Academia.eduAcademia.edu

Figure 19 – uploaded by Antonio Ficarella

See full PDFdownloadDownload figure
Fig. 19. Error distribution of the MLP networks (training 3 years, testing 2 years, prediction length 1 h).

Figure 19 Error distribution of the MLP networks (training 3 years, testing 2 years, prediction length 1 h).

Related Figures (26)

Fig. 1. Normalised hourly wind power vs hourly temperature in the month of March, year V.
Fig. 1. Normalised hourly wind power vs hourly temperature in the month of March, year V.
Fig. 4. Normalised hourly wind power vs hourly wind speed in the month of March year IV. Fig. 3. Normalised hourly wind power vs hourly wind speed in the month of September, year IV.
Fig. 4. Normalised hourly wind power vs hourly wind speed in the month of March year IV. Fig. 3. Normalised hourly wind power vs hourly wind speed in the month of September, year IV.
Fig. 2. Normalised hourly wind power vs hourly pressure in the month of March, year V.
Fig. 2. Normalised hourly wind power vs hourly pressure in the month of March, year V.
TRAINGDX = Gradient descent with momentum and adaptive learning rate backpropagation LEARNGD = Gradient descent weight and bias learning function. MSE = Mean squared error. TANSIG = Hyperbolic tangent sigmoid transfer function. PURELIN = Linear transfer function. Elman network (2 inputs) parameters used in the training for each prediction length. Table 2
TRAINGDX = Gradient descent with momentum and adaptive learning rate backpropagation LEARNGD = Gradient descent weight and bias learning function. MSE = Mean squared error. TANSIG = Hyperbolic tangent sigmoid transfer function. PURELIN = Linear transfer function. Elman network (2 inputs) parameters used in the training for each prediction length. Table 2
TRAINLM = Levenberg-Marquardt algorithm. LEARNGD = Gradient descent weight and bias learning function. MSE = Mean squared error. TANSIG = Hyperbolic tangent sigmoid transfer function. PURELIN = Linear transfer function. MLFF network parameters used in the training for each prediction length. Table 1
TRAINLM = Levenberg-Marquardt algorithm. LEARNGD = Gradient descent weight and bias learning function. MSE = Mean squared error. TANSIG = Hyperbolic tangent sigmoid transfer function. PURELIN = Linear transfer function. MLFF network parameters used in the training for each prediction length. Table 1
Fig. 9. Typical architecture of an Elman back-propagation network.
Fig. 9. Typical architecture of an Elman back-propagation network.
LEARNP = Perceptron weight and bias learning function. MSE = Mean squared error. TANSIG = Hyperbolic tangent sigmoid transfer function. PURELIN = Linear transfer function. MLP network (2 inputs) parameters used in the training for each prediction length. Table 4
LEARNP = Perceptron weight and bias learning function. MSE = Mean squared error. TANSIG = Hyperbolic tangent sigmoid transfer function. PURELIN = Linear transfer function. MLP network (2 inputs) parameters used in the training for each prediction length. Table 4
LEARNP = Perceptron weight and bias learning function. MSE = Mean squared error. TANSIG = Hyperbolic tangent sigmoid transfer function. PURELIN = Linear transfer function. MLP network (1 input) parameters used in the training for each prediction length. Table 3
LEARNP = Perceptron weight and bias learning function. MSE = Mean squared error. TANSIG = Hyperbolic tangent sigmoid transfer function. PURELIN = Linear transfer function. MLP network (1 input) parameters used in the training for each prediction length. Table 3
GBELLMF = Generalised bell-shaped built-in membership function. LINEAR = Linear function. HYBRID = Combination of backpropagation and the least-squares method to estimate membership function parameters. ANFIS parameters used in the training for each prediction length. Table 5
GBELLMF = Generalised bell-shaped built-in membership function. LINEAR = Linear function. HYBRID = Combination of backpropagation and the least-squares method to estimate membership function parameters. ANFIS parameters used in the training for each prediction length. Table 5
Fig. 11. ARMA models: normalised absolute average error vs prediction length (various combination of training-testing periods)
Fig. 11. ARMA models: normalised absolute average error vs prediction length (various combination of training-testing periods)
Fig. 12. Normalised absolute average error vs prediction length (training 3 years testing 2 years).
Fig. 12. Normalised absolute average error vs prediction length (training 3 years testing 2 years).
Fig. 13. Normalised absolute average error vs prediction length (training 4 years, testing 1 years).
Fig. 13. Normalised absolute average error vs prediction length (training 4 years, testing 1 years).
Normalised absolute average percentage error in the several testing sets. Table 7
Normalised absolute average percentage error in the several testing sets. Table 7
Fig. 14. Observed vs predicted values (training 3 years, testing 2 years, prediction length 1 h).
Fig. 14. Observed vs predicted values (training 3 years, testing 2 years, prediction length 1 h).
Fig. 16. Observed vs predicted values (training 3 years, testing 2 years, prediction length 12 h).
Fig. 16. Observed vs predicted values (training 3 years, testing 2 years, prediction length 12 h).
Fig. 17. Error distribution of some forecast models (ARMA, ANFIS, MLFF, training 3 years, testing 2 years, prediction length 1 h).
Fig. 17. Error distribution of some forecast models (ARMA, ANFIS, MLFF, training 3 years, testing 2 years, prediction length 1 h).
Fig. 18. Error distribution of the Elman networks (training 3 years, testing 2 years, prediction length 1 h). +20%] it becomes evident that in most of the cases it is appre- ciably higher with Elman neural networks, ensuring that the If the probability is focused upon that the normalised error takes in values in the ranges [—10%;+10%] and [—20%;
Fig. 18. Error distribution of the Elman networks (training 3 years, testing 2 years, prediction length 1 h). +20%] it becomes evident that in most of the cases it is appre- ciably higher with Elman neural networks, ensuring that the If the probability is focused upon that the normalised error takes in values in the ranges [—10%;+10%] and [—20%;
Probability that the normalised error takes values in the range [—10%; +10%]. Table 8
Probability that the normalised error takes values in the range [—10%; +10%]. Table 8
Probability that the normalised error takes values in the range [—20%; +20%]. Table 9
Probability that the normalised error takes values in the range [—20%; +20%]. Table 9
Related topics:
EngineeringEconomicsadaptive neuro fuzzy inference system (ANFIS)Physical ModellingWind FarmAuto Regressive Moving Average
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved