Multi Population Adaptive Inflationary Differential Evolution Algorithm

Summary: Multi Population Adaptive Inflationary Differential Evolution Algorithm (MP-AIDEA) is a population-based evolutionary algorithm for solving single-objective global optimisation problems over continuous spaces. It combines adaptive Differential Evolution with the restarting procedure of Monotonic Basin Hopping. Multiple populations are initialised in the search space and exchange information during the optimisation process. Each population starts the search for the global minimum by running a Differential Evolution algorithm. The user is not required to set the parameters of the Differential Evolution, a process that could be extremely time consuming, since the best settings of the parameters are problem-dependent. Instead, MP-AIDEA is able to automatically adapt these parameters during the optimisation process. At the end of the Differential Evolution a local search is run from the best individual of each population. Using the restarting mechanism of Monotonic Basin Hopping in combination with the Differential Evolution, the populations are able move, in a funnel structure, from one local minima to another, until the global minimum of the problem is located. MP-AIDEA implements a novel approach to avoid multiple detection of the same local minima, by restarting the population in the entire work space when it falls within the basin of attraction of an already detected minimum.Extensive tests have been conducted on the algorithm using several difficult academic test functions and real world problems, in order to assess its validity.

Some unpublished, yet interesting results obtained via MP-AIDEA can be found here.


M. Di Carlo, M. Vasile, E. Minisci, “Adaptive multi-population inflationary differential evolution“, Soft Computing volume 24, pages3861–3891, 2020

M. Di Carlo, M. Vasile, E. Minisci, “Multi-Population Adaptive Inflationary Differential Evolution Algorithm with Adaptive Local Restart“, IEEE Congress on Evolutionary Computation, CEC 2015, Sendai, Japan

M. Di Carlo, M. Vasile, E. Minisci, “Multi-Population Adaptive Inflationary Differential Evolution Algorithm“, 2014 BIOMA (Bio-Inspired Optimisation Methods and their Applications) Workshop, Ljubljana, Slovenia

E. Minisci, M. Vasile, “Adaptive Inflationary Differential Evolution“, 2014 CEC (IEEE Congress on Evolutionary Computation), Beijing, China

Vasile, Massimiliano and Minisci, Edmondo and Locatelli, Marco (2011) “An inflationary differential evolution algorithm for space trajectory optimization IEEE Transactions on Evolutionary Computation, 15 (2). pp. 267-281. ISSN 1089-778X

Timeframe: March 2014-July 2019

People: Marilena Di Carlo, Massimiliano Vasile, Edmondo Minisci

Sponsor: Airbus