Parallel Genetic Algorithms: A Survey

This report provides a wide-ranging survey of parallel genetic algorithms and reviews a number of approaches that have been adopted to parallelise them. A tutorial level introduction to genetic algorithms is given and the underlying mechanisms are described and discussed with many current developments reviewed. These broad classes of parallel genetic algorithms are considered in detail and compared with conventional sequential implementations. It is argued that significant performance benefits can be realised by using distributed population structures and selection mechanisms even when the paradigm is implemented on a sequential machine.