A Hybrid Extended Kalman Filter Based on a Parametrized FeedForward Neural Network for the Improvement of the Results of Numerical Wave Prediction Models
Aiming to develop a novel optimization model for numerical weather and wave predictions, this study proposes a hybrid approach based on the combination of Artificial Neural Networks (ANNs) and Kalman Filters (KFs). The KF technique uses fixed covariance matrices, which is inappropriate for dynamic c...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-09-01
|
| Series: | Environmental Sciences Proceedings |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2673-4931/26/1/199 |
| _version_ | 1797241111448649728 |
|---|---|
| author | Athanasios Donas George Galanis Ioannis Th. Famelis |
| author_facet | Athanasios Donas George Galanis Ioannis Th. Famelis |
| author_sort | Athanasios Donas |
| collection | DOAJ |
| description | Aiming to develop a novel optimization model for numerical weather and wave predictions, this study proposes a hybrid approach based on the combination of Artificial Neural Networks (ANNs) and Kalman Filters (KFs). The KF technique uses fixed covariance matrices, which is inappropriate for dynamic conditions where the measurement error varies due to outside factors. This work proposes an alternative method for updating the covariance matrices, depending on a selected parameter that was obtained from a parametrized FeedForward Neural Network. The FeedForward Neural Network is trained for various values of the parameter for a set of historical data to provide the best choice based on a predetermined objective. Following this procedure, the parameter is added to the hybrid Extended Kalman Filter, which improves the direct outputs of numerical forecasts by decoding the systematic error between the model’s direct output and the corresponding observations. The suggested approach was used in a number of time periods in various locations with very promising results. |
| first_indexed | 2024-04-24T18:18:08Z |
| format | Article |
| id | doaj.art-7ee5750f47f647c18024c6568e94cf6e |
| institution | Directory Open Access Journal |
| issn | 2673-4931 |
| language | English |
| last_indexed | 2024-04-24T18:18:08Z |
| publishDate | 2023-09-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Environmental Sciences Proceedings |
| spelling | doaj.art-7ee5750f47f647c18024c6568e94cf6e2024-03-27T13:37:39ZengMDPI AGEnvironmental Sciences Proceedings2673-49312023-09-0126119910.3390/environsciproc2023026199A Hybrid Extended Kalman Filter Based on a Parametrized FeedForward Neural Network for the Improvement of the Results of Numerical Wave Prediction ModelsAthanasios Donas0George Galanis1Ioannis Th. Famelis2microSENSES Laboratory, Department of Electrical and Electronics Engineering, Ancient Olive Grove Campus, University of West Attica, 12241 Athens, GreeceMathematical Modeling and Applications Laboratory, Hellenic Naval Academy, Hatzikiriakion, 18539 Piraeus, GreecemicroSENSES Laboratory, Department of Electrical and Electronics Engineering, Ancient Olive Grove Campus, University of West Attica, 12241 Athens, GreeceAiming to develop a novel optimization model for numerical weather and wave predictions, this study proposes a hybrid approach based on the combination of Artificial Neural Networks (ANNs) and Kalman Filters (KFs). The KF technique uses fixed covariance matrices, which is inappropriate for dynamic conditions where the measurement error varies due to outside factors. This work proposes an alternative method for updating the covariance matrices, depending on a selected parameter that was obtained from a parametrized FeedForward Neural Network. The FeedForward Neural Network is trained for various values of the parameter for a set of historical data to provide the best choice based on a predetermined objective. Following this procedure, the parameter is added to the hybrid Extended Kalman Filter, which improves the direct outputs of numerical forecasts by decoding the systematic error between the model’s direct output and the corresponding observations. The suggested approach was used in a number of time periods in various locations with very promising results.https://www.mdpi.com/2673-4931/26/1/199optimization of numerical forecastsextended Kalman filterfeedforward neural networksWAM-cycle 4 |
| spellingShingle | Athanasios Donas George Galanis Ioannis Th. Famelis A Hybrid Extended Kalman Filter Based on a Parametrized FeedForward Neural Network for the Improvement of the Results of Numerical Wave Prediction Models Environmental Sciences Proceedings optimization of numerical forecasts extended Kalman filter feedforward neural networks WAM-cycle 4 |
| title | A Hybrid Extended Kalman Filter Based on a Parametrized FeedForward Neural Network for the Improvement of the Results of Numerical Wave Prediction Models |
| title_full | A Hybrid Extended Kalman Filter Based on a Parametrized FeedForward Neural Network for the Improvement of the Results of Numerical Wave Prediction Models |
| title_fullStr | A Hybrid Extended Kalman Filter Based on a Parametrized FeedForward Neural Network for the Improvement of the Results of Numerical Wave Prediction Models |
| title_full_unstemmed | A Hybrid Extended Kalman Filter Based on a Parametrized FeedForward Neural Network for the Improvement of the Results of Numerical Wave Prediction Models |
| title_short | A Hybrid Extended Kalman Filter Based on a Parametrized FeedForward Neural Network for the Improvement of the Results of Numerical Wave Prediction Models |
| title_sort | hybrid extended kalman filter based on a parametrized feedforward neural network for the improvement of the results of numerical wave prediction models |
| topic | optimization of numerical forecasts extended Kalman filter feedforward neural networks WAM-cycle 4 |
| url | https://www.mdpi.com/2673-4931/26/1/199 |
| work_keys_str_mv | AT athanasiosdonas ahybridextendedkalmanfilterbasedonaparametrizedfeedforwardneuralnetworkfortheimprovementoftheresultsofnumericalwavepredictionmodels AT georgegalanis ahybridextendedkalmanfilterbasedonaparametrizedfeedforwardneuralnetworkfortheimprovementoftheresultsofnumericalwavepredictionmodels AT ioannisthfamelis ahybridextendedkalmanfilterbasedonaparametrizedfeedforwardneuralnetworkfortheimprovementoftheresultsofnumericalwavepredictionmodels AT athanasiosdonas hybridextendedkalmanfilterbasedonaparametrizedfeedforwardneuralnetworkfortheimprovementoftheresultsofnumericalwavepredictionmodels AT georgegalanis hybridextendedkalmanfilterbasedonaparametrizedfeedforwardneuralnetworkfortheimprovementoftheresultsofnumericalwavepredictionmodels AT ioannisthfamelis hybridextendedkalmanfilterbasedonaparametrizedfeedforwardneuralnetworkfortheimprovementoftheresultsofnumericalwavepredictionmodels |