Mathematical Problems in Engineering / 2016 / Article / Tab 4 / Research Article
Multistep Wind Speed Forecasting Using a Novel Model Hybridizing Singular Spectrum Analysis, Modified Intelligent Optimization, and Rolling Elman Neural Network Table 4 Parameters of the hybrid model.
Experimental parameters Default value PSO Acceleration coefficient 1.494 Inertia weight 1 Maximum number of iterations 1000 The number of particles 20 Particle velocity [−0.5, 0.5] Particle positions [−5, 5] FA The population size 20 Maximum iteration number 50 Absorption coefficient 1 Light intensity coefficient 0.2 Step size 0.5 Firefly positions [−5, 5] ENN Input layer 4 Middle layer 5 Output layer 1 Iteration time 1000 Training requirement accuracy 0.000001 Learning rate 0.1