一种基于深度学习的应急救援物资航空运输决策系统

A decision-making system for air transportation of emergency relief materials based on deep learning

摘要:

在应急救援中，救援物资航空运输的决策至关重要，但传统依赖经验专家的方法存在耗时长、效率低等问题。因此，引入深度学习技术，构建一种应急救援物资航空运输决策系统。该系统根据任务需求，针对不同型号飞行器和不同运输方式生成所有可行运输方案，并对各类可影响航空运输过程的因素量化，通过使用神经网络对任务完成耗时进行计算预测，得到所有方案完成耗时的预测结果；对不同条件下，各类因素对预测精度的影响进行验证，并使用改进的神经网络提升本系统的预测精度。结果表明：本系统与传统的经验专家决策方法相比，大幅缩短了运输方案的决策时间，提高了应急救援效率，解决了传统方法的局限性，为应急救援工作提供了更多的灵活性和可靠性。

Abstract:

In emergency rescue, the decision of air transportation of relief materials is very important, but the traditional method of relying on experienced experts has some problems such as long time consumption and low efficiency. Therefore, this paper introduces deep learning technology and constructs a decision-making system for air transportation of emergency relief materials. According to the task requirements, the system generates all feasible transportation schemes for different types of aircraft and different modes of transportation, and quantifies all kinds of factors that can affect the air transportation process. By using neural network to calculate and predict the task completion time, the prediction results of the completion time of all schemes are obtained. This paper also verifies the influence of various factors on the prediction accuracy under different conditions, and further improves the prediction accuracy of this system by using the improved neural network. Compared with the traditional empirical expert decision-making method, this system greatly shortens the decision-making time of transportation scheme and improves the emergency rescue efficiency. It solves the limitations of traditional methods and provides more flexibility and reliability for emergency rescue work..