THE RING

As International Horticultural Exhibition Hall, different from normal normal exhibition buildings, focusing more on tight relationship among indoor exhibits and outdoor garden, while the meeting the requirement to display subjects from different countries and cultures.

Comparing to arranging different scene along a monodirectional line, a Ring could match the start with the end, merging different exhibition part, giving visitors more flexibity in visiting different pieces in their own order .

After the shaped of the project was decided as a ring, Multi-Agents Algorithm is choosed to help generating the plan, by using one agent to simulate one human, the expectation of crowds' motion is calculated.

The simluation yields the expected crowds location, from which equipotential lines are generated, showing space density expected in the project.

A couple of rules are used to define single human behavior.

1. Agents will push each other when distance between two agents less than 1.5m.
2. Agents will pull each other when distance between two agents less than 20m.
3. Each agent will be pulled by landscape in its sight.
4. Each agent will be pulled by entrance.

Genetic algorithm is able to evolve the target into a better state in terms of the given 'fitness', by shuffling, inheriting 'gene' across generations. In this context, the fitness of genetic algorithm is the energy consumption calculated by Energyplus.

However, the algorithm can only push the fitness without considering any architecutre aspects. Since the output of the algorithm is not always satisfying, restrictions have been added to the 'gene' to avoid bad results.

Same idea as above, Genetic Algorithm is really good at find a good solustion based on the 'fitness', which helps a lot when designing a facade with good sustainability consideration.

On the other hand, facade should be related to people's sight, which algorithm can do nothing. Thus, the formation combine both algorithm optimization and manual intervention.