Abstract: The Peak Side Lobe Level (PSLL), sensitivity, and directivity of a radio antenna array directly determine the performance of a system in weak signal detection, anti-interference, and resolution capabilities. In applications spanning the High Frequency (HF) to Very High Frequency (VHF) bands, uniform arrays suffer from high sidelobes and grating lobes during wide-angle scanning, while traditional sparse array optimization algorithms are prone to premature convergence and local optima trapping. To address these issues, this study proposes a Multi-strategies Enhanced Grey Wolf Optimization-Particle Swarm Optimization (MEGWO-PSO) algorithm for the sparse layout optimization of two-dimensional planar arrays. The algorithm adopts a two-stage framework of global exploration and local refinement. In the first stage, an enhanced grey wolf algorithm integrated with a double-order S-shaped attenuation convergence factor, along with adaptive crossover and mutation operators, is employed to achieve efficient global search and avert premature convergence. In the second stage, the optimized search space derived from the first stage is used as the initial input, and local refinement is performed via an improved particle swarm optimization algorithm. Simulation results demonstrate that MEGWO-PSO can effectively optimize the layout of small-to-medium-sized arrays with up to 100 elements, reducing the PSLL by 9.39 dB compared with the results reported in relevant literature. For large-scale 256-element arrays, considering the key performance metrics of radio antenna arrays, the proposed method yields a PSLL that is 13.22 dB, 21.32 dB, and 14.64 dB lower than those of the standalone GWO, PSO, and GA (Genetic Algorithm) algorithms, respectively. Meanwhile, the array’s sensitivity is enhanced by 10%, 24.8%, and 25.6% relative to the three aforementioned algorithms. Furthermore, the optimized array exhibits superior PSLL suppression across different scanning angles, with reductions of 34.79 dB and 15.63 dB compared with uniform arrays and GWO-based sparse arrays, respectively. This work provides a viable technical approach and theoretical support for the development of radio telescopes with large apertures, high performance, and wide-angle scanning capabilities.