本文发展并完善了一种基于非受控解的排序、基于共享机制的小生境技术和交配约束的多目标遗传算法，在一元热力设计基础上，对一个透平级取实际压力与给定压力的差值和级的作功能力为目标函数进行了优化。算例表明该算法具有同时捕获一组解的能力，对于解决多目标优化问题具有很强的适应性和合理的优化结果，是一种有效的多目标优化方法
In this paper, a multiobjective genetic algorithm was developed onthe basis of three new ideas including non-dominated specialty, sharing mechanismof niche, and mating restriction. Multiobjective optimization for a turbinestage was performed based on one-dimensionalthermodynamic preliminary design and the objective functions were the pressuredifference between the predictive and objective ones and the stage work. Thecalculation indicates that the algorithm has a potential to obtain one setof solution with a strong suitability and rationalresults, and it is an effective method for multiobjective optimization
参考文献
| [1] | C M Fleming P J .An Overview of Evolutionary Algorithmsin Multiobjetive Optimization[J].Evolutionary Computation,1995,3(01):1-16. |
| [2] | 童彤,丰镇平,李军.遗传算法在透平叶栅多目标优化设计中的应用[J].中国电机工程学报,1999(06):74-76,88. |
| [3] | 童彤 .透平级的多目标优化设计方法及其专家系统的研究[D].西安交通大学,1999. |
| [4] | J D Multiple.Objective Optimization with VectorEvaluated Genetic Algorithms[A].Hillsdale, NJ: Lawrence Erlbaum,1985:93-100. |
| [5] | N Deb K .Multiobjective Optimization Using Nondominated Sortingin Genetic Algorithms[J].Evolutionary Computation,1995,2(03):221-248. |
- 下载量()
- 访问量()
- 您的评分：
-
10%
-
20%
-
30%
-
40%
-
50%