Bruggen tussen vervoerseconomie en verkeersmodellen: speltheoretische evenwichten in grootschalige netwerken met multi-actor prijszetting

Designing efficient pricing schemes for transportation systems has long been a central problem in both the domains of transport economics and traffic modeling. Pricing instruments such as tolls, fares, and tradable credits are among the most powerful tools available to manage demand, reduce congestion, and improve the efficiency of transport networks. Their importance has only grown as urban mobility systems continue to be more congested, and multimodal. Additionally, the number of relevant decision makers has increased: national, regional, and local governments may pursue overlapping but distinct objectives, and an increasing number of private mobility service providers who optimize their own pricing strategies have entered the market. In such settings, pricing is no longer a single-leader problem but a strategic multi-decision-maker problem, where interactions between actors can significantly alter outcomes. Addressing these challenges requires tools that are both behaviorally realistic and computationally tractable. However, despite decades of research, existing approaches struggle to address the complexity of modern urban mobility systems. Traditional transport economics models, though analytically elegant, are often restricted to highly simplified networks and cannot accommodate the behavioral and spatial heterogeneity of real systems. Conversely, full-scale traffic models capture network detail and user behavior with greater fidelity but are computationally demanding and unsuitable for direct optimization and multi-decision maker equilibration. Bridging the gap between these two branches of research remains an open challenge and is the central focus of this dissertation. This dissertation makes progress on this challenge through three interconnected contributions. First, a mathematical programming-based transport economics model is developed that moves beyond the stylized formulations common in the literature. This model endogenizes both demand and path choice user-equilibrium, while accounting for heterogeneous user classes with different Values of Time and Willingness-To-Pay. The core of this model allows computation of optimal tolls for a single government with due consideration of reaction of the travelers. On top of this foundation, game-theoretical formulations are introduced to analyze interactions between multiple governments which optimize their own tolls. Both Nash and Stackelberg competition scenarios are modelled. Applications show that competition between governments generally reduces efficiency compared to joint optimization, while taking-up leadership can mitigate losses to an extent. Case studies also demonstrate practical methods for addressing the unintended impacts due to non-uniqueness in user equilibria. Building on this foundation, the second contribution is the development of CHANAkYA, a comprehensive game-theoretical modelling framework of pricing and tradable credit charging in multi-modal transportation systems with public and private players. CHANAkYA extends the single-mode model to incorporate roads, public transport, and private mobility service providers, while accommodating stratified user demographics and a broad portfolio of instruments. These include link- and entry-based tolls, distance-based tolls, user class-based prices, and tradable credit schemes. Interactions between toll or credit charge optimizing governments and fare optimizing private mobility service providers are modelled. The framework also accounts for the fact that user-equilibrium flows are often non-unique by introducing player-specific perspectives: optimistic, pessimistic, or entropy-maximizing. These perspectives remove the ambiguity in the reaction of travelers to different prices. Through a series of policy-driven case studies inspired by Leuven, the framework demonstrates its capacity to generate nuanced insights. Results show, for instance, that a modest cordon toll combined with low public transport fares yields large efficiency gains over the baseline scenario; that age-differentiated tolling produces only marginal improvements while raising equity concerns; that tradable credits can replicate the efficiency of direct pricing instruments; and that privatization of public transport leads to efficiency losses. Collectively, these findings illustrate CHANAkYA's potential to support richer, more policy-relevant analyses of complex transport systems. The third contribution addresses the computational challenge of applying such game-theoretical models to full-scale traffic models. Toll optimization in full-scale traffic models is generally studied through surrogate modelling techniques like Bayesian Optimization and Kriging models. However, traditional surrogate modeling techniques focus on approximating an input-output relationship between design variables and the objective function of a single decision maker, and therefore cannot capture the strategic interactions that arise in multi-actor settings. To overcome this limitation, a novel metamodel-based equilibration (MBE) scheme is developed. The scheme employs CHANAkYA as a fast metamodel guiding the search for equilibrium tolls in an underlying full-scale traffic model with multiple decision makers. An iterative calibration-equilibration process ensures that the metamodel parameters are regularly updated to match sensitivities observed in the full-scale traffic model, while optimization is performed at the metamodel level. This approach shifts the computational burden away from the full-scale model, enabling equilibrium pricing to be achieved with far greater efficiency. Applications confirm its effectiveness: in single decision-maker settings, MBE outperforms Bayesian Optimization benchmarks, while in multi-actor contexts it enables, for the first time, the computation of approximate Nash and Stackelberg equilibria with full-scale traffic models. These results demonstrate the feasibility of policy-relevant multi-decision-maker analysis at realistic scales. Taken together, the three contributions form a coherent progression. The dissertation first establishes a computationally rigorous foundation for an advanced transport economics model, then generalizes this foundation into a flexible multi-modal framework which allows a portfolio of instruments and decision-makers, and finally integrates this framework with full-scale traffic models through the novel metamodel-based equilibration scheme. The resulting methods show that properly calibrated transport economics models can serve not only as stand-alone tools for illustrative case studies, but also as efficient guides for full-scale traffic models. This advances the state of the art by enabling theoretically grounded and computationally tractable analysis of multi-actor pricing problems, which until now has remained elusive. This dissertation thus provides both methodological innovations and practical tools for the design of pricing schemes in modern transportation and urban mobility systems. By bridging the complementary strengths of transport economics and traffic modeling, it offers a pathway toward more rigorous, efficient, and policy-relevant approaches to one of the central challenges in urban mobility.