Boeing737-300/400机队可靠性等级预测研究

为了研究航空公司机队可靠性等级预测问题,提出一种基于灰色关联与支持向量机(SVM)的航空公司机队可靠性等级预测模型.从某航空公司B737-300/400机队的可靠性月报中,得到影响机队可靠性等级的9个指标;应用灰色关联法对机队可靠性等级和指标进行相关性分析,就各个指标与机队可靠性等级之间的关联度进行了排序;采用基于径向基核函数(RBF)的支持向量机(SVM)建立了航空公司机队可靠性等级预测模型.以某航空公司B737-300/400机队为例进行了可靠性等级预测分析,验证了所提方法的有效性.     

基于BP神经网络误差函数优化算法的武器装备承制商信用评价模型研究

信用评价是选择武器装备承制商的重要手段.以国标为基础,结合承制商具体情况确定了信用评价指标体系.分析了传统信用评价方法的不足,对经典BP神经网络的误差函数进行优化,优化后的网络模型收敛速度更快,预测精度更高.构建BP神经网络武器装备承制商信用评价模型,仿真实验表明武器装备承制商信用评价可以选用BP神经网络模型.     

基于BP神经网络的企业信用评估模型

本文研究了企业信用评估中的模型问题.以商业企业为例,阐述了基于BP神经网络的信用评估模型的原理,通过建立指标体系,讨论基于BP神经网络的评价模型的实现,对模型的不足进行了分析,并提出改进建议.     

基于PSO-BP神经网络的外贸发展状况的评价

以我国对外贸易的实际情况为基础,建立了评价外贸发展状况的一套指标体系.使用主成分分析方法对所有评价指标进行预处理之后,借助粒子群优化算法改进了BP神经网络算法,构建了外贸发展状况的评价模型.评价模型在具体应用中,评价结果较为准确,具有一定的科学性和合理性.     

基于AHP和动量BP神经网络的工程项目承包商选择模型

运用BP神经网络技术,采用动量BP算法,构建了基于动量BP神经网络的工程项目承包商选择模型,并将AHP的评价结果作为学习样本,对BP神经网络模型进行训练和测试.结果表明,基于AHP和动量BP神经网络的工程项目承包商选择模型是可行的,该模型具有较高的自组织、自适应和自学习能力以及较强的容错功能,能够为一般的工程项目承包商选择活动提供有效的参考和依据.     

基于签派可靠度的航空公司运行品质评价方法研究

签派可靠度是反映航空公司运行品质和运营效率的顶层参数,以签派可靠度为顶层指标,根据航空公司实际运行环境,结合航空公司各主要业务部门的运行数据,建立签派可靠度评估模型,形成影响签派可靠度的二级指标体系.利用区间数层次分析法和可能度法实现各二级指标的综合权重计算和分级排序.对航空公司运行品质影响较大的二级指标,依次为日利用率,延误率,报告故障率.以国内某个航空公司为例,根据底层指标得分情况,利用幂指数法,确定签派可靠度的综合得分,最终实现该航空公司运行品质评估.研究结果表明模型能够对航空公司运行品质做出综合评价,便于发现运行决策中的薄弱环节,重点解决维修可靠性问题,可以提升航空公司的运行品质.     

基于航班时空网络模型的网络型机队随机规划方法

为了解决传统基于航班机型分配机队规划法无法准确反映枢纽航线网络下网络效应随机特性问题,将旅客收益网络优化模型合并入航班时空网络数学模型中构建航空公司网络型机队规划方法.在模拟航空公司枢纽网络运营环境(航线网络、航班计划、旅客需求与平均票价水平等)基础上,构建航班时空网络模型并划分行程结构需求情景集,以各机型飞机数为自由变量,将任一需求情景下机型飞机数等于相应自由变量视为非预期约束,构建以机队期望运营利润最大化为目标函数的网络型机队规划随机情景模型,并设计以"完成所需飞机数最少"为第一阶段,以"情景汇聚算法优化机队结构"为第二阶段的两阶段算法求解该模型.算例仿真结果表明,与基于航班机型分配的收益摊销型机队规划法相比较,网络型机队规划法的机队期望运营利润增加了17.1%;与"确定性"需求下的机队规划法相比较,网络型机队规划法的机队期望运营利润增加了1.7%.结果表明,设计的机队规划法能够更好的适应枢纽航线网络运营环境,且结果稳定性更高.     

V-fold交叉验证和BP神经网络在信用评价中的应用

研究关于公司神经网络信用评估问题的现状，提出一套甄选方法准则。用于建立适合于我国企业的信用评分指标体系；然后依据该指标体系建立了基于BP回归神经网络的信用评估模型；采用V—fold交叉验证技术，利用样本公司实际指标数据对该模型的评分效果进行了实证研究。     

基于智能控制的高速线材轧机水冷控制系统优化

针对武汉钢铁集团公司大型轧钢厂当前在高速线材生产线中存在的水冷控制系统可靠性差,轧线温度波动范围大等问题,应用智能计算理论及方法对上述工业控制系统进行系统辨识、建模以及优化.分析比较了基于梯度下降搜索BP算法、径向基函数网络、Levenberg Marquardt BP算法的前馈神经网络对SMS水冷系统的逼近精度、训练速度.研究了采用Levenberg-Marquardt BP算法的前馈神经网络在样本集和测试集上的表现,建立了基于Levenberg-Marquardt BP算法的前馈神经网络水冷控制系统模型.解决了高速线材水冷控制系统可靠性,温度控制精度问题.     

BP神经网络在上海住宅市场需求预测中的应用

人工神经网络是近期发展最快的人工智能领域研究成果之一 ,本文在介绍 BP神经网络的有关原理的基础上 ,建立了一个上海住宅市场的 BP神经网络模型 ,并通过该模型对上海住宅市场的需求进行了预测和分析 .分析结果表明人工神经网络方法在住宅市场需求预测中的应用是可行的并且是有效的 .     

Optimal dispatching strategy on an airline network after a schedule perturbation

It often happens that one or more aeroplanes from an airline fleet are taken out of operation for technical reasons and the airline has to operate on the existing network with a reduced number of planes. This paper presents the results of an effort to define a new ad hoc schedule for this situation, so that the total passenger delay on an airline network is minimized. A network is formed, in which nodes represent flights on a given airline network, and arcs are the total time losses on individual flights. The problem of determining a new routing and scheduling plan for the airline fleet is solved by branch and-bound methods. A numerical example illustrates the efficiency of the model.     

Cost evaluation of airline maintenance investigation-triggering methods

A reliable investigation-triggering mechanism is critical for airlines in managing aircraft fleet reliability. This study developed a simulation model to evaluate the performance of various investigation-triggering mechanisms in various scenarios of aircraft fleet reliability. Factors that affect performance include flight delay costs, investigation costs, the costs of corrective actions, the accuracy of investigations, the effectiveness of corrective actions, and the monthly improvement rate. Nine investigation-triggering mechanisms were evaluated. The simulation results show that the monthly improvement rate significantly influences performance. The subject airline’s original investigation-triggering mechanism is not outstanding in most scenarios. This study recommends two of the nine investigation-triggering mechanisms for airlines when they select investigation-triggering mechanisms in various scenarios. Thus, useful guidance for adjusting airlines’ investigation-triggering mechanisms can be effectively provided.     

Operational aircraft maintenance routing problem with remaining time consideration

The aircraft maintenance routing problem is one of the most studied problems in the airline industry. Most of the studies focus on finding a unique rotation that will be repeated by each aircraft in the fleet with a certain lag. In practice, using a single rotation for the entire fleet is not applicable due to stochasticity and operational considerations in the airline industry. In this study, our aim is to develop a fast responsive methodology which provides maintenance feasible routes for each aircraft in the fleet over a weekly planning horizon with the objective of maximizing utilization of the total remaining flying time of fleet. For this purpose, we formulate an integer linear programming (ILP) model by modifying the connection network representation. The proposed model is solved by using branch-and-bound under different priority settings for variables to branch on. A heuristic method based on compressed annealing is applied to the same problem and a comparison of exact and heuristic methods are provided. The model and the heuristic method are extended to incorporate maintenance capacity constraints. Additionally, a rolling horizon based procedure is proposed to update the existing routes when some of the maintenance decisions are already fixed.     

A model for designing the meteorologically most reliable airline schedule

Due to meteorological conditions certain airports in some regions have to be frequently closed during winter months. An immediate consequence is an increase in the number of cancelled flights, which is a disruption of airline schedules on the overall transportation network. In this paper a research concerning the reliability of airline scheduling as related to meteorological conditions is conducted and an indicator for quantifying the adaptability of airline schedules to meteorological conditions is proposed. A heuristic algorithm for minimizing the number of needed aircraft for given traffic volume is also presented. In case where more than one solution with the same number of engaged aircraft is possible the solution chosen is the one with the minimum number of passengers whose flights are expected to be cancelled on account of meteorological conditions. The proposed algorithm is illustrated by an appropriate numerical example.     

The impact of fleet size on performance-based incentive management

Performance-based contracting (PBC) is envisioned to lower the asset ownership cost while ensuring desired system performance. System availability, widely used as a performance metric in such contracts, is affected by multiple factors such as equipment reliability, spares stock, fleet size, and service capacity. Prior studies have either focussed on ensuring parts availability or advocating the reliability allocation during design. This paper investigates a single echelon repairable inventory model in PBC. We focus on reliability improvement and its interaction with decisions affecting service time, taking into account the operating fleet size. The study shows that component reliability in a repairable inventory system is a function of the operating fleet size and service rate. A principal-agent model is further developed to evaluate the impact of the fleet size on the incentive mechanism design. The numerical study confirms that the fleet size plays a critical role in determining the penalty and cost sharing rates when the number of backorders is used as the negative incentive scheme.     

An Iteration Method for Reliability Evaluation of a Fleet System

The fleet system considered here consists of n identical units (members) which are operated together under an operational program which specifies the schedule of the operation of the system and maintenance performed on the units in the system. Introduction of this operational program makes the reliability evaluation of the fleet system more realistic but complicated. A conventional Markov approach is effective only when a fleet system consists of a few units. This paper presents a new method for the reliability evaluation of a fleet system. This method reduces the problem of the reliability evaluation of the system to that of each unit by introducing a ‘utilization factor’. Therefore, the size of the problem is irrelevant to the number of the units in the system. An iteration method is used to obtain the unique solution of the problem.     

An overview of the issues in the airline industry and the role of optimization models and algorithms

Optimization application has revolutionized the airline industry in all phases of the planning process. One of the current issues facing the airline industry is planning under uncertainty, especially in the context of schedule disruptions. We discuss the robust models and solution algorithms that have been proposed and developed to handle the uncertain parameters. We show that stochastic programming (SP) provides an ideal paradigm for capturing the uncertainties and making robust decisions. We develop and investigate a prototype fleet assignment model formulated as a two-stage SP with recourse.     

Multifleet routing and multistop flight scheduling for schedule perturbation

Efficient and effective incidental scheduling techniques for schedule perturbation are essential to an airline carrier's operations. This research aims at developing a framework to assist carriers in fleet routing and flight scheduling for schedule perturbations in the operations of multifleet and multistop flights. The framework is based on a basic multifleet schedule perturbation model constructed as a timespace network from which strategic models are developed to research incidental scheduling. These network models are formulated as multiple commodity network flow problems. Lagrangian relaxation with subgradient methods accompanied by the network simplex method, a Lagrangian heuristic and a modified subgradient method are developed to solve the problems. A case study regarding the international operations of a major Taiwan airline carrier is presented.     

A general system theory approach to rail freight car fleet sizing

The transportation of goods from shippers to consignees is a railroad's major activity. Rail freight cars are enormously expensive and a rail vehicle fleet represents one of the largest capital resources of most railroads. Resource allocation to rail freight cars is an extraordinary complex managerial problem. This paper describes the determination of an optimal number of rail freight cars so as to satisfy the demand, on one hand, and minimize the total cost, on the other. A new mathematical model relying on optimal control theory is developed. The problem is formulated as the problem of finding an optimal regulator for a linear system, excited by Gaussian white noise, a quadratic performance index, and random initial conditions. The model has been tested on numerical examples.     

基于k-means和邻域粗糙集的航空客户价值分类研究

对客户价值进行分类, 识别重要价值客户, 对航空公司获利至关重要。本文提出了基于k-means和邻域粗糙集的航空客户价值分类模型。首先, 从客户的当前价值和潜在价值双视角出发, 建立了航空客户综合价值评价指标体系; 之后, 采用基于Elbow的k-means方法对航空客户进行聚类, 采用邻域粗糙集方法对决策系统进行指标约简, 根据约简后的决策系统完成客户价值初筛。评估前先使用SMOTE方法消除数据的不平衡性, 而后采用网格搜索组合分类器的方法对航空客户价值分类的效果进行评估和检验。最后, 根据评估结果对航空客户价值细分。文末, 对国内某航空公司的62988条真实客户记录进行了实证分析和验证, 其中, 潜在VIP客户群的分类准确率达到了92%, 从而为航空客户价值分类提供了一种新思路。