中文摘要：

建立汽轮机调节系统准确的数学模型对电力系统频率稳定的分析具有重要的应用价值。对于汽轮机调节系统的非线性参数辨识，进化算法有一定优势，但也存在着早熟和收敛速度慢的缺陷。为此提出一种基于差分一粒子群的混合优化算法，在粒子进化过程中引入差分进化操作，提高粒子进化效率。以贵州某电厂1号机组为研究对象，结合现场实测数据，采用DEPSO方法对其调节系统参数进行辨识。仿真校核结果表明，辨识得到的汽轮机调节系统模型能够较为真实地反映实际系统的功率响应特性，证明了混合算法解决此类问题的有效性。

英文摘要：

The establishment of accurate mathematical model of steam turbine regulation system has important application value for the analysis of power system frequency stability. For the nonlinear parameter identification of the steam turbine regulation system, the evolutionary algorithm has certain advantage, but there are some shortcomings of early maturing and slow convergence. A hybrid optimization algorithm based on differential evolution-particle swarm is proposed, the differential evolution operation is introduced in the process of particle evolution and particle evolution efficiency is improved. Taking No. 1 generator unit of certain power plant in Guizhou province as research object, combined with the measured data, the DEPSO method is used to identify the parameters of the steam turbine regulation system. The simulation check results show that the steam turbine regulation system identification model can reflect the power response characteristics of the actual system, which proves the validity of hybrid algorithm to solve this problem.