Optical phased arrays, as an advanced beam control technology, have wide applications in fields such as optical communication, imaging systems, and radar systems. However, due to hardware errors and environmental interference, phase errors in optical phased array systems can lead to beam pointing deviation, increased spot divergence angle, and higher sidelobe energy, thereby affecting system performance. To address this issue, this study proposes a phase error compensation method based on Particle Swarm Optimization (PSO), optimizing the phase distribution of the array to improve beam focus and reduce sidelobe energy. To verify the performance of the algorithm, a one-dimensional 128-channel optical phased array simulation model is constructed, and beam optimization under different spot size constraints is simulated and compared with the gradient descent method. The experimental results show that the PSO algorithm effectively improves the main lobe energy concentration and suppresses sidelobe energy. Compared with traditional optimization methods, PSO demonstrates stronger adaptability in high-precision requirements and complex array scales.