Research on collision identification between ships and offshore wind turbines based on neural network

With the recovery of the global economic industry, shipping development has prospered. Shipping lanes are located in various seas. However, most offshore wind farms are currently built in offshore waters close to ports and waterways, which makes a high risk of collision between wind turbines and ships. This will not only affect the normal shipping of the vessel, but also lead to structural damage to wind turbines. Ships of different tonnage suffer different severity of damage from collisions with wind turbines. The technology of artificial intelligence neural network applied to structural damage identification is gradually maturing. A BP neural network optimized by artificial bee colony algorithm (ABC) was used in this research to identity the severity of the collision damage. ABC algorithm is essentially an optimization algorithm proposed to imitate the behavior of bees, which has strong global and local search capabilities. With the numerical simulation of collision between ship and wind turbines, the type and the tonnage of the ship can be obtained, which helps to develop a specific repair and maintenance strategy. In this paper, the collisions of bulk carriers with tonnage from 2000, 3000 and 4000 tons were numerically simulated by using the finite element method. The numerical result shows that the 4000-ton collision vessel makes the natural frequency change of offshore wind turbine the most and damage is also the largest. The ABC-BP neural network can effectively identify collision damage and determine the type of ship that collided.