.png){kind=icon}
.png){kind=icon}
Return
Model order reduction for SISO and MIMO system using improved adaptive differential evolution algorithm
.png){kind=icon}
PRE
AI
DOI：10.1007/s00500-023-09489-8.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Mathematical models derived from the physical systems are usually complex and in the form of higher order differential equations. Such systems are difficult for analysis and controller synthesis. Therefore, it is desirable to develop an efficient algorithm for reducing such higher order systems to a lower order model by preserving all the significant characteristics of the original higher order system. For this, we have proposed an improved adaptive differential evolution algorithm (I-ADE), which is mixed with Routh approximation (RA) to determine the numerator and denominator coefficients of the corresponding lower order stable model (LOSM) by preserving the fundamental characteristics of a higher order stable system (HOSS). The superiority of the proposed method is illustrated by numerical test cases of single-input single-output (SISO) systems and multiple-input multiple-output (MIMO) systems. To evaluate the efficiency of proposed I-ADE algorithm, 23 benchmark functions are considered, and the results are statistically compared with nine promising meta-heuristic algorithms. The proposed I-ADE algorithm and the Routh approximation technique provide superior results in reducing the model order of SISO and MIMO systems. It achieves statistically the best rank among nine promising meta-heuristic algorithms employing the Friedman and Nemenyi Hypothesis test for optimizing 23 benchmark functions. This shows the proposed I-ADE algorithm's reliability, robustness and applicability for other optimization problems.
Keywords:
Model order reduction
Adaptive differential evolution based artificial neural network training
Routh approximation
SISO
MIMO
Journal
IF:
2.5
Papers:
1.0W
Citations:
2.1W
View more organizations