【摘要】 针对粒子群优化算法收敛速度较慢、易陷入局部最优、收敛结果不够稳定等问题，本文从初始种群、边界处理、惯性权重三个方面对传统粒子群算法进行改进，提出多策略融合粒子群算法（MSF-PSO），并通过测试函数证明了MSF-PSO可大幅提高计算速度和计算效率。将MSF-PSO应用于变压器油纸绝缘介电响应的德拜等效电路参数辨识中，计算结果表明，与其他粒子群优化算法相比，该算法获得的回复电压极化谱能更好地与现场测试获得的回复电压极化谱相吻合，进一步验证了本文所提改进算法的准确性，可为诊断变压器油纸绝缘设备老化情况提供参考。更多还原

【Abstract】 To address the issues of slow convergence speed, susceptibility to local optima, and unstable convergence results in particle swarm optimization(PSO) algorithm, this paper proposes a multi-strategy fusion particle swarm optimization(MSF-PSO) by enhancing the traditional PSO algorithm in three aspects: initial population, boundary handling, and inertia weight. Through the examination of testing functions, the MSF-PSO algorithm is proven to considerably enhance computational speed and efficiency. The MSF-PSO algorithm is applied to the identification of parameters for the Debye equivalent circuit in the dielectric response of oil-paper insulation. The computational results demonstrate that the polarization spectrum of the recovered voltage obtained by this algorithm exhibits better concordance with the polarization spectrum of the recovered voltage acquired from field tests, in comparison to other particle swarm optimization algorithms. This further validates the accuracy of the proposed method and establishes a crucial foundation for diagnosing the aging condition of transformer oil-paper insulation equipment.更多还原