Optimizing anode location in impressed current cathodic protection system to minimize underwater electric field using multiple linear regression analysis and artificial neural network methods

Design of impressed current cathodic protection (ICCP)-anode locations to minimize underwater electric field is important to increase the survivability of naval ships. However, the evaluation of all ICCP-anode location scenarios is time-consuming process even though it is conducted by numerical simulation, especially boundary element method (BEM). To solve this problem, the randomly selected cases of ICCP-anode location in 3, 4 and 5-pair, which were produced by the beta distribution model, were simulated using BEM simulations. Then, the optimized ICCP-anode location scenario from the cases was verified by statistical methods (multiple linear regression and artificial neural network). Predictions of ICCP-anode location and underwater electric field were more correlated with the artificial neural network than the multiple linear regression analysis. Also, the selected ICCP-anode locations were verified by the artificial neural network considering data interactions. Thus, the application of an artificial neural network can be more useful for designing ICCP-anode location that minimizes underwater electric fields.