Evolving Good Spread of Solutions with Improved Multi-objective Quantum-inspired Evolutionary Algorithm

This paper presents an improved multi-objective quantum-inspired evolutionary algorithm (IMQEA) for solving multi-objective optimization problems (MOPs). Different from general MQEAs, the proposed approach uses multiple observations to yield candidate solutions. In the early stage of evolution, multiple observations of a given quantum bit (Q-bit) individual can yield solutions with good diversity, which is helpful for global search. In the later stage, most Q-bits have matured, and thus multiple observations of a given Q-bit individual are similar to a local search, which improves the accuracy of solutions. Experimental results for the multi-objective 0/1 knapsack problem show that the IMQEA finds solutions close to the Pareto-optimal front and maintains a good spread of the non-dominated set.