机坪电动特种车辆运行优化调度方法研究

发布时间：2018-03-28 17:46

本文选题：电动特种车辆　切入点：决策——规划法　出处：《中国民航大学》2017年硕士论文

【摘要】：随着我国民航事业迅速发展,机场客流量逐渐增加,停机坪特种车辆数量也在增加,特种车辆调度失误导致的航班延误的影响也逐渐增大。因此,特种车辆调度好坏至关重要。2015年3月民航局机场特种车辆“油改电”工作正式启动,部分机场开始引入电动特种车辆。可预见,未来机场电动特种车辆会逐渐普及,这将给调度问题提出新的要求和挑战。由于电动特种车辆自身结构特点,其调度方法与传统燃油特种车辆有所区别。车联网技术和车辆调度智能算法为解决这一问题提供了新的思路。本文针对机坪电动特种车辆的调度问题提出决策——规划法,该调度方法将调度问题分为两层:决策层和路径规划层。决策层通过决策算法决策出适合执行任务的车辆,将决策出的任务车辆送入路径规划层。路径规划层将任务车辆执行任务归结为n个城市的TSP问题(Travelling Salesman Problem),通过组合优化算法构建满足约束条件的优化目标函数,对决策层输出的任务车辆任务行驶路径进行路径规划,求解全局最优路径。由于ID3(Iterative Dichotomiser 3)算法原理简单易懂,建树所花时间少,本文中的决策层通过ID3算法构建了电动特种车辆任务分配模型。为了提高收敛到全局最优解的概率和优化求解速度,本文提出一种改进的Hopfield神经网络算法——全反馈Hopfield神经网络(FFCHNN),通过分析其仿真结果证明该网络求解TSP问题时比传统Hopfield神经网络的性能要好,效率更高。本文采用该算法作为路径规划层对任务车辆进行路径规划的优化组合算法。最后,通过决策——规划法解决电动工具车调度实例,通过验证表明该方法能够较为快速精准决策适合执行任务的车辆,并快速获得出任务车辆最短路径,为解决机场电动特种车辆调度问题提供参考。
[Abstract]:With the rapid development of China's civil aviation industry, the airport passenger flow is increasing gradually, the number of special vehicles on the apron is also increasing, and the impact of flight delay caused by special vehicle scheduling errors is gradually increasing. Special vehicle scheduling is of great importance. In March 2015, the civil aviation bureau airport special vehicle "oil and electricity" work officially started, some airports began to introduce electric special vehicles. It is foreseeable that airport electric special vehicles will gradually become popular in the future. This will bring new requirements and challenges to the scheduling problem. Due to the structural characteristics of special electric vehicles, The method of dispatching is different from that of traditional special fuel vehicles. The technology of vehicle networking and the intelligent algorithm of vehicle scheduling provide a new way of thinking to solve this problem. In this paper, a decision-planning method is proposed for the scheduling of special electric vehicles on the aerodrome. The scheduling method divides the scheduling problem into two layers: decision layer and path planning layer. The task vehicle is put into the path planning layer. The path planning layer reduces the task execution of the mission vehicle to the TSP problem of n cities. The optimization objective function satisfying the constraint condition is constructed by combinatorial optimization algorithm. In order to solve the global optimal path, the path planning of the task vehicle path output from the decision level is carried out. Because the principle of the ID3(Iterative Dichotomiser 3 algorithm is simple and easy to understand, it takes less time to build a tree. In order to improve the probability of convergence to the global optimal solution and optimize the solution speed, the decision layer in this paper constructs the task assignment model of special electric vehicle by ID3 algorithm. In this paper, an improved Hopfield neural network algorithm, full feedback Hopfield neural network, is proposed. The simulation results show that the performance of the neural network is better than that of the traditional Hopfield neural network in solving the TSP problem. This paper uses this algorithm as the optimal combination algorithm for path planning of task vehicles. Finally, the decision-planning method is used to solve the scheduling example of electric tool vehicles. The verification results show that the proposed method can make accurate and fast decision on the vehicles that are suitable for the task and obtain the shortest path of the vehicle quickly. It provides a reference for solving the scheduling problem of special electric vehicles in the airport.
【学位授予单位】：中国民航大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：V351.35;TP183

【参考文献】