数据驱动法在船用冷藏集装箱系统模型建立中的应用  

Application of the Data-driven Method in Establishment of the System Model of Marine Refrigerated Container

作　　者：郑超瑜[1,2] 俞文胜[1,2] 陈武[1,2] 

机构地区：[1]集美大学轮机工程学院,福建厦门361021 [2]集美大学福建省船舶与海洋工程重点实验室,福建厦门361021

出　　处：《船海工程》2016年第1期120-122,136,共4页Ship & Ocean Engineering

基　　金：福建省自然科学基金项目(2013J01201);福建省科技厅项目(201205180008);福建省教育厅科技项目(JA12187);厦门市科技计划项目(3502Z20123024)

摘　　要：利用数据驱动方法避开船用冷藏集装箱制冷机组相对复杂的控制策略,建立精度较高的系统模型,采用回归模型,通过数据驱动方法训练并获取系统模型,验证结果表明,采用数据驱动法获取的系统模型能够以较高的精度预测船用冷藏集装箱系统的正常运行状态,可作为故障诊断的判断基准。The control strategy of the refrigeration unit in marine refrigerated containers is relatively complex,which can be avoided by using data-driven method to build a system model with high accuracy. A regression system model of marine refrigerated container is trained and established by data-driven method,and it is also verified. The results show that the system model obtained by data-driven method can predict the normal operation conditions of marine refrigerated container with high accuracy,which can be used as the baseline of the fault detection and diagnosis in marine refrigerated container.

关 键 词：船用冷藏集装箱 数据驱动 故障诊断 系统模型 

分 类 号：TB61[一般工业技术—制冷工程]