UAV Trajectory

该数据集包含了500个无人机生成的轨迹，每个轨迹包含201个航点，旨在为无人机路径预测和避障算法的开发提供强有力的支持。数据集被划分为70%的训练集、15%的验证集以及15%的测试集，并采用了包括均方误差（MSE）和对称平均绝对百分比误差（SMAPE）在内的多种损失函数来评估轨迹预测的效果。规模上，该数据集涉及500个无人机，每个无人机有201个航点。任务重点在于无人机轨迹预测和避障。

This dataset contains 500 drone-generated trajectories, each of which consists of 201 waypoints, aiming to provide robust support for the development of drone path prediction and obstacle avoidance algorithms. The dataset is divided into 70% training set, 15% validation set and 15% test set, and multiple loss functions including Mean Squared Error (MSE) and Symmetric Mean Absolute Percentage Error (SMAPE) are employed to evaluate the performance of trajectory prediction. In terms of scale, the dataset involves 500 drones, with each drone having 201 waypoints. The core tasks focus on drone trajectory prediction and obstacle avoidance.

搜集汇总

背景与挑战

背景概述

UAV Trajectory数据集是一个专注于无人机配送研究的模拟数据集，包含6911次在不同速度、高度和风力条件下的无人机轨迹数据。数据集通过Open Air Traffic Simulator生成，并提供了详细的轨迹属性如经纬度、高度、飞行距离等。