基于贝叶斯网络的复杂系统故障预测

针对复杂电子系统信号具有不确定性的特点,提出一种基于贝叶斯网络的故障预测模型。该模型通过对连续的信号特征进行量化处理,利用专家知识结合信号建立贝叶斯网络结构;对不同样本采用不同算法来进行网络学习,采用概率推理定量估计信号的区间预测概率,从而建立一个可推理的预测模型。将该方法应用于电源系统进行故障预测,针对不同数据样本进行实验,结果验证具有较高的区间预测率,为复杂系统的故障预测提供了新手段。     

控制系统故障检测与诊断技术的最新进展

对故障检测与诊断(FDD)技术在控制领域的最新进展进行了综述,内容主要包括三类基本方法,即:基于解析模型的FDD、基于信号处理和基于知识的故障检测与诊断方法。重点评述了近年来发展起来的信息融合方法、多Agent方法、鲁棒故障检测、网络控制系统的故障诊断等。最后介绍了故障检测与诊断技术在控制系统中的若干应用成果,并指出尚待解决的主要问题和值得进一步研究的方向。     

贝叶斯检测跟踪联合处理的检测阈值设置方法

基于贝叶斯理论的检测跟踪联合处理方法能够利用目标的运动模型在帧数据间以概率的形式对信号进行积累,从而提高系统的检测性能。然而,在Neyman-Pearson准则下,该方法很难根据系统要求的虚警概率计算检测阈值,影响了该方法的应用范围。在综合分析基于贝叶斯理论的检测跟踪联合处理方法的基础上,针对Neyman-Pearson准则,详细推导了系统虚警概率同检测阈值之间的关系,并在观测噪声为高斯白噪声的情况下给出了检测阈值的近似闭式解。利用该闭式解,可以按照系统要求的虚警概率计算检测阈值,从而使得实际系统的虚警概率满足要求。最后,计算机仿真实验验证了该闭式解的有效性。     

Robust fault detection and diagnosis for uncertain nonlinear systems

This paper considers robust fault detection and diagnosis for input uncertain nonlinear systems. It proposes a multi-objective fault detection criterion so that the fault residual is sensitive to the fault but insensitive to the uncertainty as much as possible. Then the paper solves the proposed criterion by maximizing the smallest singular value of the transformation from faults to fault detection residuals while minimizing the largest singular value of the transformation from input uncertainty to the fault detection residuals. This method is applied to an aircraft which has a fault in the left elevator or rudder. The simulation results show the proposed method can detect the control surface failures rapidly and efficiently.     

一类不确定时延网络控制系统的故障诊断

研究了具有不确定时延的网络控制系统的故障诊断问题。通过建模将一类具有不确定时延的网络控制系统等效为具有不确定性的离散线性系统,并构建了故障观测器对系统进行故障检测。基于线性矩阵不等式(LMI)的方法,将故障诊断问题转化为系统鲁棒性的设计问题,给出并证明了解存在的条件以及观测器增益矩阵的求解方法。最后通过仿真验证了该算法的有效性。     

目标数据缺失下离散动态贝叶斯网络的参数学习

离散动态贝叶斯网络参数学习的难点在于:隐藏节点的片间转移概率获得及观测数据发生不同程度缺失。针对上述问题,提出基于目标缺失数据估计的前向递归参数学习算法。该算法利用离散动态贝叶斯网络中各观测变量与隐藏变量之间的对应关系,采用支持向量机建立观测变量间的非线性函数关系完成缺失数据估计,此基础上利用完整数据集和前向递归算法完成片内和片间参数更新。以空中目标识别为仿真背景,通过与期望最大算法对比,验证了该算法的学习效率和精度两个方面的优势。     

一种改进的卫星网络故障诊断算法及其评估

星际链路的加入提高了卫星网络的连通性和自治性，也使得整个网络更为复杂。故障检测与诊断是保证此类网络容错性的一项重要措施。在前面工作的基础上，提出一种改进的基于系统级诊断的卫星网络故障诊断算法，并给出诊断正确性的证明。通过仿真对算法在一种双层LEO／MEO网络中的通信开销和诊断完全率（诊断出状态的网络节点数在网络节点总数中所占百分比）进行深入分析。结果表明，改进后的算法比原来的算法具有更好的通用性和诊断完全率。     

基于自适应神经网络的非线性系统故障诊断

针对一类模型未知及状态不可测的非线性系统,提出了基于自适应神经网络的故障诊断策略,不仅在线估计神经网络的矩阵权重,而且在线估计高斯函数的宽度和中心。该方法对系统的未知非线性特性没有特别要求,仅对神经网络提出较弱的假设条件。首先利用径向基函数(Radial Basis Function,简称RBF)神经网络构造状态观测器,估计系统的状态。然后利用另一个自适应RBF神经网络作为故障估计器,其输入是系统的估计状态(而不是系统状态),其输出为系统所发生的故障模型。利用Lyapunov稳定理论详细分析了状态误差和故障误差的收敛性,分别给出了两个神经网络的参数调整律,仿真证明了该方法的实用性和有效性。     

WPA-IGA-BP神经网络的模拟电路故障诊断

针对模拟电路渐变性故障中的特征提取困难和故障信号无法进行有效分类的问题,提出利用免疫遗传算法(immune genetic algorithm,IGA)优化反向传播(back propagation,BP)神经网络中参数寻优过程,从而实现模拟电路故障诊断.首先,采用小波包分析(wavelet package analy...     

基于多模自适应滤波的无人机控制系统故障诊断

建立了无人机控制系统传感器和执行器的全局故障和局部故障的模型，在此基础上应用多重模型自适应卡尔曼滤波方法对其传感器和执行器的各种软硬故障进行诊断，应用所建立的数学模型与方法，对无人机的三个传感器和两个执行器的局部与全局故障进行了仿真计算。在仿真过程中发现，此方法的诊断准确度高，无延迟报警，算法简单，仿真结果验证了该种方法的有效性。     

Actuator fault diagnosis of time-delay systems based on adaptive observer

1 .INTRODUCTIONOwingtotheincreasing demandfor highreliabilityinmany industrial processes , much attention has beenpaid to the problem of fault detection and diagnosis(FDD) in dynamic systems over the past twodecades. Fruitful results can be found in Refs .[1 ~3] .It is well known that faults in a dynamic systemcan take many forms . They can be actuator faults ,sensor faults ,unexpected abrupt changes of some pa-rameters or even unexpected structure changes[4 ,5].The purpose of detection…     

线性时滞系统的鲁棒故障诊断

研究了一类含有未知输入干扰和模型不确定性的时滞系统故障诊断问题。通过变换,原系统被分成两个子系统,一个子系统不受故障的影响,可以设计鲁棒观测器;另一个子系统受到故障的影响,但是它的状态可以通过测量得到。设计的观测器利用解析冗余方法能够对执行器故障和传感器故障进行诊断。仿真结果表明,算法具有良好的诊断性能。     

Robust fault diagnosis with disturbance rejection and attenuation for systems with multiple disturbances

The fault diagnosis problem is investigated for a class of nonlinear neutral systems with multiple disturbances.Time-varying faults are considered and multiple disturbances are supposed to include the unknown disturbance modeled by an exo-system and norm bounded uncertain disturbance.A nonlinear disturbance observer is designed to estimate the modeled disturbance.Then,the fault diagnosis observer is constructed by integrating disturbance observer with disturbance attenuation and rejection performances.The augmented Lyapunov functional approach,which involves the tuning parameter and slack variable,is applied to make the solution of inequality more flexible.Finally,applications for a two-link robotic manipulator system are given to show the efficiency of the proposed approach.     

Fault detection for nonlinear networked control systems based on fuzzy observer

Security and reliability must be focused on control systems firstly, and fault detection and diagnosis (FDD) is the main theory and technology. Now, there are many positive results in FDD for linear networked control systems (LNCSs), but nonlinear networked control systems (NNCSs) are less involved. Based on the T-S fuzzy-modeling theory, NNCSs are modeled and network random time-delays are changed into the unknown bounded uncertain part without changing its structure. Then a fuzzy state observer is designed and an observer-based fault detection approach for an NNCS is presented. The main results are given and the relative theories are proved in detail. Finally, some simulation results are given and demonstrate the proposed method is effective.     

Fault detection filter design for linear discrete time-varying systems with multiplicative noise

The problem of fault detection for linear discrete timevarying systems with multiplicative noise is dealt with.By using an observer-based robust fault detection filter(FDF) as a residual generator,the design of the FDF is formulated in the framework of H ∞ filtering for a class of stochastic time-varying systems.A sufficient condition for the existence of the FDF is derived in terms of a Riccati equation.The determination of the parameter matrices of the filter is converted into a quadratic optimization problem,and an analytical solution of the parameter matrices is obtained by solving the Riccati equation.Numerical examples are given to illustrate the effectiveness of the proposed method.     

一类参数不确定非线性系统的故障检测与重构

飞行器在全包络上表现出明显的气动参数不确定性,以某无人机纵向模型为研究对象,提出一种不确定参数在线估计的自适应观测器故障重构方法。首先,将系统状态方程描述为一类带时变参数的仿射非线性结构,在参数增广系统能观性分析基础上,采用增广容积卡尔曼滤波(augmented cubature Kalman filter, ACKF)算法实现气动参数在线估计,以克服鲁棒性死区故障检测方法的保守性,提高检测灵敏度。其次,将所估计参数用于自适应观测器设计,由于Lie导数分析方法保证了对象系统的能观性,故系统不必满足文献方法中的特定规范形式;在此基础上,给出了故障检测自适应阈值和故障参数调节律,并分析了估计误差的收敛性。仿真实验表明了所提方法的有效性。     

奇异时滞系统的鲁棒故障检测和分离观测器

讨论了一类不确定奇异时滞系统的鲁棒故障检测和分离观测器(RDO)设计问题,设计所有的观测器使得它们的冗余信号对非故障因素具有鲁棒性,并设计每个观测器使其冗余信号对一组特定的故障不敏感而对其余故障敏感。给出了RDO存在的充分条件及有效的设计算法,并讨论了故障集合不同划分的RDO设计之间的关系,最后给出一具体算例来说明该方法的有效性。     

Robust fault detection for a class of nonlinear network control system with communication delay

To investigate a class of nonlinear network control system, a robust fault diagnosis method is presented based on the robust state observer. To access the objective that the designed robust filter is maximally tolerant to disturbances and sensitive to fault, the robustness and stability properties of the fault diagnosis scheme are established rigorously. Using the residual vector, a fault tolerant controller is established in order to guarantee the stability of the closed-loop system, and the controller law can be obtained by solving a set of linear matrix inequalities. Then, some relevant sufficient conditions for the existence of a solution are given by applying Lyapunov stability theory. Finally, a simulation example is performed to show the effectiveness of the proposed approach.     

Fault detection for networked systems subject to access constraints and packet dropouts

This paper addresses the problem of fault detection(FD) for networked systems with access constraints and packet dropouts.Two independent Markov chains are used to describe the sequences of channels which are available for communication at an instant and the packet dropout process,respectively.Performance indexes H∞ and H_ are introduced to describe the robustness of residual against external disturbances and sensitivity of residual to faults,respectively.By using a mode-dependent fault detection filter(FDF) as residual generator,the addressed FD problem is converted into an auxiliary filter design problem with the above index constraints.A sufficient condition for the existence of the FDF is derived in terms of certain linear matrix inequalities(LMIs).When these LMIs are feasible,the explicit expression of the desired FDF can also be characterized.A numerical example is exploited to show the usefulness of the proposed results.     

Fault detection observer design for networked control system with long time-delays and data packet dropout

Focusing on the networked control system with long time-delays and data packet dropout, the problem of observerbased fault detection of the system is studied. According to conditions of data arrival of the controller, the state observers of the system are designed to detect faults when they occur in the system. When the system is normal, the observers system is modeled as an uncertain switched system. Based on the model, stability condition of the whole system is given. When conditions are satisfied, the system is asymptotically stable. When a fault occurs, the observers residual can change rapidly to detect the fault. A numerical example shows the effectiveness of the proposed method.