Research And Application Of Quantum Inspired Evolutionary Algorithm

Quantum Inspired Evolutionary Algorithm (QEA) is a EvolutionaryAlgorithm based on principles of Quantum Computing (QC). The noveloptimization method has strong robustness and is valuable for research.Based on concept and principles of QC, QEA uses a Q-bit as aprobabilistic representation and a Q-gate as a variation operator tocomplete evolutionary search. Compared with conventional EAs, QEAcan balance between exploration and exploitation more easily,additionally it is characterized by small population size, rapidconvergence and strong global search capability. As the experimentalresults show, QEA has better performances than EAs on many problems,but it is weak in solving some complex optimization problems because ittends to run into local optima. In order to make QEA solve actualproblems more effectively, in this paper, we do some further researcheson it, making the performance better and enlarging its applicationfurther.The major contributions and achievements are as follows:1. Quantum Inspired Evolutionary Algorithm based on Estimation ofDistribution (EQEA) is proposed. The novel algorithm keeps twoprobabilistic models which support each other and make the algorithmhave more rapid convergence under the circumstance of diversities. Inaddition, an adaptive rotation operator is proposed which can adjust themagnitude of angle with evolution process automatically.2. QEA is applied to solve Multiple Choice Knapsack Problem(MCKP) and Multiple Choice Multi-Dimensional Knapsack Problem(MMKP). For these special problems, a new method of making solutionsfrom quantum chromosomes and a corresponding rotation operator are designed.3. EQEA for MOPs is proposed. The new algorithm can have goodresults for Multi-Objective Knapsack Problems.