Clustering disaggregated load profiles using a Dirichlet process mixture model
The increasing availability of substantial quantities of power-use data in both the residential and commercial sectors raises the possibility of mining the data to the advantage of both consumers and network operations. We present a Bayesian non-parametric model to cluster load profiles from househo...
| 主要な著者: | , , |
|---|---|
| フォーマット: | Journal article |
| 出版事項: |
Elsevier
2015
|
| _version_ | 1826267504867016704 |
|---|---|
| author | Granell, R Axon, C Wallom, D |
| author_facet | Granell, R Axon, C Wallom, D |
| author_sort | Granell, R |
| collection | OXFORD |
| description | The increasing availability of substantial quantities of power-use data in both the residential and commercial sectors raises the possibility of mining the data to the advantage of both consumers and network operations. We present a Bayesian non-parametric model to cluster load profiles from households and business premises. Evaluators show that our model performs as well as other popular clustering methods, but unlike most other methods it does not require the number of clusters to be predetermined by the user. We used the so-called 'Chinese restaurant process' method to solve the model, making use of the Dirichlet-multinomial distribution. The number of clusters grew logarithmically with the quantity of data, making the technique suitable for scaling to large data sets. We were able to show that the model could distinguish features such as the nationality, household size, and type of dwelling between the cluster memberships. |
| first_indexed | 2024-03-06T20:55:12Z |
| format | Journal article |
| id | oxford-uuid:38fea267-d554-433b-9f36-2cee47870a04 |
| institution | University of Oxford |
| last_indexed | 2024-03-06T20:55:12Z |
| publishDate | 2015 |
| publisher | Elsevier |
| record_format | dspace |
| spelling | oxford-uuid:38fea267-d554-433b-9f36-2cee47870a042022-03-26T13:53:00ZClustering disaggregated load profiles using a Dirichlet process mixture modelJournal articlehttp://purl.org/coar/resource_type/c_dcae04bcuuid:38fea267-d554-433b-9f36-2cee47870a04Symplectic Elements at OxfordElsevier2015Granell, RAxon, CWallom, DThe increasing availability of substantial quantities of power-use data in both the residential and commercial sectors raises the possibility of mining the data to the advantage of both consumers and network operations. We present a Bayesian non-parametric model to cluster load profiles from households and business premises. Evaluators show that our model performs as well as other popular clustering methods, but unlike most other methods it does not require the number of clusters to be predetermined by the user. We used the so-called 'Chinese restaurant process' method to solve the model, making use of the Dirichlet-multinomial distribution. The number of clusters grew logarithmically with the quantity of data, making the technique suitable for scaling to large data sets. We were able to show that the model could distinguish features such as the nationality, household size, and type of dwelling between the cluster memberships. |
| spellingShingle | Granell, R Axon, C Wallom, D Clustering disaggregated load profiles using a Dirichlet process mixture model |
| title | Clustering disaggregated load profiles using a Dirichlet process mixture model |
| title_full | Clustering disaggregated load profiles using a Dirichlet process mixture model |
| title_fullStr | Clustering disaggregated load profiles using a Dirichlet process mixture model |
| title_full_unstemmed | Clustering disaggregated load profiles using a Dirichlet process mixture model |
| title_short | Clustering disaggregated load profiles using a Dirichlet process mixture model |
| title_sort | clustering disaggregated load profiles using a dirichlet process mixture model |
| work_keys_str_mv | AT granellr clusteringdisaggregatedloadprofilesusingadirichletprocessmixturemodel AT axonc clusteringdisaggregatedloadprofilesusingadirichletprocessmixturemodel AT wallomd clusteringdisaggregatedloadprofilesusingadirichletprocessmixturemodel |