A neural predictor of propeller load demand for improved control of diesel ship propulsion

The increased use of digital electronic technology for marine diesel propulsion control leads to more reliable and efficient propulsion powerplants. Neural networks have been used for the implementation of continuous propeller torque demand prediction that can be exploited by engine control preventively in order to avoid excessive overspeed at near-MCR engine running under rough sea conditions. The neural predictor has been initially validated through simulation, using data recorded during scheduled operation of a large container ship, as well as, from tests in a ship model basin. A compilation of results is presented with related commentary and comparison to a statistical prediction algorithm.