Ionosphere time series modeling using adaptive neuro-fuzzy inference system and principal component analysis

Razin, Mir Reza Ghaffari; Voosoghi, Behzad

doi:10.1007/s10291-020-0964-6

Cited by 25 publications

(11 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, existing statistical methods used to forecast regional TEC are based on relatively simple theoretical models. Among the existing approaches, Prophet time‐series forecasting (Zhai et al., 2019), auto‐regressive moving average (ARMA) (Krankowski et al., 2005; Zhang et al., 2013), statistical Holt–Winter (Elmunim et al., 2017), manifold trajectories (Moreno et al., 2018), support vector machine (Pei et al., 2019), artiﬁcial neural network (ANN) (Habarulema et al., 2007), EXtreme Gradient Boosting over Decision Trees (Zhukov et al., 2020), and hybrid methods (Feizi et al., 2020; Ghaffari & Vosooghi, 2020; Mukhtarov et al., 2014; Uwamahoro & Habarulema, 2015) are extensively used. Recently, various neural network (NN)‐based models have been developed for the prediction of regional TEC (Tebabal et al., 2018, 2019) A NN is a collection of algorithms modeled after the function of neurons in the brain.…”

Section: Introductionmentioning

confidence: 99%

Long Short‐Term Memory Neural Network for Ionospheric Total Electron Content Forecasting Over China

et al. 2021

View full text Add to dashboard Cite

• A novel extended encoder-decoder long short-term memory neural network (ED-LSTME) for ionospheric TEC forecasting over China is developed • ED-LSTME shows a strong capability in improving TEC forecasting at different geographical locations, seasons, and geomagnetic conditions • ED-LSTME is robust and outperforms all the six selected baselines when comparing with the models in terms of their performance Accepted Article This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as

show abstract

Section: Introductionmentioning

confidence: 99%

Long Short‐Term Memory Neural Network for Ionospheric Total Electron Content Forecasting Over China

et al. 2021

View full text Add to dashboard Cite

show abstract

“…Meanwhile the 4‐4‐4‐4 number of input MFs results in 256 rules. High number of rules would make the model too complex, and increases the convergence time of the model to the optimal solution (Alambeigi et al., 2016; Razin & Voosoghi, 2020).…”

Section: Resultsmentioning

confidence: 99%

Mathematical and intelligent modeling of stevia (Stevia Rebaudiana) leaves drying in an infrared‐assisted continuous hybrid solar dryer

Bakhshipour

Zareiforoush

Bagheri

2020

Food Science & Nutrition

View full text Add to dashboard Cite

Drying characteristics of stevia leaves were investigated in an infrared (IR)‐assisted continuous‐flow hybrid solar dryer. Drying experiments were conducted at the inlet air temperatures of 30, 40, and 50°C, air inlet velocities of 7, 8, and 9 m/s, and IR lamp input powers of 0, 150, and 300 W. The results indicated that inlet air temperature and IR lamp input power had significant effect on drying time (p < .05). A comparative study was performed among mathematical, Artificial Neural Networks (ANNs), and Adaptive Neuro‐Fuzzy System (ANFIS) models for predicting the experimental moisture ratio (MR) of stevia leaves during the drying process. The ANN model was the most accurate MR predictor with coefficient of determination (R2), root mean squared error (RMSE), and chi‐squared error (χ2) values of 0.9995, 0.0005, and 0.0056, respectively, on test dataset. These values of the ANFIS model on test dataset were 0.9936, 0.0243, and 0.0202, respectively. Among the mathematical models, the Midilli model was the best‐fitted model to experimental MR values in most of the drying conditions. It was concluded that artificial intelligence modeling is an effective approach for accurate prediction of the drying kinetics of stevia leaves in the continuous‐flow IR‐assisted hybrid solar dryer.

show abstract

“…Since it introduces fuzziness into the NN, the combination of this model can increase the accuracy of results. Existing research reveals that the ANFIS shows better performance than the classical NN in TEC modeling at the low solar activity [30]. Thus, this model is most likely to be used in TEC forecasting under severe weather conditions in terms of its fuzziness characteristic.…”

Section: Adaptive Neuro-fuzzy Inference System (Anfis)mentioning

confidence: 99%

“…The machine learning algorithms can automatically learn the implicit nonlinear relationship between the TEC value and the external indicators, which helps to improve the prediction performance in extreme environments. A few machine learning approaches have been used in the ionosphere prediction, such as the standard Neural Network (NN) approach [14]- [16], Long Short-Term Memory (LSTM) [17]- [19], Adaptive Neuro-Fuzzy Inference System (ANFIS) [20] etc. Their performance has been evaluated and has been compared with the GIM and IRI-2016 models and the results show that these machine learning algorithms outperform the existing models.…”

Section: Introductionmentioning

confidence: 99%

Machine Learning-Based Short-Term GPS TEC Forecasting During High Solar Activity and Magnetic Storm Periods

Han

Wang

et al. 2022

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

Precise ionospheric total electron content (TEC) is critical for many aerospace applications, and forecasting ionospheric TEC is of great significance to it. Besides, short-term prediction of TEC values fills the gap between the TEC product latency and the precision. The machine learning-based approaches are promising in solving the non-linear prediction issues, particularly suitable for short-term GPS TEC forecasting due to its complex temporal and spatial variation. In this paper, four different machine learning models, i.e., Artificial Neural Network (ANN), Long Short-Term Memory (LSTM) networks, Adaptive Neuro-Fuzzy Inference System based on Subtractive Clustering (ANFIS-SC), and Gradient Boosting Decision Tree (GBDT) are applied for forecasting ionospheric TEC in three IGS GNSS monitoring stations at the low-latitude region (16°S to 10°S). The performance of these approaches in extreme conditions is investigated, including the high solar activity and magnetic storm, which are the most challenging scenario for TEC prediction. The results show that the machine learning algorithms outperform the Global Ionospheric Map (GIM) prediction model. The prediction accuracy during the high solar activity period was improved from 37.93% to 49.28%. During the magnetic storm period, the prediction accuracy was improved from 28.16% to 67.39% . Among the machine learning algorithms, the GBDT model outperforms the rest three algorithms in ionosphere prediction scenarios, which improves the prediction accuracy by 5.6% and 12.7% than the rest three approaches on average during high solar activity (2012-2015) and magnetic storm periods respectively.

show abstract

Ionosphere time series modeling using adaptive neuro-fuzzy inference system and principal component analysis

Cited by 25 publications

References 20 publications

Long Short‐Term Memory Neural Network for Ionospheric Total Electron Content Forecasting Over China

Long Short‐Term Memory Neural Network for Ionospheric Total Electron Content Forecasting Over China

Mathematical and intelligent modeling of stevia (Stevia Rebaudiana) leaves drying in an infrared‐assisted continuous hybrid solar dryer

Machine Learning-Based Short-Term GPS TEC Forecasting During High Solar Activity and Magnetic Storm Periods

Contact Info

Product

Resources

About