Classification of Photovoltaic Failures with Hidden Markov Modeling, an Unsupervised Statistical Approach
Failure detection methods are of significant interest for photovoltaic (PV) site operators to help reduce gaps between expected and observed energy generation. Current approaches for field-based fault detection, however, rely on multiple data inputs and can suffer from interpretability issues. In co...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-07-01
|
| Series: | Energies |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1996-1073/15/14/5104 |
| _version_ | 1797406611040370688 |
|---|---|
| author | Michael W. Hopwood Lekha Patel Thushara Gunda |
| author_facet | Michael W. Hopwood Lekha Patel Thushara Gunda |
| author_sort | Michael W. Hopwood |
| collection | DOAJ |
| description | Failure detection methods are of significant interest for photovoltaic (PV) site operators to help reduce gaps between expected and observed energy generation. Current approaches for field-based fault detection, however, rely on multiple data inputs and can suffer from interpretability issues. In contrast, this work offers an unsupervised statistical approach that leverages hidden Markov models (HMM) to identify failures occurring at PV sites. Using performance index data from 104 sites across the United States, individual PV-HMM models are trained and evaluated for failure detection and transition probabilities. This analysis indicates that the trained PV-HMM models have the highest probability of remaining in their current state (87.1% to 93.5%), whereas the transition probability from normal to failure (6.5%) is lower than the transition from failure to normal (12.9%) states. A comparison of these patterns using both threshold levels and operations and maintenance (O&M) tickets indicate high precision rates of PV-HMMs (median = 82.4%) across all of the sites. Although additional work is needed to assess sensitivities, the PV-HMM methodology demonstrates significant potential for real-time failure detection as well as extensions into predictive maintenance capabilities for PV. |
| first_indexed | 2024-03-09T03:29:02Z |
| format | Article |
| id | doaj.art-fca0af90ea724c6c982ff0264d329707 |
| institution | Directory Open Access Journal |
| issn | 1996-1073 |
| language | English |
| last_indexed | 2024-03-09T03:29:02Z |
| publishDate | 2022-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Energies |
| spelling | doaj.art-fca0af90ea724c6c982ff0264d3297072023-12-03T14:58:59ZengMDPI AGEnergies1996-10732022-07-011514510410.3390/en15145104Classification of Photovoltaic Failures with Hidden Markov Modeling, an Unsupervised Statistical ApproachMichael W. Hopwood0Lekha Patel1Thushara Gunda2Sandia National Laboratories, Albuquerque, NM 87123, USASandia National Laboratories, Albuquerque, NM 87123, USASandia National Laboratories, Albuquerque, NM 87123, USAFailure detection methods are of significant interest for photovoltaic (PV) site operators to help reduce gaps between expected and observed energy generation. Current approaches for field-based fault detection, however, rely on multiple data inputs and can suffer from interpretability issues. In contrast, this work offers an unsupervised statistical approach that leverages hidden Markov models (HMM) to identify failures occurring at PV sites. Using performance index data from 104 sites across the United States, individual PV-HMM models are trained and evaluated for failure detection and transition probabilities. This analysis indicates that the trained PV-HMM models have the highest probability of remaining in their current state (87.1% to 93.5%), whereas the transition probability from normal to failure (6.5%) is lower than the transition from failure to normal (12.9%) states. A comparison of these patterns using both threshold levels and operations and maintenance (O&M) tickets indicate high precision rates of PV-HMMs (median = 82.4%) across all of the sites. Although additional work is needed to assess sensitivities, the PV-HMM methodology demonstrates significant potential for real-time failure detection as well as extensions into predictive maintenance capabilities for PV.https://www.mdpi.com/1996-1073/15/14/5104photovoltaicsfailure detectionhidden Markov modelunsupervised statistical learningclassificationoperations and maintenance |
| spellingShingle | Michael W. Hopwood Lekha Patel Thushara Gunda Classification of Photovoltaic Failures with Hidden Markov Modeling, an Unsupervised Statistical Approach Energies photovoltaics failure detection hidden Markov model unsupervised statistical learning classification operations and maintenance |
| title | Classification of Photovoltaic Failures with Hidden Markov Modeling, an Unsupervised Statistical Approach |
| title_full | Classification of Photovoltaic Failures with Hidden Markov Modeling, an Unsupervised Statistical Approach |
| title_fullStr | Classification of Photovoltaic Failures with Hidden Markov Modeling, an Unsupervised Statistical Approach |
| title_full_unstemmed | Classification of Photovoltaic Failures with Hidden Markov Modeling, an Unsupervised Statistical Approach |
| title_short | Classification of Photovoltaic Failures with Hidden Markov Modeling, an Unsupervised Statistical Approach |
| title_sort | classification of photovoltaic failures with hidden markov modeling an unsupervised statistical approach |
| topic | photovoltaics failure detection hidden Markov model unsupervised statistical learning classification operations and maintenance |
| url | https://www.mdpi.com/1996-1073/15/14/5104 |
| work_keys_str_mv | AT michaelwhopwood classificationofphotovoltaicfailureswithhiddenmarkovmodelinganunsupervisedstatisticalapproach AT lekhapatel classificationofphotovoltaicfailureswithhiddenmarkovmodelinganunsupervisedstatisticalapproach AT thusharagunda classificationofphotovoltaicfailureswithhiddenmarkovmodelinganunsupervisedstatisticalapproach |