Artificial hummingbird optimizer as a novel adaptive algorithm for identifying optimal coefficients of digital IIR filtering systems

This paper introduces the artificial hummingbird algorithm (AHA) as a novel and efficient algorithm for designing digital infinite impulse response (IIR) filtering systems. IIR filters are commonly used in various applications, and their design often involves complex error surfaces and coefficients. The AHA aims to address these challenges by optimizing the IIR model using mean squared error (MSE) cost function. Four benchmark examples are considered for evaluation by comparing both the same and reduced order cases of the IIR model. The results are compared against recent and efficient algorithms named prairie dog optimization, whale optimization algorithm, and artificial bee colony algorithm. The evaluation includes an analysis of the obtained coefficients, statistical measures, and convergence profiles. The findings show that the AHA outperforms the competing algorithms, achieving more accurate models and providing better statistical performance. Additionally, the AHA consistently converges to the lowest MSE values for each case. Further assessments are conducted using various examples from the literature, which validate the AHA's superior ability in digital IIR filtering design. The algorithm demonstrates more accurate identification and achieves lower MSE values compared to alternative methods which highlight the promising potential of the AHA for efficient and precise digital IIR filtering system design.