Efficiency studies of quantum and classical genetic algorithms

Shchetinin E.Yu., Prokopyev P.N.

Financial University under the Government of Russia

Genetic algorithms (GA) became known in the 1970s as effective numerical methods for solving various problems associated with finding the global extremum of complex multidimensional functions, optimal control and control of complex technical systems, etc.However, a number of its shortcomings are known, limiting its application. In recent years, the attention of researchers has been attracted by quantum computing and the development of quantum analogues of evolutionary algorithms, devoid of the above disadvantages and, at the same time, have advantages that allow to implement fairly complex algorithms with high speed and accuracy.

Quantum computing is a new interdisciplinary scientific field that has emerged at the intersection of computer science and quantum mechanics. In this paper, we investigate the basic properties of the quantum genetic algorithm (QA) and compare it with the genetic algorithm (GA) to identify the advantages of the computational abilities of QA to perform data processing and find optimal solutions to some control problems in a more efficient way. The article deals with quantum genetic algorithms based on a combination of quantum and classical calculations, as well as an algorithm consisting only of quantum calculations. As an example demonstrating these advantages, we have considered the classical problem of optimal solution theory – the knapsack problem.

© 2004 Designed by Lyceum of Informational Technologies №1533