Download files
Abstract
With the growing trend of transport network worldwide and to support financial and managerial decisions in the transport domain, especially in big cities, the development and application of transport planning model systems (TPMSs) are inevitable. The development of TPMSs has become an important topic of research in recent decades so that different generations of model systems have emerged over time.
Due to the huge inherent complexity of transport network behaviour, TPMSs should be developed properly to assure that the planning and management considerations are reliable. To predict the hypothetical situation of future which is surrounded with unknowns, an ideally developed TPMS should reproduce base year conditions and also should be sensitive to the policies being tested; moreover, it needs to respond logically to changes in the input. Accordingly, the performance of TPMSs is highly dependent on the quality of both estimation and calibration processes of the model system. These model systems are usually large-scale so that their development is a complex process.
Focusing on the TPMSs development and enhancement, this thesis has four aims: (1) to review the states of the practice of the TPMSs development and their practical implications, (2) to develop systematic approaches to enhance TPMSs calibration process considering both demand-side and traffic assignment models in a unified structure, (3) to formulate an integrated TPMS to have different model components in a unified structure, and (4) to formulate an emerging model component for conventional TPMSs such as activity-based models. Furthermore, this thesis includes four main chapters. Focusing on the calibration process of TPMSs, the first two main chapters introduce two different calibration models to systematically calibrate and validate large-scale TPMSs. Apart from reproducing the observed statistics, the focus of the first calibration model is on multi-objectivity nature of the calibration process and the validity of TPMSs while the focus of the second calibration model is on building a robust TPMS. The performances of the proposed calibration models are demonstrated via case studies on GTAModel V4.0 model system for the Greater Toronto-Hamilton Area (GTHA). Focusing on the asynchronisations among the conventional TPMS model components, the third main chapter explores the possibility of developing an integrated TPMS to rectify the most common problematic issues in conventional TPMSs. It formulates and calibrates novel TPMS which integrates an activity travel pattern generator and multiple traffic assignment models and then illustrates the application of the model and its capabilities using numerical experiments. The fourth main chapter does not focus on the whole structure of TPMSs in its big picture; rather, it concentrates on formulating an emerging model component suitable to be embedded in the structure of conventional TPMSs.