Optimizing telescope alignment configuration using finite element methods and machine learning

Telescope image quality was impacted by mirror misalignment caused by gravitational sag at various altitude angles, leading to surface deformation. Traditional methods of mirror alignment were inefficient. This study proposed an optimization framework for the whiffle tree support mechanism to determine the optimal bolt distances and altitude angles, minimizing Surface Form Deviation. A finite element analysis (FEA) simulation modeled mirror deformation under various support displacements and angles. Latin Hypercube Sampling (LHS) generated 2,450 design points to calculate surface deformation. Machine learning and metaheuristic algorithms were used to optimize the adjustments. Results showed significant reduction in optical Surface Form Deviation post-optimization, with the algorithm closely matching the theoretical design. FEA validation confirmed that optimized support adjustments reduced optical distortions, demonstrating the effectiveness of this integrated approach for minimizing Surface Form Deviation.