Research on Fitness Function of Tow Evolution Algorithms Using for Neutron Spectrum Unfolding

by   Rui Li, et al.

Using evolution algorithms to unfold the neutron energy spectrum, fitness function design is an important fundamental work for evaluating the quality of solution, but it has not attracted much attention. In this work, we investigated the performance of 8 fitness functions attached to genetic algorithm (GA) and differential evolution algorithm (DEA) used for unfolding four neutron spectra of IAEA 403 report. Experiments show that the fitness functions with a maximum in GA can limit the ability of population to percept the fitness change but the ability can be made up in DEA, and the fitness function with a feature penalty item help to improve the performance of solutions, and the fitness function using the standard deviation and the Chi-Squared shows the balance between algorithm and spectra. The results also show that the DEA has good potential for neutron energy spectrum unfolding. The purpose of this work is to provide evidence for structuring and modifying the fitness functions, and some genetic operations that should be paid attention were suggested for using the fitness function to unfold neutron spectra.


page 1

page 2

page 3

page 4


Learning Fitness Functions for Genetic Algorithms

A genetic algorithm (GA) attempts to solve a problem using a pool of pot...

Capturing Emerging Complexity in Lenia

This research project investigates Lenia, an artificial life platform th...

A hybrid deep-learning-metaheuristic framework for discrete road network design problems

This study proposes a hybrid deep-learning-metaheuristic framework with ...

A nested genetic algorithm strategy for the optimal plastic design of frames

An innovative strategy for the optimal design of planar frames able to r...

Sufficient Conditions for Coarse-Graining Evolutionary Dynamics

It is commonly assumed that the ability to track the frequencies of a se...

Automatically Balancing Model Accuracy and Complexity using Solution and Fitness Evolution (SAFE)

When seeking a predictive model in biomedical data, one often has more t...

A reproducibility study of "Augmenting Genetic Algorithms with Deep Neural Networks for Exploring the Chemical Space"

Nigam et al. reported a genetic algorithm (GA) utilizing the SELFIES rep...

Please sign up or login with your details

Forgot password? Click here to reset