BP神经网络对船舶远程监控系统的改进分析

为了满足公司对于远洋船舶更加有效监控的要求,应用BP神经网络对监控系统加以改进,使船舶远程监控系统发出预警信号,并能在船舶上报警.此时,相应参数识别码也能够第一时间到达岸上公司,岸上公司即能在最佳时间协助船舶对设备进行维修.BP神经网络在远程监控系统的应用分析过程中,以6缸柴油主机排气温度变化趋势为模型,利用BP神经网络良好的学习特性,建立了排气温度变化的持续升高预警模型及其他非预警模型.分析表明,此种方法适用于远洋船舶的远程故障监测及船舶系统故障预测.

A Shipboard Remote Surveillance System Improved by BP Neural Networks

So as to satisfy the requirements for remotely monitoring ocean-going ships by more effective methods, BP neural networks were applied to improve the existing monitoring system by outputting pre-warning signals instead of failure signals. Then the alarm apparatus on board ships could be actuated, and the relevant identification numbers were ready to be sent to the companies ashore. Subsequently, with the alarm apparatus actuated and the identifica-tion numbers obtained, both the ships in the ocean and the companies ashore could grasp the best chance to solve the potential problems. The exhaust temperature altering tendency of a six-cylinder diesel engine was proposed to ana-lyze the application of BP neural networks to the shipboard remote surveillance system. Due to its learning ability, BP neural networks earned the tendency,and the identification number of gradually rising exhaust temperature was distinguished. The result indicated that BP neural networks were practicable for the remote failure control and the failure prediction function.