Optimization of building placement on irregularly shaped land plots using artificial intelligence methods
Introduction. This paper addresses the problem of optimizing building placement on land plots of complex shape at the early stages of design, when initial data are limited and the requirements are specifi ed as a system of geometric constraints and quality targets. The relevance of the study stems from the high labor intensity of manually preparing layout alternatives and the need for an objective comparison of layout solutions for plots with irregular boundary geometry. The scientifi c novelty of the study lies in the formalization of the problem statement and the comparative study of heuristic and intelligent methods within a single computational protocol. Materials and methods. The experimental dataset includes 79 objects and covers the comparison of 18 AI methods for generating and selecting planning solutions. Regression models were used to select alternatives; these models were trained and verifi ed by grouping the objects. Complex-shaped areas were modeled using synthetic confi gurations. The solutions were evaluated using a composite objective function and a system of penalties for constraint violations.