| Summary: | Simulated Kalman Filter (SKF) optimization algorithm is a population-based optimizer operated mainly based on Kalman filtering. The SKF is however subjected to premature convergence problem. In this research, opposition-based learning is employed to solve the premature convergence problem in SKF. The opposition-based learning can be applied either after the solution is updated or as the prediction step in SKF. Using CEC2014 benchmark suite, it is found that the SKF with opposition-based learning outperforms the original SKF algorithm in most cases. The SKF with opposition-based learning is also applied as adaptive beamforming algorithm for adaptive array antenna. In this application, the objective is to maximize the signal to interference plus noise ratio (SINR) and results show that the SKF with opposition-based learning outperforms the existing adaptive mutated Boolean particle swarm optimization (AMBPSO)
|
|---|