Optimizes micro-turbine blade designs for improved performance in engineering applications, suggesting innovative design solutions.
Key Points
The aim is to develop a framework for the design and optimization of micro-turbine blade profiles to enhance performance while considering fabrication constraints.
Parameterization of blade morphology using 22 design variables.
Use of a physics-informed neural network surrogate model.
Implementation of a two-stage active learning strategy combining KD-tree exploration and residual-based sampling.
Multi-objective optimization via Non-dominated Sorting Genetic Algorithm II.
Achieved a 38.6% increase in rotational speed of the optimized blade.
Retained 75.1% of thrust at an inlet pressure of 0.2 MPa.
Validated the optimization framework's effectiveness through experimental results.