Building Performance Optimization of Net Zero-Energy Buildings - Modeling, Design, and Optimization of Net-Zero Energy Buildings - page 254

Fig. 5.5 Example of Hooke-Jeeves pattern search on the Broyden function
Population-based algorithms perform operations on populations of
representative building designs. Often, they are called metaheuristics due to
their nature of finding near-optimal solutions to a wide range of problems.
Two common population-based search algorithms used with BPS are
genetic, a type of evolutionary algorithm, and the particle swarm algorithm.
The first algorithm selected for discussion from the group of
population-based algorithms is the Genetic Algorithm (GA), from the EA
family. GAs have become popular due to their ease of implementation and
proven ability to solve multimodal and multi-objective problems.
Computational pseudo-evolution was first demonstrated by Goldberg
(1989) using biological inspirations. Performing genetic operations, such as
mutations and crossovers, on representations in combination with selection
operators emulate the “survival of the fittest” found in biological evolution.
Eiben and Rudolph (1999) described members of the EA family as “adaptive
systems having a “basic instinct” to increase the average and maximum