Frequency-load control based on auto-tuning neurons for ship power station

Abstract

The design of frequency-load controller greatly affects the performance of a ship power station. A common strategy to control frequency and load of the ship power station is to use speed regulator with a PI controller, these schemes require proper and continuous tuning, frequency and load of the diesel-generator are separately controlled, but frequency regulation and load distribution among parallel synchronous generators are interacted. A new parallel PID control technique based on auto-tuning neurons used as a modified hyperbolic tangent function is proposed in this paper, which specialized for frequency-load regulation of the ship power station. Simulation and experimental results show that proposed scheme can achieve an appropriate frequency-load control among parallel synchronous generators, and ensure their stable and economical performances.