摘要 :
An improved optimization strategy to the Multidisciplinary design optimization of turbine stage was presented based on Bi-Level Integrated Synthesis System 2000(BLISS2000)procedure with Variable Complexity Modeling(VCM).The strate...
展开
An improved optimization strategy to the Multidisciplinary design optimization of turbine stage was presented based on Bi-Level Integrated Synthesis System 2000(BLISS2000)procedure with Variable Complexity Modeling(VCM).The strategy keeps the basic algorithm structure of BLISS2000,cancel the process to generate approximation models,and use VCM which is improved by the two point scale function and the new updating technology in the aerodynamic analysis.For the sake of parallel computing and load balancing,the system was decomposed by structure and parallel computing was used.A hybrid algorithm Genetic Algorithm combined with Sequential Quadratic Programming method was applied to the subsystem level optimization.The approach was evaluated and validated by a turbine blade optimization problem.Multi-Disciplinary Feasible(MDF)method is used as a standard to compare the methods.The BLISS/VCM results were similar as those in MDF,but the optimization efficiency increased 5.3 times.
收起
摘要 :
An improved optimization strategy to the Multidisciplinary design optimization of turbine stage was presented based on Bi-Level Integrated Synthesis System 2000(BLISS2000)procedure with Variable Complexity Modeling(VCM).The strate...
展开
An improved optimization strategy to the Multidisciplinary design optimization of turbine stage was presented based on Bi-Level Integrated Synthesis System 2000(BLISS2000)procedure with Variable Complexity Modeling(VCM).The strategy keeps the basic algorithm structure of BLISS2000,cancel the process to generate approximation models,and use VCM which is improved by the two point scale function and the new updating technology in the aerodynamic analysis.For the sake of parallel computing and load balancing,the system was decomposed by structure and parallel computing was used.A hybrid algorithm Genetic Algorithm combined with Sequential Quadratic Programming method was applied to the subsystem level optimization.The approach was evaluated and validated by a turbine blade optimization problem.Multi-Disciplinary Feasible(MDF)method is used as a standard to compare the methods.The BLISS/VCM results were similar as those in MDF,but the optimization efficiency increased 5.3 times.
收起