1000/1000
Hot
Most Recent
Artificial neural networks (ANNs) have become key methods for achieving global climate goals. The applications of ANNs to renewable energies such as solar, wind, and tidal energy were studied.
Authors and Year |
ANN Type and Structure | Journal | Country/ Region |
I/O Setting |
Activation Function | Notes | ||
---|---|---|---|---|---|---|---|---|
Input | Output | |||||||
1 | [44] | Multilayer Perceptron (MLP) 3-4-1 |
Renewable Energy | Muppandal, India |
Wind speed (Ws), relative humidity (RH), generation hours |
Energy output of wind farms | logsig (hidden layer) purelin (output layer) |
Trained by BP algorithm Input data normalized to [0, 1] |
2 | [45] | ANN 4-X-2 |
Renewable Energy | Turkey | Longitude (lon), latitude (lat), altitude (A), measurement height |
Ws, related power | logsig (hidden and output layer) |
Trained by BP algorithm Input and Output data normalized to [0, 1] |
3 | [46] | MLP 5-10-5-1 |
Renewable Energy | Turkey | Ws, month (M) | Ws | logsig (hidden layer) purelin (output layer) |
Resilient propagation (RP) algorithm was adopted |
4 | [47] | Radial Basis Function (RBF) 1-7-2 |
Renewable and Sustainable Energy Reviews |
Iran | Ws | Proportional and integral (PI) gains | - | Use Gaussian function for hidden layer Gravitational search algorithm (GSA) is adopted |
5 | [48] | MLP 3-(2-100)-24 |
Renewable Energy | Medina city, Saudi Arabia | Mean daily Ws | Ws prediction of the next day |
tansig (hidden layer) purelin (output layer) |
Input data normalized to [0, 1] Trained by Levenberg–Marquardt (LM) BP algorithm Compared and outperforms support vector machine (SVM) SVM used Gaussian kernel 2000 days used for training and 728 days used for testing |
6 | [49] | MLP 6-7-5-1 MLP 4-7-5-1 |
Renewable and Sustainable Energy Reviews |
Alberta, Canada |
Wind power (Wp) WP1 (t − 1), WP1 (t − 2), WP1 (t − 3), WP1 (t − 4), WP1 (t − 5), WP1 (t − 6) |
Short-term forecasting of the Wp time series | tansig (hidden layer) purelin (output layer) |
Input data normalized to [−1, 1] Imperialist competitive algorithm (ICA), GA, and particle swarm optimization (PSO) are employed for training the neural network 1200 data used for training and 168 data used for testing |
7 | [50] | ANN 2-(16-32)-(16-32)-1 |
Renewable Energy | Coquimbo, Chile | Ws, wind direction (Wd) |
Turbine power | - | ADAM algorithm is adopted 103,308 data used for training and 52,560 data used for testing |
8 | [51] | RBF 2-3-1 MLP 2-4-1 ADALINE 2-4-1 |
Applied Energy | North Dakota, USA | Mean hourly Ws | Forecast value of next hourly average Ws | - | Trained by LM algorithm 5000 data used for training and 120 data used for testing |
9 | [52] | MLP 5-5-3 |
Renewable Energy | Guadeloupean archipelago, French West Indies | Ws, 30 min moving average speed | Wp (t + kt) | tansig (hidden layer) purelin (output layer) |
Bayesian regularized (BR) |
10 | [53] | ANN 7-20-1 |
Renewable Energy | China | Actual Ws, Wp | Ws | - | Trained by BP algorithm |
11 | [54] | MLP 6-7-1 |
Renewable Energy | Albacete, Spain | Wsp1, Wsp2, temperature (T) Tp2, solar cicle1, solar cicle2, Wdp1 | Ws forecast (48 h later) | - | LM algorithm is adopted |
12 | [55] | ANN 3-3-1 ANN 3-2-X ANN 3-1 ANN 2-1 |
Renewable Energy | Oaxaca, México | Previous values of hourly Ws |
Current value of Ws | - | 550 data used for training and 194 data used for testing |
13 | [56] | ANN - |
Renewable Energy | Basque Country, Spain |
Ws data in the last 3 h | Ws in 1 h | sigmoid (output layer) |
Trained by BP algorithm |
14 | [57] | MLP X-8-X |
Renewable Energy | Rostamabad, Iran | Standard deviation, average, slope |
Ws (k + l), …, Ws (k + 2), Ws (k + 1) | - | Trained by BP algorithm 672 patterns used for training |
15 | [58] | MLP 5-3-3-1 |
Communications in Nonlinear Science and Numerical Simulation |
Italy | Ws, RH, generation hours, T, maintenance hours |
Total wind energy | tansig (first hidden layer) sigmoid (second hidden layer) purelin (output layer) |
Trained by BP algorithm |
16 | [59] | MLP 6-25-1 |
Renewable Energy | Himachal Pradesh, India |
Average temperature (TAVG), maximum temperature (Tmax), minimum temperature (ax), air pressure (Pair), solar irradiance (G), A |
Average daily Ws for 11 H.P. locations |
- | Trained by LM algorithm Scaled conjugate gradient (SCG) algorithm is adopted Input and target data are normalized to [−1, 1] 60% data used for training, 20% used for testing, and 20% used for validation |
17 | [60] | MLP 4-15-15-1 |
Applied Energy |
Nigeria | lat, lon, A, M | Mean monthly Ws | tansig (hidden layers) purelin (output layer) |
SCG and LM algorithms are adopted Input and target data normalized to [−1, 1] |
18 | [61] | MLP 14-15-1 |
WSEAS Transactions on Systems | Portugal | Average hourly values of Ws | Average hourly Ws | - | Trained BP algorithm 87.75% patterns used for training, 9.75% used for validation, and 2.5% used for testing |
19 | [62] | MLP 5-6-6-6-2 |
- | Cyprus | M, mean monthly values of Ws at two levels (2 and 7 m) |
Mean monthly values of Ws of a third station |
tansig (hidden layer) logsig (output layer) |
Trained by BP algorithm 90% patterns used for training and 10% patterns used for testing |
20 | [63] | MLP 9-10-1 |
Energy Conversion and Management |
Marmara, Turkey |
9 stations Ws | Ws | - | Trained by BP algorithm |
21 | [64] | MLP 4-8-1 |
Theoretical and Applied Climatology | Tabriz, Azerbaijan, Iran |
Pair, air temperature (Tair), RH, precipitation |
Monthly Ws | logsig (hidden layer) purelin (output layer) |
Trained by LM algorithm Input and output data normalized to [0, 1] 75% of data used for training and 25% used for testing |
22 | [65] | MLP 31-63-31 |
Knowledge-Based Systems |
Minqin, China | Historical daily average Ws during March previous year |
Daily average Ws during March target year |
tansig (hidden layer) logsig (output layer) |
Trained by BP algorithm |
23 | [66] | MLP X-25-1 |
2014 4th IEEE International Conference on Information Science and Technology |
Colorado, USA | T, RH, Wd, wind gust, pressure (P), historical Ws | Ws | tansig (hidden layer) |
Trained by BP with momentum 1000 input/output pairs used for training and 200 input/output pairs used for testing |
Authors and Year |
ANN Type and Structure |
Journal | Country/ Region |
I/O Setting | Activation Function | Notes | ||
---|---|---|---|---|---|---|---|---|
Input | Output | |||||||
1 | [74] | RBF 6-11-24 RBF 6-15-24 |
Solar Energy | Huazhong, China | G (t + 1), Ws (t + 1), Tair (t + 1), RH (t + 1), t, power (Pw) (t) | Pw1 (t + 1), Pw2 (t + 1), …, Pw24 (t + 1) | - | k-fold (validation) Input and output data normalized to [0, 1] |
2 | [75] | MLP 2-3-1 |
Renewable Energy | Jaen, Spain | G, module cell temperature (TC) |
G, ambient temperature (Ta) | - | Trained by LM BP algorithm |
3 | [76] | MLP 3-3-1 MLP 4-3-1 |
Energy | Corsica Island, France Bastia Ajaccio |
RH, sunshine duration (S), nebulosity (Y) Y, S, P, differential pressure (DGP) |
Global radiation (GR) |
tansig (hidden layer) purelin (output layer) |
Trained by LM algorithm Input data normalized to [−1, 1] 80% data used for training, 10% for validation, and 10% used for testing |
4 | [77] | MLP 8-3-1 |
Solar Energy | Ajaccio, Corsica Island, France |
Clearness index (KT) KTt−1, KTt−2, KTt−3, KTt−4, KTt−5, KTt−6, KTt−7, KTt−8 | Daily global solar radiation (GSR) |
purelin (output layer) | Trained by LM algorithm Use Gaussian function for hidden layer Input data normalized to [0, 1] 80% data used for training, 10% for validation, and 10% used for testing |
5 | [78] | MLP 3-11-17-24 |
Solar Energy | Trieste, Italia | G, Tair, hour or day (t) | G1 (t + 1), G2 (t + 1), …, G24 (t + 1) | - | Trained by LM BP Algorithm k-fold validation Input and output data normalized to [−1, 1] |
6 | [79] | RBFN (2-3-4)- (4-5-7)-1 MLP (2-3-4)- (2-3-5)-1 |
Energy | Al-Medina, Saudi Arabia | Tair, S, RH, t | Daily global solar radiation (GD) |
- | 1460 data used for training and 365 data used for testing |
7 | [80] | MLP 6-5-1 |
Applied Energy |
Turkey | lat, lon, A, M, S, T | G | logsig (hidden layer) | Trained by BP algorithm SCG, Pola–Ribiere conjugate gradient (CGP), and LM algorithms are adopted Input and output data normalized to [−1, 1] |
8 | [81] | MLP 3-20-1 |
Renewable and Sustainable Energy Reviews |
Morocco | lon, lat, A | Mean annual and monthly G | - | Trained by BP algorithm Input and output data normalized to [0, 1] |
9 | [82] | MLP 2-36-1 MLP 3-20-1 |
Energy Sources, Part A: Recovery, Utilization, and Environmental Effects |
Abha, Saudi Arabia | Tair, RH, hour or day (t) | Diffuse solar radiation (DSR) |
logsig (hidden layer) | Trained by BP algorithm 1462 days used for training and 250 days used for testing |
10 | [83] | MLP 5-8-1 |
Expert Systems with Applications |
Anatolia, Turkey |
lat, lon, A, S, average cloudiness | G | tansig (hidden layer) purelin (output layer) |
Trained by BP algorithm |
11 | [84] | MLP 7-5-1 |
Applied Energy |
Nigeria | lat, lon, A, M, S, T, RH | G | tansig (hidden layer) purelin (output layer) |
SCG and LM algorithms are adopted Input data normalized to [−1, 1] 11,700 datasets used for training and 5850 datasets used for validation and testing |
12 | [85] | MLP 4-X-1 |
International Journal of Photoenergy | Malaysia | lat, lon, day or hour (t), S | KT | logsig (hidden layer) | Trained by BP algorithm |
13 | [86] | MLP 4-4-1 |
International Journal of Computer Applications | India | lat, lon, S, A | G | tansig (hidden layer) purelin (output layer) |
LM algorithm is adopted |
14 | [87] | MLP 5-40-1 |
Energy | Egypt | GSR, like long-wave atmospheric emission, Tair, RH, P | Diffuse fraction (KD) |
sigmoid (output layer) | Trained by BP algorithm |
15 | [88] | MLP 7-15-1 |
Solar Energy | Jaen, Spain | t (day), t (hour), KT, hourly clearness index (kt) kt−1, kt−2, kt−3, S | Solar radiation maps |
- | Trained by BP algorithm (with momentum and random presentations) Input data normalized to [0, 1] |
16 | [89] | MLP 6-X-1 |
Solar Energy | Helwan, Egypt | Wd, Ws, Ta, RH, cloudiness, water vapor | G | sigmoid (output layer) | Trained by LM BP algorithm Input data normalized to [0, 1] |
17 | [90] | MLP 2-X-1 MLP 3-X-1 MLP 3-X-X-1 |
Solar Energy | Athalassa, Cyprus |
S, theoretical sunshine duration (S0d), M, Tmax |
GD | tansig (hidden layer) | Trained by BP algorithm 90% data used for training and 10% used for testing |
18 | [91] | MLP 7-9-1 |
Renewable Energy | India | lat, lon, A, M, S, rainfall ratio, RH |
KT | tansig (hidden layer) purelin (output layer) |
Trained by BP algorithm |
19 | [92] | MLP 2-5-1 |
Energy Policy | China | Kt, S (%) | Monthly mean daily KD | tansig (hidden layer) purelin (output layer) |
Trained by BP algorithm TRAINLM algorithm is adopted Input and output data normalized to [0, 1] |
20 | [93] | MLP 6-15-1 |
Solar Energy | Uganda | S, Tmax, Total Cloud Cover (TCC), lat, lon, A | Monthly average daily GSR on a horizontal surface |
tansig (hidden layer) purelin (output layer) |
Trained by LM BP algorithm Input data normalized to [−1, 1] |
21 | [94] | MLP 6-6-1 |
Applied Energy |
Turkey | lat, lon, A, M, DSR, mean beam radiation | G | logsig (hidden layer) purelin (output layer) |
SCG and RP algorithms are adopted |
22 | [95] | GRNN 6-1.0-1 |
Energy | Turkey | lat, lon, A, surface emissivity (ε4), surface emissivity (ε5), land surface temperature |
G | - | - |
23 | [96] | MLP 7-4-1 |
Energy Conversion and Management |
Iran | Tmax, Tmin, RH, VP, total precipitation, Ws, S |
GSR | logsig (hidden layer) purelin (output layer) |
Trained by BP algorithm 65 months used for training and 7 months used for testing |
24 | [97] | ANN 6-6-1 |
Applied Energy | Turkey | lat, lon, A, M, S, T | G | logsig (hidden layer) | SCG, CGP, and LM algorithms are adopted Trained by BP algorithm Input and output data normalized to [−1, 1] |
25 | [98] | MLP 3-6-1 |
Renewable Energy | Khuzestan, Iran | Tmax, Tmin, extra-terrestrial radiation (Ra) |
GSR | logsig (hidden layer) | Trained by LM BP algorithm Input data normalized to [0, 1] 70% data used for training and 30% patterns used for testing |
26 | [99] | MLP 5-3-1 |
Energy Procedia |
Bechar, Algeria |
M, t (day), t (hour), T, RH |
GSR | tansig (hidden layer) purelin (output layer) |
Trained by LM BP algorithm 81% data used for training and 19% used for testing |
27 | [100] | MLP 9-11-1 |
Renewable and Sustainable Energy Reviews |
Republic of Indonesia |
T, RH, S, Ws, precipitation, lon, lat, A, M |
GSR | - | Trained by BP algorithm |
28 | [101] | RBF 4-50-2 |
Energy Sources, Part A: Recovery, Utilization, and Environmental Effects | Saudi Arabia | Ta, RH, GSR, t | DSR, direct normal radiation (DNR) |
- | Use Gaussian function for hidden layer 1460 values used for training and 365 values used for testing |
29 | [102] | MLP 8-15-1 |
Renewable Energy | Sultanate of Oman | Location (L), M, P, T, VP, RH, WS, S | GR | - | Trained by BP algorithm |
Authors and Year |
ANN Type and Structure |
Journal | Country/ Region |
I/O Setting | Activation Function | Notes | ||
---|---|---|---|---|---|---|---|---|
Input | Output | |||||||
1 | [104] | ANN 28-15-4 28-9-4 28-4-4 28-7-4 |
Journal of Atmospheric and Oceanic Technology | Gulf of Maine, Gulf of Alaska, Gulf of Mexico | 7 days of significant H | 6, 12, 18, 24 h forecast | logsig (hidden and output layer) |
Input data normalized to [0, 1] Conjugate gradient algorithm with Fletcher–Reeves is adopted |
2 | [105] | MLP 3-5-5-2 |
Ocean Engineering |
Bombay, India | Deep water wave height (Ho), wave energy period (Te) |
Breaking wave height (Hb), water depth at the time of breaking (db) | sigmoid (output layer) |
Trained by BP algorithm Input and output data normalized to [0, 1] |
3 | [106] | MLP 48-97-24 |
Ocean Engineering |
Ireland | 48 h history wave parameters |
H and zero-up- crossing peak wave period (Tp) over hourly intervals from 1 h to 24 h |
logsig (hidden layer) purelin (output layer) |
Trained by resilient BP algorithm |
4 | [107] | MLP 6-5-1 |
Proceedings of the Institution of Civil Engineers-Maritime Engineering |
Anzali, Iran | H, Tp | Energy flux (Fe) over horizon of 1 to 12 h |
sigmoid (output layer) |
Conjugate gradient algorithm is adopted 80% data used for training and 20% used for testing |
5 | [108] | MLP 2-4-3 MLP 4-4-4 |
Ocean Engineering |
Karwar, India | Ws | 3-hourly values of H and average cross-period | - | Trained by BP algorithm 80% data used for training and 20% data used for testing |
6 | [109] | Deep Neural Network (DNN) 6-64-32-32-1 |
Ocean Engineering |
Pacific and Atlantic coasts and the Gulf of Mexico |
H, Te, Fe, weighted average period, Tp, Ws, Wd |
Fe, Te, H | - | SCG BP algorithm is adopted Input data normalized to [0, 1] 75% data used for training and 25% data used for testing |
7 | [110] | MLP 3-300-300-2 |
Ocean Engineering |
Lake Michigan, United Sates of America | Wind field, db, ice coverage |
H, Te | ReLU (hidden layer) |
Stochastic gradient-based algorithm is adopted 80% data used for training and 20% data used for testing |
8 | [111] | MLP 1-x-1 |
Marine Structures | Goa, India | H | Fe | sigmoid (output layer) |
Trained by BP cascade correlation algorithms 80% patterns used for training and 20% patterns used for testing |
9 | [112] | MLP 6-5-1 |
Ocean Engineering |
Persian Gulf | Ht, Ht−1, Ht−2, Utcos(Φt − θ), Ut−1cos(Φt – 1 − θt), Ut−2cos(Φt − 2 − θ2) |
H for the next 3, 6, 12, 24 h | sigmoid (output layer) |
Conjugate gradient and LM algorithms are adopted 80% data used for training and 20% data used for testing |
10 | [113] | MLP 3-4-4-1 |
Applied Soft Computing |
Spain | H, Te, θm | Fe | tansig (hidden layer) purelin (output layer) |
Trained by BP algorithm 67% data used for training and 33% data used for testing |
11 | [114] | MLP 3-3-1 |
Ocean Engineering |
Lake Superior, USA | Ws, weather station index (W) |
H | sigmoid (hidden and output layer) |
Trained by BP algorithm Input and output data normalized to [−1, 1] Compared with SVM, Bayesian networks, and adaptive neuro-fuzzy inference system (ANFIS) 345 patterns used for training and 54 patterns used for testing |
12 | [115] | MLP 5-2-1 |
Renewable Energy | Brazil | Wind shear velocity (U) U1, U2, Un, Y (t − 1), Y (t − i) |
Wave energy potential | tansig (hidden layer) purelin (output layer) |
Trained by LM BP algorithm 90% data used for training and 10% data used for testing |
13 | [116] | MLP X-15-1 |
Applied Ocean Research |
Canary Islands, Spain | H, Tp | Predict Fe | tansig (hidden layer) purelin (output layer) |
Gradient descent with momentum and BP algorithm are adopted 89% data used for training and 11% data used for testing Input and output data normalized to [−1, 1] |
14 | [117] | MLP 4-4-1 |
Applied Ocean Research |
India | H values of the preceding 3, 6, 12, and 24th hour | H subsequent 3, 6, 12 and 24th hour | - | Trained by LM BP algorithm 60% data used for training and 40% data used for testing |
15 | [118] | RBF 21-13-1 MLP 21-9-1 |
Marine Structures | India | H(1–21) | H(SW3) | - | Use Gaussian function for hidden layer BP, SCG, conjugate gradient Powell–Beale (CGB), Broyden–Fletcher–Goldfarb (BFG), and LM algorithms are adopted 80% data used for training and 20% data used for testing |
16 | [119] | MLP 8-4-1 MLP 2-2-1 |
Ocean Engineering |
Taiwan | Significant wave height (H1/3), highest one-tenth wave height (H1/10), highest wave height (Hmax), mean wave height (Hmean) (stations A and B) |
H1/3 (station C) |
sigmoid (output layer) |
Trained by BP algorithm Input data normalized to [0, 1] |
17 | [120] | MLP 2-5-1 |
Marine Structures | Yanam, India | Ht, Ht−1 | Ht+1 | - | Trained by BP algorithm Conjugate gradient and cascade correlation algorithms are adopted 80% data used for training and 20% data used for testing |
18 | [121] | ANN 9-1-1 ANN 4-1-1 ANN 9-8-1 ANN 9-1-1 |
Applied Ocean Research | Ratnagiri, Pondicherry, Gopalpur, Kollam, India |
t − 24, t − 21, t − 18, t − 15, t − 12, t − 9, t − 6, t − 3 |
t + 24 (24 h ahead predicted error) |
logsig (hidden layer) purelin (output layer) |
Trained by LM algorithm Input data normalized to [0, 1] 70% data used for training and 15% used for validation and testing |
19 | [122] | MLP 4-9-3 MLP 4-7-1 MLP 2-5-1 MLP 4-8-1 |
Applied Ocean Research | Lake Ontario, Canada/USA | Ws, Wd, fetch length, wind duration | H, Tp, (wave direction) Θ |
tansig/ sigmoid (hidden layer) purelin (output layer) |
Trained by BP algorithm 10-fold cross- validation used Input data normalized to [0, 1] 611 data used for training and 326 data used for testing |
20 | [123] | MLP 1-3-1 |
Ocean Engineering |
Lake Superior, Canada/USA | Ws | H | sigmoid (transfer function) | Compared and outperforms with model tree 4045 data used for training and 3259 data used for testing |