随着低空经济的快速发展，城市低空空域飞行器数量急剧增加，空飘气球等无意识空飘物与无人机的碰撞风险日益凸显。针对该问题，本文构建了一套完整的无人机与空飘气球碰撞概率评估模型，综合考虑了城市低空风场特性、空飘气球动力学模型与运动学特征以及von Kármán湍流扰动等因素。首先通过对实测风场数据的多模型拟合对比，确定了风速幂律模型与风向线性模型的组合作为建立平均风场的最优表达形式；基于流体力学原理建立了包含浮力、附加质量效应和空气阻力等主要作用力的空飘气球动力学方程；并采用von Kármán能谱和附加垂直阵风描述低空大气湍流特性；在此基础上，提出了基于蒙特卡洛方法的碰撞概率评估框架。并通过大规模随机仿真给出了碰撞概率的点估计、标准误差与置信区间等统计指标。在基准场景下，碰撞概率的点估计值为8.58%，95%置信区间为[8.03%，9.13%]，标准误差为0.28%，结果表明该模型具有较高的精度与可信度。进一步的参数敏感性分析表明：湍流强度的增强可显著降低碰撞风险；无人机半径与碰撞概率呈强正相关；而气球载荷与碰撞概率呈显著负相关。综上，本文所建评估框架可为无人机在含空飘气球等空飘物环境下的自主避撞与路径规划提供了可量化依据，同时可为城市低空空域针对空飘物的风险阈值设定与安全管控策略提供数据支撑与方法参考。

With the rapid development of low-altitude economy, the number of aerial vehicles in urban low-altitude airspace has increased dramatically, and the risk of collisions between unconscious aerial objects such as free-floating balloons and unmanned aerial vehicles (UAVs) has become increasingly prominent. To address this issue, this paper constructs a comprehensive collision probability assessment model for UAVs and free-floating balloons, taking into account the characteristics of urban low-altitude wind fields, the dynamic and kinematic models of free-floating balloons, and factors such as von Kármán turbulence disturbances. Firstly, through a multi-model fitting comparison of measured wind field data, the combination of the wind speed power law model and the wind direction linear model was determined as the optimal expression for establishing the average wind field. Based on the principles of fluid mechanics, a dynamic equation for free-floating balloons was established, incorporating major forces such as buoyancy, added mass effect, and air resistance. The von Kármán energy spectrum and additional vertical gusts were used to describe the characteristics of low-altitude atmospheric turbulence. On this basis, a collision probability assessment framework based on the Monte Carlo method was proposed. Through large-scale random simulations, statistical indicators such as point estimates, standard errors, and confidence intervals for collision probability were provided. In the benchmark scenario, the point estimate of collision probability was 8.58%, with a 95% confidence interval of [8.03%, 9.13%] and a standard error of 0.28%, indicating that the model has high accuracy and reliability. Further parameter sensitivity analysis showed that increased turbulence intensity significantly reduces collision risk; UAV radius is strongly positively correlated with collision probability; while balloon payload is significantly negatively correlated with collision probability. In summary, the assessment framework established in this paper provides a quantifiable basis for autonomous collision avoidance and path planning of UAVs in environments containing free-floating balloons and other aerial objects, and also offers data support and methodological references for setting risk thresholds and safety control strategies for aerial objects in urban low-altitude airspace.