Condition monitoring method for marine engine room equipment based on machine learning

技术指标	数值	技术指标	数值
缸径/mm	980	最大额定转速/(r·min⁻¹)	94
行程/mm	2 660	最大平均指示压力/bar	18.2
活塞面积/m²	0.754 3	最高爆发压力/bar	140.1
整机重量/t	2 100	涡轮增压器	3×TPL85-B11
最大功率/kW	40 055	发火顺序	1-7-2-5-4-3-6

Table 2. Comparison between simulation results and shop test data

View table

View in Article

Table 2. Comparison between simulation results and shop test data

柴油机负荷/%	结果	功率/kW	油耗/（g·kW⁻¹·h⁻¹）	气缸最高爆发压力/bar	气缸压缩压力/bar	涡轮转速/（r ·min⁻¹）	扫气箱压力/bar	排气管温度/K
25	仿真值	10 105	186.04	74.04	47.95	4485	1.33	579.29
	实验值	10 014	186.37	73.60	47.40	4291	1.32	577.17
	误差/%	0.92	−0.18	0.59	1.16	4.53	0.54	0.37
50	仿真值	20 226	179.60	99.43	72.62	7896	2.07	593.68
	实验值	20 028	179.46	98.00	72.00	7782	2.05	600.17
	误差/%	0.99	0.08	1.46	0.86	1.47	0.89	−1.08
75	仿真值	30345	175.03	127.91	100.13	9 710	2.86	611.63
	实验值	30041	176.01	128.00	99.90	9 670	2.87	614.57
	误差/%	1.01	−0.55	−0.07	0.23	0.41	−0.40	−0.48
100	仿真值	40462	177.69	138.40	125.67	10 943	3.58	660.83
	实验值	40055	177.98	139.40	126.70	10 946	3.63	663.90
	误差/%	1.02	−0.16	−0.71	−0.81	−0.03	−1.39	−0.46

Table 3. Simulation datasets
View table
View in Article
Table 3. Simulation datasets
类别工况特征个数数量
1 正常 15 400
2 压缩机故障 15 100
3 空冷机故障 15 100
4 喷油定时错误 15 100

Table 4. The accuracy FDR and FAR under different hybrid fault monitoring schemes
View table
View in Article
Table 4. The accuracy FDR and FAR under different hybrid fault monitoring schemes
方法 FDR/% FAR/%
PCA-OS 81.2 1.63
PCA-RC 81.1 1.62
PCA-iforest 81.3 1.6
MDS-OS 85 1.24
MDS-RC 85.5 1.23
MDS-iforest 87.1 1.2
LLE-OS 92.1 1.1
LLE-RC 93.1 8.5
LLE-iforest 93.4 9
TSNE-OS 94.9 7.5
TSNE-RC 96.1 6
TSNE-iforest 98.5 3

Tools

Get Citation

Copy Citation Text

Ruihan WANG, Hui CHEN, Cong GUAN. Condition monitoring method for marine engine room equipment based on machine learning[J]. Chinese Journal of Ship Research, 2021, 16(1): 158

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites