Energy Informatics: Third Energy Informatics Academy Conference, EI.A 2023, Campinas, Brazil, December 6–8, 2023, Proceedings, Part I (Lecture Notes in Computer Science, 14467)

This two-volume set LNCS 14467-14468 constitutes the proceedings of the First Energy Informatics Academy Conference, EI.A 2023,held in Campinas, Brazil, in December 2023. The 39 full papers together with 8 short papers included in these volumes were carefully reviewed and selected from 53 submissions. The conference focuses on the application of digital technology and information management to facilitate the global transition towards sustainable and resilient energy systems. Preface Organization Contents – Part I Contents – Part II AI Methods in Energy Managing Anomalies in Energy Time Series for Automated Forecasting 1 Introduction 2 Related Work 3 Strategies for Managing Anomalies in Energy Time Series Forecasting 4 Evaluation 4.1 Data and Inserted Synthetic Anomalies 4.2 Applied Anomaly Detection Methods 4.3 Applied Anomaly Compensation Method 4.4 Anomaly-Free Baseline Strategy 4.5 Applied Forecasting Methods 4.6 Experimental Setting 4.7 Results 5 Discussion 6 Conclusion References Illuminating Metaheuristic Performance Using Vortex MAP-Elites for Risk-Based Energy Resource Management 1 Introduction 2 Proposed Methodology 2.1 Risk-Based Problem Formulation 2.2 Vortex MAP-Elites 3 Experiment Parameters 3.1 13-Bus Distribution Network 3.2 MAP-Elites Settings 4 Experiments Results 5 Conclusions References Comparing Manual vs Automatic Tuning of Differential Evolution Strategies for Energy Resource Management Optimization 1 Introduction 2 Proposed Methodology 2.1 Problem Formulation 2.2 Differential Evolution Strategies 2.3 Iterated Racing 3 Case Study 3.1 33-Bus Distribution Network 3.2 Irace Parameterization 4 Results and Discussion 4.1 Auto-tuning Experiments 4.2 Manual vs. Automatic Tuning ERM Results 5 Conclusions References Standard Energy Data Competition Procedure: A Comprehensive Review with a Case Study of the ADRENALIN Load Disaggregation Competition 1 Introduction 2 Review of Data Competition Procedures 2.1 Official Website 2.2 Hosting Platforms 2.3 Competition Descriptions 2.4 Timeline and Stages 2.5 Competition Durations 2.6 Data and Starter Kit 2.7 Submission and Evaluation 3 Case Study 3.1 Competition Scope Definition 3.2 Official Website 3.3 Hosting Platforms 3.4 Competition Timeline, Stages, and Durations 3.5 Data and Starter Kit 3.6 Submission and Evaluation 3.7 Competition Descriptions 4 Discussion and Conclusion References Deep HarDec: Deep Neural Network Applied to Estimate Harmonic Decomposition 1 Introduction 2 Harmonic Estimation Methods and Correlated Studies 3 The Deep HarDec Method 3.1 Dataset Generation for Training, Validation and Testing 3.2 Exhaustive Search over Parameters Values for DNN 3.3 Creating the DNN Model 3.4 Controlled Rectifier Model for Deep HarDec Evaluation 4 Results and Discussions 4.1 Selective Compensation of Estimated Harmonics by Deep HarDec 4.2 Discussions, Advantages, and Limitations of the Deep HarDec 5 Conclusions References Automating Value-Oriented Forecast Model Selection by Meta-learning: Application on a Dispatchable Feeder 1 Introduction 2 Related Work 3 Meta-learning Framework for Forecast Model Selection 3.1 Components of the Proposed Framework 3.2 Usage of the Proposed Framework 4 Applying the Proposed Framework: Application on a Dispatchable Feeder 4.1 Application Dispatchable Feeder 4.2 Applying the Proposed Framework 5 Evaluation 5.1 Experimental Setup 5.2 Results 6 Discussion 7 Conclusion A Appendix A.1 Optimisation Problems A.2 Implementation A.3 Training and Test Data Sets A.4 Input Features References Data-Driven Smart Buildings Early Stage Design Methodology for Energy Efficiency in Buildings Using Asynchronous Distributed Task Queues Framework 1 Introduction 2 Methodology 3 Conclusions References A Real-Time Non-Invasive Anomaly Detection Technique for Cooling Systems 1 Introduction 2 Methodology 2.1 Identifying the Anomalies 2.2 Identifying a Cause of the Anomaly 2.3 Identifying the Faulty Component 3 Evaluation 3.1 Simulation 3.2 Experimentation 4 Related Work 5 Conclusion References A Novel Approach for Climate Classification Using Agglomerative Hierarchical Clustering 1 Introduction 2 Methodology 2.1 Research Strategy 2.2 Preparation of Data 2.3 Clustering the Data 2.4 Calculation of Score 3 Analysis of Climate Classification of USA 3.1 Clustering of Cities of USA 3.2 Building Energy Simulation 3.3 Calculation of Score 4 Results and Discussion 4.1 Limitations 5 Conclusion References Smart Data-Driven Building Management Framework and Demonstration 1 Introduction 2 Methodology 2.1 Overview of the Proposed Framework 2.2 Multi-source Data Available in Buildings 2.3 Semantic Data Integration Schema 2.4 AI Engine 2.5 Smart 3D Interactive Building Management Platform 3 Case Study 3.1 Introduction of the Target Chiller Plant 3.2 Development of Digital Twin and Semantic Model 3.3 Test of AI-Enabled Chiller Sequencing Control Strategy 4 Conclusion References Advances in Machine-Learning Based Disaggregation of Building Heating Loads: A Review 1 Introduction 2 Background 2.1 Scope 2.2 Related Works and Contributions 2.3 Outline of the Paper 3 Methodology 3.1 Datasets 4 Disaggregation of Buildings’ Heating Loads 4.1 Traditional Methods and Shallow Algorithms 4.2 Deep Supervised Learning 4.3 Reinforcement Learning 5 Evaluation of Datasets and Requirements 6 Conclusion References Incorporating Resilience into the IoT-Based Smart Buildings Architecture 1 Introduction 2 IoT-Based Smart Building 3 The Fundamental of Resilience Architecture 4 Related Works on Challenges and Efforts in IoT Resilience 5 Incorporating Resilience into the Design 6 Conclusion References Energy and Industry 4.0 Impact of Setpoint Control on Indoor Greenhouse Climate Modeling 1 Introduction 2 Methodology 2.1 Sensitivity Analysis (SA) 3 Results 4 Discussion 4.1 Control and Reaction of the Indoor Climate 4.2 Energy Consumption Contribution 4.3 Linearization of the Inputs 4.4 Limitation of the Model 5 Conclusion References A Modifiable Architectural Design for Commercial Greenhouses Energy Economic Dispatch Testbed 1 Introduction 1.1 Related Works 2 Methodology 3 High-Level Stakeholders’ Requirements Identification 3.1 Architectural Design 4 Experimental Setup 4.1 Greenhouse Energy System Configuration 4.2 Multi-objective Problem Definition 4.3 Optimization Configuration 4.4 External Inputs 5 Results 6 Discussion 7 Conclusion and Future Work References Business Models for Digitalization Enabled Energy Efficiency and Flexibility in Industry: A Survey with Nine Case Studies 1 Introduction 2 Related Works 2.1 Digital Solutions for Energy Efficiency and Flexibility 2.2 Business Models for Energy Efficiency and Flexibility 3 Methodology 3.1 Data Collection 3.2 Data Analysis – Business Model Canvas 3.3 Evaluation Index for Business Models 4 Case Studies 5 Results 5.1 Business Model Analysis 5.2 Value of Business Model Evaluation 6 Discussion 6.1 Strengths and Weaknesses 6.2 Relationships Between BMC Components 6.3 Recommendations on Digital Solution Development for Enabling Energy Efficiency and Flexibility in Industry 7 Conclusion References Identifying Best Practice Melting Patterns in Induction Furnaces: A Data-Driven Approach Using Time Series K-Means Clustering and Multi-criteria Decision Making 1 Introduction 2 Background 2.1 Energy Efficiency in Foundry Production 3 Methodology 3.1 Clustering Algorithm Selection 3.2 Multi-criteria Decision Making 4 Case Study 5 Results 5.1 Time-Series K-means Clustering of Melting Profiles 5.2 Multi-criteria Decision Making for Best Practice Melting Profile Identification 5.3 Energy Efficiency Potential for Best-Practice Operation 6 Discussion 7 Conclusion References Machine Learning Applied to Industrial Machines for an Efficient Maintenance Strategy: A Predictive Maintenance Approach 1 Introduction 2 Training/Testing Dataset 3 Proposed Methodology 3.1 Artificial Neural Network Training 3.2 Random Forest Training 4 Results and Discussion 5 Conclusion References Author Index