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 specified 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 scientific 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 verified by grouping the objects. Complex-shaped areas were modeled using synthetic configurations. The solutions were evaluated using a composite objective function and a system of penalties for constraint violations.

Results. The results identify AI models that are optimal for ranking alternatives under a fixed computational budget and provide a stable ordering of solutions. For Random Forest, the achieved R² values are in the range of 0.995–0.999. A practical compromise between optimality and admissibility was identified, with the proportion of admissible solutions being significantly affected by the selection method and mode, ranging from approximately 39 to 99 % in certain scenarios.

Conclusions. The study identified optimal AI models that enable the generation of planning solutions, the calculation of quality and admissibility indicators, and the comparison of methods using unified evaluation rules. This supports the early-stage assessment and management of investment and construction project indicators.

1. Shumov V.N. Transformation of construction technologies under the influence of AI. Symbol of Science: International Scientific Journal. 2024; 1(12-2):87-98. EDN STOEYQ. (rus.).

2. Leptyukhova O.Yu. Optimisation of residential elements of the planning structure (calculation and graphic stage). Biosphere Compatibility: Human, Region, Technologies. 2024; 3(47):13-21. DOI: 10.22227/2311-1518.2024.3.13-21. EDN GTGTCD. (rus.).

3. Manin P.A., Epifantseva L.R. Robotic design in construction. Industrial and Civil Engineering. 2025; 4:68-72. DOI: 10.33622/0869-7019.2025.04.68-72. EDN GCKMFS. (rus.).

4. Kievskiy I.L., Semenov S.A., Krutyakov A.Y., Grishutin I.B., Maksimov S.O., Kargashin M.E. et al. Artificial intelligence in urban planning: modern methods and their prospects. Industrial and Civil Engineering. 2025; 8:77-82. DOI: 10.33622/0869-7019.2025.08.77-82. EDN GPNWYL. (rus.).

5. Mirza M., Osindero S. Conditional Generative Adversarial Nets. arXiv. 2014. DOI: 10.48550/arXiv.1411.1784

6. Goodfellow I., Pouget-Abadie J., Mirza M., Xu B., Warde-Farley D., Ozair S. et al. Generative Adversarial Nets. Advances in Neural Information Processing Systems. 2014; 27.

7. Kingma D.P., Welling M. Auto-Encoding Variational Bayes. arXiv. 2013. DOI: 10.48550/arXiv.1312.6114

8. Ho J., Jain A., Abbeel P. Denoising Diffusion Probabilistic Models. arXiv. 2020. DOI: 10.48550/arXiv.2006.11239

9. Vaswani A., Shazeer N., Parmar N., Uszkoreit J., Jones L., Gomez A.N. et al. Attention Is All You Need. arXiv. 2017.

10. Adonina A., Denisova M. Architectural and urban planning activity and artificial intelligence. Innovative Project. 2024; 9(15):78-82. DOI: 10.17673/IP.2024.9.15.10. EDN SROKHY. (rus.).

11. Zhong C., Shi Y., Cheung L.H., Wang L. AI-Enhanced Performative Building Design Optimization and Exploration. CAADRIA Proceedings. 2024; 1:59-68. DOI: 10.52842/conf.caadria.2024.1.059

12. Rahbar M., Mahdavinejad M., Markazi A.H.D., Bemanian M. Architectural Layout Design through Deep Learning and Agent-Based Modeling: A Hybrid Approach. Journal of Building Engineering. 2022; 47:103822. DOI: 10.1016/j.jobe.2021.103822. EDN VIVSDE.

13. Deb K., Jain H. An Evolutionary Many-Objective Optimization Algorithm Using Reference-Point-Based Nondominated Sorting Approach, Part I: Solving Problems With Box Constraints. IEEE Transactions on Evolutionary Computation. 2014; 18(4):577-601. DOI: 10.1109/tevc.2013.2281535

14. Todinov M. Reliability and Risk Models: Setting Reliability Requirements. 2nd ed. Wiley, 2015; 406. DOI: 10.1002/9781118873199

15. Rhee J., Veloso P. A Physics-Based Constraint Solver for the Site Layout Optimization of Non-convex Buildings with Multiple Requirements. Communications in Computer and Information Science. 2023; 594-606. DOI: 10.1007/978-3-031-37189-9_39

16. Knecht K., König R. Generating Floor Plan Layouts with K-d Trees and Evolutionary Algorithms. GA2010 — 13th Generative Art Conference. 2010; 238-253.

17. Klar M., Langlotz P., Aurich J.C. A Framework for Automated Multiobjective Factory Layout Planning Using Reinforcement Learning. Procedia CIRP. 2022; 112:555-560. DOI: 10.1016/j.procir.2022.09.099. EDN CNDQKA.

18. Gureev M.V., Makarov A.N. Model of forecasting of material resources and estimated cost at early stages of life cycle of construction objects. Vestnik MGSU [Monthly Journal on Construction and Architecture]. 2024; 19(11):1835-1849. DOI: 10.22227/1997-0935.2024.11.1835-1849. EDN ZCNAQS. (rus.).

19. Lathong K., Wisaeng K. An innovative hybrid machine learning techniques for predicting construction cost estimates. International Journal for Computational Civil and Structural Engineering. 2024; 20(3):69-83. DOI: 10.22337/2587-9618-2024-20-3-69-83. EDN KIPHPW.

20. Nindartin A., Park S.Ju., Lee K.T., Kim Ju.H., Rostiyanti S.F. Prediction of cost contingency in construction projects by introducing machine learning algorithms. Journal of Civil Engineering and Management. 2025; 31(8):860-880. DOI: 10.3846/jcem.2025.24913. EDN OTWXUA.