Transport modelling: macro and micro simulation for the studied case of Funchal

The work done in this thesis attempts to demonstrate the importance of using models that can predict and represent the mobility of our society. To answer the proposed challenges two models were examined, the first corresponds to macro simulation with the intention of finding a solution to the frequency of the bus company Horários do Funchal, responsible for transport in the city of Funchal, and some surrounding areas. Where based on a simplified model of the city it was possible to increase the frequency of journeys getting an overall reduction in costs. The second model concerns the micro simulation of Avenida do Mar, where currently is being built a new roundabout (Praça da Autonomia), which connects with this avenue. Therefore it was proposed to study the impact on local traffic, and the implementation of new traffic lights for this purpose. Four possible situations in which was seen the possibility of increasing the number of lanes on the roundabout or the insertion of a bus lane were created. The results showed that having a roundabout with three lanes running is the best option because the waiting queues are minimal, and at environmental level this model will project fewer pollutants. Thus, this thesis presents two possible methods of urban planning. Transport modelling is an area that is under constant development, the global goal is to encourage more and more the use of these models, and as such it is important to have more people to devote themselves to studying new ways of addressing current problems, so that we can have more accurate models and increasing their credibility.

The combination of neural estimates in prediction and decision problems

In this dissertation, different ways of combining neural predictive models or neural-based forecasts are discussed. The proposed approaches consider mostly Gaussian radial basis function networks, which can be efficiently identified and estimated through recursive/adaptive methods. Two different ways of combining are explored to get a final estimate – model mixing and model synthesis –, with the aim of obtaining improvements both in terms of efficiency and effectiveness. In the context of model mixing, the usual framework for linearly combining estimates from different models is extended, to deal with the case where the forecast errors from those models are correlated. In the context of model synthesis, and to address the problems raised by heavily nonstationary time series, we propose hybrid dynamic models for more advanced time series forecasting, composed of a dynamic trend regressive model (or, even, a dynamic harmonic regressive model), and a Gaussian radial basis function network. Additionally, using the model mixing procedure, two approaches for decision-making from forecasting models are discussed and compared: either inferring decisions from combined predictive estimates, or combining prescriptive solutions derived from different forecasting models. Finally, the application of some of the models and methods proposed previously is illustrated with two case studies, based on time series from finance and from tourism.