Chiller Fault Research Articles

A new chiller fault diagnosis method under the imbalanced data environment via combining an improved generative adversarial network with an enhanced deep extreme learning machine

The existing chiller fault diagnosis approaches often ignore the problem of data imbalance of chiller, which leads to low accuracy in diagnosing minority class fault samples. To conquer this issue, this paper proposes an improved generative adversarial network (IGAN) with an enhanced deep extreme learning machine (EDELM) method. Firstly, to better learn the latent structure of chiller fault data, the multi-head attention (MHA) mechanism is integrated into the traditional generative adversarial network (GAN) method to generate new samples that are more in line with the distribution of minority class fault samples for the purpose of obtaining a rebalanced dataset. Secondly, to fully handle the nonlinear features hidden in the massive chiller data, the deep extreme learning machine (DELM) basic classifier is trained on the rebalanced dataset. To enhance more attention to the misclassified samples, the adaptive boosting (AdaBoost) ensemble strategy is employed to train multiple DELM basic classifiers by updating the sample weights following the classification results through the iterative rounds. The voting weight of the current DELM basic classifier is given according to its fault diagnosis accuracy. Finally, multiple DELM basic classifiers are ensembled according to their voting weights to obtain the final ensemble classifier. The pattern of the snapshot sample is determined through the weighted voting strategy. Detailed experimental results based on the research project 1043 (RP-1043) conducted by the American society of heating, refrigeration, and air conditioning engineers (ASHRAE) confirm the effectiveness of the proposed IGAN-EDELM approach under imbalanced data environments.

Chiller fault diagnosis based on improved variational autoencoder and co-training framework: A case study of insufficient samples

Implementing efficient automatic fault diagnosis is critical for saving energy and minimizing financial losses in the heating ventilation air-conditioning (HVAC) systems of commercial buildings. However, the limited quantity and weak features of fault samples acquired during HVAC operations hinder the effectiveness of conventional machine learning-based fault diagnosis methodologies. This paper proposes a method based on an improved conditional variational autoencoder (MCVAE) and co-training (CT) ideology to address the issue of insufficient training samples. Initially, we employ MCVAE to synthesize an extensive dataset of chiller fault samples from the original training dataset. Subsequently, the beneficial samples for training our fault diagnostic classifier, namely high-quality samples, are selected from the generated dataset using the CT-based framework. Finally, the selected high-quality samples are merged into the original training dataset to train the ultimate fault classifiers. Experimental results demonstrate that our proposed method outperforms in effectiveness and efficiency compared to recently published methods. For instance, in the case of fault level 1 compared to the suboptimal model, our approach exhibits improvements of 2.41% when each type has 5 fault samples.

An interpretable feature selection method integrating ensemble models for chiller fault diagnosis

Timely diagnosis of chiller faults is crucial for indoor comfort and building energy saving with the number of sensors being minimized. To address this challenge and select the top important features for chiller diagnosis (FD), a novel feature selection method (FSM) with interpretability is proposed for chiller FD. The proposed method integrates four ensemble models, namely, random forest, XGBoost, LightGBM, and CatBoost in the feature selection process. A feature sets fusion strategy is designed based on SHAP analysis and a novel normalized feature weight calculation (NFWC) method. An interpretability analysis is performed using SHAP. The proposed NFWC method integrates the weights of four feature subsets based on SHAP values. Subsequently, the method of joint correlation analysis of feature weights is leveraged to eliminate redundant features. To evaluate the effectiveness of the proposed FSM, ten models, including traditional machine learning and deep learning, are used for FD performance validation. The experiment results show that the top ten features selected by the proposed FSM have obvious advantages compared to existing methods, especially in the early stage of FD. This work effectively eliminates 84% of redundant features for chiller FD, significantly improving the efficiency of HVAC system maintenance.

Interpretable chiller fault diagnosis based on physics-guided neural networks

Chillers are critical for building system energy efficiency and occupant comfort. Various methods have been developed for diagnosing chiller faults, with deep learning being a common choice due to its strong end-to-end learning capability. But in the absence of physical knowledge, deep learning models cannot provide a diagnosis outcome in accordance with the mechanisms behind faults, resulting in models lacking interpretability. In this regard, this paper proposes an interpretable chiller fault diagnosis method based on physics-guided neural networks. Through a comprehensive analysis of the chiller’s working process and fault evolution, a key physical indicator related to the condenser fouling fault is derived. Then a corresponding physical inconsistency loss function is developed to construct a condenser fouling fault-oriented diagnostic model. The proposed method has been demonstrated to be effective in reducing physical inconsistency while retaining diagnostic performance. The proposed method is conducive to applying the neural network approach in real operational scenarios and can serve as a reference for future research.

Performance Evaluation of Chiller Fault Detection and Diagnosis Using Only Field-Installed Sensors

Owing to the rapid expansion of data science, data-driven methods have emerged as a dominant trend in chiller fault detection and diagnosis (FDD). Most of these methods prioritize feature selection to achieve optimal diagnostic performance. However, on-site research indicates a common installation of a limited number of sensors, coupled with a necessity to minimize diagnostic costs. This discrepancy between existing research’s feature selection principles and the current on-site sensor installation status presents a significant challenge. To facilitate the practical implementation of data-driven methods in real chiller units, this study addresses a critical question: under the constraint of limited on-site sensor installations, what is the optimal performance achievable by data-driven methods and their improved versions? To answer this, only features derived from commonly installed sensors on field chillers are chosen as indicators for typical chiller faults. The FDD performance of six frequently used data-driven methods, namely, back-propagation neural network, convolutional neural network, support vector machine, support vector data description, Bayesian network, and random forest, along with their improved versions, is comprehensively evaluated and validated using experimental data, considering four evaluation metrics. The conclusions drawn in this paper provide valuable insights for users/manufacturers with limited or no budget, detailing the best achievable diagnostic performance for each typical fault and offering guidance for those aiming to further enhance FDD performance.

A novel semi-supervised fault diagnosis method for chillers based on neighbor-optimized graph convolutional network

The fault diagnosis of chillers is of great significance in reducing building energy consumption and extending the operational lifespan of refrigeration equipment. Several popular machine learning-based fault diagnosis methods rely heavily on many labeled samples. However, such samples are difficult to obtain in practice due to sparse fault data and high labeling costs. This limits the application of ML-based fault diagnosis methods based on supervised learning. To reduce the dependence on labeled samples, this paper proposes a novel chiller fault diagnosis method based on neighbor-optimized graph convolutional network. The method improves the utilization efficiency of unlabeled samples by mining the spatio-temporal relationship between a large number of unlabeled samples and a limited number of labeled samples. And it dynamically adjusts the graph's structure by optimizing the number of adjacent samples in the correlation graph to obtain better diagnostic results. Its effectiveness is validated on the authoritative dataset ASHRAE RP-1043 and a more challenging dataset of real-world chillers in a building. Experimental results show that the proposed method can achieve better diagnostic performance than the state-of-the-art methods.

Improved convolutional neural network chiller early fault diagnosis by gradient-based feature-level model interpretation and feature learning

For chillers, fault diagnosis (FD) is important for maintaining system reliability and performance. Deep learning methods, such as convolutional neural network (CNN), have been widely studied for chiller FD for its more significant diagnosis accuracy. But CNN model with deep layers and complex structures is black-box and difficult to interpret, which would greatly limit its practical FD applications for chillers. Traditional CNN model interpretation method may be not sensitive to interpret the chillers systematic faults especially at their early stages. Hence, to further obtain better interpretation of the CNN FD model, this study proposed a high-sensitivity gradient-based interpretation method. The proposed method adopts a softsign-forward-ReLU-backward manner to interpret the CNN model from the prospective of fault-discriminative feature, which localizes the fault-related feature variables and visualizing the diagnosis criteria for the CNN identified faults. The ASHRAE research project 1043 (RP-1043) chiller fault dataset was used to validate the proposed model interpretation and explanation method with higher feature-level sensitivity for the incipient fault. Based on the feature-level explanation and feature learning results, different feature combinations were investigated to improve diagnosis accuracy of some early-stage faults. If only small sizes of training data were available for modelling, 17 fault-related features were selected from the original 64 features to re-develop the CNN model and achieved diagnosis accuracy improvement of 9% for the early-stage improper refrigerant charge faults at most.

Digital twin model for chiller fault diagnosis based on SSAE and transfer learning

The equipment of chiller systems is characterized by a complex mechanical structure and operating environments that vary widely, resulting in high failure rates, energy waste, and costly maintenance. Traditional fault diagnosis methods suffer from low levels of digitalization and intelligence. Therefore, timely and accurate detection and diagnosis of the operating status of chiller systems are critical. To address the above issues, this study proposes a digital twin (DT) model for chiller system fault diagnosis based on stacked sparse auto-encoder (SSAE) and transfer learning (TL). First, a chiller system digital twin mapping model is constructed utilizing digital twin technology. Then, the SSAE model is constructed to provide real-time defect diagnosis and fault result validation. Finally, to address the limited chiller system data in practical applications, the knowledge of fault diagnosis acquired in the source domain is transferred to the target domain via TL. Experimental results demonstrate that the SSAE model outperforms back propagation (BP), recurrent neural network (RNN) and long short-term memory (LSTM) model in term of accuracy. Furthermore, using TL achieves a diagnostic accuracy of over 90% for different degrees of fault severity, with a significant decrease in cross-entropy loss compared to no TL, confirming the effectiveness of the TL method.

Novel data-pulling-based strategy for chiller fault diagnosis in data-scarce scenarios

Data-driven fault detection and diagnosis (FDD) for chillers depends heavily on the quantity of effective fault data (labelled). Plenty of labelled data is needed for a robust FDD model. Under data-scarce scenario when just a few labelled data are available, the model is easily biased. This study proposes a novel data pulling (DP) strategy based on a semi-supervised method to deal with this problem. Different from the traditional data cleaning idea, DP is concentrated on finding correct data to use instead of trying to identify all the noise to delete. System-level faults and component-level faults are treated separately and differently according to their distinct characteristics. Validations on seven types of chiller faults show that with scarce data, overfitting is easy to occur, random forest (RF) needs not be optimized, and the quality of the supplemented data is normally more important than the quantity. DP strategy can pull out data of high-confidence level, 99.87% for component-level faults and 100.00% for system-level faults. With just 20 labelled data for each fault, compared with the supervised model and the semi-supervised models without DP, the accuracy of the proposed method is improved by 1.77%∼4.63%, and the performance for refrigerant leakage is especially promoted.

Imbalanced data based fault diagnosis of the chiller via integrating a new resampling technique with an improved ensemble extreme learning machine

Fault diagnosis of the chiller is essential to guarantee chiller's safe operation and reduce building energy consumption. However, the existing fault diagnosis methods rarely consider the chiller's imbalanced data conditions, which always leads to low diagnosis accuracy of the minority class samples. To figure out the issue of imbalanced fault pattern data during the chiller fault diagnosis, a hybrid resampling based improved extreme learning machine (HRIELM) is developed in our work. A new hybrid resampling technique (HRT) is first proposed to balance the majority and minority classes in the imbalanced fault pattern datasets. The HRT is then carried out repetitively to obtain multiple diverse rebalanced training datasets using different benchmark datasets. Subsequently, various primary extreme learning machine (ELM) models are established utilizing these rebalanced training datasets. Furthermore, an improved ELM model based on a novel selective ensemble learning strategy is presented to enhance the effectiveness of the chiller fault diagnosis. The basic ELM models with superior performance for diagnosing the neighbor subset constructed according to the fault snapshot sample are first picked out, and a weight factor is further defined to build the final selective ensemble ELM model. Finally, the fault pattern of the snapshot sample is identified according to the weighted voting mechanism. Detailed experimental results on the ASHRAE Research Project RP-1043 experimental datasets certify the effectiveness of the presented HRIELM scheme for the chiller fault diagnosis under the imbalanced data environments.

Study on Chiller Fault Detection and Diagnosis Method Based on KNN Algorithm and ANOVA

Study on Chiller Fault Detection and Diagnosis Method Based on KNN Algorithm and ANOVA

Augmented data driven self-attention deep learning method for imbalanced fault diagnosis of the HVAC chiller

The chiller fault diagnosis is of great significance to maintain the normal operation of the HVAC system and indoor comfort. Due to the difficulty in collecting the chiller’s fault data, we usually face the data imbalance problem of less fault data and more normal data. To tackle this problem and further improve the fault diagnosis accuracy, this paper proposes a novel augmented data driven self-attention based deep learning model for the chiller fault diagnosis. Firstly, to solve the data imbalance problem, a stable synthetic minority oversampling technique (SSMOTE) is presented to generate the artificial fault data, which are then mixed with the raw fault data to achieve data augmentation. Further, to enhance the diagnosis accuracy, the self-attention mechanism based temporal convolutional network (STCN) is developed to classify normal and fault datasets. The developed STCN is realized by stacking the modified blocks which can dynamically pay attention to different data information through combining the self-attention mechanism and the traditional residual block together. What is more, in the STCN, the skip connection structure is added to the self-attention mechanism to solve the gradient disappearance problem. Finally, detailed experiments and comparisons are performed. Experimental results show that, compared with other data augmentation methods, the SSMOTE can complete a large scale expansion of the minority fault datasets, and its augmented data have better effect on handling the data imbalance problem. Moreover, the skip self-attention mechanism based deep learning model can achieve better diagnosis accuracy compared with some popular deep and shallow models, such as the LSTM, ELM and SVM, etc.

Achieving energy savings through artificial-intelligence-assisted fault detection and diagnosis: Case study on refrigeration systems

This study developed an innovative auto-scale transfer learning (ASTL) technology for artificial intelligence (AI) assisted fault detection and diagnosis (FDD). Openly available data on water chiller faults, namely the RP-1043 data set, were used to train the developed ASTL system. The data of water chillers were then transferred to refrigeration systems for FDD. AI-assisted FDD was conducted with an internet of things (IoT) system to complete a case study on 100 refrigeration systems located in 39 different places. The power of these refrigeration systems ranged from 1/4 to 10 hp, and the froze products such as fresh foods, pickled foods, dairy products, and meat products. The practical case studies conducted in this research indicates that energy savings can be achieved by conducting AI-assisted FDD for refrigeration systems. The energy savings for equipment with a power of 1/4–10 hp were 13.43%–26.83%. The application of AI-assisted FDD in refrigeration systems resulted in higher energy conservation than did its application in HVAC systems. In addition, AI-assisted FDD effectively reduced the transportation costs and personnel costs required for equipment maintenance, and the annual cost of the AI platform and the installation cost of the IoT system had the total return of 40.92%.

Unsupervised outlier detection using neural network-based mixtures of probabilistic principal component analyzers for building chiller fault diagnosis

The application of fault detection and diagnosis (FDD) for chillers is very crucial for ensuring the smooth operation, stability, and durability of HVAC systems. Recent FDD models with outstanding diagnostic performance have been proposed for chiller systems. However, the high performance of FDD models is only guaranteed in condition observed data for FDD model construction are not contaminated, i.e., do not contain outliers, which is inevitable during data recording process. To resolve this problem, this paper proposes a novel outlier detection (OD) method named NN-MPPCA. The proposed method uses a neural network (NN) to model a mixture of probabilistic principal component analyzers (MPPCA) framework which parameters are updated via back-propagation. The combining mechanism of MPPCA guarantees NN-MPPCA to fit well complex data distributions, and unseen instances with low reconstruction probability under MPPCA framework are detected as outliers. Comprehensive experimental results of exhaustive comparisons have demonstrated the superiority of NN-MPPCA over other state-of-the-art OD algorithms as well as its effectiveness for chiller FDD application. Under the scenario that datasets contain 20% outliers (i.e., both local outliers and global outliers), the outlier detection performance of NN-MPPCA is highest with F1-scores achieving from 95.4% to 99.7%. Also, with NN-MPPCA applied in the data purification step, FDD models can achieve a diagnostic accuracy up to 99.5%; while they can only reach a diagnostic accuracy of 82.6% at most if contaminated data are directly used.

Across working conditions fault diagnosis for chillers based on IoT intelligent agent with deep learning model

In the real application of chiller fault diagnosis models, the various working conditions may cause the different data distribution between test data and training data, which may lead to the lower diagnosis accuracy of models. In this paper, a novel fault diagnosis method for chillers across working conditions based on the domain knowledge-assisted Deep Extreme Learning Machine (DELM) is proposed. Firstly, the actual normal and performance degradation operation data of the chiller is collected through the Internet of Things Intelligent Agent (IOTIA). The Random Forest (RF) is used to determine the importance of features selected by domain knowledge analysis, and different feature subsets are selected as the model input. Then, the model of the DELM is used for fault diagnosis of chillers across varying temperature and load rate working conditions. Finally, experiments show that the proposed model achieves outstanding effect under the designed seven experimental across working conditions, indicating that the model has good generalization ability and is suitable for fault diagnosis across working conditions of chillers.

Enhanced chiller faults detection and isolation method based on independent component analysis and k-nearest neighbors classifier

The fault detection and isolation (FDI) of centrifugal chiller system is essential for improving energy efficiency and reducing energy consumption. Considering the non-Gaussianity of chiller measurement, this paper proposes an effective non-Gaussian chiller FDI method based on independent component analysis (ICA) and k-nearest neighbor (KNN) classifier. First of all, an enhanced fault detection method is developed by combining exponentially weighted moving average and ICA (EWICA). The exponentially weighted moving average is employed to develop a dynamic threshold scheme, which reduces the false alarm rate (FAR) and improves the fault detection rate (FDR) of chiller faults. Then, based on the KNN with cosine similarity metric, a novel fault isolation method using the direction of the residual vector is presented to isolate the chiller faults. To further improve the isolation performance, the number of neighbors (k) is optimized by 10-fold cross validation. Finally, the proposed FDI method is validated by using a data set from the ASHRAE Research Project 1043 (RP-1043). The FDI performance of the proposed method is comprehensively analyzed and compared with six state-of-the-art methods. Both of detection and isolation results suggest that the EWICA-based method delivers superior performance for chiller FDI.

Chiller faults detection and diagnosis with sensor network and adaptive 1D CNN

Computer-empowered detection of possible faults for Heating, Ventilation and Air-Conditioning (HVAC) subsystems, e.g., chillers, is one of the most important applications in Artificial Intelligence (AI) integrated Internet of Things (IoT). The cyber-physical system greatly enhances the safety and security of the working facilities, reducing time, saving energy and protecting humans’ health. Under the current trends of smart building design and energy management optimization, Automated Fault Detection and Diagnosis (AFDD) of chillers integrated with IoT is highly demanded. Recent studies show that standard machine learning techniques, such as Principal Component Analysis (PCA), Support Vector Machine (SVM) and tree-structure-based algorithms, are useful in capturing various chiller faults with high accuracy rates. With the fast development of deep learning technology, Convolutional Neural Networks (CNNs) have been widely and successfully applied to various fields. However, for chiller AFDD, few existing works are adopting CNN and its extensions in the feature extraction and classification processes. In this study, we propose to perform chiller FDD using a CNN-based approach. The proposed approach has two distinct advantages over existing machine learning-based chiller AFDD methods. First, the CNN-based approach does not require the feature selection/extraction process. Since CNN is reputable with its feature extraction capability, the feature extraction and classification processes are merged, leading to a more neat AFDD framework compared to traditional approaches. Second, the classification accuracy is significantly improved compared to traditional methods using the CNN-based approach.

Knowledge mining for chiller faults based on explanation of data-driven diagnosis

Data-driven model is considered to be an efficient and convenient diagnosis method for refrigeration systems, especially in Big Data Era, but the black box characteristic has always been a criticism and hindered its application. For this problem, SHapley Additive exPlanation (SHAP) method is explored to unveil the black box and the knowledge of system-level chiller faults has been mined. Two ensemble models with excellent diagnostic performance have been investigated including Random Forest (RF) and Light Gradient Boosting Machine (LightGBM). Detailed analysis shows that the sub-cooling at the condenser outlet (TRC_sub) and the condenser approaching temperature (TCA) have a great contribution to the detection of refrigerant leakage and overcharge with different margin, which is consistent with but standing out from the prior knowledge. In practical applications, temperature and pressure sensors for lubricating oil are recommended for a refrigeration system with lubricant due to the sensitive variation with system change (failure or condition altering), which can be a supplement to the traditional experience. The data-driven model is, indeed, an effective way to discover the hidden knowledge of a data set from all dimensions to upgrade the traditional experience. Furthermore, the local explanation by individual samples and the explanation comparison between the RF and the LightGBM model show that the primary features on which the model(s) depends to diagnose the same fault are basically the same, while the auxiliary features may vary slightly. The consensus further proves that the primary features and the related rules mined out in this study can be trusted.

Fault detection and diagnosis using tree-based ensemble learning methods and multivariate control charts for centrifugal chillers

Fault detection and diagnosis (FDD) of centrifugal chillers plays an essential role in reducing energy consumption and ensuring safe operation of heating, ventilation and air conditioning systems. In the existing chiller FDD strategy, Data-driven based system modelling methods and statistical analysis-based detection techniques have been widely discussed and applied in recent years. This paper studies the performance of tree-based ensemble learning methods in chiller modelling and analyses multivariate control charts' detection effect. Predictions from the reference model are used in tandem with multivariate control charts to isolate faulty behavior from normal behavior, and then specific failures are identified through the diagnosis model when anomalies are detected. Firstly, in order to establish the optimal chiller reference model and diagnosis model, three tree-based ensemble learning algorithms, namely (i) random forest (RF), (ii) extreme gradient boosting (XGBoost), and (iii) light gradient boosting machine (LightGBM), are evaluated and compared with commonly used support vector machine (SVM). Secondly, Hotelling's T2, multivariate cumulative sum (MCUSUM) and multivariate exponentially weighted moving average (MEWMA) are introduced and their performance in chiller fault detection is tested. Finally, the proposed LightGBM-MEWMA method is validated by the experimental data of ASHRAE RP-1043. Results show that the proposed method effectively identifies seven common chiller faults. Especially at low severity levels, the average detection rate is 88.71%, and the diagnosis accuracy rate is 82.3%.

End-to-end chiller fault diagnosis using fused attention mechanism and dynamic cross-entropy under imbalanced datasets

Fault diagnosis techniques play an increasingly important role in the operation and maintenance of smart city systems. Artificial intelligence improves the efficiency of chiller system fault diagnosis, and greatly reduces the energy consumption of urban buildings. The existing intelligent fault diagnosis methods of chiller mostly rely on balanced training datasets; lacking fault samples makes these methods incompetent to extract reliable features to recognize abnormal machine conditions, resulting in the degraded performance. To overcome the deficiencies of reported studies, a new method, called end-to-end chiller fault diagnosis, is proposed using a fused attention mechanism and dynamic cross-entropy. Firstly, a one-dimensional convolution network (1D-CNN) and long-short term memory (LSTM) are combined to capture the spatial-temporal features from the original data directly. Afterwards, a fused attention mechanism is developed to further refine the extracted features to increase the contribution of crucial features and achieve high-quality diagnostic information mining. Finally, the dynamic cross-entropy (DCE) is designed for updating the imbalance factor in real-time, with more focus on the hard-classified types. The experimental analysis results demonstrate the feasibility and superiority of the proposed method in identifying chiller system faults with imbalanced datasets.