In the last decades, an increasing number of global optimization algorithms has been proposed to solve geophysical inverse problems. Indeed, an inverse problem can be posed as an optimization problem where the function to be optimized, usually called objective function, misfit function or fitness, provides an estimate of the difference between observed data and synthetic data computed by a trial model. In the framework of probabilistic global optimization methods, some algorithms use statistical distributions inspired by physical processes to get suggestion on which solution candidate has to be tested next, as for example the Simulated Annealing algorithms, which select the next solution candidate according to the Boltzmann probability factor of atom configurations of cooling metals. Other algorithms, such as the Genetic Algorithms, are instead inspired by Darwinian evolution processes and treat solution candidates as individuals that compete in a virtual environment controlled by the genetic mechanisms of mutation, reproduction and crossover. In this work, we present and discuss some applications of a new hybrid global optimization algorithm, GPA, recently proposed by the authors for quantitative interpretations of potential field data. The proposed approach includes a mutation operator in a controlled random search algorithm and its effectiveness has been proved on several synthetic and field data concerning single self-potential (SP) anomalies. Here, the results of a numerical study focused on multiple self-potential anomalies are shown for two of the main application fields of the SP method, i.e. for volcanic and soil contamination risk.

An innovative global optimization algorithm for identifying the source parameters of potential fields. Application to multiple self-potential anomalies

PIEGARI E.;DI MAIO R.;CARBONARI R.;VITAGLIANO E.;MILANO L.
2018

Abstract

In the last decades, an increasing number of global optimization algorithms has been proposed to solve geophysical inverse problems. Indeed, an inverse problem can be posed as an optimization problem where the function to be optimized, usually called objective function, misfit function or fitness, provides an estimate of the difference between observed data and synthetic data computed by a trial model. In the framework of probabilistic global optimization methods, some algorithms use statistical distributions inspired by physical processes to get suggestion on which solution candidate has to be tested next, as for example the Simulated Annealing algorithms, which select the next solution candidate according to the Boltzmann probability factor of atom configurations of cooling metals. Other algorithms, such as the Genetic Algorithms, are instead inspired by Darwinian evolution processes and treat solution candidates as individuals that compete in a virtual environment controlled by the genetic mechanisms of mutation, reproduction and crossover. In this work, we present and discuss some applications of a new hybrid global optimization algorithm, GPA, recently proposed by the authors for quantitative interpretations of potential field data. The proposed approach includes a mutation operator in a controlled random search algorithm and its effectiveness has been proved on several synthetic and field data concerning single self-potential (SP) anomalies. Here, the results of a numerical study focused on multiple self-potential anomalies are shown for two of the main application fields of the SP method, i.e. for volcanic and soil contamination risk.
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: http://hdl.handle.net/11588/739934
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact