Automated generation of optimised building envelope: simulation based multi-objective process using evolutionary algorithm

Abstract

This paper seeks to investigate methodology on the seamless integration for constructing a climate-conscious building envelope. The focus is targeted to provide architectural designers with an intelligent way to develop their design alternatives in the early phase of a project. Improved environmental performance can be achieved by using an optimised building envelope based on local data analysis. For tight integration, coupling of MATLAB and Rhino was employed through coding for real-time information exchanges. EnergyPlus and ECOTECT were implemented in the modelling process to simulate environmental performance. This process considers two cost functions: solar radiation and energy. Respective performance check-up runs into Multi-Objective Programming (MOP). Decision-making in MOP is optimised by using an Evolutionary Algorithm (EA), and culminated with Pareto efficiency. A sample experiment was conducted to test the methodology and results. The city, Moscow, having a conflicting claim of energy use reduction and increment of solar radiation due to its frigid local weather, was selected. Variants of adapted forms in the urban setting were visualised though the EA process in Rhino, and finalised to each dissimilar objective. The Pareto frontier method compromised the final models which were fitted best to energy and solar radiation each. The result shows how a form of building envelope can be adapted in a specific local context. This procedure serves to put forward a new user-oriented methodology and its application, intensifying the practical viability of performance-based building design in the future.