The existence of significant uncertainties in the models and systems required for trajectory prediction represents a major challenge for the Air traffic Management (ATM) system. Weather can be considered as one of the most relevant sources of uncertainty. Understanding and managing the impact of these uncertainties is necessary to increase the predictability of the ATM system. State-of-the-art probabilistic forecasts from Ensemble Prediction Systems are employed to characterize uncertainty in the wind and potential convective areas. A robust optimal control methodology to produce efficient and predictable aircraft trajectories in the presence of these uncertainties is presented. Aircraft motion is assumed to be at a constant altitude and variable speed, considering BADA4 as the aircraft performance model. A set of Pareto-optimal trajectories is obtained for different preferences among predictability, convective risk, and average cost index running a thorough parametric study on a North Atlantic crossing use case. Results show that the cost of reducing the arrival time window by 10 s. is between 100 and 200 kg or 3 and 6 min., depending on the cost-index. They also show that reducing the exposure to convection by 50 km is on the order of 5 and 10 min. or 100 and 200 kg. of average fuel consumption.

轨迹预测所需的模型和系统中存在重大不确定性，这对空中交通管理（ATM）系统构成了重大挑战。天气可以被认为是最不确定的因素之一。了解和管理这些不确定性的影响对于提高ATM系统的可预测性是必要的。Ensemble Prediction Systems的最新概率预报用于表征风和潜在对流区域的不确定性。提出了一种鲁棒的最优控制方法，可以在存在这些不确定性的情况下产生有效且可预测的飞机轨迹。将BADA4作为飞机性能模型，假定飞机运动处于恒定高度和可变速度。针对可预测性，对流风险和平均成本指数之间的不同偏好，获得了一组帕累托最优轨迹，并对北大西洋跨越用例进行了全面的参数研究。结果表明，将到达时间窗口减少10 s的成本。介于100到200公斤或3到6分钟之间，具体取决于成本指标。他们还表明，将对流暴露减少50 km大约需要5到10分钟。或100和200公斤。平均油耗。他们还表明，将对流暴露减少50 km大约需要5到10分钟。或100和200公斤。平均油耗。他们还表明，将对流暴露减少50 km大约需要5到10分钟。或100和200公斤。平均油耗。