Enzyme action optimizer based infinite impulse response filter identification through a comprehensive benchmark...

Area of Science:

• Digital Signal Processing
• Computational Intelligence
• Optimization Algorithms

Background:

• Identifying infinite impulse response (IIR) filter models is challenging due to nonlinearities.
• Accurate and stable identification requires advanced optimization techniques.

Purpose of the Study:

• To evaluate the enzyme action optimizer for adaptive IIR filter identification.
• To assess its performance across various system complexities and orders.

Main Methods:

• Testing the enzyme action optimizer on four benchmark IIR systems.
• Comparing its performance against starfish optimization, hippopotamus optimizer, and grey wolf optimization.
• Evaluating results using mean squared error, mean absolute error, standard deviation, and convergence speed.

Main Results:

• Enzyme action optimizer achieved perfect reconstruction (zero MSE) in full-order, low-order systems.
• It consistently showed lower errors and stable convergence in reduced-order scenarios.
• The algorithm outperformed comparative methods in high-order and asymmetric systems for accuracy and speed.

Conclusions:

• Enzyme action optimizer is highly effective for complex IIR identification.
• Its bio-inspired design with adaptive parameters balances exploration and exploitation.
• The algorithm shows potential for broader optimization applications in signal processing.