摘要
Lennad-Jones团簇结构优化是在物理化学和材料科学领域经常遇到的优化问题,由于原子之间的相互作用具有高度的复杂性,因此常用来检验算法性能。量子动力学是一种具有完整数学框架且非常有效的理论模型,量子动力学框架(QDF)已经被证明在复杂问题中具有很强的竞争力。在不引入任何先验知识的基础下,使用QDF在Lennard-Jones团簇结构优化问题中进行求解,探讨QDF在具有高度复杂的解空间的团簇问题中的性能。将QDF与差分进化算法、骨架烟花算法进行对比,比较最低能量、能量分布和时间精度等。实验结果表明,在Lennard-Jones团簇结构优化问题中,可接受时间内QDF取得优于其他两种算法的最低能量。QDF是一种具有潜力的解决框架。
Lennard-Jones cluster structure optimization is a common optimization problem in the fields of physical chemistry and materials science.With the complex interactions between atoms,it serves as a benchmark for evaluating algorithm performance.Quantum Dynamics Framework(QDF)is highly competitive in complex problems.Without introducing any prior knowledge,QDF was used to solve the Lennard-Jones cluster structure optimization problem,exploring the performance of QDF in cluster problems with highly complex solution spaces.The QDF was compared with the differential evolution algorithm and the bare bones fireworks algorithm in terms of lowest energy,energy distribution,and time accuracy.The experimental results show that QDF achieves the lowest energy in the acceptable time than the other two algorithms in the Lennard-Jones cluster optimization problem.The research results indicate that QDF is a potential framework for the Lennard-Jones cluster structure optimization problem.
作者
王旭
王鹏
吕林桃
WANG Xu;WANG Peng;LYU Lintao(College of Computer Science and Engineering,Southwest Minzu University,Chengdu 610041,China)
出处
《成都信息工程大学学报》
2024年第1期56-60,共5页
Journal of Chengdu University of Information Technology
