Abstract
Shafting alignment is crucial for marine propulsion systems and may affect the safety and stability of ship operations. Air spring vibration isolation systems (ASVISs) for marine shafting can help control the shafting alignment state by actively adjusting air spring pressures while effectively reducing the mechanical noise. However, how to accurately control the alignment state of marine shafting with air spring vibration isolation system remains a challenge. To address this issue, a digital twin (DT)-driven alignment control method is proposed in this paper. First, we design a digital twin prediction model based on the neural network to describe the data mapping relationship between the air spring pressures and shafting alignment state. Then, based on the prediction model, we transform the shafting alignment control problem into a non-linear optimization problem in which our objective is to minimize the alignment error while balancing the load on different air springs. To obtain the optimal air spring pressures, the genetic algorithm is introduced to solve the optimization problem, fully exploiting its global search capacity. Moreover, in order to achieve the optimized pressures, a soft-constrained controller based on proportional-integral-derivative (PID) algorithm is developed to accurately generate specific control policies based on the monitoring data. Finally, the feasibility and the effectiveness of the proposed alignment control method is verified with a real ASVIS.