替代模型
响应面法
推力
刀(考古)
实验设计
工程类
可靠性(半导体)
径向基函数
计算机科学
控制理论(社会学)
数学优化
数学
结构工程
机械工程
人工智能
人工神经网络
机器学习
统计
功率(物理)
物理
控制(管理)
量子力学
作者
Tao Bian,Liangliang Liu,Wenhao Cai,Wen Jiang,Zhiwen Liu
标识
DOI:10.1177/14750902231188393
摘要
This paper proposes a multi-objective optimization design method for shaftless rim-driven thruster (RDT) based on the ISIGHT platform. The pitch ratio, the blade area ratio and the advance coefficient of RDT were considered as the design optimization variables, the thrust and efficiency of the RDT were the optimization objectives. The multi-objective optimization design method was based on the surrogate module and the optimization module in the ISIGHT platform. Two analytical methods (response surface methodology (RSM) and radial basis function Model (RBF)) were used to build the surrogate model. The Muti-Island GA optimization algorithm was adopted in the optimization module. The paper indicates that both RSM model and RBF model are feasible to build the surrogate model, the RBF model has better accuracy and reliability than RSM model. The blade obtained by RBF method has larger thrust and smaller torque than that of the blade obtained by RSM method. For the large advance coefficient ( J > 0.6), the efficiency of the blade obtained by RBF method is slightly higher than that of the blade obtained by RSM method.
科研通智能强力驱动
Strongly Powered by AbleSci AI