基于ABC-BP神经网络的海上风机碰撞损伤识别

Collision Damage Identification of Offshore Wind Turbines Based on ABC-BP Neural Network

为降低海上风机与近海船舶发生碰撞的风险,避免产生结构性损伤,以单桩式海上风机为研究对象,开展海上风机与近海船舶碰撞损伤识别研究。通过模拟2 000 t、3 000 t和4 000 t等3种吨位的船舶与单桩风机碰撞,利用经人工蜂群(ABC)算法优化的反向传播(BP)神经网络实现对海上风机损伤程度的识别,继而识别碰撞船舶的吨位。将识别结果与采用其他神经网络的识别结果相对比,结果表明,ABC-BP神经网络在海上风机结构损伤识别中的准确率能达到99%,均方根误差仅为1.82,可为相应维修策略的制定提供技术支持。

In order to reduce the risk of collision between offshore wind turbines and offshore ships and to avoid structural damage,a research on collision damage identification between offshore wind turbines and offshore ships is carried out using monopile offshore wind turbines as the research object.By simulating the collision between three tonnage ships of 2000 t,3000 t and 4000 t and monopile wind turbines,the Backpropagation(BP)neural network optimized by the Artificial Bee Colony(ABC)algorithm is used to identify the degree of damage to offshore wind turbines,and then the tonnage of the colliding ship is identified.Comparing the recognition results with those of other neural networks,the results show that the accuracy of the ABC-BP neural network in identifying the structural damage of offshore wind turbines can reach 99%,and the root mean square error is only 1.82,which can provide technical support for formulating appropriate maintenance strategies.