| Titre : | Lot based solution for the prediction of civil engineering building |
| Auteurs : | Imane MERIZIG, Auteur ; Sadek Labib Terrissa, Directeur de thèse |
| Type de document : | Monographie imprimée |
| Editeur : | Biskra [Algérie] : Faculté des Sciences Exactes et des Sciences de la Nature et de la Vie, Université Mohamed Khider, 2021 |
| Format : | 1 vol. (78 p.) / ill. / 29 cm |
| Langues: | Anglais |
| Mots-clés: | IoT, Cloud computing, Edge computing, machine learning, smart civil engineering, mosquitto, MLP, Linear Regression, RMSE. |
| Résumé : | Internet of things (IoT), Cloud computing (CC), Edge computing (EC) and machine learning (ML) plays an important role in Smart Civil Engineering (SCE).Internet of things (IoT) is the branch of Information technology which generates a connectivity link between “Internet” and actual physical “Things”, cloud and edge computing are a new computing paradigms, their purpose is to offering access to services, save and monitor transmitted data from IoT in any place and at any time, and machine learning is a branch in artificial intelligent (AI) which aim to predict and diagnosis the results.The main objective of our system is IoT based for the prediction of Civil Engineering buildings, using Raspberry Pi and sensors to monitor buildings of civil engineering, then, we send data to Edge computing utilize the protocol IoT mosquitto to pre-processing and filtering data. After these, we save the send data from edge computing in cloud computing to do the prediction and diagnosis results, using the best models regression of machine learning Multi- Layer Perceptron (MLP) Regression and Linear regression (LR), which have the lowest value of root mean squared error (RMSE). |
| Sommaire : |
General introduction
1 Chapter 01: Basic concepts Introduction 6 1.Smart civil engineering (SCE) 6 1.1.Definition of smart civil engineering 7 1.2.The reason for the emergence of smart civil engineering 7 1.3.Definition of structural health monitoring (SHM) 8 1.4.Features of smart civil engineering 8 1.4.1.Systems are connected 8 1.4.2.The use of sensors 8 1.4.3.Automation 9 1.4.4.Data 9 1.5.Applications of Smart Civil Engineering 9 2.Technologies used 10 2.1.Cloud Computing 11 2.1.1.Definitions of Cloud Computing 11 2.1.2.Characteristics of Cloud Computing 11 2.1.3.Types of Cloud Computing 12 2.1.3.1.Public Cloud 12 2.1.3.2.Private Cloud 13 2.1.3.3.Community Cloud 13 2.1.3.4.Hybrid Cloud 13 2.2.Internet of things (IoT) 13 2.2.1.Definition of Internet of things (IoT) 14 2.2.2.Characteristics of IoT 14 2.2.3.Architecture of IoT 15 2.2.3.1.Perception layer 16 2.2.3.2.Network layer 16 2.2.3.3.Application layer 16 2.2.4.Components of IoT 16 2.2.4.1.Sensors 16 2.2.4.2.Networks 17 2.2.5.Protocol of IoT 17 2.2.5.1.Message Queuing Telemetry Transport (MQTT) 17 2.2.5.1.1.Definition of MQTT 17 2.2.5.1.2.MQTT control packet 17 2.2.5.1.3.MQTT Fundamentals 18 2.2.6.Applications of IoT 19 2.3.Edge computing (EC) 20 2.3.1.Definition of Edge computing 20 2.3.2.Architecture of edge computing 21 2.3.3.Characteristics of Edge Computing 22 2.3.3.1.Dense Geographical Distribution 22 2.3.3.2.Location Awareness 22 2.3.3.3.Low Latency 23 2.3.3.4.Open architectures 23 2.3.3.5.Data pre-processing and filtering 23 2.3.4.Benefits of Edge Computing 23 2.3.4.1.Eliminates Latency 23 2.3.4.2.Saves Bandwidth 24 2.3.4.3.Reduces Congestion 24 2.3.5.Applications of Edge computing 24 Conclusion 25 Chapter 02: Machine learning and previous studies Introduction 27 1.Machine learning (ML) 27 1.1.Definition of Machine learning 27 1.2.Architecture of machine learning 28 1.3.Types of Machine learning 30 1.3.1.Supervised Learning 31 1.3.1.1.Classification 31 1.3.1.2.Regression 31 1.3.2.Unsupervised Learning 32 1.3.2.1.Clustering 32 1.3.3.Reinforcement Learning 33 1.4.Machine Learning Algorithms 33 1.4.1.Gradient Boosting Regressor (GBR) 33 1.4.2.K-Nearest Neighbor 34 1.4.3.Multi-layer perceptron regression 34 1.4.4.Linear regression 35 1.5.Activation function 36 1.5.1.Function of Rectified Linear Unit (Relu) 36 1.5.2.Function of Heaviside 36 1.5.3.Function of sigmoid 37 1.6.Applications of machine learning 37 2.Previous studies 38 2.1.Article 01 38 2.1.1.Content 38 2.1.2.The hierarchical structure of early-alarm system 39 2.1.2.1.Sensing layer 39 2.1.2.2.IoT layer 39 2.1.2.3.Cloud computing layer 40 2.1.2.4.Application layer 40 2.1.3.Result 40 2.1.4.Conclusions 40 2.2.Article 02 41 2.2.1.Content 41 2.2.2.Result 42 2.2.3.Conclusions 42 2.3.Article 03 42 2.3.1.Content 42 2.3.2.Results 43 2.3.3.Conclusions 43 Conclusion 44 Chapter 03: Design of system Introduction 46 1.Problematics 46 2.Objectives 46 3.Significance of Study 47 4.Architecture of system 47 4.1.IoT layer 49 4.2.Edge computing layer 49 4.3.Cloud computing layer 50 5.System hardware design 50 5.1.Components of publisher part 52 5.1.1.Raspberry Pi 52 5.1.2.AM2302 sensor 52 5.1.3.HC-SR04 sensor 53 5.2.WI-FI Standard 53 5.3.Mosquitto 54 5.4.Components of subscriber part 54 5.4.1.SQLite Browser 54 5.4.2.Machine learning 54 5.4.2.1.Linear Regression (LR) 55 5.4.2.2.Multi-layer perceptron Regressor (MLPRegressor) 56 Conclusion 58 Chapter 04:Implementation and experimental results Introduction 60 1.Environments and tools 60 1.1.Hardware configuration 60 1.2.Development environment 60 1.3.Programming language and packages 61 1.4.The implementation of system 63 1.4.1.Connectivity part (IoT) 63 1.4.1.1.Publisher part 63 1.4.1.2.Subscriber part 64 1.4.2.Prediction part of machine learning (ML) 65 1.4.2.1.Results obtained and discussion 66 a.Results obtained for model LR in Ka dataset 67 b.Results obtained for model MLP in Kp dataset 69 Conclusion 75 General conclusion 77 |
| Type de document : | Mémoire master |
Disponibilité (1)
| Cote | Support | Localisation | Statut |
|---|---|---|---|
| MINF/606 | Mémoire master | bibliothèque sciences exactes | Consultable |
