In this paper, a novel approach is proposed for the reduced order modelling of linear time invariant (LTI) systems. The proposed approach is a combination of modified Big bang big crunch (BBBC) optimization algorithm and Pade approximation technique. The beauty of the proposed approach is that the selection of solution space for BBBC algorithm is not entirely random, but structured via the use of Pade approximation approach. Hence, two principal criticisms of soft computing algorithms, i.e., random choice of solution space and larger simulation time are averted in the proposed technique. The proposed technique is substantiated via four different numerical examples from literature and compared with existing model order reduction (MOR) techniques. The concept of controller design is introduced via application of fractional order internal model control technique for load frequency control of power systems. Further, BBBC algorithm is employed to tune a boiler loop in power station. The results convey the efficiency and powerfulness of the proposed technique.

在本文中，提出了一种新颖的方法来对线性时不变（LTI）系统进行降阶建模。所提出的方法是改进的大爆炸（BBBC）优化算法和Pade逼近技术的结合。所提出的方法的优点在于，BBBC算法的解空间选择不是完全随机的，而是通过使用Pade逼近方法进行结构化的。因此，在所提出的技术中避免了对软计算算法的两个主要批评，即解决方案空间的随机选择和较大的仿真时间。所提出的技术通过文献中的四个不同数值示例得到了证实，并与现有的模型降阶（MOR）技术进行了比较。通过分数阶内部模型控制技术在电力系统负荷频率控制中的应用，引入了控制器设计的概念。此外，BBBC算法被用来调整电站中的锅炉回路。结果传达了所提出技术的效率和强大性。