In the new optimisation script, the convergence to a sensible set of 'optimal' thermal networks is very slow. This is likely due to the high level of randomness with which the buildings are connected to one or the other network. The non-dominated solutions, therefore, often include networks that connect buildings that are very far apart and make very little sense at first sight.
Therefore, new mutation and crossover algorithms should be introduced that take into consideration the spatial distribution of buildings when testing new network layouts.
In the new optimisation script, the convergence to a sensible set of 'optimal' thermal networks is very slow. This is likely due to the high level of randomness with which the buildings are connected to one or the other network. The non-dominated solutions, therefore, often include networks that connect buildings that are very far apart and make very little sense at first sight.
Therefore, new mutation and crossover algorithms should be introduced that take into consideration the spatial distribution of buildings when testing new network layouts.