تحقیقات منابع آب ایران

تحقیقات منابع آب ایران

بررسی فرامدل هیبریدی سه‌گانه MLP- PSO- ARIMA به‌منظور پیش‌بینی شاخص FDSD (مطالعه موردی: استان خوزستان)

نوع مقاله : مقاله پژوهشی

نویسنده
استادیار، گروه مهندسی احیاء مناطق خشک و کوهستانی، دانشکده منابع طبیعی، دانشگاه تهران، کرج، ایران
چکیده
پژوهش حاضر با هدف بررسی عملکرد فرامدل هیبریدی سه‌گانه ماشین- کاتالیزور- جنکینز MLP-PSO-ARIMA در پیش‌بینی شاخص فراوانی روزهای همراه با طوفان گرد و غبار، در هفت ایستگاه منتخب در استان خوزستان در طول دوره آماری 50 سال (2019- 1970) و با استفاده از داده‎های دید افقی و کدهای سازمان جهانی هواشناسی انجام شده ‌است. نتایج حاصل از فرامدل هیبریدی سه‌گانه فوق با استفاده از شاخص‌های R، RMSE، MAE و NS با مدل‌‌های انفرادی MLP و ARIMA و همچنین فرامدل‌های هیبریدی MLP-PSO، ARIMA-PSO و MLP-ARIMA مورد مقایسه قرار گرفته است. تمامی مدل‌های مذکور بیشترین عملکرد خود را در ترکیب‌های فصلی اول و دوم نمایش دادند. لذا می‌توان نتیجه‌ گرفت که استفاده از یک و یا دو فصل پیشین به‌منظور پیش‌بینی شاخص FDSD در فصل‌های آتی در استان خوزستان نتایج بهتر و دقیق‌تری به‌همراه دارد و بکارگیری فصول سوم و چهارم، سبب بهبود نتیجه پیش‌بینی نخواهد شد. از طرف دیگر، دقت مدل شبکه عصبی پرسپترون چندلایه از مدل باکس- جنکینز ARIMA به‌منظور پیش‌بینی گرد و غبار استان خوزستان بیش‌تر شده است. همچنین، ترکیب مدل MLP با مدل ARIMA سبب افزایش دقت نسبت به مدل‌های انفرادی MLP و ARIMA شد ولی رشد دقت آن آنچنان معنی‌دار نبود. از طرف دیگر، مدل هیبرید سه‌گانه سبب افزایش رشد معنی‌دار دقت نسبت به مدل هیبرید دو‌گانه فوق شده است.
 
کلیدواژه‌ها

موضوعات


عنوان مقاله English

A Comprehensive Analysis of the Triple-Hybrid Metamodel of MLP-PSO-ARIMA for Forecasting the FDSD Index: A Case Study of Khuzestan Province

نویسنده English

Mohammad Ansari Ghojghar
Assistant Professor, Department of Rehabilitation of Arid and Mountainous Regions Engineering, Faculty of Natural Resources, University of Tehran, Karaj, Iran.
چکیده English

This study aims to evaluate the performance of the triple-hybrid metamodel of MLP-PSO-ARIMA in forecasting the frequency of dust storm days (FDSD) index across seven selected stations in Khuzestan Province during a 50-year statistical period (1970–2019). The results of the proposed triple-hybrid metamodel was compared against the standalone MLP and ARIMA models, as well as the hybrid models of MLP-PSO, ARIMA-PSO, and MLP-ARIMA, using performance metrics including R, RMSE, MAE, and NS. All the tested models demonstrated their highest accuracy during the first and second seasonal combinations. Accordingly, it was concluded that utilizing data from one or two preceding seasons yield more accurate predictions of the FDSD index for subsequent seasons in Khuzestan Province, whereas incorporating data from the third and fourth seasons does not enhance forecasting performance. Moreover, the multilayer perceptron (MLP) neural network outperformed the Box-Jenkins ARIMA model in predicting dust storm events in the region. While combining the MLP and ARIMA models improved the accuracy compared to their standalone counterparts, the improvement was not statistically significant. In contrast, the proposed triple-hybrid metamodel exhibited a statistically significant enhancement in accuracy over the dual-hybrid models.
 

کلیدواژه‌ها English

Dust Storm, Optimization Algorithm, Triple-Hybrid, Artificial Neural Network, Box-Jenkins
Ahmadi H & Mousavi S F (2018) Comparative study of machine learning algorithms for drought forecasting in arid regions. Environmental Modelling & Software 104:1-12 (In Persian)
Baghbanan P, Ghavidel Y, & Farajzadeh M (2020) Spatial analysis of spring dust storms hazard in Iran. Theoretical and Applied Climatology 139(3-4):1447–1457 (In Persian)
Bahmani S & Najafi F (2017) Hybrid AI models for flood forecasting: Case study of Iran. Natural Hazards 87(1):57-71 (In Persian)
Bakhshizadeh A, Sadeghi K, Ahmadi S, & Royaei J (2023) Damage identification in long-span cable-stayed bridges under multiple support excitations. International Journal of Civil Engineering 21(8):1275–1290 (In Persian)
Borzou F, Zolfaghari H, Masoompour Samakosh J, & Sahraei J (2021) Spatial analysis of dust storms in Iran based on climatic and vegetation characteristics. Geography and Environmental Sustainability 11(1):1–23 (In Persian)
Chafi Z S & Afrakhte H (2021) Short-term load forecasting using neural network and particle swarm optimization (PSO) algorithm. Wiley Online Library
Dargahian F & Pourhashemi M (2022) Detection of climate change based on cold temperature indices in the forest ecosystem of Central Zagros. Forest Research and Development 8(2):197–215 (In Persian)
Dargahian F, Ghasemiaryan Y, & Heydarnejad S (2024) Revealing the change of the frost indices as a consequence of climate change: A case study of Northern Zagros forests (Azerbaijan). Iranian Journal of Forest and Range Protection Research 22(1):30–46 (In Persian)
Ebrahimi H & Mohammadi J (2020) Application of extreme learning machine for rainfall-runoff simulation. Hydrological Processes 34(21):3938-3949 (In Persian)
Ebrahimi-Khusfi Z, Taghizadeh-Mehrjardi R, & Nafarzadegan A R (2021) Accuracy, uncertainty, and interpretability assessments of ANFIS models to predict dust concentration in semi-arid regions. Environmental Science and Pollution Research 28(6):6796–6810 (In Persian)
Hamidi M, Kavianpour M R, & Shao Y (2014) Numerical simulation of dust events in the Middle East. Aeolian Research 13:59–70 (In Persian)
Hamidianpour M, Jahanshahi S M A, Kaskaoutis D G, Rashki A, & Nastos P G (2021) Climatology of the Sistan Levar wind: Atmospheric dynamics driving its onset, duration and withdrawal. Atmospheric Research 260:105711 (In Persian)
Heydarian P, Ashrafi K, Mohebali M, Kia E B, Aryaeipour M, Sharafi A C, ... & Rokni M B (2017) Seroprevalence of human fasciolosis in Lorestan Province, western Iran, in 2015–16. Iranian Journal of Parasitology 12(3):389 (In Persian)
Mohammed F K, Beckles D M, & Opadeyi J (2018) Characterization, source apportionment, and human health risk assessment of polycyclic aromatic hydrocarbons (PAHs) in road dust of a small island state in the Caribbean. Human and Ecological Risk Assessment: An International Journal 24(7):1852–1871
Omidvar K & Omidi Z (2013) The analysis of dust phenomenon in the southern and central Fars province. The Journal of Geographical Research on Desert Areas 1(1):85–114 (In Persian)
Pourgholam-Amiji M, Ansari Ghojghar M, & Ahmadaali K (2021) Prediction of dust storms in Khuzestan province using artificial neural networks. Nivar 45(114–115):56–75 (In Persian)
Rahmati O, Panahi M, Ghiasi S S, Deo R C, Tiefenbacher J P, Pradhan B, ... & Bui D T (2020) Hybridized neural fuzzy ensembles for dust source modeling and prediction. Atmospheric Environment 224:117320 (In Persian)
Samadi, M, Darvishi Boloorani A, Alavipanah S K, Mohamadi H, & Najafi M S (2014) Global dust detection index (GDDI): A new remotely sensed methodology for dust storms detection. Journal of Environmental Health Science and Engineering 12:1–14
Shahsavani A, Naddafi K, Haghighifard N J, Mesdaghinia A, Yunesian M, Nabizadeh R, ... & Goudarzi G (2012) The evaluation of PM10, PM2.5, and PM1 concentrations during the Middle Eastern Dust (MED) events in Ahvaz, Iran, from April through September 2010. Journal of Arid Environments 77:72–83 (In Persian)
Sobhani B, Safarian Zengir V, & Faizollahzadeh S (2020) Modeling and prediction of dust in western Iran. Physical Geography Research 52(1):17–35 (In Persian)
Socha J (2008) Effect of topography and geology on the site index of Picea abies in the West Carpathian, Poland. Scandinavian Journal of Forest Research 23(3):203–213
Tosan M, Nourani V, Kisi O, & Dastourani M (2025) Evolution of ensemble machine learning approaches in water resources management: A review. Earth Science Informatics 18(2):1–36
Tositti L, Brattich E, Cassardo C, Morozzi P, Bracci A, Marinoni A, ... & Zappi A (2022) Development and evolution of an anomalous Asian dust event across Europe in March 2020. Atmospheric Chemistry and Physics 22(6):4047–4073
Zhang L & Li Q (2018) Application of support vector machines for flood prediction. Hydrological Sciences Journal 63(9):1482-1495
 

  • تاریخ دریافت 25 آذر 1403
  • تاریخ بازنگری 26 فروردین 1404
  • تاریخ پذیرش 03 اردیبهشت 1404