摘要
现有舰艇的动力系统一部分为核动力装置,相对于传统的动力装置,系统复杂度﹑对安全性及可靠性要求都大大提高。一旦核动力装置系统出现故障,需要在规定的时间内及时诊断排除故障。依据经验库及人的故障检测方法已不适用于核动力装置系统,需要一种高可靠及自动智能化的故障排除方法。本文对出现故障的核动力装置数据进行融合,结合神经网络,设计时效性及自动化程度高的动力系统检测系统,最后进行仿真实验。
Most of the existing dynamic systems of the marine military vessels are updated to the nuclear power plant,and the system are complexity, security and reliability requirements are greatly improved compared with the traditional power system. Once the nuclear power plant system have fault, need to diagnose and remove faults within the prescribed time. According to the experience and the fault detection method is not applicable to nuclear power plant system, requires a high reliability and automatic intelligent troubleshooting methods. In this paper, first analyze the Fault data of nuclear power system, combined with the neural network, design the power testing system with high efficiency and high degree of automation.
出处
《舰船科学技术》
北大核心
2017年第2X期49-51,共3页
Ship Science and Technology
关键词
数据融合
神经网络
故障诊断
data fusion
neural network
fault diagnosis
