双闭环比值控制系统传感器故障诊断方法

doi:10.12382/bgxb.2022.0969

摘要/Abstract

摘要：

双闭环比值控制系统由于主从回路滞后效应和闭环回路调节作用,使其故障诊断较为复杂。针对该系统传感器故障难以识别的问题,通过分析反映系统动态特性的实时数据,提出一种基于数据驱动的故障诊断方法。利用拉伊达准则标定检测阈值建立故障检测模型;利用传感器故障传播的滞后特性建立故障定位模型;利用突变故障的幅值特性建立了故障估计模型;利用故障调节动态过程的变化趋势建立了故障分离模型;通过在线数据标定获得了故障诊断动态模型。通过实验验证该方法的有效性和高精度。研究结果表明,所提方法适用于一般双闭环比值控制系统传感器的实时故障诊断,其中故障检测准确率可达98.5%,故障定位准确率可达98.5%,故障估计准确率可达99.28%,故障分离准确率可达98%。

关键词: 传感器, 故障诊断, 双闭环比值控制系统, 数据驱动, 液位系统

Abstract:

The fault diagnosis of double closed loop ratio control system is more complicated due to the lag effect of master slave loop and the regulating effect of closed loop. Aiming at the problem that sensor faults are difficult to be identified in this system, a data-driven fault diagnosis method is proposed by analyzing the real-time data reflecting the dynamic characteristics of the system. In this method, a fault detection model is established by calibrating the detection threshold using Pauta criterion; A fault location model is established based on the hysteresis characteristic of sensor fault propagation; a fault estimation model is established based on the amplitude characteristics of abrupt faults; a fault separation model is established by using the change trend of dynamic process of fault regulation; and a dynamic model of fault diagnosis is obtained by online data calibration. The effectiveness and high precision of the proposed method are verified by experiments. The results show that the proposed method is suitable for the real-time fault diagnosis of sensors in general double closed loop ratio control system. The accuracy of fault detection can reach 98.5%, the accuracy of fault location can reach 98.5%, the accuracy of fault estimation can reach 99.28%, and the accuracy of fault separation can reach 98%.

Key words: sensor, fault diagnosis, double closed loop ratio control system, data driven, liquid-level system

中图分类号:

TP277

王伽豪, 那文波, 刘志威, 昝琪, 王铮. 双闭环比值控制系统传感器故障诊断方法[J]. 兵工学报, 2024, 45(4): 1252-1263.

WANG Jiahao, NA Wenbo, LIU Zhiwei, ZAN Qi, WANG Zheng. Sensor Fault Diagnosis Method of Double Closed Loop Ratio Control System[J]. Acta Armamentarii, 2024, 45(4): 1252-1263.

图/表 20

图1 双闭环比值控制系统方框图

Fig.1 Block diagram of double closed loop ratio control system

图2 传感器单点故障仿真曲线

Fig.2 Simulation curve of sensor single point fault

图3 传感器两点故障仿真曲线

Fig.3 Simulation curve of sensor two-point fault

图4 实时故障诊断流程

Fig.4 Flow chart of real-time fault diagnosis

图5 双闭环比值控制系统故障诊断验证实验模型

Fig.5 Verification experimental model for fault diagnosis of double closed loop ratio control system

表1 无故障最大液位波动值

Table 1 Maximum liquid level fluctuation value without fault

序号	主传感器最大波动值/cm	从传感器最大波动值/cm
1	0.0868	0.1357
2	0.1032	0.1121
3	0.0957	0.1032
4	0.1023	0.0953
5	0.1229	0.1418
6	0.1321	0.1231
7	0.0957	0.0957
8	0.0822	0.0872
9	0.0752	0.0621
10	0.1012	0.0912

图6 主传感器故障检测

Fig.6 Fault detection of main sensor

图7 从传感器故障检测

Fig.7 Fault detection of follow sensor

图8 主传感器单点故障定位

Fig.8 Single fault location of main sensor

图9 从传感器单点故障定位

Fig.9 Single fault location of follow sensor

图10 传感器双重故障定位

Fig.10 Double fault location of sensor

图11 加性故障大小估计

Fig.11 Additive fault size estimation

图12 乘性故障大小估计

Fig.12 Multiplicative fault size estimation

图13 主传感器故障分离模型

Fig.13 Fault isolation model of main sensor

图14 从传感器故障分离模型

Fig.14 Fault isolation model of follow sensor

表2 主传感器20%内单点故障分离结果

Table 2 The result of single fault isolation of main sensor within failure intensity of 20%

强度/%	真实值	预测值	强度/%	真实值	预测值
1.5	加性	加性	1.5	乘性	乘性
2	加性	乘性	2	乘性	乘性
4	加性	加性	4	乘性	乘性
6	加性	加性	6	乘性	乘性
8	加性	加性	8	乘性	乘性
10	加性	加性	10	乘性	乘性
14	加性	加性	14	乘性	乘性
16	加性	加性	16	乘性	乘性
20	加性	加性	20	乘性	乘性

表3 从传感器20%内单点故障分离结果

Table 3 The result of single fault isolation of follow sensor within failure intensity of 20%

强度/%	真实值	预测值	强度/%	真实值	预测值
1.5	加性	乘性	1.5	乘性	乘性
2	加性	乘性	2	乘性	乘性
4	加性	加性	4	乘性	乘性
6	加性	加性	6	乘性	乘性
8	加性	加性	8	乘性	乘性
10	加性	加性	10	乘性	乘性
14	加性	加性	14	乘性	乘性
16	加性	加性	16	乘性	乘性
20	加性	加性	20	乘性	乘性

表4 从传感器18%内两点故障分离结果

Table 4 The result of double fault isolation of follow sensor within failure intensity of 18%

强度/%	真实值	预测值	强度/%	真实值	预测值
2	加性	乘性	2	乘性	乘性
4	加性	加性	4	乘性	乘性
6	加性	加性	6	乘性	乘性
8	加性	加性	8	乘性	乘性
10	加性	加性	10	乘性	乘性
14	加性	加性	14	乘性	乘性
18	加性	加性	18	乘性	乘性

表5 不同方法诊断精度对比

Table 5 Comparison of diagnostic accuracies of different methods%

诊断方法	故障检测	故障定位	故障估计	故障分离
WL方法^[18]				82.5
CNN方法^[19]				93.6
SVM方法^[20]	98.2			97.9
KNN方法^[21]		98.0
本文方法	98.5	98.5	99.3	98.0

表6 不同方法复杂度对比

Table 6 Comparison of complexities of different methods

诊断方法	空间复杂度	时间复杂度
WL方法^[18]	O(N³)	O(Nlog ₂N)
CNN方法^[19]	O(N)	O(N²)
SVM方法^[20]	O(N³)	O(N²)
KNN方法^[21]	O(N³)	O(N²)
本文方法	O(N)	O(N)

参考文献 21

[1]	李良钰, 苏铁熊, 马富康, 等. 基于集合经验模态分解-支持向量机的高压共轨系统故障诊断方法[J]. 兵工学报, 2022, 43(5):992-1001. doi: 10.12382/bgxb.2021.0155
	LI L Y, SU T X, MA F K, et al. Fault diagnosis method of high pressure common rail system based on ensemble empirical mode decomposition support vector machine[J]. Acta Armamentarii, 2022, 43(5):992-1001. (in Chinese)
[2]	王英敏, 崔涛, 张付军, 等. 基于局部线性模型树模型的柴油机进气系统漏气和堵塞故障诊断[J]. 兵工学报, 2017, 38(8):1457-1468. doi: 10.3969/j.issn.1000-1093.2017.08.001
	WANG Y M, CUI T, ZHANG F J, et al. Fault diagnosis of air leakage and blockage in diesel engine intake system based on local linear model tree model[J]. Acta Armamentarii, 2017, 38(8):1457-1468. (in Chinese)
[3]	李志农, 胡志峰, 毛清华, 等. 非线性调频模态分解-同步提取变换方法及其在滚动轴承故障诊断中的应用[J]. 兵工学报, 2021, 42(6):1324-1330.
	LI Z N, HU Z F, MAO Q H, et al. Nonlinear frequency modulation mode decomposition synchronous extraction transformation method and its application in rolling bearing fault diagnosis[J]. Acta Armamentarii, 2021, 42(6):1324-1330. (in Chinese)
[4]	FU F Z, XUE T, WU Z G, et al. A fault diagnosability evaluation method for dynamic systems without distribution knowledge[J]. IEEE Transactions on Cybernetics, 2022, 52(6):5113-5123.
[5]	GAO Z W, CECATI C, DING S X. A survey of fault diagnosis and fault-tolerant techniques—Part II: fault diagnosis with knowledge-based and hybrid/active approaches[J]. IEEE Transactions on Industrial Electronics, 2015, 62(6):3768-3774.
[6]	ZHIRABOK A N, SHUMSKY A E, ZUEV A V. Fault diagnosis in linear systems via sliding mode observers[J]. International Journal of Control, 2021, 94(2):327-335.
[7]	HAN X J, HU Y R, XIE A H, et al. Quadratic-Kalman-filter-based sensor fault detection approach for unmanned aerial vehicles[J]. IEEE Sensors Journal, 2022, 22(19):18669-18683.
[8]	XIAO Y J, HAN F R, DING Y H, et al. Research on fault diagnosis method of rapier loom based on the fusion of expert system and fault tree[J]. Journal of Intelligent and Fuzzy Systems, 2021, 41(2):3429-3441.
[9]	KONG X Y, GE Z Q. Deep learning of latent variable models for industrial process monitoring[J]. IEEE Transactions on Industrial Informatics, 2022, 18(10):6778-6788.
[10]	SONG J, SHANG W Z, AI S J, et al. Model and data-driven combination: a fault diagnosis and localization method for unknown fault size of quadrotor UAV actuator based on extended state observer and deep forest[J]. Sensors, 2022, 22(19):7355.
[11]	XING Y Q, WANG B W, GONG Z C, et al. Data-driven fault diagnosis for PEM fuel cell system using sensor pre-selection method and artificial neural network model[J]. IEEE Transactions on Energy Conversion, 2022, 37(3):1589-1599.
[12]	GOU B, XU Y, XIA Y, et al. An intelligent time-adaptive data-driven method for sensor fault diagnosis in induction motor drive system[J]. IEEE Transactions on Industrial Electronics, 2019, 66(12):9817-9827.
[13]	XIA Y, XU Y, GOU B. Current sensor fault diagnosis and fault-tolerant control for single-phase PWM rectifier based on a hybrid model-based and data-driven method[J]. Iet Power Electronics, 2020, 13(18):4150-4157.
[14]	JI H Q. Statistics Mahalanobis distance for incipient sensor fault detection and diagnosis[J]. Chemical Engineering Science, 2020, 230(16):116233.
[15]	SUN Y Y, LIU S X, ZHAO T T, et al. A new hydrogen sensor fault diagnosis method based on transfer learning with LeNet-5[J]. Frontiers in Neurorobotics, 2021, 15:664135.
[16]	LUO X J, FONG K F, SUN Y J, et al. Development of clustering-based sensor fault detection and diagnosis strategy for chilled water system[J]. Energy and Buildings, 2019, 186(1):17-36.
[17]	ZHU P C, QIAN H, CHAI T T. Research on early fault warning system of coal mills based on the combination of thermodynamics and data mining[J]. Transactions of The Institute of Measurement and Control, 2020, 42(1):55-68.
[18]	任海锋, 潘宏侠. 基于多分形特征的枪械自动机裂纹故障诊断[J]. 兵工学报, 2018, 39(3):457-462. doi: 10.3969/j.issn.1000-1093.2018.03.006
	REN H F, PAN H X. Fault diagnosis of gun automata cracks based on multifractal features[J]. Acta Armamentarii, 2018, 39(3):457-462. (in Chinese)
[19]	WANG S H, XIANG J W, ZHONG Y T, et al. Convolutional neural network-based hidden Markov models for rolling element bearing fault identification[J]. Knowledge-Based Systems, 2018, 144(12):65-76.
[20]	ZHANG X Y, LIANG Y T, ZHOU J Z, et al. A novel bearing fault diagnosis model integrated permutation entropy, ensemble empirical mode decomposition and optimized SVM[J]. Measurement, 2015, 69(3):164-179.
[21]	SWETAPADMA A, YADAV A. A novel single-ended fault location scheme for parallel transmission lines using k-nearest neighbor algorithm[J]. Computers and Electrical Engineering, 2018, 69(5):41-53.