Abstract: A Competition Quantum Evolutionary Algorithm（CQEA） is proposed in this paper. Every particle changes its position towards to global best and the mean value of the population at each time step. By comparing the two homologous particles it chooses the better one as the next generation. Meanwhile, a kind of self-adapting variation is introduced. For the worse particle, the variable probability is low; for the better one, it is relatively high. Finally, five test functions are used to analyze the property of CQEA. Simulation results show that CQEA has better performance compared with PAQEA and NVCQEA.

Key words: quantum evolutionary algorithm, evolve under double directions, phase angle encoded, self-adapting variation

摘要： 针对连续空间数值优化问题，提出了一种竞争型量子进化算法。粒子每次向全局最优和种群均值两个方向分别进化，从而得到两个子粒子。根据“优胜劣汰”原则选择适应度较高者作为下一代。同时，为了保证粒子的多样性，引入了一种自适应变异机制：对适应度较低的粒子以较高概率进行变异，而对适应度较高粒子以较低概率进行扰动。通过5个标准测试函数验证了算法的性能。仿真结果表明，与PAQEA及NVCQEA相比，该算法收敛速度快，收敛精度高，稳定性好。

关键词: 量子进化算法, 双方向进化, 相位角编码, 自适应变异