面向船舶的基于WPA-BP与虚拟现实的机舱安全监测

Ship-oriented research on cabin safety monitoring based on WPA-BP and virtual reality technology

摘要:

当前国内外对于船舶机舱的智能化诊断研究较浅,且远程监测技术难以满足船舶安全运行的高效需求。因此,研究创新地设计了一种船舶机舱安全监测系统。利用优化后的狼群算法对反向传播神经网络进行改进,并建立了机舱柴油机故障诊断模型。同时,结合虚拟现实技术进行了船舶机舱安全监测系统的设计。实验验证显示,研究构建的故障诊断模型对柴油机五种故障类型的平均诊断准确率高达99.102%。安全监测系统对故障识别准确率比其他系统提高了5.34%,且响应时间为0.048ms。结果表明,利用优化后的狼群算法改进反向传播神经网络的权值和阈值可有效提高故障诊断模型的诊断结果,且收敛效果更理想。研究提出的机舱安全监测系统在船舶安全运行方面具有积极的应用价值,对船舶安全运行具有一定的促进作用。

Abstract:

The current research on intelligent diagnosis of ship"s engine room at home and abroad is shallow, and the remote monitoring technology is difficult to meet the efficient demand of ship"s safe operation. Therefore, the research innovatively designs a ship nacelle safety monitoring system. The optimized wolf pack algorithm is used to improve the back-propagation neural network, and the nacelle diesel engine fault diagnosis model is established. Meanwhile, the design of the ship cabin safety monitoring system was carried out by combining virtual reality technology. Experimental validation shows that the average diagnostic accuracy of the fault diagnosis model constructed by the study is as high as 99.102% for five fault types of diesel engines. The safety monitoring system improved the fault identification accuracy by 5.34% compared with other systems, and the response time was 0.048 ms. The results show that improving the weights and thresholds of the back-propagation neural network by using the optimized wolf pack algorithm can effectively improve the diagnostic results of the fault diagnostic model, and the convergence effect is more satisfactory. The cabin safety monitoring system proposed by the study has positive application value in the safe operation of ships and has a certain role in promoting the safe operation of ships.