EUVIMEDEUVIMED entdecken
Crossref

Soft Computing Methods for Groundwater Level Prediction: A Survey

Tarakanta Sahoo, Pravat Malik, Pallishree Mahapatra

International Journal on Advanced Computer Theory and Engineering · 2024

DOI: 10.65521/ijacte.v13i1.912 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal on Advanced Computer Theory and Engineering, 2024. Verfasst von Tarakanta Sahoo und 2 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Population growth and pollution are causinggroundwater depletion in developing countries likeIndia. Monitoring groundwater levels is crucial foreffective water resource management. Accurateestimation of groundwater resources is essential, butgroundwater modeling is inherently non-linear andcannot be solved using traditional methods alone.Consequently, soft computing technologies such asFuzzy Logic, Genetic Algorithms, and Artificial NeuralNetworks (ANN) are gaining importance in hydrologicalstudies.Fuzzy Logic is adept at handling imprecise andambiguous datasets, while ANN, inspired by humanlearning, can learn from examples and adjust weightsaccordingly. Genetic Algorithms, mimicking naturalevolutionary processes, offer innovative solutions. Thisthesis investigates the development of Fuzzy Logic (FL),ANN, and other methodologies like FPSO in predictinggroundwater levels. Four models are evaluated,incorporating different combinations of groundwaterrecharge and discharge as inputs, with groundwater levelas the output.ANN is trained, tested, and validated using groundwaterdatasets to identify the most effective model forgroundwater level prediction. FL works optimally withtwo inputs, while ANN performs better with moreinputs. Fuzzy interval optimization within FL hastraditionally been challenging, often relying on hit-ormissapproaches. However, Genetic Algorithms offer asolution by adjusting fuzzy interval length, withmethodologies like Wang and Mendel's rule basiscreation enhancing performance.The fuzzy Genetic Algorithm method for estimatinggroundwater levels outperforms traditional FL methods,providing more accurate predictions. Creating fuzzy setsin FL without specialized knowledge poses a challenge,prompting the development of computational methods.This thesis proposes a methodology based on centraltendency concepts to determine the appropriate numberof fuzzy sets.This approach effectively identifies intervals and fuzzysets for fuzzy time series forecasting. Chennai'sreservoir rainfall is modeled using a central tendenciesbasedfuzzy approach, serving as a benchmark challengefor fuzzy time series analysis. Comparative evaluationsagainst other methods demonstrate the superiority of theproposed computational technique, yielding promisingresults for benchmark datasets.

Zitieren

APA

Sahoo, T., Malik, P., & Mahapatra, P. (2024). Soft Computing Methods for Groundwater Level Prediction: A Survey. International Journal on Advanced Computer Theory and Engineering. https://doi.org/10.65521/ijacte.v13i1.912

Vancouver

Sahoo T, Malik P, Mahapatra P. Soft Computing Methods for Groundwater Level Prediction: A Survey. International Journal on Advanced Computer Theory and Engineering. 2024. doi:10.65521/ijacte.v13i1.912

BibTeX
@article{Sahoo2024,
  author = {Sahoo, Tarakanta and Malik, Pravat and Mahapatra, Pallishree},
  title = {Soft Computing Methods for Groundwater Level Prediction: A Survey},
  journal = {International Journal on Advanced Computer Theory and Engineering},
  year = {2024},
  doi = {10.65521/ijacte.v13i1.912}
}
Kostenlos starten →