A Quality-Based Nonlinear Fault Diagnosis Framework Focusing on Industrial Multimode Batch Processes

This work presents a framework for fault detection and diagnosis of multimode industrial processes. It is shown that the proposed method to generate features for the faults provides a generalization that can allow the fault diagnosis in modes for which no training was performed. The proposed method is applied to a sophisticated simulator of a continuous stirred tank reactor (CSTR). Experimental results suggest the usefulness of this approach for the detection of faults under operating conditions not considered during training phase.

Fault detection and diagnosis for non-linear processes empowered by dynamic neural networks

An effective process monitoring system serves as an early warning system for influences affecting the chemical plant and helps plant operator to devise remedial actions to mitigate the adverse effects. However, the design of such system presents challenges such as complex cause-effect correlations, imprecise process model and novelty identifiability. In this work, a two-step fault detection and diagnosis framework is presented. This framework utilizes boundary models developed from mass and energy balances for each section of the chemical plant. The fault detection step consists of a fuzzy inference system (FIS) to analyze the balances and identify the faulty section if the balances deviate from the normal boundary. Then, multiple adaptive neuro-fuzzy inference system (ANFIS) classifiers are constructed to diagnose the exact root causes of bad performance. The combination of boundary models and FIS provides fault isolation of the faulty plant section even when novel faults have occurred. Utilization of multiple ANFIS classifiers reduces the complexity of the networks and improves the proficiency of the process monitoring system. The proposed scheme is applied on a model of a large scale industrial process.

Achieving asymptotic and concurrent fault diagnosis in hydraulic system remains a challenging endeavor due to the inherent attributes of the hidden occurrence, simultaneous manifestation, coupling, and limited sample size. To address the above issues, this paper proposes a hierarchical multi-output fault detection and diagnosis framework, namely, HMDF, based on a hierarchical learning strategy to leverage an improved convolutional neural network (CNN) and support vector machine (SVM). Both a multi-channel CNN and a multi-branch CNN are employed to extract and downscale features collected by the sensors at diverse sampling frequencies first, and then, such features are subsequently subjected to classification using SVM. The hierarchical learning strategy enables the identification of different fault states, both at the component and the intra-component level. Additionally, a modified whale optimization algorithm is also utilized to optimize the classification process of SVM. Extensive experiments are conducted to test the proposed HMDF with the hydraulic system datasets. Results show that HMDF achieves a diagnostic accuracy of up to 98.9% for the dataset, surpassing traditional methods reliant on manual extraction of time–frequency features, and it also exhibits superior classification performances with a small sample size. The HMDF is expected to offer a generalized framework for the multi-output fault detection and diagnosis in hydraulic systems and other complex components.

Buildings are subject to faults in their heating, ventilation and air-conditioning (HVAC) systems that can lead to excessive energy wastage, poor indoor climate, equipment failures and high maintenance costs. Field studies have shown that employing fault detection, diagnosis and prognosis (FDDP) tools followed up with equipment services and corrections can help achieve up to 40% of energy savings within the HVAC system and improve indoor climate, increase equipment lifecycle and reduce maintenance costs. The increasing adoption of building automation systems (BAS), Internet of Things (IoT) and other smart technologies in recent years have allowed large amounts of data to be continuously collected from building systems. This data-rich environment, along with the surge in data analytics and machine learning tools, has made cost-effective data-driven FDDP strategies possible. Compared to purely physics-based methods, data-driven methods require less explicit knowledge of the underlying physical system, thus are often easier to develop, and can learn certain intricate relationships that exist among data. Within the reported data-driven FDDP methods, there exists a few research gaps: 1) data imputation methods that leverage mutual information from correlated measurements to defy poor data quality from BAS have not been utilized efficiently; 2) there lacks a systematic and scalable fault diagnosis framework that incorporates probabilistic temporal relationships to track fault evolution; 3) existing fault diagnosis strategies typically focus on traditional rule-based control strategies and their scalability for advanced control strategies such as Guideline 36 have not been explored yet; 4) active threats information such as cyber-attacks, are typically not incorporated in an FDDP framework; 5) fault prognosis strategies to preemptively identify gradual faults for predictive maintenance have rarely been studied. This research attempts to address the above-mentioned research gaps through the following: Data Imputation: reported data imputation methods that are suitable for handling and repairing multi-source BAS data are evaluated. Data collected from a medium-sized, mixed-use institution building situated in Stockholm, Sweden and a small commercial building simulated in a laboratory setup is used to evaluate five different data imputation methods. Results demonstrate that incorporating time-lagged cross-correlations within the k-Nearest Neighbor (kNN) method helps to significantly improve the imputation accuracy and minimize the impact of repaired data on data-driven algorithms. Dynamic Bayesian Network (DBN)-based Framework for Cyber-Physical Fault Diagnosis: a DBN framework with discretized conditional probabilities parameters to represent the temporal relationships among building measurements is developed. Both domain knowledge and machine learning methods are used to develop the DBN structure and parameter model. The developed framework is evaluated for both traditional rule-based and Guideline 36 controls using datasets from a real building, a laboratory building, and a virtual testbed. Results show that the developed DBN framework is effective in diagnosing and isolating faults in systems even with different control strategies. The framework also successfully distinguishes whether system abnormalities originate from cyber-attacks or naturally occurring physical faults. Potential future direction to improve fault isolation using modified DBN topological structure is also reported in this study. DBN-based Framework for Fault Prognosis: an extension of the DBN framework in conjunction with Robust Multivariate Temporal (RMT) variate selection is proposed for fault prognosis. The RMT variate selection is used to extract localized temporal features from high dimensional datasets to determine the best inputs for training forecasting models. The expected fault-free behavior of multiple target variates, selected using domain knowledge, is forecasted using incoming data. The prediction errors generated from the forecasting phase are used as evidences in the DBN inference to estimate future fault probabilities. Gradual faults simulated in the virtual testbed are used to evaluate the prognosis framework. Results show that the developed framework is effective in prognosing gradual faults by leveraging the trending growth on the prediction errors. The research presented in this thesis contributes to the overall objective of developing a robust and cost-effective DBN-based framework for fault diagnosis and prognosis of building HVAC systems. Potential solutions to other existing challenges of implementing data-driven FDDP strategies, such as obtaining high-quality datasets, handling and repairing missing data, establishing a baseline model for detecting abnormalities despite other disturbances such as weather and internal conditions changes, and extracting temporal features from timeseries data are also examined.

This study explores the application of an adaptive Gaussian Mixture Regression (GMR) framework for fault detection and diagnosis (FDD) across three distinct HVAC systems: passive chilled beams (PCB), radiant slabs (RS), and conventional Variable Air Volume (VAV) systems. This investigation addresses the growing need for robust and adaptable fault detection methods capable of maintaining high diagnostic precision across various HVAC configurations, which exhibit unique operational characteristics and fault signatures.The methodology follows three crucial steps: system modeling, feature selection, and FDD. First, the GMR model is used to predict the flow and the heating and cooling capacities of each HVAC system. Accurate prediction of these parameters is crucial, as they are important features during the FDD process. Next, a feature selection algorithm is implemented to identify the optimal features for fault diagnosis for each HVAC system. The selected features are then compared across the three systems to identify similarities and differences. Understanding these similarities is valuable for developing generalized fault detection strategies that can be adapted to multiple HVAC configurations, thereby improving the scalability and versatility of the FDD framework. The performance of the FDD model is evaluated by testing it against various fault scenarios in each HVAC system in the presence of both known and unknown faults. The evolving nature of the GMR model allows it to update its parameters and integrate new Gaussian components when encountering previously unknown states, thus maintaining high diagnostic accuracy over time.The feature selection process highlights both common and unique features across the HVAC systems. Common features include variables related to flow rates and heating/cooling capacity of the system which are critical for diagnosing faults in any HVAC system. Unique features, specific to each system's operational principles, further enhance the model's diagnostic precision. For instance, slab surface temperature is critical for radiant slabs, while supply air temperature deviation from the setpoint is more significant for the VAV system. In terms of fault prediction accuracy, the GMR model shows remarkable effectiveness across all three HVAC systems. The fault prediction accuracy was above 90% for all three systems, with the highest performance observed in the radiant slabs, followed by the passive chilled beams and the VAV system. The evolving capability of the GMR model plays a significant role in maintaining this high accuracy, as it allows the model to adapt to new fault patterns and operational changes over time.

Two neural network models are compared using a process fault detection and diagnosis problem. Process fault detection and diagnosis using steady-state information is a nonlinear pattern recognition problem. In such problems the measurement pattern vectors are noisy and collection of a complete training data set is time consuming and in some cases impossible. The training data were obtained from a simulated chemical process. The process consists of a reactor and a distillation column. The dynamically capacity allocating network models were found to have the best performance as compared to backpropagation-type network models. These networks can also be used for continuously learning systems and therefore the difficulties of training data collection are avoided. >

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.

A wavelet neural network assisted framework for active fault detection and diagnosis of process systems

A COMPARISON OF IDENTIFICATION ALGORITHMS FOR RAPIDLY TIME-VARYING PARAMETERS ON A REAL PROCESS

Online fault detection and diagnosis (FDD) enables Unmanned Aerial Vehicles (UAVs) to take informed decisions upon actuator failure during flight, adapting their control strategy or deploying emergency systems. Despite the camera being a ubiquitous sensor on-board of most commercial UAVs, it has not been used within FDD systems before, mainly due to the nonexistence of UAV multi-sensor datasets that include actuator failure scenarios. This paper presents a knowledge-based FDD framework based on a lightweight LSTM network and a single layer neural network classifier that fuses camera and Inertial Measurement Unit (IMU) information. Camera data are pre-processed by first computing its optical flow with RAFT-S, a state-of-the-art deep learning model, and then extracting features with the backbone of MobileNetV3-S. Short-Time Fourier Transform is applied on the IMU data for obtaining their time-frequency information. For training and assessing the proposed framework, UUFOSim was developed: an Unreal Engine-based simulator built on AirSim that allows the collection of high-fidelity photo-realistic camera and sensor information, and the injection of actuator failures during flight. Data were collected in simulation for the Bebop 2 UAV with 16 failure cases. Results demonstrate the added value of the camera and the complementary nature of both sensors with failure detection and diagnosis accuracies of 99.98% and 98.86%, respectively.

Dynamic model-based fault diagnosis for (bio)chemical batch processes

A batch process is finite in duration and can be separated into two stages: startup and production. We develop a methodology to monitor a batch process during the startup stage to reduce the length of the startup stage. We focus on processes that are characterized by multiple process parameters and product characteristics. Because of the complex interdependencies characterizing the process parameters and product characteristics, it is more effective to evaluate them simultaneously. To address the multivariate nature of the process we use a multivariate statistical model: PLS (Projection to Latent Structures). PLS has been applied to several applications in statistical process monitoring. We present a new application of PLS to the startup stage of a batch process. Iterative adjustments made during startup in search of an acceptable production zone consume considerable amounts of material, labor and equipment time. We develop a monitoring procedure to reduce the time as well as the number of iterations and adjustments needed for startup. A PLS model is constructed, using baseline data, to characterize the relationship among process parameters during good production. The startup stage is monitored using the PLS characterization to determine if the process is consistent with good production. We illustrate the improved startup operations with an example from a batch process in filament extrusion, the application that motivates this work. Copyright © 2001 John Wiley & Sons, Ltd.

Although cyclical operation systems are relatively widespread in practice (notably in the realm of physical separations, for example, pressure‐swing adsorption and chromatography), the development of specific fault detection mechanisms has received little attention compared to the extensive efforts dedicated to continuous or batch processes. Here, a novel geometric approach for process fault detection is proposed. Specifically, a time‐explicit multivariable representation of data collected from the process, which provides a natural framework for defining “normal” operation and the corresponding confidence regions is developed. On this basis, a two‐step fault detection approach is proposed, based on detecting intercycle variations to locate a faulty cycle, and intracycle changes to determine the exact timing of a fault. The theoretical developments are illustrated with two simulation case studies. © 2017 American Institute of Chemical Engineers AIChE J, 63: 2719–2730, 2017

The fault detection and diagnosis of continuous process is very important for the production safety and product quality. Owing to its no need to know much about the process mechanism and exact process model ,the data driven method, typically the principal component analysis (PCA) has attracted much attention by chemical researchers for monitoring process. PCA is powerful in fault detection ,however, it has difficulties in diagnosing fault correctly in complex process. In this paper, the multiblock principal component analysis (MBPCA) is applied for fault detection and diagnosis in continuous process, which uses the integral PCA to detect fault and block contribution and variables contribution to diagnose fault. The simulations on the Tennessee Eastman process show that the proposed method can not only detect fault quickly ,but also find the fault location exactly.