detection and diagnosis of multiple dependent faults in HVAC systems using machine learning techniques
Bibliographic record
Abstract
The building sector accounts for about 40% of the total annual energy consumption in the United States and 25% in Canada. Therefore, it is so essential to design and operate energy-efficient smart buildings. About 15 to 30% of the energy in the commercial buildings would be wasted if the heating, ventilation, and air conditioning (HVAC) systems are not maintained regularly, or they are inappropriately controlled, and if the system degradation has not been detected at early stages. Therefore, the building performance should be monitored in real-time using the Building Automation System (BAS), in order to detect any potential fault in the system and diagnose the sources of malfunction in the HVAC systems. \nMachine learning (ML) models which are developed from BAS trend data without information about the physical system, proved in the past, good performance for analysis the linear and non-linear systems. Therefore, ML models due to their capacity for distinguishing the faulty from normal operation are applied in this thesis for multiple dependent fault detection and diagnosis (MDFDD). In this thesis, ML models are proposed for MDFDD of sensors in an air handling unit (AHU) of an institutional building located in the Concordia University campus. \nTwo approaches are proposed with application of experimental and synthetic data sets: 1) combination of machine learning models with rule-based technique, 2) classification machine learning models. \n \n1) ML models (e.g., Support vector regression (SVR)), developed from building automation system (BAS) trend data, predict air temperature of two target sensors, under normal operation conditions without known problems. The fault symptom is detected when the residual of measured and predicted values exceed the threshold. The recurrent neural network (RNN) models predict the normal operation values of regressor sensors, which are compared with measurements, as the first step for the identification of fault symptoms. Rule-based models are used for fault diagnosis of sensors or equipment. \n2) The optimized classification ML models (e.g., shallow artificial neural network (ANN), deep ANN, K-Nearest Neighbor (KNN), decision tree classification, random forest classification, support vector machine (SVM), Naïve Bayes, and principal component analysis) are developed for MDFDD. ML algorithms parameters are optimized over RandomizedSearch method by varying the length of training dataset, input time lags, and relevant parameters of each ML models. \nResults from the two data set types of an existing building show the high quality of proposed method for the detection and diagnosis of the multiple dependent faults.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.002 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".