Volterra filter modelling of non-linear system using Artificial Electric Field algorithm assisted Kalman filter and...

Area of Science:

• Control Systems Engineering
• Signal Processing
• Computational Intelligence

Background:

• Conventional Kalman Filter (KF) parameter tuning presents challenges, often leading to system divergence.
• Meta-heuristic algorithms offer potential solutions for optimizing KF performance.

Purpose of the Study:

• To enhance the identification efficiency of non-linear systems.
• To address the parameter tuning and divergence problems inherent in the conventional KF.
• To introduce a novel approach using the Artificial Electric Field (AEF) algorithm to optimize KF parameters.

Main Methods:

• The proposed method converts the system identification model into a measurement problem.
• The Artificial Electric Field (AEF) algorithm optimizes the KF parameters by incorporating a fitness function with KF equations.
• The optimized KF is then used for non-linear system identification.

Main Results:

• The AEF-assisted Kalman filter (AEF-KF) demonstrated improved performance across five distinct non-linear models under noisy conditions.
• Evaluations considered parameter estimation error, Mean Squared Error (MSE), and fitness percentage.
• The AEF-KF showed superior convergence speed and computational efficiency compared to KF, Kalman Smoother (KS), and other stochastic algorithms.

Conclusions:

• The proposed AEF-KF method is effective and robust for non-linear system identification.
• The approach significantly improves upon existing methods in terms of accuracy and efficiency.
• The AEF-KF shows practical applicability for real-world non-linear benchmark systems.