To enhance the performance of the Dynamic Window Approach (DWA) algorithm in obstacle avoidance applications for unmanned aerial vehicles (UAVs), this paper proposes a fusion and improvement strategy. By combining the DWA algorithm with the Optimal Reciprocal Collision Avoidance (ORCA) algorithm, the deficiencies of the DWA algorithm in dynamic obstacle avoidance are compensated for. Additionally, by incorporating UAV speed margin references into the speed evaluation and setting weight coefficients that vary based on the number of obstacles, the static obstacle avoidance effectiveness of the algorithm is improved. This paper first analyzes the principles of the DWA and ORCA algorithms, describes their processes, and explains the improved design of the weight coefficients for the DWA algorithm. Secondly, it summarizes the advantages and disadvantages of both algorithms and analyzes the feasibility of their fusion. Subsequently, this paper introduces the mechanism of the fused algorithm: the DWA algorithm is used to obtain feasible and optimal trajectories through the dynamic speed window, and the ORCA algorithm performs secondary obstacle avoidance optimization on the speed under these conditions and updates the UAV′s state information, forming a closed loop for continuous speed optimization. The fused DWA-ORCA algorithm combines the speed and practicality advantages of the DWA algorithm with the precision of the ORCA algorithm. Simulation analysis demonstrates that the improved fused DWA algorithm is capable of efficiently completing obstacle avoidance tasks for UAVs.