船舶柴油机突变工况转速自适应瞬态控制OA

A transient simulation model for marine diesel engines was established based on inter-cycle state transfer and transient accumulation.The model was calibrated using data from diesel engine bench tests,achieving an overall accuracy exceeding 95%.A back propagation(BP)neural network was integrated with proportional-integral-derivative(PID)control to autonomously develop an adaptive governor.Subsequently,transient performance analysis under varying exhaust back pressures was conducted for load mutation conditions involving a three-step load surge and a single-step load rejection.The engine speed fluctuations under control by different governors were comparatively investigated.The results demonstrate that,compared to the traditional PID governor,the adaptive governor reduced the instantaneous speed regulation rate by 37.94%and decreased the speed stabilization time by 37.48%.The adaptive speed regulator can effectively solve the problem of speed control under rapidly changing conditions of marine diesel engines.Furthermore,research on the impact of the learning rate and momentum term on governor performance indicates that selecting a smaller learning rate coupled with a larger momentum term is more suitable for marine diesel engines experiencing frequent load fluctuations.