Synergising AI-Driven Kinetic Facades and Machine Learning for Thermal Load Mitigation in Iraq's Administrative Infrastructure: A Predictive Simulation Approach

الملخص

Background: Iraq suffers from extremely high temperatures in summer, where temperatures often rise above 50°C. Consequently, demands on HVAC have affected the national power grid of Iraq. Conventional forms of insulation that do not move become ineffective because of the region’s changing solar angles and dust storms.

Objective: An intelligent building envelope system responsive to users' propositions proposed by the study. The paper aims to assess the operation of a Kinetic Facade controlled by a Machine Learning (ML) algorithm for maximising the indoor thermal comfort while consuming minimal energy in Baghdad’s administrative buildings.

Methodology: A parametric model of a typical office building was created using Rhino and Grasshopper. Using local EPW data, environmental simulations were carried out through Ladybug and Honeybee. An ML model, which is based on Python, was deployed to control the movements of the facade at the right time using predictive or analytic tools, considering solar gain, natural light, and protection during peak dust.

Results: According to initial calculations, the annual cooling loads of the building may be reduced by 32% to 38%, depending on the curtain wall system. In addition, 70% of the working hours had UDI lighting as per the requirement, which resulted in reduced usage of artificial lights.

Conclusion: AI and kinetic architecture give a probable, highly efficient solution for the energy crisis in Iraq. This research establishes a framework for the next generation of “smart” sustainable buildings in arid zones by changing the passive environmental response to active use.