centrifugal Impeller[C]//2014:V02DT42A013,ASME Turbo Expo 2014.2014.
[9]赖晨光,陆茂桂,张海林.基于数据挖掘的皮卡货车导流板减阻研究[J].重庆理工大学学报(自然科学),2017,31(2):1-6.
[10]黄解军,潘和平,万幼川.数据挖掘技术的应用研究[J].计算机工程与应用,2003,39(2):45-48.
[11]张春华,王阳.数据挖掘技术、应用及发展趋势[J].现代情报,2003,23(4):47-48.
[12]郭慧东,王玮,夏明超.基于数据挖掘的风电机组变桨系统劣化状态在线辨识方法[J].中国电机工程学报,2016,36(9):2389-2397.
[13]李蔚,俞芸萝,盛德仁,等.基于动态数据挖掘的热力参数传感器故障诊断[J].振动、测试与诊断,2016,36(4):694-699.
[14]王文正,郑鹍鹏,陈功,等.数据挖掘技术在飞行试验数据分析和气动参数辨识中的应用研究[J].空气动力学学报,2016,34(6).
[15]SUMEET P,NADER F.Self Organizing Maps(SOM)for Design Selection in Multi-Objective Optimization Using mode FRONTIER[J].SAE,2008(1):1-10.
[2]胡林峰.柴油机对燃油喷射系统的要求和喷油系统的发展趋势[J].现代车用动力,2002(4):1-4.
[3]DUAN Menghua,LAI Chenguang.Design space exploration on Combustion Chamber international[J].Conference on Flow Dynamics,2015(11):701-702.
[4]袁方恩,林学东,田维,等.缩口燃烧室中气流特性与燃油喷雾匹配对柴油机燃烧及排放的影响[J].吉林大学学报(工学版),2011(5):629-634.
[5]邹林君.基于Kriging模型的全局优化方法研究[D].武汉:华中科技大学,2011:16-17.
[6]KAVEH R,NAGARAJAN G.Optimization of Diesel Engine Operating Parameters Using a Response Surface Method[J].SAE Technical Paper,2010(1):1262.
[7]赖晨光,陈小雄.基于遗传算法某汽车外形空气动力学优化[J].重庆理工大学学报(自然科学),2016,30(4):1-5.
[8]GUO Z,ZHOU Z,SONG L,et al.Aerodynamic Analysis and Multi-Objective Optimization Design of a High Pressure Ratio Centrifugal Impeller[C]//2014:V02DT42A013,ASME Turbo Expo 2014.2014.
[9]赖晨光,陆茂桂,张海林.基于数据挖掘的皮卡货车导流板减阻研究[J].重庆理工大学学报(自然科学),2017,31(2):1-6.
[10]黄解军,潘和平,万幼川.数据挖掘技术的应用研究[J].计算机工程与应用,2003,39(2):45-48.
[11]张春华,王阳.数据挖掘技术、应用及发展趋势[J].现代情报,2003,23(4):47-48.
[12]郭慧东,王玮,夏明超.基于数据挖掘的风电机组变桨系统劣化状态在线辨识方法[J].中国电机工程学报,2016,36(9):2389-2397.
[13]李蔚,俞芸萝,盛德仁,等.基于动态数据挖掘的热力参数传感器故障诊断[J].振动、测试与诊断,2016,36(4):694-699.
[14]王文正,郑鹍鹏,陈功,等.数据挖掘技术在飞行试验数据分析和气动参数辨识中的应用研究[J].空气动力学学报,2016,34(6).
[15]SUMEET P,NADER F.Self Organizing Maps(SOM)for Design Selection in Multi-Objective Optimization Using mode FRONTIER[J].SAE,2008(1):1-10.