Optimization Algorithms Supporting the Cost-Optimal Analysis: the Behavior of PSO

This work is within the wide context of simulation-based optimization methods applied to the cost-optimal analysis of nearly-zero energy buildings, with the objective of studying the behavior of the particle swarm optimization algorithm (PSO) in solving cost-optimal problems. After the presentation of the features of the involved design variables and of the resulting design space, the paper focuses on the application of PSO, implemented by coupling GenOpt to TRNSYS, to a typical cost-optimal problem for a single-family home. The algorithm performance related to different sets of algorithm parameters was analyzed and classified according to defined metrics. Best results are reached with a small number of particles and higher cognitive acceleration.