This paper introduces an innovative joint optimization scheme to improve the energy efficiency and task completion rate of UAVs when performing tasks, especially in terms of communication service quality and perception error constraints. The scheme integrates communication and perception modules of independent frequency bands and adopts state switching optimization and multi-objective resource allocation technology to meet all user needs. The system uses the improved Hungarian algorithm to achieve efficient task allocation, combines the A* algorithm with model predictive control (MPC) for dynamic path planning, and uses sequential quadratic programming (SQP) to optimize the hovering height and transmission power of each UAV to ensure the overall optimal performance. Experimental results show that compared with traditional methods, this scheme not only improves the total communication rate, but also performs well in multiple key indicators, proving the effectiveness of cluster coordination mechanism in overcoming the problems of single-machine field of view angle limitation, power constraint and collision avoidance. This provides new ideas and technical support for the application of UAVs in complex environments.