Scholarlink

<p>This research proposal employs an upgraded residualization approach to approximate higher-order multi-input multi-output (MIMO) interconnected power system (IPS) models. The simplified strategy of the offered work is valuable for model order reduction (MOR) of the IPS models. The focus is on generating reduced-order systems (ROSs) by residualizing a balanced realization’s less visible and controlled states. Nevertheless, this technique does not keep the original Markov parameters (MakPms) and time moments (TiMts), leading to a considerable loss of accuracy in the high-frequency range. This article uses balanced residualization to get the ROS denominator polynomial to ensure stability. The numerator coefficients of the condensed model are generated by matching suitable TiMts and MakPms of the IPS models with the desired ROS. Time-domain characteristics (TDCs) such as rise time and settling time are considered when comparing the previous ROSs researchers obtained for IPS models using different methods. To improve readability, a comparative analysis is additionally provided. It takes into account various performance error criteria (PEC) such as root mean square error (RMSE), integral of squared error (ISE), integral of time absolute error (ITAE), integral of absolute error (IAE) and is provided. The percentage enhancement in TDC and PEC is tabulated, and bar graphs were drawn to help improve comprehension. The usefulness of the suggested strategy for order reduction of the IPS models is demonstrated by the tabulated values of TDCs and PEC, as well as step and Bode graphs. In addition, the proposed method’s effectiveness and efficiency are illustrated using an eighty-fourth-order system and controller design.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Enhanced Residualization Method for Model Reduction of Practical Systems and Application in Controller Design

  • 期刊论文

  • 发表日期：

    2026年2月28日

查看全文
  • Bala Bhaskar Duddeti,
  • Asim Kumar Naskar

摘要

This research proposal employs an upgraded residualization approach to approximate higher-order multi-input multi-output (MIMO) interconnected power system (IPS) models. The simplified strategy of the offered work is valuable for model order reduction (MOR) of the IPS models. The focus is on generating reduced-order systems (ROSs) by residualizing a balanced realization’s less visible and controlled states. Nevertheless, this technique does not keep the original Markov parameters (MakPms) and time moments (TiMts), leading to a considerable loss of accuracy in the high-frequency range. This article uses balanced residualization to get the ROS denominator polynomial to ensure stability. The numerator coefficients of the condensed model are generated by matching suitable TiMts and MakPms of the IPS models with the desired ROS. Time-domain characteristics (TDCs) such as rise time and settling time are considered when comparing the previous ROSs researchers obtained for IPS models using different methods. To improve readability, a comparative analysis is additionally provided. It takes into account various performance error criteria (PEC) such as root mean square error (RMSE), integral of squared error (ISE), integral of time absolute error (ITAE), integral of absolute error (IAE) and is provided. The percentage enhancement in TDC and PEC is tabulated, and bar graphs were drawn to help improve comprehension. The usefulness of the suggested strategy for order reduction of the IPS models is demonstrated by the tabulated values of TDCs and PEC, as well as step and Bode graphs. In addition, the proposed method’s effectiveness and efficiency are illustrated using an eighty-fourth-order system and controller design.