Читать книгу Security Issues and Privacy Concerns in Industry 4.0 Applications - Группа авторов - Страница 26
1.3.4 Machine Learning Models in SWMS
ОглавлениеVarious literature surveys on SWMS have been summarized in Table 1.1 [19-25]. Monitoring the water quality and classifying them according to the contamination level is performed through ML methods along with IoT. When the water is classified under impure category, the level of contamination has to be tested. In water parameters such as chloride, sulphate and alkalinity contents were analyzed. With the presence of these chemicals, water quality is predicted through neural networks [19]. Big data and Artificial Intelligence (AI)–based Support Vector Machine (SVM) play an important role in categorization of water. When drinking water is analyzed, an ML-based prediction method is implemented and also IoT sensors are deployed in video-surveillance for the classification of polluted water and clean water [20]. In 2020, an intelligent water management system has been designed and implemented through Thingspeak cloud platform, where the water leakage has been detected via Blynk application [21]. Also, water metering is attached through which the amount of water consumed can be measured in real time.
Table 1.1 Research on IoT-based SWMS.
| Research | Purpose | Device/method used | Models |
| Water Contamination [19, 24] | Water Contamination Assessments | ML with Fast Fourier Transform | SVM and Color Layout Descriptor |
| Water Quality Parameters [19, 25] | Water Contamination and Quality Analysis | Neural Network, ML-based classification, IoT devices | SVM, IoT sensor models |
| Drinking Water [10, 22] | Drinking Water Analysis | ML-based prediction and classification | Decision Tree, K-Nearest Neighbour, SVM |
| Water Level [21, 23] | Water Level Detection | IoT device | Raspberry Pi |
| Water Meter [20] | Water usage measurements | IoT device, WSN | Arduino and NodeMCU |
