多速率传感器融合系统故障检测方法

139 浏览量更新于2024-08-27 收藏 484KB PDF 举报

"这篇研究论文探讨了在多速率传感器融合系统中如何在多种不确定性情况下进行故障检测。在分布式和/或异构系统中，不同传感器的采样率往往不一致，同时，系统建模可能面临复杂的环境，如随机噪声、未知输入（UI）和设备故障等多重不确定性。作者提出了一种针对带有UI、已知方差随机噪声以及作用于执行器和传感器上的故障的多速率传感器融合系统的故障检测问题。此外，他们还提出了一种新的多速率观测器（MRO）设计，并将其提升至单一框架下，以解决这些挑战。" 详细说明: 该论文主要关注的是在具有多速率传感器融合的复杂系统中的故障检测。在现实世界的应用中，例如航空航天、自动化和工业控制等领域，传感器通常以不同的速率提供数据，这增加了故障检测的复杂性。论文指出，由于系统可能同时受到随机噪声、未知输入和设备故障等多种不确定性的影响，传统的故障检测方法可能不足以应对这些挑战。作者提出的方法着重于设计一种多速率观测器，这种观测器能够处理来自不同采样速率传感器的数据。观测器的目标是实时估计系统的状态，从而有效地检测和隔离可能的故障。通过将多速率数据集成到一个统一的框架中，可以更准确地识别出由于采样率不匹配而可能被忽视的异常行为。此外，考虑到随机噪声的存在，论文假设其具有已知的协方差，这意味着可以通过统计方法来量化和处理这种不确定性。对于未知输入（UI），即那些无法直接观测到但对系统状态有影响的因素，作者也考虑了它们对故障检测的影响。同样，他们还考虑了执行器和传感器可能出现的故障，这些都是系统可靠性的重要因素。论文的创新之处在于提出的新多速率观测器设计，它旨在克服由于采样率差异和不确定性带来的问题，提供一个更全面的故障检测策略。通过这样的方法，系统可以更加灵敏地识别和响应潜在的故障情况，从而提高整个系统的稳定性和安全性。这篇研究论文为处理多速率传感器融合系统中不确定性下的故障检测提供了一个新的理论框架，对于实际工程应用具有重要的指导价值。通过构建能够处理不同采样率和多种不确定性的观测器，可以更好地确保系统的健康运行，并提高故障管理的效率。

IET Control Theory & Applications

Brief Paper

Fault detection for multi-rate sensor fusion

under multiple uncertainties

ISSN 1751-8644

Received on 14th July 2014

Accepted on 5th December 2014

doi: 10.1049/iet-cta.2014.1134

www.ietdl.org

Hang Geng

1,2

,Yan Liang

1,2

, Linfeng Xu

1,2

School of Automation, Northwestern Polytechnical University, Xi’an, People’s Republic of China

Key Laboratory of Information FusionTechnology, Ministry of Education, Xi’an, People’s Republic of China

E-mail: liangyan@nwpu.edu.cn

Abstract: In multi-sensor fusion, it is hard to guarantee that all sensors have an identical sampling rate, especially in the

distributive and/or heterogeneous case. Meanwhile, system modelling may face the coexistence of multiple uncertainties

including stochastic noise, unknown input (UI) and faults in complex environment. To this end, the authors propose the

problem of fault detection for multi-rate sensor fusion systems subject to UI, stochastic noise with known covariance, and

faults imposed on the actuator and sensors. Furthermore, the new form of multi-rate observer (MRO) is presented and lifted

to the single-rate one with causality constraint for parameter design. Observer parameters are determined optimally in

pursuit of the UI decoupling and maximum noise attenuation under the causality constraint. Differing from the traditional

observer, the proposed MRO is time varying, that is, its parameters need recursive computation and hence has better

adaptability to the effect of uncertainties. Finally, a multi-rate residual generator is constructed via a hypothesis test in

which the threshold is adaptively designed. A numerical example is given to show the effectiveness of their proposed

method.

1 Introduction

In some complex systems, it is often unrealistic or sometimes

impossible to guarantee that all physical signals operate at one sin-

gle rate [1]. For example, for signals with different bandwidths,

better trade-offs between performance and implementation cost

can be obtained using A/D and D/A converters at different rates.

On the other hand, for processed/estimated quantities, sometimes

users may specify rates which are different from the sampling

rates of sensors. Therefore researches on multi-rate sensor fusion

arise.

State estimation approach based on dual-rate sensor measure-

ments was proposed in [2] and extended to deal with the asyn-

chronous multi-rate sensor measurements, where the ratio between

the sampling rates of different sensors was allowed to be any pos-

itive integer [3]. Accounting for out of sequence data and latent

data, general asynchronous fusion estimation methods were put

forward for multi-sensor systems [4, 5]. Since different dynamic

models always have different frequency properties in multiple

model systems, the fast-rate sensor measurement was thus com-

pressed to a slow rate sub-situation with little or no accuracy

degradation in low-frequency models, and hence results in multi-

rate interacting multiple model estimators [6] and its target-tracking

application with out-of-sequence GMTI data [7] or distributed

fusion [8]. In the case that the updating rate of state estimates

is different from the measurement sampling rate, the wavelet-

transform-based and the optimal H

∞

-based estimation schemes

were proposed in [9, 10], respectively. For a four-rate estimation

problem including the state updating rate, measurement sampling

rate, estimate updating rate and the estimate output rate, Liang

et al. [11] presented a linear minimum variance (LMV) estima-

tor. For systems existing measurement missing or packet losses,

Liang et al. [12] and Zhang et al. [13] presented a multi-rate

∞

method and a two-stage distributed fusion estimation method,

respectively. When such packet dropouts are multiple, the optimal

LMV centralised and decentralised fusion estimators are designed

in [14]. In addition, Zhang et al. [15] concerned the multi-sensor

fusion estimation scheme for wireless sensor networks with non-

uniform estimation rates and provided two fusion algorithms that

can fuse available local estimates generated at different time

scales, allowing estimation rates at different sensors to be different

from each other. In general, no fault is considered in the above

research.

Concerning the fault detection (FD) problem in multi-rate sen-

sor systems with unknown input (UI) in the state equation, the

lifting technique was used to convert the multi-rate system into a

linear time-invariant (LTI) model with slow sampling rate, based

on which residual generators were designed using state observers

[16, 17], but the residual can only be updated at the same rate as

the LTI model resulting in a slow-rate FD scheme. For fast-rate

residual generation, an observer was designed for each set of syn-

chronous measurements resulting in a bank of observers that run

simultaneously at different rates [18]. For multi-rate systems with

UI both in the state equation and output equation, by optimising a

performance index which maximises the effect of the fault on the

residual, at the same time, minimises the inﬂuence of the UI on the

residual, Izada et al. [19] presented a parity-space-based residual

generator, and Izada et al. [20] proposed different H

∞

optimal and

causal residual generators. For the case of norm-bounded UI [21],

an observer-based FD ﬁlter (FDF) was considered as a residual

generator and the residual was obtained by solving an H

∞

−

optimisation problem.

However, all the above researches are limited to the scope of

deterministic systems, that is, the stochastic noise is not considered

either in the state equation or output equation although random

noise is common to state evolvement and sensor sampling in com-

plex environment. In other words, an interesting but still open

problem is FD of multi-sensor fusion in the presence of UI and

noise.

In this paper, the FD problem for the multi-rate sensor fusion

system with both UI and noise is proposed. First, we proposed a

new multi-rate observer (MRO) to generate the multi-rate resid-

ual, and the MRO parameters are determined through decoupling

the residual with the UI and minimising the effect of the noise

on the residual in the minimum-mean-square error (MMSE) sense.

The disturbance decoupling and noise attenuation conditions are

found to be closely related with sensor rates and the causality con-

straint because of the multi-rate nature. The residual is evaluated

via a hypothesis test in which the detection threshold is adaptively

determined.

IET Control Theory Appl., 2015, Vol. 9, Iss. 11, pp. 1709–1716

下载后可阅读完整内容，剩余7页未读，立即下载

夏影影

粉丝: 318
资源: 914

多速率传感器融合系统故障检测方法

The joint optimal filtering and fault detection for multi-rate sensor fusion under unknown inputs

signal 6 (SIGABRT), code -1 (SI_QUEUE), fault addr --------

signal 8 (SIGFPE), code -6 (SI_TKILL), fault addr --------

signal 6 (SIGABRT), code -6 (SI_TKILL), fault addr --------错误的原因是什么？

signal 0 (SIGABRT), code -2 (SI_TIMER from pid 5763, uid 0), fault addr --------

基于知识图谱的SDN网络故障参考文献

No such file or directory: 'E:\\codes\\Bearing-fault-Diagnosis-based-on-deep-learning\\T_SNE\\cwt_picture\\train\\0-0.jpg

帮我找一点轿厢里面的数据集，给我下载链接

haspuserstep 微软终止代码page-fault-in-nonpaged-area

最新资源