Reaching an optimum floor allocation plan is a complex process, it involves a wide variety of design criteria. the huge number of possible and desired solutions can be a load of work. the use of generative algorithms in producing these alternatives make the process easier and achievable. the possibility to engage a different kind of optimisation thresholds and design milestones made the design process as accurate as engineering production, and at the same time allowed for innovative and novel aesthetically beautiful architectural solutions.
Plan Permutation and Optimization
in this research, I am investigating the process of using evolutionary algorithms on plan permutation and re-allocating of a number of architectural spaces on a 2D & 3D environments. the explorative research involves parametric design approach and optimisation algorithms together to reach what is called Optimum plan. this research was a continuity of my previous thesis.
the 10 rectangular spaces are the first version of the architectural composition, in this research, I started to play with spaces entities without playing with the diversity of sizes or rotations. the main algorithm works on replacing the location of each space rapidly and then compare between one and each other space, that happens continuously until the distance between spaces become minimum. At the same time, the algorithm test the overall area intersected between spaces, and it makes sure its minimum too; we called it Adjacency allocation. Finally, we make sure to set the average between the different optimisation goals. to allow for a minimum optimisation time, especially with large gene pool engaged in te mutation and solution populations, all these are part of te evolutionary algorithms principles which the article will not be enough to mention all.