Parameter Setting of Quantum-Inspired Genetic Algorithm Based on Real Observation

摘要

Parameter setting, especially the angle of Q-gate, has much effect on the performance of quantum-inspired evolutionary algorithm. This paper investigates how the angle of Q-gate affects the optimization performance of real-observation quantum-inspired genetic algorithm. Four methods, including static adjustment methods, random adjustment methods, dynamic adjustment methods and adaptive adjustment methods, are presented to bring into comparisons to draw some guidelines for setting the angle of Q-gate. Comparative experiments are carried out on some typical numerical optimization problems. Experimental results show that real-observation quantum-inspired genetic algorithm has good performance when the angle of Q-gate is set to lower value.