| Summary: | The process of planning and scheduling the flights of an airline consists of several steps, some of which are prepared several months in advance. Even though, having a great plan is as important as keeping it, this task can be quite demanding due to unexpected events (disruptions) that can occur close to the day of operation. Such problems can lead to delays and / or cancellation of flights, if nothing is done to prevent it. In the Laboratório de Inteligência Artificial e Ciência da Computação (LIACC) is being developed a project called Multi-Agent System for Disruption Management (MASDIMA), in collabo- ration with TAP Portugal, as part of an automatic negotiation on Cooperative Distributed Problem Solving (CDPS), applied to the scenario of Airlines Operation Control Center (AOCC) through a multi-agent system for disruption management. The aim of this dissertation is to incorporate, in MASDIMA system, an additional software layer, on the set of agents responsible by the generation, analysis and decision regarding new solutions, so they can learn from the past. Therefore, it is being investigated a way of having the system solve current problems based on its knowledge of similar situations occurred in the past and already solved. In order for this to become a reality, it will be used Case-based Reasoning (CBR). Using this methodology, we will be able to resolve problems, learning from the past, on the AOCC. The intention in this dissertation is to show that the introduction of learning from the past in MASDIMA system will maintain the quality of the solutions presented, and, at the same time, decrease the average response time of the system to a new problem and increase its trust. The achievement of these objectives will be analyzed using metrics such as the average response time of the system to a new case, and determining the quality of the proposed solutions, i.e. comparing the results to previously produced solutions by the system MASDIMA and TAP actual AOC.
|
|---|