Search

Since the classification methods mentioned in previous studies are currently unable to meet the accuracy requirements for fault diagnosis in large-scale chemical industries, these methods are gradually being eliminated and rarely used. This research offers a probabilistic neural network (PNN) based on feature selection and a bio-heuristic optimizer as a fault diagnostic approach for chemical industries using artificial intelligence. The sample characteristics are initially simplified using heuristic feature selection and support vector machine recursive feature elimination (SVM-RFE). Using PNN as the principal classifier of the fault diagnostic model and employing a modified salp swarm algorithm (MSSA) linked with the bio-heuristic optimizer to optimize the hidden smoothing factor (σ) of PNN further improves the classification performance of PNN. The MSSA introduces the Lévy flight method, greatly enhancing exploration capabilities and convergence speed compared to the standard SSA. To validate the engineering application of the suggested method, a PSO-SVM-REF-MSSA-PNN model is created, and TE process data are utilized in tests. The model's performance is evaluated by comparing its accuracy and F1-score to other regularly used classification models. The results indicate that the data samples selected by PSO-SVM-RFE features simplify and eliminate redundant features more effectively than other feature selection techniques. The MSSA algorithm's optimization capabilities surpass those of conventional optimization techniques. The PNN network is more suitable for fault detection and classification in the chemical industry. The three considerations listed above make it evident that the proposed approach might greatly help identify TE process problems. [ABSTRACT FROM AUTHOR]

Chemical processes usually exhibit complex, high-dimensional and non-Gaussian characteristics, and the diagnosis of faults in chemical processes is particularly important. To address this problem, this paper proposes a novel fault diagnosis method based on the Bernoulli shift coyote optimization algorithm (BCOA) to optimize the kernel extreme learning machine classifier (KELM). Firstly, the random forest treebagger (RFtb) is used to select the features, and the data set is optimized. Secondly, a new optimization algorithm BCOA is proposed to automatically adjust the network hyperparameters of KELM and improve the classifier performance. Finally, the optimized feature sequence is input into the proposed classifier to obtain the final diagnosis results. The Tennessee Eastman (TE) chemical process have been collected and used to verify the effectiveness of the proposed method. A comprehensive comparison and analysis with widely used algorithms is also performed. The results demonstrate that the proposed method outperforms other methods in terms of classification accuracy. The average diagnosis rate of 21 faults is found to be 89.32%. [ABSTRACT FROM AUTHOR]

The operating environments of ship pipeline valves are complicated. A single sensor may struggle to comprehensively characterize the valve fault features. Pressure variations in the pipeline network may result in changes of the feature distribution of the valve leakage acoustic emission (AE) signals, which makes it challenging for traditional data-driven detection models to achieve satisfactory generalization performance. Unfortunately, there has been limited attention in the current literature on multi-sensor cross-domain fault diagnosis for the ship pipeline valves. To address this research gap, this study aims to develop a novel multi-sensor cross-domain fault diagnosis method. Firstly, the time-frequency-multisqueezing transform (TFMST) is employed for time-frequency analysis of the valve leakage AE signals. By accurately describing harmonic and pulse components in the AE signals, the time-frequency domain features of the valve leakage can be extracted. Subsequently, a multi-sensor deep transfer learning model based on multi-channel fusion convolutional neural network (MCFCNN) is proposed, which utilizes the domain adaptation strategy to minimize the maximum mean discrepancy between source domain data and target domain data, significantly improving the robustness of the fault detection to variations of operating environments. Experimental testing has been carried out to evaluate the performance of the proposed method, and the results indicate effective valve leakage detection. • A multi-sensor deep transfer learning fault diagnosis framework is proposed for leakage monitoring. • TFMST is utilized to generate a concentrated TFR of leakage sound emission signals from valve leaks. • A multi-sensor information fusion cross-domain fault diagnosis model is established. [ABSTRACT FROM AUTHOR]

The health condition monitoring of rotating machinery can avoid the disastrous failure and guarantee the safe operation. The vibration-based fault diagnosis shows the most attractive character for fault diagnosis of rotating machinery (FDRM). Recently, Lempel-Ziv complexity (LZC) has been investigated as an effective tool for FDRM. However, the LZC only performs single-scale analysis, which is not suitable to extract the fault features embedded in vibrational signal over multiple scales. In this paper, a novel complexity analysis algorithm, called hierarchical Lempel-Ziv complexity (HLZC), was developed to extract the fault characteristics of rotating machinery. The proposed HLZC method considers the fault information hidden in both low-frequency and high-frequency components, resulting in a more accurate fault feature extraction. The superiority of the proposed HLZC method in detecting the periodical impulses was validated by using simulated signals. Meanwhile, two experimental signals were utilized to prove the effectiveness of the proposed HLZC method in extracting fault information. Results show that the proposed HLZC method had the best diagnosing performance compared with LZC and multi-scale Lempel-Ziv complexity methods. [ABSTRACT FROM AUTHOR]

This paper presents a hybrid machine‐learning method based on oil‐immersed power transformer fault diagnosis Probability Neural Network (PNN) optimized via a Multi‐Verse Optimizer (MVO) algorithm. PNN is a radial basis function prefeedback neural network based on Bayesian decision theory. It has strong fault tolerance and has significant advantages in pattern classification. However, the performance of PNN is greatly affected by the hidden‐layer unit‐smoothing factor, and the classification result is affected. MVO is a metaheuristic algorithm with strong global convergence. Therefore, the smoothing factor of MVO‐optimized PNN (MVO‐PNN) can effectively improve the fault diagnosis ability. Recent studies have demonstrated the MVO algorithm. We utilize an experiment about the oil data in the power transformer in Jiangxi Province, China. The results show that MVO‐PNN can significantly improve the accuracy of power transformer fault classification and is more efficient than the Cuckoo search algorithm, Bat algorithm, Genetic Algorithm optimization, and other algorithms capabilities in some cases. © 2020 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc. [ABSTRACT FROM AUTHOR]

In the chemical industry, fault diagnosis is a challenging task due to the complexity of chemical equipment. This paper proposes a machine learning‐based approach to achieve the goal of fault diagnosis. First, in order to reduce the impact of redundant features, support vector machine recursive feature elimination (SVMRFE) is used to select important features. The trained probabilistic neural network (PNN) is then used for fault diagnosis. Considering that the diagnostic performance is affected by its hidden layer element smoothing factor (σ), the modified bat algorithm (MBA) is used to optimize the PNN to obtain optimal global parameter values. The MBA adopts a better optimization mechanism than the basic algorithm and achieves excellent global convergence. It can globally optimize the smoothing factor, which effectively improves the fault diagnosis ability of the PNN. During the testing of the Tennessee Eastman (TE) process data set, we evaluate the performance of the proposed model by comparing the F1‐score and accuracy of the different methods. The charts provided describe the fault diagnostic results and classification for the different models. The results indicate that the MBA has a better optimization ability than other traditional optimization algorithms. At the same time, the combination method proposed in this paper is also superior to others and can significantly improve the accuracy of TE process fault diagnosis. [ABSTRACT FROM AUTHOR]

For fault detection and diagnosis in large‐scale industrial systems, traditional methods have a low classification accuracy, which is an issue. This paper proposes a fault diagnosis method based on the modified cuckoo search algorithm (MCS) to optimize the probabilistic neural network (PNN). The random forest treebagger (RFtb) is used to reduce the data feature and the PNN is trained for fault diagnosis and classification. However, in order to address the problem that the parameters of PNN easily fall into the local optimal value, the MCS algorithm is introduced to globally optimize the hidden layer element smoothing factor (σ) in the PNN. The MCS algorithm uses a parameters update and a better optimization mechanism to achieve excellent global convergence and to effectively improve the fault diagnosis capability of the model. During the testing process using the Tennessee Eastman (TE) process dataset, the performance of the proposed model is assessed by comparing the accuracy and the F1‐score of different methods. Graphs are presented that depict fault classification and diagnostic results for the different models. The results show that the MCS algorithm has a better optimization ability than the traditional optimization algorithm and the proposed combination method can significantly improve the accuracy of the TE process fault diagnosis. [ABSTRACT FROM AUTHOR]

This article presents an innovative approach for fault diagnosis based on an optimized generalized regression neural network (GRNN) by integrating with dissolved gas analysis, cuckoo search algorithm (CSA), and rough set theory (RS). In the proposed method, the high dimensioned data will be simplified and reduced by RS to generate better features or attributes for the GRNN input. Meanwhile, to enhance the network performance, the smoothing factor of GRNN is optimized by CSA with Levy flight, which leads to a good global convergence. As a consequence, CSA can provide a good solution to effectively improve the fault diagnosis performance. To validate and demonstrate the proposed method, we applied it to a real‐world fault diagnosis application, power transformer fault diagnosis, by comparing the results with those of other methods. From the experimental results obtained from the evaluation, it is obvious that the proposed fault diagnosis method enabled with RS‐CSA‐GRNN can provide a useful solution for power transformer fault diagnosis because it outperformed other GRNN‐based methods that deployed different optimizing algorithms such as the particle swarm optimization and genetic algorithm. © 2019 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc. [ABSTRACT FROM AUTHOR]

Abstract This paper presents a machine learning-based approach to power transformer fault diagnosis based on dissolved gas analysis (DGA), a bat algorithm (BA), optimizing the probabilistic neural network (PNN). PNN is a radial basis function feedforward neural network based on Bayesian decision theory, which has a strong fault tolerance and significant advantages in pattern classification. However, one challenge still remains: the performance of PNN is greatly affected by its hidden layer element smooth factor which impacts the classification performance. The proposed approach addresses this challenge by deploying the BA algorithm, a kind of bio-inspired algorithm to optimize PNN. Using the real data collected from a transformer system, we conducted the experiments for validating the performance of the developed method. The experimental results demonstrated that BA is an effective algorithm for optimizing PNN smooth factor and BA-PNN can improve the fault diagnosis performance; in turn, and the machine learning-based model (BA-PNN) can significantly enhance the accuracies of power transformer fault diagnosis. [ABSTRACT FROM AUTHOR]

Dissolved gas analysis (DGA) based in insulating oil has become a more mature method in the field of transformer fault diagnosis. However, due to the complexity and diversity of fault types, the traditional modeling method based on oil sample analysis is struggling to meet the industrial demand for diagnostic accuracy. In order to solve this problem, this paper proposes a probabilistic neural network (PNN)-based fault diagnosis model for power transformers and optimizes the smoothing factor of the pattern layer of PNN by the improved gray wolf optimizer (IGWO) to improve the classification accuracy and robustness of PNN. The standard GWO easily falls into the local optimum because the update mechanism is too single. The update strategy proposed in this paper enhances the convergence ability and exploration ability of the algorithm, which greatly alleviates the dilemma that GWO is prone to fall into local optimum when dealing with complex data. In this paper, a reliability analysis of thirteen diagnostic methods is conducted using 555 transformer fault samples collected from Jiangxi Power Supply Company, China. The results show that the diagnostic accuracy of the IGWO-PNN model reaches 99.71%, which is much higher than that of the traditional IEC (International Electrotechnical Commission) three-ratio method. Compared with other neural network models, IGWO-PNN also has higher diagnostic accuracy and stability, and is more applicable to the field of transformer fault diagnosis. [ABSTRACT FROM AUTHOR]

Since it is difficult for the traditional fault diagnosis method based on dissolved gas analysis (DGA) to meet today's engineering needs in terms of diagnostic accuracy and stability, this paper proposes an artificial intelligence fault diagnosis method based on a probabilistic neural network (PNN) and bio-inspired optimizer. The PNN is used as the basic classifier of the fault diagnosis model, and the bio-inspired optimizer, improved salp swarm algorithm (ISSA), is used to optimize the hidden layer smoothing factor of PNN, which stably improves the classification performance of PNN. Compared with the traditional SSA, the sine cosine algorithm (SCA) and disruption operator are introduced in ISSA, which effectively improves the exploration capability and convergence speed. To verify the engineering applicability of the proposed method, the ISSA-PNN model was developed and tested using sensor data provided by Jiangxi Province Power Supply Company. In addition, the method is compared with machine learning methods such as support vector machine (SVM), back propagation neural network (BPNN), multi-layer perceptron (MLP), and traditional fault diagnosis methods such as the international electrotechnical commission (IEC) ratio method. The results show that the proposed method has a strong learning ability for complex fault data and has advantages in accuracy and robustness compared to other methods. [ABSTRACT FROM AUTHOR]

The burgeoning prognostic and health management (PHM) engineering technology with superior performance has lately received extensive attention in the academic circle. Nevertheless, the various types of faults of the power transformer often lead to less accurate predictions and the instability of the power system. To address these problems, a power transformer PHM model with a hybrid machine learning method-approach is proposed in this paper. The model uses intelligent sensors to obtain dissolved gas analysis (DGA) data for fault diagnosis of the power transformer system, so as to compress the complexity of features (gas types) in the power transformer. In particular, to enhance the robustness of the model, we adopt a modified differential evolution whale optimization algorithm (MDE-WOA) to optimize the probabilistic neural network (PNN), namely, the classification performance of the model is improved by updating the smoothing factor (σ) of PNN. In addition, compared with other optimization algorithms, the MDE-WOA algorithm has a lower complexity and more stable optimization process. Finally, we evaluate this model with real world data from the power transformer sensor in Jiangxi province, China. The results indicated that the proposed algorithm could achieve the highest diagnostic accuracy in the fourth iteration, its accuracy having reached 98.86%. Therefore, the proposed PNN parameter optimization meta heuristic algorithm could effectively enhance the accuracy and efficiency of the power transformer fault diagnosis. [ABSTRACT FROM AUTHOR]

The running state of a geared transmission system affects the stability and reliability of the whole mechanical system. It will greatly reduce the maintenance cost of a mechanical system to identify the faulty state of the geared transmission system. Based on the measured gear fault vibration signals and the deep learning theory, four fault diagnosis neural network models including fast Fourier transform–deep belief network model, wavelet transform–convolutional neural network model, Hilbert-Huang transform–convolutional neural network model, and comprehensive deep neural network model are developed and trained respectively. The results show that the gear fault diagnosis method based on deep learning theory can effectively identify various gear faults under real test conditions. The comprehensive deep neural network model is the most effective one in gear fault recognition. [ABSTRACT FROM AUTHOR]

Prognostics and Health Management (PHM) is an emerging technique which can improve the availability and efficiency of equipment. A series of related optimization of the PHM system has been achieved due to the growing need for lowering the cost of maintenance. The PHM system highly relies on data collected from its components. Based on the theory of machine learning, this paper proposes a bio-inspired PHM model based on a dissolved gas-in-oil dataset (DGA) to diagnose faults of transformes in power grids. Specifically, this model applies Bat algorithm (BA), a metaheuristic population-based algorithm, to optimize the structure of the Back-propagation neural network (BPNN). Furthermore, this paper proposes a modified Bat algorithm (MBA); here the chaos strategy is utilized to improve the random initialization process of BA in order to avoid falling into local optima. To prove that the proposed PHM model has better fault diagnostic performance than others, fitness and mean squared error (MSE) of Bat-BPNN are set as reference amounts to compare with other power grid PHM approaches including BPNN, Particle swarm optimization (PSO)-BPNN, as well as Genetic algorithm (GA)-BPNN. The experimental results show that the BA-BPNN model has increased the fault diagnosis accuracy from 77.14% to 97.14%, which is higher than other power transformer PHM models. [ABSTRACT FROM AUTHOR]

An emerging prognostic and health management (PHM) technology has recently attracted a great deal of attention from academies, industries, and governments. The need for higher equipment availability and lower maintenance cost is driving the development and integration of prognostic and health management systems. PHM models depend on the smart sensors and data generated from sensors. This paper proposed a machine learning-based methods for developing PHM models from sensor data to perform fault diagnostic for transformer systems in a smart grid. In particular, we apply the Cuckoo Search (CS) algorithm to optimize the Back-propagation (BP) neural network in order to build high performance fault diagnostics models. The models were developed using sensor data called dissolved gas data in oil of the power transformer. We validated the models using real sensor data collected from power transformers in China. The results demonstrate that the developed meta heuristic algorithm for optimizing the parameters of the neural network is effective and useful; and machine learning-based models significantly improved the performance and accuracy of fault diagnosis/detection for power transformer PHM. [ABSTRACT FROM AUTHOR]