基于滤波器的风能转换系统传感器故障诊断与隔离

需积分: 10 111 浏览量更新于2024-08-13 收藏 1.53MB PDF 举报

"这篇研究论文主要探讨了风能转换系统（WECS）的传感器故障诊断和隔离问题，采用了基于滤波器的创新方法。文章指出，随着风能转换系统的运行可靠性变得日益重要，对这类系统的故障检测和诊断技术的研究变得至关重要。作者们来自中国多个知名的电力和工程技术学院，包括华北电力大学、珠海供电局、太原理工大学和北方工业大学等。在研究中，首先建立了一个考虑了典型传感器故障的变速风能转换系统物理模型。由于风速和测量噪声的非高斯特性，研究团队提出了一种改进的熵优化准则来设计针对WECS的滤波器。这种方法通过广义密度演化方程（GDEE）来分析估计误差、非高斯噪声和滤波器增益之间的相互作用，这与传统的熵滤波方法有所区别。接下来，利用滤波器产生的残差信号，结合评估决策规则，开发了一套传感器故障检测和隔离算法。这种方法的优越性在于能够有效地识别和定位WECS中的传感器故障。仿真结果验证了该方法的有效性，表明它能够在实际操作环境中成功地检测和隔离传感器故障。论文的贡献在于提供了一种适用于非高斯噪声环境的故障诊断策略，这对于提升风能转换系统的稳定性和效率具有重要意义。这一研究成果不仅有助于预防和减少因传感器故障导致的系统性能下降，还可能推动风能行业的故障预测和健康管理技术的进步。"

Jianhua Zhang

State Key Laboratory of Alternate Electrical Power

System With Renewable Energy Sources,

North China Electric Power University,

Beijing 102206, China

Jing Xiong

Zhuhai Power Supply Bureau,

Zhuhai 519000, China

Mifeng Ren

College of Information Engineering,

Taiyuan University of Technology,

Taiyuan 030024, China

e-mail: renmifeng@126.com

Yuntao Shi

Key Lab of Field Bus and

Automation of Beijing,

North China University of Technology,

Beijing 100144, China

Jinliang Xu

The Beijing Key Laboratory of New

and Renewable Energy,

North China Electric Power University,

Beijing 102206, China

Filter-Based Fault Diagnosis

of Wind Energy Conversion

Systems Subject to

Sensor Faults

The operational reliability of wind energy conversion systems (WECSs) has attracted a

lot of attention recently. This paper is concerned with sensor fault detection (FD) and iso-

lation problems for variable-speed WECSs by using a novel ﬁltering method. A physical

model of WECS with typical sensor faults is ﬁrst built. Due to the non-Gaussianity of

both wind speed and measurement noises in WECSs, an improved entropy optimization

criterion is then established to design the ﬁlter for WECSs. Different from previous

entropy-ﬁltering results, the generalized density evolution equation (GDEE) is adopted to

reveal the relationship among the estimation error, non-Gaussian noises, and the ﬁlter

gain. The sensors FD and isolation algorithms are then obtained by evaluating the deci-

sion rule based on the residual signals generated by the ﬁlter. Finally, simulation results

show that the sensor faults in WECSs can be detected and isolated effectively by using

the proposed method. [DOI: 10.1115/1.4032827]

Keywords: wind energy conversion systems, fault diagnosis, minimum error entropy,

sensor faults

1 Introduction

Due to the rapid development and strong requirement of wind

turbine technology, fault diagnosis of the WECSs has received a

lot of attention in recent years [1]. The reliability of more than

6000 modern onshore wind turbines and their subassemblies have

been investigated [2].

Because of the direct interaction with environment, which

makes it susceptible to noise, aging, manufacturing defects, etc.,

sensors are vulnerable to faults. Therefore, sensor fault diagnosis

has attracted a lot of attention and some strategies have been pro-

posed [3–13]. Generally speaking, these sensor fault diagnosis

approaches can be mainly classiﬁed into two groups: (1) observer

or ﬁlter-based method [3] and (2) statistic approach based on the

Bayesian theorem, Monte Carlo approach, likelihood method, and

hypothesis test technique [4].

These observer or ﬁlter-based sensor fault diagnosis methodol-

ogies can diagnose sensor faults through the analysis and dis-

posal of the residual generated by observer or ﬁlter [5–7]. An

approach for detecting and isolating sensor fault was presented

and applied into aircraft by using an extended Kalman ﬁlter [5].

An adaptive sliding mode observer was developed for sensor

fault diagnosis in an industrial gas turbine with parameter uncer-

tainties [6]. The sensor fault of a microelectromechanical system

was diagnosed by an agile and efﬁcient extended Kalman ﬁlter

[7]. The observer or ﬁlter-based fault diagnosis methods applied

in WECSs proved effective as well [8–13]. Liu et al. [8] and

Wei and Liu [9] utilized mixed H



observer to diagnose

power sensor fault and blade root bending moment sensor fault

in various speed wind turbine systems, respectively. Both Odg-

aard and Stoustrup [10] and Chen et al. [11] employed the

unknown input observer for sensor and process FD in WECSs.

In Ref. [12], a robust fuzzy fault-tolerant control algorithm was

used to deal with sensor faults, wind disturbances, and paramet-

ric uncertainties of variable-speed WECSs based on a group of

observers, which supposed that only one sensor failed at each

instant. A bank of observers were applied to detect multiple sen-

sor faults occurred simultaneously in doubly fed induction gener-

ators [13].

The main limitation of the above method is that it needs the

exact system model, however, it is difﬁcult to obtain exact

models due to nonlinearities, uncertainties, and multivariable

coupling existed in practical industrial processes. Hence, the

statistic approach might be an alternative fault diagnosis strat-

egy. If the statistical properties of the noise, together with the

way it affects the plant outputs, are known (or can be reason-

ably approximated or assumed), the FD and isolation problem

can then be formulated in the framework of statistical

decision-making. In Ref. [14], a sensor FD and identiﬁcation

approach for steam generators in thermal power plants was

presented. Based on the identiﬁed model, a data-driven fault

detector was then formulated. Subspace identiﬁcation techni-

ques have been adopted for sensor FD of wind turbine in Ref.

[15]. Luo et al. [16] extended the data-driven scheme to fault-

tolerant control of WECSs.

The above-mentioned approaches presented for WECSs hardly

obtained satisﬁed diagnosis results in practical WECSs subject to

non-Gaussian disturbances induced by both wind speed and mea-

surement noises. A few improved ﬁlter-based fault diagnosis

scheme has been proved to be an efﬁcient method to solve FD and

isolation problems for nonlinear and non-Gaussian systems

recently. Guo and coworkers [17] and [18] presented ﬁltering

algorithms for stochastic distribution systems based on minimum

Corresponding author.

Contributed by the Dynamic Systems Division of ASME for publication in the

OURNAL OF DYNAMIC SYSTEMS,MEASUREMENT, AND CONTROL. Manuscript received

April 28, 2015; ﬁnal manuscript received December 21, 2015; published online April

1, 2016. Assoc. Editor: Ryozo Nagamune.

Journal of Dynamic Systems, Measurement, and Control JUNE 2016, Vol. 138 / 061008-1

2016 by ASME

Downloaded From: http://dynamicsystems.asmedigitalcollection.asme.org/ on 12/24/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

下载后可阅读完整内容，剩余9页未读，立即下载

weixin_38613548

粉丝: 4
资源: 934

基于滤波器的风能转换系统传感器故障诊断与隔离

电子政务-基于DSP的风力发电机组无传感器故障诊断.zip

三个并联电压源换流器的故障诊断MATLAB仿真设计.doc

行业-电子政务-用于控制功率变换器的控制器和方法及电力收集系统.zip

MATLAB实现风电机组故障诊断技术探究

对数在工程学中的应用：信号处理和控制系统，构建智能系统

STM32单片机电源系统设计指南：从原理到应用，打造高效电源系统

电力拖动系统中的电源质量改善技术

电机控制系统逆变器技术详解：原理、拓扑结构及应用实践

数据分析高级培训：客户体验分析-课件

【微电网潮流】分布式电源微电网潮流【Matlab仿真 7357期】.zip

最新资源