The atrium, as the core space of modern university libraries, is crucial for providing ample natural lighting and creating a comfortable spatial experience. It is also key to achieving the building’s green and low-carbon goals. However, designing the atrium of a university library faces a significant challenge: finding the right balance between ensuring good natural lighting and effectively reducing the energy consumption of the air conditioning system. This study aims to explore this balance and provide architects with various feasible design schemes. Firstly, a parametric typical model of the atrium space was established by researching 36 university libraries. Based on the Grasshopper platform in Rhino, the typical model was simulated for natural lighting and energy consumption, and the Wallacei plugin was used to couple the optimization parameters of the typical model with the optimization target parameters. The multi-objective optimization experiment of the typical model was carried out with the objectives of maximizing spatial daylight autonomy and the percentage of useful daylight illuminance, as well as minimizing air conditioning energy consumption. The experiment generated 2000 optimization solutions, and the analysis of the historical solution set revealed the complex non-linear relationship between optimization parameters and performance indicators. Moreover, three Pareto optimal solutions were selected as representative design schemes, providing valuable references for architects when designing the spatial form of the atrium.
In the early stage of architectural design, addressing the challenges posed by negative spaces in high-density urban environments is crucial for enhancing spatial efficiency and building sustainability. Multiple studies employed digital methods and tools to address these issues, such as parametric design, simulation, and genetic algorithms, to investigate architectural generation approaches for urban negative spaces. This article proposes an integrated design process that involves finding the location and form of negative spaces, generating solutions using slime mold and wasp algorithms, and optimizing and analyzing solutions using the Wallacei plugin in Grasshopper. This comprehensive approach underscores the potential of parametric design to yield a multitude of solutions while also acknowledging the convergence challenges encountered during simulations, particularly in optimizing for optimal sunlight exposure during the winter solstice and minimal solar radiation in the summer. Analyzing the optimization goals and parameter values of the 15th Pareto optimal solution in the 100th generation reveals: (1) a higher number of units leads to positive correlation growth in both objectives; (2) within a certain number of units, parametrically generated solutions facilitate the convergence of optimization goals, yielding optimal outcomes. Therefore, factors such as the range of unit quantities and proportions need consideration during early-stage parametric design and simulation. This study explores a design methodology for negative spaces in high-density urban cities, validating the feasibility of various mainstream generation methods and offering insights for future research.