﻿ 船舶推进系统组合预测算法研究
 舰船科学技术  2022, Vol. 44 Issue (20): 112-115    DOI: 10.3404/j.issn.1672-7649.2022.20.022 PDF

Research on combination prediction algorithm of ship propulsion system
LIU Hai-li
Wuhan Institute of Shipbuilding Technology, Wuhan 430050, China
Abstract: This research will use the combined prediction algorithm as a tool to study the reliability of the ship propulsion system. It is hoped that through the analysis of its theory, the practical application will be discussed, in order to minimize the risk of failure of the ship propulsion system and contribute to the healthy development of the ship industry. Provide reliability research, so that the ship propulsion system can be further improved and perfected in subsequent production applications.
Key words: ship propulsion system     combined prediction     algorithm
0 引　言

1 组合预测算法体系构建原则

1.1 整体性和系统性

1.2 科学性和合理性

1.3 重要性和代表性

1.4 有效性和敏感性

2 基于组合预测的船舶推进系统影响因素分析 2.1 影响因素的选择

2.2 各因素的影响作用分析

1）驾驶室操纵台的影响

2）集控室操纵台

3 故障预测实证分析 3.1 TOPSIS原理

 $S_i^ + = \sqrt {\sum\limits_{j = 1}^n {{{({x_{ij}} - x_j^ + )}^2}} } \text{。}$ (3-1)

 $S_i^ - = \sqrt {\sum\limits_{j = 1}^n {{{({x_{ij}} - x_j^ - )}^2}} } \text{。}$ (3-2)

 $C_i^{} = \frac{{S_i^ - }}{{S_i^ + + S_i^ - }} \text{。}$ (3-3)
3.2 组合预测建模原理

 ${\tilde X_j} = {w_1}{x_{1j}} + {w_2}{x_{2j}} + \cdots + {w_m}{x_{mj}} = \sum\limits_{i = 1}^m {{w_i}{x_{ij}}} \text{。}$ (3-4)

3.3 基于TOPSIS原理的故障预测组合权重确定方法

 $S({X^ + },{X_i}) = \sqrt {\sum\limits_{j = 1}^n {{{({x_{ij}} - x_j^ + )}^2}} } \;\;\; i = 1,2, \cdots ,m \text{。}$ (3-5)

 ${\lambda _i} = {\raise0.7ex\hbox{$1$} \mathord{\left/ {\vphantom {1 {S({X^ + },{X_i})}}}\right.} \lower0.7ex\hbox{${S({X^ + },{X_i})}$}}\;\;\; {i = 1,2, \cdots ,m} \text{。}$ (3-6)

 ${w_i} = {{{\lambda _i}} \mathord{\left/ {\vphantom {{{\lambda _i}} {\sum\limits_{k = 1}^m {{\lambda _k}} }}} \right. } {\sum\limits_{k = 1}^m {{\lambda _k}} }}\;\;\; {i = 1,2, \cdots ,m} \text{。}$ (3-7)

3.4 模型精度检验方法

 ${\bar \Delta _i} = \frac{1}{n}\sum\limits_{k = 1}^n {||\frac{{{Y_k} - {X_{ik}}}}{{{Y_k}}}||} \text{。}$ (3-8)

3.5 故障预测结果分析

 $Y = ({\text{63}}{\text{.1 ,65}}{\text{.2, 72}}{\text{.8 ,81}}{\text{.2 ,88}}{\text{.7 , 93}}{\text{.5 ,99}}{\text{.1}}) 。$

 图 1 船舶推进系统发生故障变化趋势 Fig. 1 Trend of ship propulsion system failure

 ${\tilde y_1} = {\text{54}}{\text{.7286}} + {\text{6}}{\text{.4464}} t ，\;\;\; t = 1,2,3, \cdots \text{。}$ (3-9)

 ${\bar \Delta _1} = \frac{1}{7}\sum\limits_{k = 1}^7 {|\frac{{{Y_k} - {X_{1k}}}}{{{Y_k}}}|} = {\text{0}}{\text{.0174}} \text{。}$ (3-10)

 ${\tilde y_2}(t) = {\text{ (803}}{\text{.3578)}}{\text{e}}^{\text{0}{\text{.0807}}(t - 1))}{\text{ - 740}}{\text{.2578}} \text{。}$ (3-11)

 ${\bar \Delta _2} = \frac{1}{7}\sum\limits_{k = 1}^7 {|\frac{{{Y_k} - {X_{2k}}}}{{{Y_k}}}|} = {\text{0}}{\text{.0167}} \text{。}$ (3-12)

 ${\bar \Delta _3} = \frac{1}{7}\sum\limits_{k = 1}^7 {|\frac{{{Y_k} - {X_{3k}}}}{{{Y_k}}}|} = {\text{0}}{\text{.0242}} \text{。}$ (3-13)

 $w = ({\text{0}}{\text{.4023 0}}{\text{.3498 0}}{\text{.2479)}} \text{。}$ (3-14)

 ${\bar \Delta _4} = \frac{1}{7}\sum\limits_{k = 1}^7 {|\frac{{{Y_k} - {X_{4k}}}}{{{Y_k}}}|} = {\text{0}}{\text{.0150}} \text{。}$ (3-15)

 图 2 故障预测特征值误差分布图 Fig. 2 Error distribution of fault prediction eigenvalue

 图 3 平均相对误差分布图 Fig. 3 Average relative error distribution diagram

4 船舶推进系统的维护与修理建议 4.1 船舶推进控制系统运行维护

4.2 推进船舶核心设备运行维护

5 结　语

 [1] 刘翔. 船舶电力推进系统故障诊断与预测技术综述[J]. 船电技术, 2020, 40(6): 30-33+38. DOI:10.3969/j.issn.1003-4862.2020.06.002 [2] 余国虎. 船舶综合电力推进系统的特征分析[J]. 电子技术, 2021, 50(4): 172-173. [3] 王俊龙, 袁伟. 船舶推进系统可靠性分析研究[J]. 四川兵工学报, 2020, 41(7): 208-212. [4] 冯明, 刘艳年. 船舶推进系统故障诊断和预测关键技术研究[J]. 舰船科学技术, 2022, 44(3): 107-110. [5] 张刚. 船舶推进系统安装技术工法研究与质量控制[J]. 船舶物资与市场, 2021(3): 91-92. DOI:10.19727/j.cnki.cbwzysc.2021.03.042 [6] 张玉龙, 吴炜, 周建辉. 基于灰色系统理论的推进轴系负荷预测分析[J/OL]. 中国舰船研究: 1−7[2022-10-18].