An overview on fault diagnosis and nature-inspired optimal control of industrial process applications

Fault detection, isolation and optimal control have long been applied to industry. These techniques have proven various successful theoretical results and industrial applications. Fault diagnosis is considered as the merge of fault detection (that indicates if there is a fault) and fault isolation (that determines where the fault is), and it has important effects on the operation of complex dynamical systems specific to modern industry applications such as industrial electronics, business management systems, energy, and public sectors. Since the resources are always limited in real-world industrial applications, the solutions to optimally use them under various constraints are of high actuality. In this context, the optimal tuning of linear and nonlinear controllers is a systematic way to meet the performance specifications expressed as optimization problems that target the minimization of integral- or sum-type objective functions, where the tuning parameters of the controllers are the vector variables of the objective functions. The nature-inspired optimization algorithms give efficient solutions to such optimization problems. This paper presents an overview on recent developments in machine learning, data mining and evolving soft computing techniques for fault diagnosis and on nature-inspired optimal control. The generic theory is discussed along with illustrative industrial process applications that include a real liquid level control application, wind turbines and a nonlinear servo system. New research challenges with strong industrial impact are highlighted.     

Optimal State Estimation and Fault Diagnosis for a Class of Nonlinear Systems

This study proposes a scheme for state estimation and,consequently,fault diagnosis in nonlinear systems.Initially,an optimal nonlinear observer is designed for nonlinear systems subject to an actuator or plant fault.By utilizing Lyapunov's direct method,the observer is proved to be optimal with respect to a performance function,including the magnitude of the observer gain and the convergence time.The observer gain is obtained by using approximation of Hamilton-Jacobi-Bellman(HJB)equation.The approximation is determined via an online trained neural network(NN).Next a class of affine nonlinear systems is considered which is subject to unknown disturbances in addition to fault signals.In this case,for each fault the original system is transformed to a new form in which the proposed optimal observer can be applied for state estimation and fault detection and isolation(FDI).Simulation results of a singlelink flexible joint robot(SLFJR)electric drive system show the effectiveness of the proposed methodology.     

Actuator fault-tolerant control design: Demonstration on a three-tank-system

Fault-tolerant control or reconfigurable control systems are generally based on a nominal control law associated with a fault detection and isolation module. A general review of techniques dealing with this problem is given and a new fault-tolerant control approach is presented. This method is based on the on-line estimation of an eventual fault and the addition of a new control law to the nominal control law in order to reduce the fault effect once this fault is detected and isolated. The performances of this method depend on the time delay between the occurrence of the fault and its detection and isolation. A modified approach is then proposed in order to avoid the problems generated by delays, false alarms or non-detection inherent to diagnosis techniques. These methods are applied to a pilot plant and their performances are compared and discussed.     

基于级联观测器的多输入多输出非线性系统的故障诊断

将块观测方法应用于非线性系统的故障检测和分离.首先给出了非线性系统的块观测形式,针对传感器故障和执行器故障对非线性系统进行分块,得到了带有故障系统的块观测器形式.利用滑模观测器实现系统状态观测,得到观测器误差;利用所设计的观测器对非线性系统进行故障的诊断和分离;采用等效输出注入概念重构了故障信号,使得多变量输入输出非线性系统的故障诊断问题得到了解耦;针对异步电动机系统实现了传感器故障的分离.仿真结果证明了算法的有效性.     

System identification applied to stiction quantification in industrial control loops: A comparative study

A comparative study of different models and identification techniques applied to the quantification of valve stiction in industrial control loops is presented in this paper, with the objective of taking into account for the presence of external disturbances. A Hammerstein system is used to model the controlled process (linear block) and the sticky valve (nonlinear block): five different candidates for the linear block and two different candidates for the nonlinear block are evaluated and compared. Two of the five linear models include a nonstationary disturbance term that is estimated along with the input-to-output model, and these extended models are meant to cope with situations in which significant nonzero mean disturbances affect the collected data. The comparison of the different models and identification methods is carried out thoroughly in three steps: simulation, application to pilot plant data and application to industrial loops. In the first two cases (simulation and pilot plant) the specific source of fault (stiction with/without external disturbances) is known and hence a validation of each candidate can be carried out more easily. Nonetheless, each fault case considered in the previous two steps has been found in the application to a large number of datasets collected from industrial loops, and hence the merits and limitations of each candidate have been confirmed. As a result of this study, extended models are proved to be effective when large, time varying disturbances affect the system, whereas conventional (stationary) noise models are more effective elsewhere.     

Design of fault diagnosis filters and fault-tolerant control for aclass of nonlinear systems

This paper presents a set of algorithms for fault diagnosis and fault tolerant control strategy for affine nonlinear systems subjected to an unknown time-varying fault vector. First, the design of fault diagnosis filter is performed using nonlinear observer techniques, where the system is decoupled through a nonlinear transformation and an observer is used to generate the required residual signal. By introducing an extra input to the observer, a direct estimation of the time-varying fault is obtained when the residual is controlled, by this extra input, to zero. The stability analysis of this observer is proved and some relevant sufficient conditions are obtained. Using the estimated fault vector, a fault tolerant controller is established which guarantees the stability of the closed loop system. The proposed algorithm is applied to a combined pH and consistency control system of a pilot paper machine, where simulations are performed to show the effectiveness of the proposed approach     

非线性HAMMERSTEIN系统的预测控制及其pH过程应用

本文基于非线性离散Hammerstein模型,开发了一种非线性Hammerstein系统预测控制(Non-Linear Hammerstein Predic- tive Control,NLHPC)算法。遵循预测控制策略,该算法利用Hammerstein模型进行输出预测。理论分析结果表明,该算法不仅具有好的稳定性和鲁棒性,而且其自身具有积分作用。在一台工业PC机上实现了该NLHPC算法,并用于具有强非线性的酸碱中和过程实验装置pH值的控制。实验结果表明NLHPC有着比工业界常用的非线性PID控制(nonlinear PID,NL-PID)更好的控制性能。     

基于SαS分布参数估计的系统鲁棒故障检测

因为复杂系统难以建立精确的数学模型,基于模型的故障检测方法在实际复杂控制系统中应用时往往难以获得很好的效果。针对这类数学模型未知的非线性系统,提出了一种基于SαS分布参数估计的系统故障检测方法。首先应用预测方法对系统输出序列进行预测建模,利用预测误差放大信号的脉冲突变,然后利用SαS分布的参数估计方法对预测误差序列的参数α进行估计,获得α的变化曲线,根据α的变化可以直观地判断出故障的发生。该方法对大幅值的有色噪声污染的信号仍然有很好的检测鲁棒性。以轴承系统的故障检测为例进行仿真实验,通过分析轴承振动信号故障条件下α曲线的变化情况,判断轴承的故障状态。仿真结果证实了该方法有效且可行。     

Sensor and actuator fault diagnosis of systems with discrete inputs and outputs

The paper describes a method for detecting and identifying faults that occur in the sensors or in the actuators of dynamical systems with discrete-valued inputs and outputs. The model used in the diagnosis is a stochastic automaton. The generalized observer scheme (GOS), which has been proposed for systems with continuous-variable inputs and outputs some years ago, are developed for discrete systems. This scheme solves the diagnostic problem as an observation problem, which is set up here for discrete-event systems. As the system under consideration is described by a stochastic automaton rather than a differential equation, the mathematical background and the diagnostic algorithms obtained are completely different from the well-known observers developed for continuous-variable systems. The GOS is extended here by a fault detection module to cope with plant faults that are different from actuator or sensor faults. The diagnostic algorithm consists of two steps, the first detecting the existence of a fault and the second isolating possible sensor or actuator faults or identifying plant faults. The results are applied to quantized systems whose discrete inputs and outputs result from a quantization of the continuous-variable input and output signals. Experimental results illustrate the proposed diagnostic method.     

Closed-loop model validation based on the two-model divergence method

This paper is concerned with model validation and detection of parameter changes under closed-loop conditions. Two closed-loop model validation algorithms are developed based on the two-model divergence method. The algorithms are only sensitive to the plant changes that affect closed-loop performance. The first algorithm is sensitive to changes in both plant and disturbance dynamics. A novel feature of the second algorithm is that it is only sensitive to the changes in plant dynamics, regardless of changes in disturbance dynamics. Under certain conditions, the proposed algorithms can also be applied for fault detection, e.g. detection of actuator or sensor fault, etc. The developed algorithms are evaluated by simulations as well as experimental applications on a pilot scale process.     

Design of a fault diagnosis scheme for a class of singular nonlinear systems

A new fault detection and diagnosis approach is developed in this paper for a class of singular nonlinear systems via the use of adaptive updating rules. Both detection and diagnostic observers are established, where Lyapunov stability theory is used to obtain the required adaptive tuning rules for the estimation of the process faults. This has led to stable observation error systems for both fault detection and diagnosis. A simulated numerical example is included to demonstrate the use of the proposed approach and encouraging results have been obtained.     

Design of a fault diagnosis scheme for a class of singular nonlinear systems

A new fault detection and diagnosis approach is developed in this paper for a class of singular nonlinear systems via the use of adaptive updating rules. Both detection and diagnostic observers are established, where Lyapunov stability theory is used to obtain the required adaptive tuning rules for the estimation of the process faults. This has led to stable observation error systems for both fault detection and diagnosis. A simulated numerical example is included to demonstrate the use of the proposed approach and encouraging results have been obtained.     

Fault isolation for nonlinear dynamic systems based on parameter intervals

The study of fault detection and isolation for nonlinear dynamic systems has been receiving significant attention. Up to now few literatures pay attention to the speed of fault isolation. However, it is a crucial problem for the design of the fault-tolerant control (FTC) of the nonlinear dynamic systems. In this article a new method of fault isolation for nonlinear dynamic systems is proposed. The method is based on the monotonous characteristic of the prediction error of the observer with respect to singular parameter difference between the system and the observer. The proposed method has the advantage of the methods based on adaptive observers that fits a large kind of nonlinear dynamic systems, while it does not have their disadvantage that take a long time to identify the system parameter: Therefore the fault isolation of this method is quicker. The performance of the method is illustrated by simulation results using a nonlinear dynamic model of an alcoholic fermentation process.     

Hybrid causal analysis combining a nonparametric multiplicative regression causality estimator with process connectivity information

Industrial processes are often subjected to abnormal events such as faults or external disturbances which can easily propagate via the process units. Establishing causal dependencies among process measurements has a key role in fault diagnosis due to its ability to identify the root cause of a fault and its propagation path. This paper proposes a hybrid nonlinear causal analysis based on nonparametric multiplicative regression (NPMR) for identifying the propagation of an oscillatory disturbance via control loops. The NPMR causality estimator addresses most of the limitations of the linear model-based methods and it can be applied to both bivariate and multivariate estimations without any modifications to the method parameters. Moreover, the NPMR-based estimations can be used to pinpoint the root cause of a fault. The process connectivity information is automatically integrated into the causal analysis using a specialized search algorithm. Thereby, it enables to efficiently tackle industrial systems with a high level of connectivity and enhance the quality of the results. The proposed approach is successfully demonstrated on an industrial board machine exhibiting oscillations in its drying section due to valve stiction and. The NPMR-based estimator produced highly accurate results with relatively low computational effort compared with the linear Granger causality and other nonlinear causality estimators.     

基于神经网络的非线性系统故障检测及容错控制方法

利用神经网络的非线性建模能力，提出了一种非线性系统的故障检测及容错控制方法。在本方法中，首先应用神经网络设计故障估计器，在线估计系统故障向量，实现故障检测；在此基础上，引入补偿控制器，消除故障对系统运行的影响，从而实现容错控制。同时基于Ｌｙａｐｕｎｏｖ方法进行了稳定性分析。     

基于观测器的多智能体系统一致性控制与故障检测

针对一类受到未知干扰的非线性多智能体系统,提出了一种鲁棒一致性控制与故障检测算法.首先,针对每个智能体系统设计了一个未知输入非线性观测器.然后,基于观测器的状态估计信息,设计了鲁棒一致性控制协议.控制协议保证了给定的干扰抑制性能指标.接着,考虑智能体出现故障的情形,采用自适应阈值法,提出了一种分布式故障检测算法.最后,以多个直流电机驱动的单摆系统为例进行了仿真实验,仿真结果表明了一致性控制与故障检测算法的有效性.     

Design of an auto-associative neural network by using design of experiments approach

Advanced monitoring systems enable integration of data-driven algorithms for various tasks, for e.g., control, decision support, fault detection and isolation (FDI), etc. Due to improvement of monitoring systems, statistical or other computational methods can be implemented to real industrial systems. Algorithms which rely on process history data sets are promising for real-time operation especially for online process monitoring tasks, e.g., FDI. However, a reliable FDI system should be robust to uncertainties and small process deviations, thus, false alarms can be avoided. To achieve this, a good model for comparison between process and model is needed and for easier FDI implementation, the model has to be derived directly from process history data. In such cases, model-based FDI approaches are not very practical. In this paper a nonlinear statistical multivariate method (nonlinear principal component analysis) was used for modeling, and realized with auto-associative artificial neural network (AANN). A Taguchi design of experiments (DoE) technique was used and compared with a classic approach, where according to the analysis best AANN model structure was chosen for nonlinear model. Parameters that are important for neural network’s performance have been included into a joint orthogonal array to consider interactions between noise and control process variables. Results are compared to AANN design recommendations by other authors, where obtained nonlinear model was designed for reliable fault detection of very small faults under closed-loop conditions. By using Taguchi DoE robust design on AANN, an improved and reliable FDI scheme was achieved even in case of small faults introduced to the system. The accuracy and performance of AANN and FDI scheme were tested by experiments carried out on a real laboratory hydraulic system, to validate the proposed design for industrial cases.     

Subspace identification for FDI in systems with non-uniformly sampled multirate data

This paper proposes a novel subspace approach towards direct identification of a residual model for fault detection and isolation (FDI) in a system with non-uniformly sampled multirate (NUSM) data without any knowledge of the system. From the identified residual model, an optimal primary residual vector (PRV) is generated for fault detection. Furthermore, by transforming the PRV into a set of structured residual vectors, fault isolation is performed. The proposed algorithms have been applied to an experimental pilot plant with NUSM data for sensor FDI, where different types of faults are successfully detected and isolated, fully validating the practicality and utility of the developed theory.     

A neuro-fuzzy online fault detection and diagnosis algorithm for nonlinear and dynamic systems

This paper presents a new fault detection and diagnosis approach for nonlinear dynamic plant systems with a neuro-fuzzy based approach to prevent developing of fault as soon as possible. By comparison of plants and neuro-fuzzy estimator outputs in the presence of noise, residual signal is generated and compared with a predefined threshold, the fault can be detected. To diagnose the type, size, time and fault conditions, are used analytical approach and neural network for tracking fault developing online. The neuro-fuzzy nets are compared with some other identification methods in application of power plant gas turbine. Faults are considered in two forms, step, and ramp shape. This work was implemented with real data from gas turbine of Kazeroun (Iran) power plant (Mitsubishi unit) and result is presented to demonstrate the benefits of the proposed method.     

Fault detection and diagnosis approach based on nonlinear parity equations and its application to leakages and blockages in the drying section of a board machine

This study aims at providing a fault detection and diagnosis (FDD) approach based on nonlinear parity equations identified from process data. Process knowledge is used to reduce the process nonlinearity from high to low-dimensional nonlinear functions representing common process devices, such as valves, and incorporating the monotonousness properties of the dependencies between the variables. The fault detection approach considers the obtained process model to be nonlinear parity equations, and fault diagnosis is carried out with the standard structured residual method. The applicability of the approach to complex flow networks controlled by valves is tested on the drying section of an industrial board machine, in which the key problems are leakages and blockages of valves and pipes in the steam–water network. Nonlinear model equations based on the mass balance of different parts of the network are identified and validated. Finally, fault detection and diagnosis algorithms are successfully implemented, tested, and reported.