Climate change and growing urban populations are increasingly putting pressure on cities to reduce their carbon emissions and transition towards efficient and renewable energy systems. This challenges in particular urban planners, who are expected to integrate technical energy aspects and balance them with the conflicting and often elusive needs of other urban actors. This thesis explores how multicriteria decision analysis, and in particular multiobjective optimization techniques, can support this task. While multiobjective optimization is particularly suited for generating efficient and original alternatives, it presents two shortcomings when targeted at large, intractable problems. First, the problem size prevents a complete identification of all solutions. Second, the preferences required to narrow the problem size are difficult to know and formulate precisely before seeing the possible alternatives. Interactive optimization addresses both of these gaps by involving the human decision-maker in the calculation process, incorporating their preferences at the same time as the generated alternatives enrich their understanding of acceptable tradeoffs and important criteria. For interactive optimization methods to be adopted in practice, computational frameworks are required, which can handle and visualize many objectives simultaneously, provide optimal solutions quickly and representatively, all while remaining simple and intuitive to use and understand by practitioners.

Accordingly, the main objective of this thesis is:

To develop a decision support methodology which enables the integration of energy issues in the early stages of urban planning.

The proposed response and main contribution is SAGESSE (Systematic Analysis, Generation, Exploration, Steering and Synthesis Experience), an interactive multiobjective optimization decision support methodology, which addresses the practical and technical shortcomings above. Its innovative aspect resides in the combination of (i) parallel coordinates as a means to simultaneously explore and steer the alternative-generation process, (ii) a quasi-random sampling technique to efficiently explore the solution space in areas specified by the decision maker, and (iii) the integration of multiattribute decision analysis, cluster analysis and linked data visualization techniques to facilitate the interpretation of the Pareto front in real-time.

Developed in collaboration with urban and energy planning practitioners, the methodology was applied to two Swiss urban planning case-studies: one greenfield project, in which all buildings and energy technologies are conceived ex nihilo, and one brownfield project, in which an existing urban neighborhood is redeveloped. These applications led to the progressive development of computational methods based on mathematical programming and data modeling (in the context of another thesis) which, applied with SAGESSE, form the planning support system URBio. Results indicate that the methodology is effective in exploring hundreds of plans and revealing tradeoffs and synergies between multiple objectives. The concrete outcomes of the calculations provide inputs for specifying political targets and deriving urban master plans.