In expert systems, complex optimization problems are usually nonlinear, nonconvex, multimodal and discontinuous. As an efficient and simple optimization algorithm, particle swarm optimization(PSO) has been widely applied to solve various real optimization problems in expert systems. However, avoiding premature convergence and balancing the global exploration and local exploitation capabilities of the PSO remains an open issue. To overcome these drawbacks and strengthen the ability of PSO in solving complex optimization problems, a modified PSO using adaptive strategy called MPSO is proposed. In MPSO, in order to well balance the global exploration and local exploitation capabilities of the PSO, a chaos-based non-linear inertia weight is proposed. Meanwhile, to avoid the premature convergence, stochastic and mainstream learning strategies are adopted. Finally, an adaptive position updating strategy and terminal replacement mechanism are employed to enhance PSO’s ability to solve complex optimization problems in expert systems. 30 complex CEC2017 benchmark functions are utilized to verify the promising performance of MPSO, experimental results and statistical analysis indicate that MPSO has competitive performance compared with 16 state-of-the-art algorithms. The source code of MPSO is provided at https://github.com/lhustl/MPSO .

在专家系统中，复杂的优化问题通常是非线性，非凸，多峰和不连续的。作为一种高效，简单的优化算法，粒子群优化算法（PSO）已被广泛应用于解决专家系统中的各种实际优化问题。但是，避免过早收敛并平衡PSO的全球勘探和本地开发能力仍然是一个悬而未决的问题。为了克服这些缺点并增强PSO解决复杂优化问题的能力，提出了一种采用自适应策略MPSO的改进PSO。在MPSO中，为了很好地平衡PSO的全球勘探和局部开发能力，提出了一种基于混沌的非线性惯性权重。同时，为了避免过早收敛，采用随机和主流学习策略。最后，采用自适应位置更新策略和终端替换机制来增强PSO解决专家系统中复杂优化问题的能力。利用30种复杂的CEC2017基准功能来验证MPSO的有前途的性能，实验结果和统计分析表明，MPSO与16种最新算法相比具有竞争优势。MPSO的源代码位于https://github.com/lhustl/MPSO。实验结果和统计分析表明，MPSO与16种最新算法相比具有竞争优势。MPSO的源代码位于https://github.com/lhustl/MPSO。实验结果和统计分析表明，MPSO与16种最新算法相比具有竞争优势。MPSO的源代码位于https://github.com/lhustl/MPSO。