Data-Driven Leak Localization in Urban Water Distribution Networks Using Big Data for Random Forest Classifier
In the present paper, a Random Forest classifier is used to detect leak locations on two different sized water distribution networks with sparse sensor placement. A great number of leak scenarios were simulated with Monte Carlo determined leak parameters (leak location and emitter coefficient). In o...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-03-01
|
| Series: | Mathematics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2227-7390/9/6/672 |
| _version_ | 1797540520784822272 |
|---|---|
| author | Ivana Lučin Bože Lučin Zoran Čarija Ante Sikirica |
| author_facet | Ivana Lučin Bože Lučin Zoran Čarija Ante Sikirica |
| author_sort | Ivana Lučin |
| collection | DOAJ |
| description | In the present paper, a Random Forest classifier is used to detect leak locations on two different sized water distribution networks with sparse sensor placement. A great number of leak scenarios were simulated with Monte Carlo determined leak parameters (leak location and emitter coefficient). In order to account for demand variations that occur on a daily basis and to obtain a larger dataset, scenarios were simulated with random base demand increments or reductions for each network node. Classifier accuracy was assessed for different sensor layouts and numbers of sensors. Multiple prediction models were constructed for differently sized leakage and demand range variations in order to investigate model accuracy under various conditions. Results indicate that the prediction model provides the greatest accuracy for the largest leaks, with the smallest variation in base demand (62% accuracy for greater- and 82% for smaller-sized networks, for the largest considered leak size and a base demand variation of <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mo>±</mo><mn>2.5</mn><mo>%</mo></mrow></semantics></math></inline-formula>). However, even for small leaks and the greatest base demand variations, the prediction model provided considerable accuracy, especially when localizing the sources of leaks when the true leak node and neighbor nodes were considered (for a smaller-sized network and a base demand of variation <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mo>±</mo><mn>20</mn><mo>%</mo></mrow></semantics></math></inline-formula> the model accuracy increased from 44% to 89% when top five nodes with greatest probability were considered, and for a greater-sized network with a base demand variation of <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mo>±</mo><mn>10</mn><mo>%</mo></mrow></semantics></math></inline-formula> the accuracy increased from 36% to 77%). |
| first_indexed | 2024-03-10T13:02:24Z |
| format | Article |
| id | doaj.art-dc6bf656431c4b12b7ee70eb86a3b655 |
| institution | Directory Open Access Journal |
| issn | 2227-7390 |
| language | English |
| last_indexed | 2024-03-10T13:02:24Z |
| publishDate | 2021-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Mathematics |
| spelling | doaj.art-dc6bf656431c4b12b7ee70eb86a3b6552023-11-21T11:26:32ZengMDPI AGMathematics2227-73902021-03-019667210.3390/math9060672Data-Driven Leak Localization in Urban Water Distribution Networks Using Big Data for Random Forest ClassifierIvana Lučin0Bože Lučin1Zoran Čarija2Ante Sikirica3Faculty of Engineering, University of Rijeka, Vukovarska 58, 51000 Rijeka, CroatiaFaculty of Engineering, University of Rijeka, Vukovarska 58, 51000 Rijeka, CroatiaFaculty of Engineering, University of Rijeka, Vukovarska 58, 51000 Rijeka, CroatiaFaculty of Engineering, University of Rijeka, Vukovarska 58, 51000 Rijeka, CroatiaIn the present paper, a Random Forest classifier is used to detect leak locations on two different sized water distribution networks with sparse sensor placement. A great number of leak scenarios were simulated with Monte Carlo determined leak parameters (leak location and emitter coefficient). In order to account for demand variations that occur on a daily basis and to obtain a larger dataset, scenarios were simulated with random base demand increments or reductions for each network node. Classifier accuracy was assessed for different sensor layouts and numbers of sensors. Multiple prediction models were constructed for differently sized leakage and demand range variations in order to investigate model accuracy under various conditions. Results indicate that the prediction model provides the greatest accuracy for the largest leaks, with the smallest variation in base demand (62% accuracy for greater- and 82% for smaller-sized networks, for the largest considered leak size and a base demand variation of <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mo>±</mo><mn>2.5</mn><mo>%</mo></mrow></semantics></math></inline-formula>). However, even for small leaks and the greatest base demand variations, the prediction model provided considerable accuracy, especially when localizing the sources of leaks when the true leak node and neighbor nodes were considered (for a smaller-sized network and a base demand of variation <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mo>±</mo><mn>20</mn><mo>%</mo></mrow></semantics></math></inline-formula> the model accuracy increased from 44% to 89% when top five nodes with greatest probability were considered, and for a greater-sized network with a base demand variation of <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mo>±</mo><mn>10</mn><mo>%</mo></mrow></semantics></math></inline-formula> the accuracy increased from 36% to 77%).https://www.mdpi.com/2227-7390/9/6/672leak localizationwater distribution networkrandom forestprediction modelingbig data |
| spellingShingle | Ivana Lučin Bože Lučin Zoran Čarija Ante Sikirica Data-Driven Leak Localization in Urban Water Distribution Networks Using Big Data for Random Forest Classifier Mathematics leak localization water distribution network random forest prediction modeling big data |
| title | Data-Driven Leak Localization in Urban Water Distribution Networks Using Big Data for Random Forest Classifier |
| title_full | Data-Driven Leak Localization in Urban Water Distribution Networks Using Big Data for Random Forest Classifier |
| title_fullStr | Data-Driven Leak Localization in Urban Water Distribution Networks Using Big Data for Random Forest Classifier |
| title_full_unstemmed | Data-Driven Leak Localization in Urban Water Distribution Networks Using Big Data for Random Forest Classifier |
| title_short | Data-Driven Leak Localization in Urban Water Distribution Networks Using Big Data for Random Forest Classifier |
| title_sort | data driven leak localization in urban water distribution networks using big data for random forest classifier |
| topic | leak localization water distribution network random forest prediction modeling big data |
| url | https://www.mdpi.com/2227-7390/9/6/672 |
| work_keys_str_mv | AT ivanalucin datadrivenleaklocalizationinurbanwaterdistributionnetworksusingbigdataforrandomforestclassifier AT bozelucin datadrivenleaklocalizationinurbanwaterdistributionnetworksusingbigdataforrandomforestclassifier AT zorancarija datadrivenleaklocalizationinurbanwaterdistributionnetworksusingbigdataforrandomforestclassifier AT antesikirica datadrivenleaklocalizationinurbanwaterdistributionnetworksusingbigdataforrandomforestclassifier |