Biased reduction method by combining improved modified pole clustering and improved Pade approximations

• We propose a biased method for model order reduction of any linear dynamic system.
• It generates k−k− number of reduced order models for the kth−kth− order reduction.
• It maintains the stability, transient and steady state values of the original model.
• Decreasing error index will increase the performance of reduced models.

This paper presents a mixed method to reduce the order of the linear high order dynamic systems by combining improved modified pole clustering technique and improved Pade approximations. The denominator of the reduced order model is computed by improved modified pole clustering while the numerator coefficients are obtained by improved Pade approximations. The proposed method is competent in generating ‘k’ number of reduced order models from the original high order system. The superiority of the proposed method is concluded through numerical examples taken from literature and compared with existing well- known order reduction methods.