An improved artificial rabbits optimization for accurate and efficient infinite impulse response system identification

Identifying models with Infinite Impulse Response (IIR) is crucial in signal processing and system identification. This paper addresses the challenges of IIR model identification by proposing an improved version of the Artificial Rabbits Optimization (ARO) algorithm called improved ARO (IARO). The IARO algorithm integrates an adaptive local search mechanism and an experience-based perturbed learning strategy as two key enhancements to improve the effectiveness of ARO. These additions aim to address the loss of accuracy during iterations and improve the algorithm's ability to exploit promising areas. Four benchmark examples of different IIR plants are considered, and the performance of the proposed IARO algorithm is compared to existing competitive methods. The results consistently demonstrate that IARO outperforms ARO in accuracy and convergence for system identification across all orders of IIR systems. Visual analysis, convergence curves, coefficient comparison, and statistical metrics comparison all validate the superiority of the IARO algorithm. Additionally, the Wilcoxon signed-rank test results provide further statistical evidence supporting the superior performance of IARO. The comprehensive analysis showcases the efficacy and effectiveness of the IARO algorithm in accurately identifying IIR systems. This work represents a significant advancement in IIR system identification, offering a superior methodology for accurate and efficient system modeling.