We present a novel method for optimizing surface modulation structures of solid targets to enhance ion beam yield via the Target Normal Sheath Acceleration (TNSA) mechanism. For the initially set conical target, we used a random walk sampling algorithm to optimize the microstructure on the conical surface. With only 36 simulations, we achieved the optimized structure, significantly improving the temperature and yield of hot electrons, and forming a more robust longitudinal sheath electric field. Two-dimensional Particle-In-Cell (PIC) simulations demonstrated that the optimized irregular modulation structure achieved an ion beam yield of 34% higher than that of the initial cone structure, and a fivefold enhancement compared to planar targets. This study validates the effectiveness of random walk sampling in multi-parameter optimization for laser-driven ion acceleration, offering a effective approach for designing high performance targets in experimental applications.

Laser-driven ion acceleration,Random Walk