<p>UAV path planning is crucial for ensuring flight safety and mission success and can be formulated as a complex, high-dimensional constrained optimization problem. Traditional swarm intelligence algorithms often exhibit slow convergence, limited solution accuracy, and a loss of population diversity when addressing such problems. To overcome these limitations, this paper proposes an improved dung beetle optimization (IDBO) algorithm, which integrates multiple coordinated mechanisms to enhance the original dung beetle optimizer. Specifically, adaptive inertia weights and a nonlinear dynamic adjustment factor are incorporated to balance global exploration and local exploitation across different evolutionary stages. Additionally, a random multi-strategy boundary control mechanism is devised to enhance the update quality of boundary individuals and mitigate the risk of premature convergence. Furthermore, elite quasi-opposition-based learning and a role-switching mechanism are integrated to increase population diversity and reactivate stagnated individuals. To evaluate the proposed method's performance, IDBO is compared with 11 optimization algorithms on the CEC2017 and CEC2020 benchmark suites. Experimental results show that IDBO achieves competitive overall performance in terms of convergence speed, solution accuracy, and stability. Furthermore, in offline, static, three-dimensional UAV path-planning simulation scenarios under a predefined composite cost model, IDBO achieves favorable total path costs. These results indicate that the proposed IDBO can serve as a competitive optimizer for numerical benchmark problems and offline static 3D UAV path-planning simulations.</p>