Analysis of higher‐order gas turbine plant in real time would be tedious and expensive. In order to overcome this complexity, reduced‐order model for 5001M heavy‐duty gas turbine rated 18.2 MW has been obtained by Routh approximation, clustering technique, modified pole clustering, eigen permutation, Mihailov criterion, and Padé approximation algorithms. The step responses are obtained using MATLAB/Simulink and compared based on time domain specifications and performance index criteria. It indicates that the mixed method, namely, Routh approximation–Padé approximation algorithm–based reduced‐order model, retains the original characteristics. Further, particle swarm optimization (PSO) algorithm has also been applied to develop an optimal reduced‐order model. Based on the dynamic response against the load disturbance and set point variations, PSO‐based reduced‐order model has been identified as an optimal reduced‐order model for heavy‐duty gas turbine. The reduced‐order model proposed in this paper will be suitable for analyzing the dynamic behavior of heavy‐duty gas turbine plants in real‐time environment.

中文翻译：

---

利用粒子群优化技术开发燃气轮机最优降阶模型

实时分析高阶燃气轮机设备既繁琐又昂贵。为了克服这种复杂性，已经通过劳斯逼近，聚类技术，改进的极点聚类，特征置换，Mihailov准则和Padé逼近算法获得了额定值为18.2 MW的5001M重型燃气轮机的降阶模型。使用MATLAB / Simulink获得阶跃响应，并根据时域规范和性能指标标准进行比较。它表明混合方法，即基于Routh近似-Padé近似算法的降阶模型，保留了原始特征。此外，粒子群优化（PSO）算法也已用于开发最佳降阶模型。根据对负载干扰和设定点变化的动态响应，基于PSO的降阶模型已被确定为重型燃气轮机的最佳降阶模型。本文提出的降阶模型将适合于分析实时环境中重型燃气轮机的动态行为。