Particle swarm optimization (PSO), a classic swarm intelligence algorithm with excellent performance in solving various optimization problems, has widespread applications in UAV path planning. However, PSO often suffers from slow convergence. To address the slow convergence and local optima issues of traditional PSO, a hybrid approach that combines deep deterministic policy gradient algorithms with PSO is proposed, which uses neural networks to assess the particle swarm’s state. This led to the development of the Fusion of DDPG and the Particle Swarm Optimization (FDPSO) algorithm. To evaluate the effectiveness of the proposed FDPSO algorithm, three experiments were carried out within a Python simulation environment, evaluating the convergence speed and flight cost for both individual and multiple UAVs across various obstacle environments. The results indicate that the FDPSO algorithm achieves faster convergence and lower flight cost than to traditional PSO dose.