The combined use of wavelet transform and black box models in reservoir inflow modeling

Abstract

In the study presented, different hybrid model approaches are proposed for reservoir inflow modeling from the meteorological data (monthly precipitation, one-month-ahead precipitation and monthly mean temperature data) by the combined use of discrete wavelet transform (DWT) and different black box techniques. Multiple linear regression (MLR), feed forward neural networks (FFNN) and least square support vector machines (LSSVM) were considered as the black box methods. In the modeling strategy, meteorological input data were decomposed into wavelet sub-time series at three resolution levels and ineffective sub-time series were eliminated by Mallows' C p based all possible regression method. As a result of all possible regression analyses, 2-months mode of time series of monthly temperature (D1_T t ), 8-months mode of time series (D3_T t ) of monthly temperature and approximation mode of time series (A3_T t ) of monthly temperature were eliminated. Remained effective sub-time series were used as the inputs of MLR, FFNN and LSSVM. When the performances of the training and testing periods were compared, it was observed that the DWT-FFNN conjunction model has better results in terms of mean square errors (MSE) and determination coefficients (R 2 ) statistics. The discrete wavelet transform approach also increased the accuracy of multiple linear regression and least squares support vector machines.

Streamflow Forecasting and Estimation Using Least Square Support Vector Regression and Adaptive Neuro-Fuzzy Embedded Fuzzy c-means Clustering

“…Hemmati-Sarapardeh et al (2014) determined the reservoir oil viscosity using LSSVR approach. However, there are limited studies in the literature related to application of LSSVR in water resources (Guo et al 2011;Hwang et al 2012;Kisi 2012Kisi , 2013Okkan and Serbes 2013). Guo et al (2011) used LSSVR for prediction of reference evapotranspiration and found promising results.…”

“…He showed that LSSVR model performed better than the NN and other models. Okkan and Serbes (2013) used and proposed LSSVR with wavelet in reservoir inflow modeling. Despite the reported success of employing LSSVR in numerous studies, to the best of our knowledge the capability of LSSVR has not yet been examined in terms of streamflow forecasting.…”

Comparative Study of the Three Models (ANN-PMC), (DWT-ANN-PMC) and (MLR) for Prediction of the Groundwater Level of the Surface Water Table in the Saïss Plain (North of Morocco)

“…With the impetus to predict the depths of groundwater in 64 wells dug in the Munijhara micro-basin at Orissa, Similarly, several researchers have used ANN to predict groundwater levels in confined aquifers [13; 14], in leaking aquifers [15; 16] and in multi-layer aquifers. In recent years, the conjunction of wavelet transformation and ANN techniques has been successfully implemented in hydrological applications [18][19][20][21][22][23][24]. The wavelet transform is another technique which can analyze a signal in time and frequency in order to overcome the disadvantages of the conventional Fourier transform.…”

River Stage Modeling by Combining Maximal Overlap Discrete Wavelet Transform, Support Vector Machines and Genetic Algorithm

“…Recently, hybrid time series modeling approaches utilizing wavelet transform have been one of the research themes studied actively in the hydrological field [8][9][10][11][12][13][14]. In terms of signal analysis, the wavelet transform is a signal decomposition method which splits an original signal into sub-signals, including detail and approximation (smooth) components.…”

“…Kisi et al [11] predicted short-and long-term air temperatures using wavelet-based genetic programming. Okkan and Serbes [12] modeled reservoir inflow using a combination of DWT and black box models including ANNs, multiple linear regression and least square support vector machines (LS-SVMs). …”