Search

The timely fault diagnosis of HVAC systems is important for building energy saving, equipment maintenance and indoor comfort. The Gaussian mixture model method has rarely been studied in the fault diagnosis application of HVAC systems. Therefore, a novel fault diagnosis strategy is proposed based on the Gaussian mixture model (GMM) method for the variable refrigerant flow air-conditioning system. To reduce excessive input variables resulting in large model complexity and long running time, the principal component analysis approach (PCA) is used to perform data dimensionality reduction. Therefore, the fault diagnosis model combining the Gaussian mixture model and principal component analysis is established, which is evaluated using the four types of faults of the variable refrigerant flow system. These faults include refrigerant undercharge, refrigerant overcharge, outdoor unit fouling and four-way reversing valve faults. Experiments are carried out under three heating conditions. Results show that the PCA-GMM approach can effectively reduce the running time. Especially for the VVV type model, the running time is reduced from 176.78 s to 15.18 s. Meanwhile, established PCA-GMMs still have good fault diagnosis correct rates when the input data dimension is reduced. Especially, some PCA-GMMs have fault diagnosis correct rates of over 99% when the number of principal components exceeds 7.

Short-term energy consumption prediction has fundamental importance in many HVAC system management tasks, such as demand-side management, short-term maintenance, etc. Currently, the prevailing data-driven techniques, especially supervised machine learning methods, are widely applied for short-term energy consumption prediction. Deep reinforcement learning (DRL), as the state-of-the-art machine learning techniques, have been applied for HVAC system control, but rarely for energy consumption prediction. In this paper, a DRL algorithm, namely Deep Deterministic Policy Gradient (DDPG), is firstly introduced for short-term HVAC system energy consumption prediction. Moreover, Autoencoder (AE), which is powerful in processing data in their raw form, is incorporated into DDPG method to extract the high-level features of state space and optimize the prediction model. The operation data of the ground source heat pump (GSHP) system of an office building in Henan province, China is used to train and assess the proposed models. The results demonstrate that the proposed DDPG based models can achieve better prediction performance than common supervised models like BP Neural Network and Support Vector Machine. This study is an enlightening work which may inspire other researchers to tap the potential of DRL algorithms in this field.

Identifying abnormal energy consumption in the variable refrigerant flow (VRF) system is important for its energy efficiency enhancement and energy saving. The energy performances of the VRF system are extremely distinct under various operation conditions. In order to evaluate the VRF’s energy performance, this study proposes a data-mining-based method to identify the abnormal energy consumption under different operation conditions. The local outlier factor (LOF) algorithm is used to distinguish transient data and three operation conditions are partitioned through clustering analysis. Besides, the correlation analysis is employed for key variables extraction. Then, the energy consumption is forecasted by support vector regression (SVR). The residuals of the measured energy values and the predicted values are applied in support vector data description (SVDD) algorithm to develop the responding D statistic as abnormal energy identification threshold. Finally, experiment data of various refrigerant charge level (RCL) are used to validate the proposed method. The methodology is sensitive to identifying abnormal energy consumption in VRFs at various RCLs. Nearly 100% of the abnormal energy consumption data can be accurately identified at some refrigerant charge levels. Operation condition partitioning can definitely enhance the identification efficiency.

This paper proposed a novel fault diagnosis strategy of the variable refrigerant flow (VRF) system based on expert rules for the first time. The VRF fault diagnosis rules (VFDR) are obtained through the expert knowledge and characteristics of the VRF system. The VFDR includes a total of 22 expert rules, 10 rules for the outdoor unit and other 12 rules for the indoor unit. The proposed VFDR can help maintenance personnel to identify and eliminate VRF faults in time. In addition to obtaining the coefficients of the sensor regression model, the VFDR does not require a training process and is computationally simple, making it easily embedded in the building’s automatic control system to achieve online fault diagnosis. The proposed fault diagnosis strategy is validated with nine faults of the VRF system under cooling mode. These faults contain temperature sensor faults, the system fault and indoor unit faults. The diagnosis correct rate (DCR) is used to evaluate the performance of the VFDR. Through expert rules fault diagnosis strategy, the DCRs of all faults exceed 70% and the overall DCR of all faults is 85.13%. The results show that faults of VRF system are well diagnosed by fault diagnosis strategy based on expert rules.

Energy demand prediction is able to improve the energy efficiency and energy savings of the residential buildings. In this article, the prediction models are developed based on ensemble learning methods which combined extreme gradient boosting, extreme learning machine, multiple linear regression with support vector regression. According to the importance analysis of the random forest method, the optimal set of feature variables is selected. Besides, a historical energy comprehensive variable named EWMA was added in the prediction models to improve the prediction accuracy. A ground source heat pump for residential buildings located in Henan, China, is selected as a model for examining 2-hour ahead heating load forecasting. Results showed that the proposed prediction model based on ensemble learning could reduce the MAE of the testing set prediction result, which ranged from 29.1% to 70%.

The fault diagnosis of air-conditioning systems is of great significance to the energy saving of buildings. This study proposes a novel fault diagnosis approach for building energy saving based on the deep learning method which is deep belief network, and its application potential in the air conditioning fault diagnosis field is investigated. Then, a parameter optimization selection strategy is developed for model optimization. Four kinds of faults of the variable flow refrigerant system under heating mode are used to evaluate the performance of the models. The fault diagnosis results show that the deep belief network model with initial parameters can be used to diagnose the faults of the variable flow refrigerant system. Through the parameter optimization selection strategy, the fault diagnosis correct rate of the optimized model is 97.7%, which is improved by 5.05% compared with the model with initial parameters. The number of hidden layers of the deep belief network model is selected to be 2 layers. This result indicates that the fault diagnosis for variable flow refrigerant systems may not require a very deep model. Additionally, the performance of the optimized deep belief network model is compared with that of the traditional back propagation neural network, and the former is better. This finding also shows that the unsupervised restricted Boltzmann machine layer for data feature reconstruction can improve the fault diagnosis performance.

VRF (Variable Refrigerant Flow) system is an efficient and energy-saving air-conditioning system. The VRF system performance is significantly affected by the quantity of refrigerant (charge). Both insufficient and excessive refrigerant charge will lead a dramatic decline of air conditioning system performance. Accordingly, it is necessary to forecast refrigerant charge in existing systems. This study proposed a charge fault diagnosis method based on expert modification C5.0 decision tree. Initially, C5.0 decision tree is used to diagnosis VRF system charge faults in nine different conditions. Furthermore, in order to reduce error rate of the decision tree, expert experience is used to modify the decision tree. And temperature differential variables are used to improve the diagnosis result. The result s of the C5.0 decision tree using temperature differential variables shows fault diagnosis error rates under expert experience guidance are 10% lower than without expert experience. Besides, this method an increased accuracy in classifies between excessive and normal charge states.

Energy demand prediction of building heating is conducive to optimal control, fault detection and diagnosis and building intelligentization. In this study, energy demand prediction models are developed through machine learning methods, including extreme learning machine, multiple linear regression, support vector regression and backpropagation neural network. Seven different meteorological parameters, operating parameters, time and indoor temperature parameters are used as feature variables of the model. Correlation analysis method is utilized to optimize the feature sets. Moreover, this paper proposes a strategy for obtaining the thermal response time of building, which is used as the time ahead of prediction models. The prediction performances of extreme learning machine models with various hidden layer nodes are analyzed and contrasted. Actual data of building heating using a ground source heat pump system are collected and used to test the performances of the models. Results show that the thermal response time of the building is approximately 40 min. Four feature sets are obtained, and the performances of the models with feature set 4 are better. For different machine learning methods, the performances of extreme learning machine models are better than others. In addition, the optimal number of hidden layer nodes is 11 for the extreme learning machine model with feature set 4.

Building heating/cooling consumption prediction is of great importance for HVAC system management tasks, such as optimal operation/control strategies, demand and supply management, abnormal energy diagnosis, etc. Compared to traditional methods, data-driven methods have received a lot of attention due to their flexibility and efficiency. In particular, this paper investigates the potential of data partitioning techniques in improving prediction performance of ultra-short-term building heating load prediction. Specifically, with three proposed statistical attributes of 32 days considered by clustering analysis, similar daily operation patterns of pumps (OPPs) in a water-source heat pump system (WSHPS) were identified stepwise. Afterward, the sub-models based on different OPPs were developed by machine learning methods and their performance were compared to the general model without data partitioning. In additional, an operation tree was constructed to predict daily OPPs based on historical weather conditions and available date information. With the assistance of the operation tree, the proposed method can be applied in online prediction. Based on the validation, it can be concluded that the introduction of OPPs-clustering can improve the performance of building heating load prediction.

Fault diagnosis for centrifugal chillers is very important for saving energy and maintaining optimal operating conditions. A fault diagnosis method for centrifugal chillers is proposed based on the associative classification (AC) algorithm, which constructs an associative classifier by excavating strong rules between fault classes and physical attributes. First, association rules with significant support and high confidence values are discovered. Instead of the Apriori algorithm, FP-growth is adopted to accelerate association rule mining. Second, only association rules named class association rules (CARs) whose consequents are limited to fault classes are preserved. Third, pruned CARs are obtained by means of ranking CARs and pruning the redundant rules according to the concept of “higher rank”. Fourth, a limited number of rules are selected out of pruned CARs based on the AC algorithm to construct an associative classifier. This approach is validated using experimental centrifugal chiller data from the ASHRAE Research Project 1043 (RP-1043). Results demonstrate that this proposed AC-based approach can effectively identify seven common chiller faults at both low and high severity levels and the average correct fault diagnosis ratio can be examined up to 86.3%.

Energy consumption models play an integral part in energy management and conservation, as it pertains to buildings. It can assist in evaluating building energy efficiency, in carrying out building commissioning, and in identifying and diagnosing building system faults. This review takes an in-depth look at energy-demand prediction models for buildings in that it delves into recent developments in building energy approaches used to predict energy usage. By enlisting current approaches to the modelling of buildings, methods for building energy simulations can be categorized into four level classes as follows: (i) data-driven approaches; (ii) physics-based approaches; (iii) large scale building energy forecasting approaches; and (iv) hybrid approaches. The focus of this review is to target the data-driven approach and large-scale building energy predicting-based approaches. Here the data driven approaches can be categorized by (1) artificial neural network based approaches; (2) clustering based approaches; (3) statistical and machine learning-based approaches; and (4) support vector machine based approaches. From there, the type of data-driven based approach is further grouped by (a) benchmarking models; (b) energy-mapping models; (c) energy forecasting models; and (d) energy profiling models. Large-scale building-energy prediction techniques is then categorized as follows: (1) white-box based approaches; (2) black-box based approaches, and (3) grey-box based approaches. The current study explores first-rate data-driven based approaches about building energy analysis for industrial, commercial, domestic, etc., within a rural and urban setting. This review paper is based on the necessity of identifying points of departure and research opportunities for urban and rural-level analyses of building level energy performance. A variety of issues are explored which include: energy performance metrics; end-use of different building types; multiple levels of granularity; and urban and rural scales. Each technique encompasses a variety of input information as well as varying calculations or simulation models along with furnishing contrasting outcomes that suggest a variety of usages. A thorough review of each technique is presented in this study. This review highlights strengths, shortcomings, and purpose of the methods of numerous data-mining based approaches. A comprehensive review of energy forecasting models that are specified in the literature part is also provided.

This review paper focuses on the various modeling practices for the identification and apprehension of non-technical losses. The modeling practices are extremely vital to develop, upsurge energy performance, examine and foresee the performance of power transmission & distribution of the electrical system. The data mining based models are innovative and have the subsistence to examine an enormous potential of energy consumption records and performing area profile for preparing housing zone directing the electricity effective living. In this concern, support vector machine model, which classifies illegal customers is a form of advanced mix evolutionary neural network model. Optimum-path forest clustering process is activated to recognize legitimate and irregular profiles of an industry as well as commercial customers to find out theft of electricity. Real time state estimation technique determines a state approximation method in the actual stage for every conversion (transformation) point. Aforementioned allows us to regulate the parts to the maximum extent of non-technical losses through the radial distribution method. The support vector machine with genetic algorithm advances a hybrid method for the non-technical loss investigation and provide an automated assistance to dominate the electricity theft. This model is simplified version of support vector machine. Decision tree and Bayesian regularization networks are appropriated to identify the several kinds of patterns of losses in the electrical system. These practices have been accompanied concerning testing and validation for power system losses in the experimental laboratory. It operates in an influence tool intended to expedite the investigators and scientists. It assists short of spending a massive amount of money, time and energy in experimental events. Prior fabrication modeling methods are remarkably significant in replication of diverse kinds of solar electrical systems. Accordingly, this study concentrates on the base of modeling methods not only saves time but, also preserves the monetary investment in the electrical system. The benefit and imminent opportunity of modeling practices are also conferred in the review paper.

An intelligent fault diagnosis network for variable refrigerant flow air conditioning system is proposed in this study. The network is developed under the foundation of bayesian belief network theory, which comprises two main elements: the structure and parameters. The structure obtained by machine learning and experts’ experiences illustrates the relationships among faults and physical variables from the qualitative prospective, and its parameters (including prior probability distribution and conditional distribution) describe the uncertainty between them quantitatively. Once the structure and parameters are determined, the posterior probability distribution which can be used to complete fault diagnosis and isolation will be calculated by some algorithms. In comparison with other fault diagnosis approaches, the proposed approach can make full use of performance information. Moreover, it is more reasonable and precise to express the relationship between faults and variables rather than Boolean variables. Evaluation was conducted on a variable refrigerant flow air conditioning system, which demonstrated that this strategy is effective and efficient.

This paper proposes an improved decision tree (DT)-based fault diagnosis method for practical variable refrigerant flow (VRF) system. The proposed method is a three-stage method combining DT model with virtual sensor-based fault indicators (FIs). First, FIs are developed based on the virtual sensor (VS) theory for VRF faults, i.e., condenser air-side fouling (Fouling), refrigerant undercharge (RU) and overcharge (RO). Second, FIs are employed as additional input variables to induct a DT-based classification model classification and regression tree (CART). Third, the FIs-CART classification model is used to diagnose on-line data. Validation is conducted using two different datasets, the experimental testing dataset and the on-line testing dataset. Results indicate that the method correctly isolates the three faults i.e., Fouling, RU and RO. The improved DT method is also compared with three tree-based data-driven methods including CART, random forest (RF) and generalized boosted regression (GBM). Comparative results reveal that the proposed method has better fault diagnosis performance for both the experimental and the on-line testing datasets.

A fault detection and diagnosis (FDD) strategy is critical for the refrigerant charge amount (RCA) fault since improper RCA may affect the operational performance of a variable refrigerant flow system. The author’s former work proposes a FDD strategy for the RCA fault. However, three aspects of the former FDD strategy need improvement, i.e. model performance evaluation, more feature information preservation and fault diagnosis accuracy (FDA), especially for the undercharge fault. Firstly, with regard to the model performance evaluation, the concept of a confidence space is proposed to evaluate the FDD model. Secondly, principle component analysis (PCA) is used to reduce the dimension of all feature variables to improve the computational efficiency while preserving almost all feature information. Finally, in order to improve the FDA for the undercharge fault, a dual neural network model for the RCA fault diagnosis strategy has been adopted. The results show that a confidence space can effectively reflect the reliability of fault diagnosis, and the PCA reduces nearly half of the dimension while preserving more than 97% of the feature information, more importantly, the dual neural network improves the correct classification ratio (CCR) more than 9% for three classes (undercharge, normal charge, overcharge), with CCR for the undercharge fault improving by 26.8%.

Energy demand prediction of building heating is conducive to optimal control, fault detection and diagnosis and building intelligent. In this paper, the prediction models are developed using machine learning methods including extreme learning machine (ELM), multiple linear regression, support vector regression and BP neural network. The feature variable sets are optimized through correlation analysis and supplementing indoor temperature. Besides, this paper proposed a strategy to determine the time ahead of prediction model. The thermal response time of building is used as the prediction time step of model. The prediction performances of ELM models with different hidden layer nodes are analyzed and contrasted. The actual data of the building heating using ground source heat pump system are collected and used to test the performances of the models. The results show that the thermal response time of the building is about 40 minutes. Four feature sets are obtained and performances of models with FS4 are better. For different machine learning methods, the performances of ELM models are better than others. In addition, the optimal number of hidden layer nodes is 11 for the ELM model with FS4.

This paper presents the optimized back propagation neural network (BPNN) method for fault diagnosis of the variable refrigerant flow air conditioning (VRF) system in the heating mode. A feature variable set optimization approach of diagnosis models is proposed based on data mining method. First, the correlation analysis method is used to eliminate redundant variables. Then, the association rule mining method is used to optimize the feature set (FS) selection. Five FSs (FS1-FS5) are obtained by optimized feature variable selection. FS1 is the original set. FS2 is the set obtained by correlation analysis and FS3-FS5 are the sets obtained by the association rule mining. The fault diagnosis models with different FSs are evaluated using four fault experiments which include outdoor unit heat exchanger air-side fouling, four-way reversing valve fault, refrigerant undercharge and refrigerant overcharge faults. The results show that the correlation analysis method can effectively eliminate redundant variables and the association rule mining method is feasible to optimize the FSs for fault diagnosis. The BPNN-FS5 model shows the best fault diagnosis performance of the VRF system in the heating mode, whose fault diagnosis correct rate has increased from 88.71% to 96.40% and hit rates of four faults are higher than 90%.

Analysis and application for real-time operational data of buildings is important for energy management. But original data inevitably contains a number of outliers and which usually lead to a significant negative impact on performance of data-based models. In order to eliminate the influence of outliers and improve the robustness of data-based models, this study employs three methods (Boxplot, local outlier factor (LOF) and PCOut) to identify the potential outlying observations in original data set. For purpose of evaluating the outlier detection performance of these three methods, four SVR-based electricity consumption prediction models, Original-SVR, BOX-SVR, LOF-SVR and PCO-SVR, are established. And the performance indexes (RE, RMSE and RSE) of the models are compared and analyzed. The results show that the accuracy of electricity consumption prediction is improved with the help of Boxplot and LOF methods for outlier detection, but PCOut method reduces the accuracy compared with the Original-SVR model. Further study indicates that these observations which repeatedly identified as outlying by Boxplot and LOF methods are the most likely to be abnormal, and when these samples are removed from training data set, the RMSE falls to 2.76 from 6.44 and the RSE falls to 0.11 from 0.58 during testing course.

The outdoor air-side fouling fault is almost inevitable for the practical variable refrigerant flow system. Fouling fault grows gradually and naturally causing increasingly energy penalty and performance degradation to the system. Implementing a proper and reliable fault detection and diagnosis strategy is crucial for practical systems keeping away from fouling fault and maintaining optimal operations. However, traditional model-based and data-driven methods cannot work in practical systems due to the lack of critical sensors and interpretation for model reliability, respectively. Therefore, this study proposes a data mining approach to identify and isolate fouling faults using only built-in sensors. Density-based spatial clustering of applications with noise(DBSCAN) is used for data pre-processing. The classification and regression tree(CART)-based classifier is employed for fault detection. Based on Pearson’s correlation analysis, a multiple linear regression(MLR)-based fault indicator is developed for fault isolation. A case study on a 29.8 kW cooling capacity R410A variable refrigerant flow system is used to validate the proposed strategy. Four levels of foulings are experimentally investigated under three cooling conditions. Results reveal that the proposed strategy correctly identifies 98% of fouling data using only three built-in sensor measurements. It also isolate fouling data from both normal and refrigerant charge fault data.

The enhancement of fault detection and diagnosis (FDD) strategy for air-conditioning system is always a complex difficulty. In previous studies, the virtual refrigerant charge (VRC) sensor method and principal component analysis (PCA) based exponentially-weighted moving average (EWMA) method were proposed to identify refrigerant charge faults for variable refrigerant flow (VRF) systems, respectively. However, both methods had defects in some cases. On the basis of complementary advantages, this study employs the VRC model to detect the undercharge faults as it shown outstanding efficiency on identifying undercharge cases. Similarity, the PCA-EWMA model is used to detect the overcharge faults, since it is very sensitive to the little variation in the overcharge situations. Further, a novel online refrigerant charge fault diagnosis strategy is proposed based on two fault detection methods, i.e. VRC method and PCA-EWMA method. The new hybrid model overcomes the defects of two previous methods appropriately and well inherits the advantages of both. Finally, the robustness of the proposed refrigerant charge fault diagnosis strategy is verified using the experimental data and online data collected from different type of VRF systems.

Sensor faults of air conditioning systems are harmful to optimal control strategies and system performance resulting in poor control of the indoor environment and waste of energy. In order to improve the fault detection and diagnosis (FDD) performance, this paper presents an enhanced sensor fault detection and diagnosis method based on the Satizky-Golay (SG) method and principal component analysis (PCA) method for the VRF system, namely SG-PCA method. Due to the volatility of the original data set of VRF system, the original data are smoothed using SG method at first. Then, the smoothed data are used for PCA model training and fault detection and diagnosis. In order to determine parameters of the SG method, an optimization index is proposed, which is calculated by the signal to noise ratio (SNR), the standard deviation (SD) and the self-detection efficiency. This SG-PCA method for VRF system sensor FDD is validated using field operation data of the VRF system. Various sensor faults at different fault levels are introduced. The results have showed that the SG-PCA method can significantly improve the fault detection and diagnosis performance compared to conventional PCA method.

Developing fault detection and diagnosis (FDD) for the variable refrigerant flow (VRF) system is very important for saving energy and improving reliability of the equipment. For indoor unit (IDU) faults of VRF system, it is especially necessary to detect faults and identify which IDU is faulty. Therefore, this paper has proposed a fault detection strategy based on modularized PCA method, which cannot only detect faults but also specify the faulty IDU. Fault detection models are established respectively for outdoor unit (ODU) and IDUs of VRF system using the modularized PCA method. Then, the expert-based multivariate decoupling strategy with six variables for VRF system is developed to isolate faults. Four common faults are taken into account for VRF system, which include two IDUs faults (electronic expansion valve fault and IDU air-side fouling), one ODU fault (reversing valve stick) and one system fault (refrigerant undercharge). The proposed FDD strategy is evaluated by experimental data of four faults. The test results have shown that modularized PCA-based fault detection method and rule-based diagnosis method are effective for the four typical faults in VRF system. Therefore, it is quite suitable for FDD of VRF system.

This paper proposed a universal virtual variable-speed compressor power (VVCP) sensor for VRF system based on 20-coefficient model. Compressor power can be obtained by the VVCP sensor using three input parameters (frequency, condensing temperature and evaporation temperature) which are measured by system itself. The performance of the proposed VVCP sensor is evaluated using experiments data. Experimental conditions include cooling, heating and non-standard refrigerant charge levels. The result shows that the mean square percentage errors (MSPE) are 9.9% under cooling conditions and 7.96% under heating conditions. The MSPE are 9.88% at steady state and 9.75% at dynamic state when the VVCP sensor is applied under nine different refrigerant charge levels. It demonstrated that the proposed VVCP sensor can obtain compressor power under cooling, heating and non-standard refrigerant charge levels, which could be applied to do operational monitoring and fault detection and diagnosis for VRF system at low cost.

Variable refrigerant flow systems account for a considerable portion of energy consumption in buildings. In order to improve the energy efficiency and estimate the energy saving potentials of variable refrigerant flow systems this study proposes a data mining based method to identify and interpret the power consumption patterns and associations. Two descriptive data mining algorithms, clustering analysis and association rules mining, are used for data partitioning and association mining. The proposed method consists of four phases: data pre-processing, data partitioning, data association mining and knowledge interpretation. Experimental data collected from a tested variable refrigerant flow system in the standard psychrometer testing room are pre-processed and prepared to examine the proposed method. Three time independent influential factors: part load ratio, refrigerant charge level and cooling condition are analyzed. Results show that the method is able to help identify energy consumption patterns and extract energy consumption rules in variable refrigerant flow systems. Three distinct energy consumption patterns are identified: undercharge fault, low and high part load ratio conditions. For compressor operation frequency switch control and refrigerant undercharge patterns, the energy saving potentials could be estimated by making comparisons between energy patterns and rules in a top-down way.

This paper presents an entire sensor fault detection and diagnosis (FDD) for screw chiller system using the support vector data description (SVDD) algorithm. It has advantages of solving problems on describing non-linear and non-Gaussian distributed data. A distance-based D-statistic plot is employed to detect sensor faults. Based on the distance transformation in mathematical way, a new distance variation-based DV-contribution plot is proposed to diagnose the sensor fault. The screw chiller field measured data is utilized to train the SVDD model via a hybrid parameter tuning approach combined the grid search and the 10-fold cross validation. Six typical sensor faults are introduced for validation, i.e. positive and negative biases, positive and negative drifts, precision degradation and complete failure. Test results of both D-statistic and DV-contribution plots show that the SVDD-based method has good FDD results for the six sensor faults. Furthermore, the proposed DV-contribution plot shows more accurate fault diagnosis results compared with the principal component analysis (PCA)-based Q-contribution plot.

s Sensors play essential roles in industrial automatic control systems. The faulty or inaccurate sensors may cause uncomfortable thermal environments, shortened component lifetime and energy consumption loss. Considering the condition-adaptive issue of principal component analysis (PCA) models in fault diagnosis, a data-driven optimized statistical model applied for sensor fault detection and diagnosis (FDD) is proposed in the paper: the subtraction clustering and k-means clustering are combined to identify and classify modeling measurements of unsteady operating conditions. Sensor measurements from a real water source heat pump air-conditioning system is tested and the result shows that the clustering-based statistical model can enhance the ability of dealing with data of multiple operation conditions compared with the traditional PCA model; Different statistical indexes show sensitivity difference in detecting faults in the case of same sensors and same faults.