基于多轮升维策略与改进量子行为粒子群优化算法的热导率函数估计方法

将改进的量子行为粒子群优化算法应用于材料热导率函数估计问题中，并提出了一种多轮升维策略对算法的搜索过程进行优化，形成了一种鲁棒性强且高效的反演方法。通过数值实验测试了该方法在测量误差以及系统误差下的表现，并对不同粒子群优化算法的性能进行了比较研究。结果表明，采用的反演方法能够在较大的搜索范围与反演维度下稳定收敛，对测量误差的敏感度较低；提出的多轮升维策略能够使各类粒子群优化算法在热导率函数估计问题中的搜索效率得到提升。

Thermal conductivity function estimation approach based on improved quantum-behavior particle swarm optimization algorithm

In this paper,the improved quantum-behavior particle swarm optimization algorithm is applied to the estimation of thermal conductivity function of materials,and a multi-round upgrading strategy is proposed to optimize the search process of the algorithm,which forms a robust and efficient inversion method. The performance of this method under measurement errors and system errors is tested by numerical experiment,and the performance of different particle swarm optimization algorithms is compared and studied. The results show that the inversion method presented in this paper can converge stably in a large search range and dimension,and has low sensitivity to measurement errors.The multi-round upgrading strategy can improve the searching efficiency of various kinds of particle swarm optimization algorithms in thermal conductivity function estimation problems.