Groundwater Flow Rate Prediction From Geo‐Electrical Features Using Support Vector Machines
2022
Kouadio, Kouao Laurent | Kouame, Loukou Nicolas | Drissa, Coulibaly | Mi, Binbin | Kouamelan, Kouamelan Serge | Gnoleba, Serge Pacôme Déguine | Zhang, Hongyu | Xia, Jianghai
Unsuccessful drillings are issues in groundwater exploration using electrical resistivity profiling (ERP) and vertical electrical sounding (VES). Many geophysical companies spend a lot of money without obtaining the flow rate (FR) required during the campaigns for drinking water supply (CDWS). To solve this problem, we applied the support vector machines (SVMs) to real‐world data to predict the FRs before any drilling operations. First, from the ERP and VES, the features such as shape, type, power, magnitude, pseudo‐fracturing index, and ohmic‐area were defined including the geology of the survey area. Second, the FRs were categorized into four classes (dry: FR0 (FR = 0), unsustainable: FR1 (0 < FR ≤ 1), and productive boreholes: FR2 (1 < FR ≤ 3) and FR3 (FR > 3 m³/hr)) and associated with the features to compose two separated data sets: a multiclass data set (D $\mathcal{D}$) for common prediction during the CDWS and a binary data set Db ${\mathcal{D}}_{b}$ (FR < FR2, FR ≥ FR2) addressed to the population living in a rural area. Features were vectorized and data were transformed before feeding to the SVM algorithms. As a result, the SVM models performed 77% of good predictions on D $\mathcal{D}$ and 83% on Db ${\mathcal{D}}_{b}$. Better performances with the optimal hyper‐parameters in D $\mathcal{D}$ (81.61%) and Db ${\mathcal{D}}_{b}$ (87.36%) were achieved using the polynomial and radial basis function kernels respectively. Furthermore, the learning curves have shown that the performance scores on D $\mathcal{D}$ can be improved if larger training data becomes available (275 test samples at least) while it is not necessarily so for Db ${\mathcal{D}}_{b}$. As a benefit, the proposed approach could minimize the rate of unsuccessful drillings during future CDWS.
显示更多 [+] 显示较少 [-]