基于支持向量机的输气管道泄漏压降信号智能识别方法

doi:10.16085/j.issn.1000-6613.2021-2326

摘要/Abstract

摘要：

输气管道干线气液联动阀根据检测管道压降速率、持续时间判断管道是否发生泄漏和自动关断阀门。该方法难以识别小孔泄漏等压降速率低于关断阈值的事故工况和压缩机抽吸等正常运行工况。以相国寺储气库集注干线为对象，通过仿真获得与管道泄漏、压缩机抽吸及截断阀紧急截断3种工况相关的压降速率信号，基于支持向量机建立了管道泄漏信号识别模型。提出混沌映射与自适应惯性权重的教与学优化算法，获得了模型中惩罚因子C和核函数参数g的最优值。利用相国寺储气库铜相线600组数据验证表明，优化后的模型：①对3种工况识别准确率为98.5%，较优化前提升了4.2%；②对于当量直径为50~125mm的小孔泄漏识别准确率为100%，提升了对小孔泄漏信号识别的准确性；③对压缩机抽吸和截断阀紧急截断工况识别的准确率分别为96.7%和100%；④当泄漏孔径小于50mm、压降速率小于0.01MPa/min时，阀室检测到的压降速率信号特征相近，此时建议使用气液联动阀与SCADA系统监测数据综合判断。

关键词: 天然气管道, 截断阀, 压降速率, 支持向量机, 优化算法

Abstract:

The gas-liquid linkage valve of the trunk line of the gas pipeline determines whether the pipeline leaks and automatically shuts off the block valve according to the pressure drop rate and its duration of the pipeline. This method is difficult to identify the accident condition that the pressure drop rate of small aperture leakage is lower than the shutdown threshold and the normal operation condition such as compressor pumping suction. Taking the gathering and injection trunk line of Xiangguosi Gas Storage as the object, the pressure drop rate signals related to the three working conditions of pipeline leakage, compressor pumping suction and emergency shut off of block valve were obtained by simulation and a pipeline leakage signals recognition model was established based on support vector machine. The teaching-learning-based -optimization algorithm of chaotic mapping and adaptive inertia weight was proposed, and the optimal values of penalty factor C and kernel function parameter g in the model were obtained. The verification results of 600 sets of data on Tongxiang pipeline of Xiangguosi Gas Storage were obtained. Firstly, the identification accuracy of the optimized model for the three working conditions was 98.5%, which was 4.2% higher than that of previous optimization. Secondly, the optimized model had 100% recognition accuracy for small aperture leakage with equivalent diameter of 50—125mm, which improved the accuracy of small aperture leakage signals recognition. Thirdly, the accuracy of the optimized model for the identification of the compressor pumping suction and the block valve emergency shut off conditions were 96.7% and 100%, respectively. Lastly, when the leakage aperture was less than 50mm and the pressure drop rate was less than 0.01MPa/min, the pressure drop rate signal characteristics detected in the valve chamber were similar. It is suggested to use the gas-liquid linkage valve and the SCADA system monitoring data to comprehensively judge.

Key words: natural gas pipeline, block valve, rate of pressure drop, support vector machine, optimized algorithm

中图分类号:

贾文龙, 孙溢彬, 汤丁, 陈家文, 雷思罗, 李长俊. 基于支持向量机的输气管道泄漏压降信号智能识别方法[J]. 化工进展, 2022, 41(9): 4713-4722.

JIA Wenlong, SUN Yibin, TANG Ding, CHEN Jiawen, LEI Siluo, LI Changjun. Intelligent recognition method for pressure drop signals of gas pipeline leakage based on support vector machine[J]. Chemical Industry and Engineering Progress, 2022, 41(9): 4713-4722.

图/表 16

参考文献 20

1	李星局. 输气管道线路截断阀控制逻辑与阈值研究[D]. 北京: 中国石油大学(北京), 2019.
	LI Xingju. Research on control logic and threshold of automatic block valves in gas pipelines[D]. Beijing: China University of Petroleum (Beijing), 2019.
2	吴家勇, 李海娜, 王立坤, 等. 管道泄漏监测系统的性能评价指标[J]. 油气储运, 2017, 36(2): 209-213.
	WU Jiayong, LI Haina, WANG Likun, et al. Performance evaluation indices of pipeline leak monitoring systems[J]. Oil & Gas Storage and Transportation, 2017, 36(2): 209-213.
3	ZUO Lili, JIANG Fangmei, JIN Bin, et al. Value setting for the rate of pressure drop of automatic line-break control valves in natural gas pipelines[J]. Journal of Natural Gas Science and Engineering, 2015, 26: 803-809.
4	杨毅, 向敏, 孙晓波, 等. 输气管道泄漏后截断阀压降速率计算分析[J]. 油气田地面工程, 2020, 39(1): 40-44.
	YANG Yi, XIANG Min, SUN Xiaobo, et al. Calculation and analysis of pressure drop rate of block valve after leakage of gas transmission pipeline[J]. Oil-Gas Field Surface Engineering, 2020, 39(1): 40-44.
5	赫德明, 郑洪峰, 徐华龙, 等. 输气管道气液联动执行机构压降速率设定值与持续时间的研究[J]. 现代化工, 2016, 36(2): 194-196, 198.
	HE Deming, ZHENG Hongfeng, XU Hualong, et al. Setting value and duration of pressure drop rate of pneumatic-hydraulic actuator[J]. Modern Chemical Industry, 2016, 36(2): 194-196, 198.
6	廖钰朋, 贾文龙, 赵雯嘉, 等. 往复式压缩机抽吸工况下储气库注采干线压降速率研究[J]. 管道技术与设备, 2021(4): 1-5.
	LIAO Yupeng, JIA Wenlong, ZHAO Wenjia, et al. Study on pressure drop rate of main line of gas reservoir injection and production under pumping condition of reciprocating compressor[J]. Pipeline Technique and Equipment, 2021(4): 1-5.
7	孟令雅, 李玉星, 宋立群, 等. 输气管道泄漏音波传播特性及监测定位[J]. 天然气工业, 2010, 30(11): 74-79, 123.
	MENG Lingya, LI Yuxing, SONG Liqun, et al. Acoustic propagation characteristics, position monitoring and locating of gas transmission pipeline leakage[J]. Natural Gas Industry, 2010, 30(11): 74-79, 123.
8	焦敬品, 李勇强, 吴斌, 等. 基于BP神经网络的管道泄漏声信号识别方法研究[J]. 仪器仪表学报, 2016, 37(11): 2588-2596.
	JIAO Jingpin, LI Yongqiang, WU Bin, et al. Research on acoustic signal recognition method for pipeline leakage with BP neural network[J]. Chinese Journal of Scientific Instrument, 2016, 37(11): 2588-2596.
9	张瑞程, 王新颖, 胡磊磊, 等. 基于一维卷积神经网络的燃气管道泄漏声发射信号识别[J]. 中国安全生产科学技术, 2021, 17(2): 104-109.
	ZHANG Ruicheng, WANG Xinying, HU Leilei, et al. Acoustic emission signal identification of gas pipeline leakage based on one-dimensional convolution neural network[J]. Journal of Safety Science and Technology, 2021, 17(2): 104-109.
10	郝永梅, 杜璋昊, 杨文斌, 等. 基于改进ELMD和多尺度熵的管道泄漏信号识别[J]. 中国安全科学学报, 2019, 29(8): 105-111.
	HAO Yongmei, DU Zhanghao, YANG Wenbin, et al. Pipeline leakage signal identification based on improved ELMD and multi-scale entropy[J]. China Safety Science Journal, 2019, 29(8): 105-111.
11	逯雯雯. 基于改进EMD和SVM的输油管道泄漏信号处理方法研究[D]. 东营: 中国石油大学(华东), 2019.
	LU Wenwen. Research on oil pipeline leakage signal processing method based on improved EMD and SVM[D]. Dongying: China University of Petroleum (Huadong), 2019.
12	NING Fangli, CHENG Zhanghong, MENG Di, et al. Enhanced spectrum convolutional neural architecture: an intelligent leak detection method for gas pipeline[J]. Process Safety and Environmental Protection, 2021, 146: 726-735.
13	ZADKARAMI M, SHAHBAZIAN M, SALAHSHOOR K. Pipeline leak diagnosis based on wavelet and statistical features using Dempster-Shafer classifier fusion technique[J]. Process Safety and Environmental Protection, 2017, 105: 156-163.
14	LI Zhenlin, ZHANG Haifeng, TAN Dongjie, et al. A novel acoustic emission detection module for leakage recognition in a gas pipeline valve[J]. Process Safety and Environmental Protection, 2017, 105: 32-40.
15	LIU Cuiwei, LI Yuxing, FANG Liping, et al. Experimental study on a de-noising system for gas and oil pipelines based on an acoustic leak detection and location method[J]. International Journal of Pressure Vessels and Piping, 2017, 151: 20-34.
16	DIAO Xu, JIANG Juncheng, SHEN Guodong, et al. An improved variational mode decomposition method based on particle swarm optimization for leak detection of liquid pipelines[J]. Mechanical Systems and Signal Processing, 2020, 143: 106787.
17	王明达, 韦永健, 王建军, 等. 基于SVM-BN的天然气长输管道燃气轮机故障预警方法研究[J]. 天然气工业, 2017, 37(11): 112-118.
	WANG Mingda, WEI Yongjian, WANG Jianjun, et al. A SVM-BN based early warning method for gas turbines of natural gas long-distance pipelines[J]. Natural Gas Industry, 2017, 37(11): 112-118.
18	孙凤山, 范孟豹, 曹丙花, 等. 基于混沌映射与差分进化自适应教与学优化算法的太赫兹图像增强模型[J]. 仪器仪表学报, 2021, 42(4): 92-101.
	SUN Fengshan, FAN Mengbao, CAO Binghua, et al. The terahertz image enhancement model based on adaptive teaching-learning based optimization algorithm with chaotic mapping and differential evolution[J]. Chinese Journal of Scientific Instrument, 2021, 42(4): 92-101.
19	盛晓晨, 史旭东, 熊伟丽. 改进粒子群优化的极限学习机软测量建模方法[J]. 计算机应用研究, 2020, 37(6): 1683-1687.
	SHENG Xiaochen, SHI Xudong, XIONG Weili. Soft sensor modeling of extreme learning machine based on improved particle swarm optimization[J]. Application Research of Computers, 2020, 37(6): 1683-1687.
20	敖永才, 师奕兵, 张伟, 等. 自适应惯性权重的改进粒子群算法[J]. 电子科技大学学报, 2014, 43(6): 874-880.
	AO Yongcai, SHI Yibing, ZHANG Wei, et al. Improved particle swarm optimization with adaptive inertia weight[J]. Journal of University of Electronic Science and Technology of China, 2014, 43(6): 874-880.

标准或管线	压降速率/MPa·min^-1	持续时间/s
美国Shafer气液联动阀	0.15~0.30	—
BIFFI气液联动阀	0.20~0.30	—
《输气管道线路阀室技术与运行管理规定》	0.01~0.04	60~180
西气东输二线、陕京三线、相国寺储气库干线	0.15	120~180
安徽省某输气干线	0.08	120
陕京一线	0.06	120

标准或管线	压降速率/MPa·min^-1	持续时间/s
美国Shafer气液联动阀	0.15~0.30	—
BIFFI气液联动阀	0.20~0.30	—
《输气管道线路阀室技术与运行管理规定》	0.01~0.04	60~180
西气东输二线、陕京三线、相国寺储气库干线	0.15	120~180
安徽省某输气干线	0.08	120
陕京一线	0.06	120

项目	f₁	f₂	f₃	f₄	f₅	f₆
GA
均值	4.8×10^-1	3.9×10^-1	1.1×10^-1	3.9×10^-1	1.5×10^-1	3.4×10^-1
方差	3.1×10^-1	2.5×10^-1	5.4×10^-3	8.5×10^-2	6.6×10^-3	1.5×10^-1
收敛步数	724	222	448	821	533	650
PSO
均值	1.5×10^-230	3.4×10^-229	4.0×10^-116	8.9×10^-16	0	1.7×10^-1
方差	0	0	3.3×10^-231	1.0×10^-62	0	1.4×10^-1
收敛步数	—	—	—	120	256	73
TLBO
均值	0	0	1.2×10^-199	8.9×10^-16	0	0
方差	0	0	0	1.0×10^-62	0	0
收敛步数	824	846	—	90	325	89
ITLBO
均值	0	0	0	8.9×10^-16	0	0
方差	0	0	0	1.0×10^-62	0	0
收敛步数	523	527	747	56	110	50

项目	f₁	f₂	f₃	f₄	f₅	f₆
GA
均值	4.8×10^-1	3.9×10^-1	1.1×10^-1	3.9×10^-1	1.5×10^-1	3.4×10^-1
方差	3.1×10^-1	2.5×10^-1	5.4×10^-3	8.5×10^-2	6.6×10^-3	1.5×10^-1
收敛步数	724	222	448	821	533	650
PSO
均值	1.5×10^-230	3.4×10^-229	4.0×10^-116	8.9×10^-16	0	1.7×10^-1
方差	0	0	3.3×10^-231	1.0×10^-62	0	1.4×10^-1
收敛步数	—	—	—	120	256	73
TLBO
均值	0	0	1.2×10^-199	8.9×10^-16	0	0
方差	0	0	0	1.0×10^-62	0	0
收敛步数	824	846	—	90	325	89
ITLBO
均值	0	0	0	8.9×10^-16	0	0
方差	0	0	0	1.0×10^-62	0	0
收敛步数	523	527	747	56	110	50

分类器模型	优化参数结果		平均分类准确率/%
分类器模型	惩罚系数C	核函数参数g	平均分类准确率/%
GA-SVM	81.6	33.4	88.9
PSO-SVM	74.1	16.6	91.1
TLBO-SVM	91.6	18.3	94.3
ITLBO-SVM	89.7	96.4	98.5