基于ASQ松弛约束的航站楼资源自适应分配优化策略研究

Operation Resource Manage Systems (ORMS) for hub airports generally employ a rule engine to schedule and allocate resources like boarding gates, checking counters and etc., based on security constraints and business contracts. Under the influences of uncertainty of flight dynamic and randomness of inbound and outbound services, there are big contradictions between usage of resource and KPIs of airport operation. Measured result of Pudong International airport shows that the unqualified rate is up to 75%..Our research tries to collect sample data related to passenger service by a variety of indoor wireless sensing technologies. Using these data, a passenger service queuing model is established with multiple external factors (such as flight schedule, aircraft types, routes, passenger ratio, individual characteristics of passengers and etc.). Based on GEP’s powerful function discovery capabilities and Boosting’s function precision push capabilities, the accuracy of passenger service queuing model can be continuously calibrated. According to KPIs of airport operation, a set of new optimized resource allocation rules is formed by the model and put into ORMS rule engine to take effect within certain time period so that ORMS obtains the feature of periodic self-optimization.

国内大型枢纽机场的资源分配系统普遍采用基于安全约束和商务契约的规则驱动技术实现值机柜台等楼内资源的分配。然而，实际运营中航班和气象等不确定因素使传统资源分配策略无法满足面向旅客服务的KPI考核指标要求，自然也不能向旅客提供高水平的机场服务质量ASQ。以上海浦东机场为例，实测结果表明，候机楼内资源分配结果在KPI上的不合格率高达75%。.针对此问题，本项目拟提出基于ASQ松弛约束的楼内资源自适应分配优化方案。通过多种室内感知技术实现旅客服务样本数据采集，结合生产数据挖掘出含多个外部因素（航班时刻/机型/航线/旅客比例）的旅客服务排队模型，利用GEP和Boosting的函数推进学习能力，不断校准排队模型精度，最终基于ASQ的机场KPI指标，结合现有分配结果共同得出优化的资源分配规则，送入规则引擎使其在本周期内生效，从而使航站楼资源分配策略具有周期性的自适应特性。

国内大型枢纽机场的资源分配系统普遍采用基于安全约束和商务契约的规则驱动技术实现值机柜台等楼内资源的分配。然而，实际运营中航班和气象等不确定因素使传统资源分配策略无法满足面向旅客服务的KPI考核指标要求，自然也不能向旅客提供高水平的机场服务质量ASQ。.针对此问题，本项目拟提出基于ASQ松弛约束的楼内资源自适应分配优化方案。从枢纽机场旅客满意度及服务质量考核体系测评技术，枢纽机场旅客服务无线定位数据优化策略研究，民航机场旅客典型服务行为数据采集等多个层面展开研究，并根据研究过程中的具体需求适当扩充部分研究内容，完成了既定的研究内容，取得了预期的研究成果，达到了项目研究预期目标。共完成研究论文 11篇，EI检索期刊3篇，EI检索会议2篇，其他文章6篇。