Optimisation of Urban Energy Demand Using an Evolutionary Algorithm
Computer modelling at the urban scale is an increasingly vibrant area of research activity which aims to support designers to optimise the performance of new and existing urban developments. But the parameter space of an urban development is infinitely large, so that the probability of identifying an optimal configuration of urban design variables with say energy minimisation as a goal function is correspondingly small. To resolve this we have coupled a micro-simulation model of urban energy flows CitySim with a new evolutionary algorithm (EA): a hybrid of the CMA-ES and HDE algorithms. In this paper we present the means of coupling the EA and CitySim and identify a subset of urban design variables that have been parameterised. We then present results from application of this new methodology to minimise the energy demand of part of a case-study district in the city of Basel, Switzerland. The papers closes by discussing work that is planned to further increase the scope of this new methodology for optimising urban sustainability.