Use of Genetic Programming for the Estimation of CODLAG Propulsion System Parameters

In this paper, the publicly available dataset for Combined Diesel-Electric and Gas (CODLAG) propulsion system was used to obtain symbolic expressions estimation of fuel flow, ship speed, starboard propeller torque, port total torque using genetic programming (GP) algorithm. The consists 11,934 samples that were divided into training testing portions in an 80:20 ratio. portion which consisted 9548 train GP algorithm propeller, respectively. After obtained 2386 measure performance terms coefficient correlation (R2) Mean Absolute Error (MAE) metric, Based on each case three best selected with without decay state coefficients. From conducted investigation, highest R2 lowest MAE values achieved coefficients while have slightly lower performance.