A novel fault diagnosis method based on CNN and LSTM and its application in fault diagnosis for complex systems

Fault diagnosis plays an important role in actual production activities. As large amounts of data can be collected efficiently and economically, data-driven methods based on deep learning have achieved remarkable results of fault diagnosis of complex systems due to their superiority in feature extraction. However, existing techniques rarely consider time delay of occurrence of faults, which affects the performance of fault diagnosis. In this paper, by synthetically considering feature extraction and time delay of occurrence of faults, we propose a novel fault diagnosis method that consists of two parts, namely, sliding window processing and CNN-LSTM model based on a combination of Convolutional Neural Network (CNN) and Long Short-Term Memory Network (LSTM). Firstly, samples obtained from multivariate time series by the sliding window processing integrates feature information and time delay information. Then, the obtained samples are fed into the proposed CNN-LSTM model including CNN layers and LSTM layers. The CNN layers perform feature learning without relying on prior knowledge. Time delay information is captured with the use of the LSTM layers. The fault diagnosis of the Tennessee Eastman chemical process is addressed, and it is verified that the predictive accuracy and noise sensitivity of fault diagnosis can be greatly improved when the proposed method is applied. Comparisons with five existing fault diagnosis methods show the superiority of the proposed method.

     

A holonic architecture for the global road transportation system

In recent years the development of automated traffic systems has been gaining increasing attention, and a substantial amount of effort has been invested in trying to find a solution to problems associated to road transport. Among these problems are road accidents caused by human-related factors, such as tiredness, loss of control, a slow reaction time, limited field of view, etc. Another transport-related problem is that of loss of time which may be caused by a slow driving speed due to weather conditions, road conditions, visibility, traffic congestion, etc. In this paper, we present a distributed architecture and the underline distributed algorithm for solving the global road transportation system, which is being developed by several European Universities.     

Development of holonic manufacturing execution systems

Rapid changes of market demands and pressures of competition require manufacturers to maintain highly flexible manufacturing systems to cope with a complex manufacturing environment. To meet these requirements, this work adopts the concepts of holon and holarchy to design manufacturing systems. Holon and holarchy are derived from the studies of social organizations and living organisms and possess the properties of intelligence, autonomy, cooperation, reconfigurability, and extensibility. Moreover, advanced manufacturing systems also require the properties of security certification and failure recovery. Based on the requirements of these properties, a systematic approach is proposed to develop a holonic manufacturing execution system (HMES) for the semiconductor industry. This systematic approach starts with a system analysis by collecting domain requirements and analyzing domain knowledge. The HMES Holarchy is designed by the procedure of constructing an abstract object model based on domain knowledge, partitioning application domain into functional holons, identifying generic parts among functional holons, developing the Generic Holon, defining holarchy messages and the holarchy framework of HMES, and finally designing functional holons based on the Generic Holon. It is believed that this proposed systematic approach provides a novel and efficient way to design HMES.     

An integrated architecture for fault diagnosis and failure prognosis of complex engineering systems

Complex engineering systems, such as aircraft, industrial processes, and transportation systems, are experiencing a paradigm shift in the way they are operated and maintained. Instead of traditional scheduled or breakdown maintenance practices, they are maintained on the basis of their current state/condition. Condition-Based Maintenance (CBM) is becoming the preferred practice since it improves significantly the reliability, safety and availability of these critical systems. CBM enabling technologies include sensing and monitoring, information processing, fault diagnosis and failure prognosis algorithms that are capable of detecting accurately and in a timely manner incipient failures and predicting the remaining useful life of failing components. If such technologies are to be implemented on-line and in real-time, it is essential that an integrating system architecture be developed that possesses features of modularity, flexibility and interoperability while exhibiting attributes of computational efficiency for both on-line and off-line applications. This paper presents a .NET framework as the integrating software platform linking all constituent modules of the fault diagnosis and failure prognosis architecture. The inherent characteristics of the .NET framework provide the proposed system with a generic architecture for fault diagnosis and failure prognosis for a variety of applications. Functioning as data processing, feature extraction, fault diagnosis and failure prognosis, the corresponding modules in the system are built as .NET components that are developed separately and independently in any of the .NET languages. With the use of Bayesian estimation theory, a generic particle-filtering-based framework is integrated in the system for fault diagnosis and failure prognosis. The system is tested in two different applications—bearing spalling fault diagnosis and failure prognosis and brushless DC motor turn-to-turn winding fault diagnosis. The results suggest that the system is capable of meeting performance requirements specified by both the developer and the user for a variety of engineering systems.     

线性系统的鲁棒故障诊断

本文讨论了线性系统关于执行器故障的鲁棒诊断问题.应用混沌同步方法,能够将线性系统关于执行器故障的鲁棒诊断问题转化为极点配置问题或线性矩阵不等式的求解问题.该方法能够直接重构故障信息而不是通过产生和评价残差信号得到故障信息,从而避免了产生和评价残差信号的复杂性.     

Collaborative reconfiguration mechanism for holonic manufacturing systems

Although holonic manufacturing systems (HMS) are recognized as a paradigm to cope with the changes in manufacturing environment based on a flexible architecture, development of reconfiguration mechanism is required to realize the advantages of HMS. Finding a solution from scratch to deal with changes in HMS is not an appropriate approach as it may lead to chaos at the shop floor. The objective of this paper is to propose a viable design methodology to achieve effective reconfiguration in HMS based on the cooperation of holons. We formulate and study a holarchy reconfiguration problem and define an impact function to characterize the impact of resource failures on different holons in a holarchy. A collaborative reconfiguration algorithm based on the impact function is proposed to effectively reconfigure the systems to achieve minimal cost solutions.     

Multidisciplinary fault diagnosis of complex engineering systems: A case study of nuclear power plants

Fault diagnosis is a key process in ensuring complex engineering system safety. It often requires collaborative and multidisciplinary efforts. This study seeks to understand the process of multidisciplinary fault diagnosis in complex engineering systems and the key human factors issues that impair this process. Data were collected from multidisciplinary diagnostic activities conducted in the commissioning phase of nuclear power plants (NPPs). In the first phase, we proposed a process model based on a combination of literature review, specialist interviews, and field observations. In phase two, the influencing issues identified in the first phase were assessed through a survey with 117 NPP commissioning specialists. Five factors influencing multidisciplinary fault diagnosis were identified: cognitive artifacts, diagnosis biases, preparation for multidisciplinary diagnosis, information sharing and collaborative reasoning, and collaborative decision-making. The significances of each factor were compared. The results provide guidance for the development of improvement measures to enhance the performance of multidisciplinary fault diagnosis.Practitioner summaryThe processes and influencing issues of multidisciplinary fault diagnosis during the commissioning phase of nuclear power plants were studied with field observations, interviews of 28 specialists, and a survey of 117 specialists. Five major influencing factors were identified, and their influences were compared.     

非线性系统的故障诊断技术

文中对非线性系统的故障诊断方面问题给予了归纳总结,指出了基于数学模型方法,基于信号处理方法和基于知识的方法在实现非线性系统故障诊断的基本思想,并进一步指出了各各非线性系统故障诊断方法及可能的发展方向。     

Collaborative composition of processes in holonic manufacturing systems

Although it provides a flexible architecture to deal with changes and uncertainties, holonic manufacturing systems (HMS) also pose challenges in design and implementation. A challenge is to design a problem solving environment to guide the holons in HMS such that the decisions made by the individual holons as a whole collaboratively compose the production processes for the tasks. The objective of this paper is to propose a problem solving environment for the composition of minimal cost processes in HMS. Given a specific product type and due date, the problem is to dynamically compose a complete process in HMS to process the product by the due date. To achieve the objective, we combine multi-agent systems technology with Petri nets in this paper. We first propose architecture and a two-layer contract net protocol to describe the negotiation between order holons, product holons and resource holons in HMS. To determine whether it is feasible for a set product holons and resource holons to compose a process for an order holon, we propose Petri net models to capture the workflows and activities in product holons and resource holons, respectively. To optimize the cost in the composed manufacturing process, the concept of activity-based costing is adopted to assign the cost of the activities to Petri nets. The advantages of our design methodology include specification of workflows and resource activities with PNML to minimize the cost. To realize the two-layer contract net protocol relies on a mechanism for a holon to discover the services provided by other holons. We proposed a scheme for publication and discovery of holon services based on FIPA compliant multi-agent system platform. Based on the service publication and discovery scheme, interactions between holons with the two-layer contract net protocol can be realized. We develop a HMS system to solve the process composition problem.     

复杂系统故障诊断的新型滤波方法研究

针对复杂系统故障诊断问题,采用一种相关噪声统计特性未知情况下的粒子滤波方法,该方法给出了相关噪声的统计模型,并详细推导了相关噪声统计特性的分布函数。在新方法的框架内采用噪声未知统计特性和系统故障状态的联合递推估计,为复杂系统的故障诊断问题提供了一种新的思路。仿真实验验证了该方法对于复杂系统故障诊断中的有效性和精确性。     

Robust fault diagnosis for non-linear difference-algebraic systems

A new robust fault diagnosis method based on linear matrix inequality (LMI) for non-linear difference-algebraic systems (DAS) with uncertainties is proposed. Based on the known nominal model of DAS, it firstly constructs an auxiliary system consisting of a difference equation and an algebraic equation, then converts the problem of fault identification into the problem of parameter estimation, and finally realizes fault identification using an LMI method. This method can not only detect, isolate and identify faults for DAS, but also give the upper bounds of fault identification error. Simulation indicates that it can give satisfactory diagnostic results for both abrupt and incipient faults.     

Neural-network-based robust fault diagnosis in robotic systems

Fault diagnosis plays an important role in the operation of modern robotic systems. A number of researchers have proposed fault diagnosis architectures for robotic manipulators using the model-based analytical redundancy approach. One of the key issues in the design of such fault diagnosis schemes is the effect of modeling uncertainties on their performance. This paper investigates the problem of fault diagnosis in rigid-link robotic manipulators with modeling uncertainties. A learning architecture with sigmoidal neural networks is used to monitor the robotic system for any off-nominal behavior due to faults. The robustness and stability properties of the fault diagnosis scheme are rigorously established. Simulation examples are presented to illustrate the ability of the neural-network-based robust fault diagnosis scheme to detect and accommodate faults in a two-link robotic manipulator.     

Decentralized fault diagnosis approach without a global model for fault diagnosis of discrete event systems

Diagnosability property ensures that a predefined set of faults are diagnosable by a centralized diagnoser built using a global model of the system, while co-diagnosability guarantees that these faults are diagnosed in decentralized manner using a set of local diagnosers. A fault must be diagnosed by at least one local diagnoser by using its proper local observation of the system. The aim of using decentralized diagnosis approaches is to overcome the space complexity and weak robustness of centralized diagnosis approaches while at the same time preserving the diagnostic capability of a centralized diagnosis. However, co-diagnosability property is stronger than diagnosability property. If a system is co-diagnosable, then it is diagnosable, while a diagnosable system does not ensure that it is co-diagnosable. Therefore, the challenge of decentralized diagnosis approaches is to perform local diagnosis and to verify that it is equivalent to the centralized one without the need for a global model. In this paper, an approach is proposed to obtain co-diagnosable decentralized diagnosis structure of discrete event systems without the use of a global model. This approach is based on the synchronization of local diagnosis decisions in order to solve the ambiguity between local diagnosers. This synchronization allows obtaining local diagnosis equivalent to the global one without the use of a global model.     

Monitoring and fault diagnosis of hybrid systems.

Many networked embedded sensing and control systems can be modeled as hybrid systems with interacting continuous and discrete dynamics. These systems present significant challenges for monitoring and diagnosis. Many existing model-based approaches focus on diagnostic reasoning assuming appropriate fault signatures have been generated. However, an important missing piece is the integration of model-based techniques with the acquisition and processing of sensor signals and the modeling of faults to support diagnostic reasoning. This paper addresses key modeling and computational problems at the interface between model-based diagnosis techniques and signature analysis to enable the efficient detection and isolation of incipient and abrupt faults in hybrid systems. A hybrid automata model that parameterizes abrupt and incipient faults is introduced. Based on this model, an approach for diagnoser design is presented. The paper also develops a novel mode estimation algorithm that uses model-based prediction to focus distributed processing signal algorithms. Finally, the paper describes a diagnostic system architecture that integrates the modeling, prediction, and diagnosis components. The implemented architecture is applied to fault diagnosis of a complex electro-mechanical machine, the Xerox DC265 printer, and the experimental results presented validate the approach. A number of design trade-offs that were made to support implementation of the algorithms for online applications are also described.     

Incipient fault diagnosis of dynamical systems using onlineapproximators

Detection of incipient (slowly developing) faults is crucial in automated maintenance problems where early detection of worn equipment is required. In this paper, a general framework for model-based fault detection and diagnosis of a class of incipient faults is developed. The changes in the system dynamics due to the fault are modeled as nonlinear functions of the state and input variables, while the time profile of the failure is assumed to be exponentially developing. An automated fault diagnosis architecture using nonlinear online approximators with an adaptation scheme is designed and analyzed. A simulation example of a simple nonlinear mass-spring system is used to illustrate the results     

复累量切片在故障诊断中的应用

复数三阶累量具有不同的定义形式,定义形式不同,其耦合性质也不相同,则由其产生的复数累量切片也包含了信号间不同的耦合信息,实验中利用这些不同的定义方式,对特定的故障信号和正常信号同时采用相同的耦合方式进行故障判别。同时定义了复数三阶累量切片谱,通过切片谱对正常信号和故障信号的耦合特征进行了分析。实验结果表明,随着耦合方式的不同,同一故障的诊断正确率也不相同,从而为特定的故障判别提供了一种可供选择的方法。     

Sensor selection for fault diagnosis in uncertain systems

ABSTRACT

Finding the cheapest, or smallest, set of sensors such that a specified level of diagnosis performance is maintained is important to decrease cost while controlling performance. Algorithms have been developed to find sets of sensors that make faults detectable and isolable under ideal circumstances. However, due to model uncertainties and measurement noise, different sets of sensors result in different achievable diagnosability performance in practice. In this paper, the sensor selection problem is formulated to ensure that the set of sensors fulfils required performance specifications when model uncertainties and measurement noise are taken into consideration. However, the algorithms for finding the guaranteed global optimal solution are intractable without exhaustive search. To overcome this problem, a greedy stochastic search algorithm is proposed to solve the sensor selection problem. A case study demonstrates the effectiveness of the greedy stochastic search in finding sets close to the global optimum in short computational time.     

A study on neuro-fuzzy systems for fault diagnosis

Fault diagnosis can be facilitated by using either quantitative or qualitative information of the system monitored. This paper presents a novel approach to integrate quantitative and qualitative information in fault-diagnosis, based on the use of neuro-fuzzy systems. In this approach the diagnostic signals (residuals) are generated and evaluated via a B-Spline functions network. The configuration adopted allows the designer to both extract and include symbolic knowledge from the trained network to provide reliable diagnostic information. The effectiveness of the proposed diagnosis strategy is illustrated through a simulation study of a nonlinear two-tank system.