Chaos time series prediction based on membrane optimization algorithms.

Area of Science:

• Computational Intelligence
• Time Series Analysis
• Electromagnetic Spectrum Management

Background:

• Accurate prediction of electromagnetic environment parameters is crucial for effective spectrum management.
• Chaos time series analysis is essential for understanding complex dynamic systems.

Purpose of the Study:

• To develop a novel prediction model for chaos time series using membrane computing optimization.
• To optimize phase space reconstruction and least squares support vector machine parameters simultaneously.
• To forecast band occupancy rates in the FM broadcasting and interphone bands.

Main Methods:

• Utilized a membrane computing optimization algorithm to tune parameters (τ, m) for phase space reconstruction.
• Optimized parameters (γ, σ) for the least squares support vector machine (LS-SVM) model.
• Applied the developed model to predict frequency modulation (FM) broadcasting and interphone band occupancy rates.

Main Results:

• The proposed model demonstrated superior performance in both single-step and multi-step predictions.
• Evaluated model accuracy using normalized mean square error (NMSE), root mean square error (RMSE), and mean absolute percentage error (MAPE).
• Experimental results confirmed the model's effectiveness and superiority over conventional methods.

Conclusions:

• The membrane computing optimization algorithm effectively enhances chaos time series prediction accuracy.
• The developed model provides a reliable tool for predicting electromagnetic environment parameters, aiding spectrum management decisions.
• The model's superior performance validates its applicability and advantages in forecasting band occupancy rates.