17th International Conference on Soft Computing Models in Industrial and Environmental Applications (SOCO 2022): Salamanca, Spain, September 5–7, 2022, Proceedings

Description

This book contains accepted papers presented at SOCO 2022 conference held in the beautiful and historic city of Salamanca (Spain), in September 2022. Soft computing represents a collection or set of computational techniques in machine learning, computer science, and some engineering disciplines, which investigate, simulate, and analyze very complex issues and phenomena. After a thorough peer-review process, the 17th SOCO 2022 International Program Committee selected 64 papers which are published in these conference proceedings and represent an acceptance rate of 60%. In this relevant edition, a particular emphasis was put on the organization of special sessions. Seven special sessions were organized related to relevant topics such as machine learning and computer vision in Industry 4.0; time series forecasting in industrial and environmental applications; optimization, modeling, and control by soft computing techniques; soft computing applied to renewable energy systems; preprocessing big data in machine learning; tackling real-world problems with artificial intelligence. The selection of papers was extremely rigorous to maintain the high quality of the conference. We want to thank the members of the program committees for their hard work during the reviewing process. This is a crucial process for creating a high-standard conference; the SOCO conference would not exist without their help. Preface Organization General Chair International Advisory Committee Program Committee Chairs Program Committee Special Sessions Machine Learning and Computer Vision in Industry 4.0 Program Committee Time Series Forecasting in Industrial and Environmental Applications Program Committee Optimization, Modeling, and Control by Soft Computing Techniques Program Committee Soft Computing Applied to Renewable Energy Systems Program Committee Preprocessing Big Data in Machine Learning Program Committee Tackling Real-World Problems with Artificial Intelligence Program Committee SOCO 2022 Organizing Committee Chairs SOCO 2022 Organizing Committee Contents Decision Support and Deep Learning Anomaly Detection of Security Threats to Cyber-Physical Systems: A Study 1 Introduction 2 Statistical Analysis 3 Literature Analysis 3.1 CPS Security Design 3.2 Anomaly Detection/Threat Detection in CPS 4 Outstanding Challenges 5 Conclusions References Predictive Maintenance for Maintenance-Effective Manufacturing Using Machine Learning Approaches 1 Introduction 2 State-of-the-Art 3 Training/Testing Dataset 4 Proposed Methodology 4.1 Gradient Boosting Training 4.2 Support Vector Machine Training 5 Results and Discussion 6 Conclusions References Estimation of Lamb Weight Using Transfer Learning and Regression 1 Introduction 2 Image Acquisition and Data Preparation 3 Proposed Architecture 4 Experimental Results 5 Conclusions References UAV Simulation for Object Detection and 3D Reconstruction Fusing 2D LiDAR and Camera 1 Introduction 2 Related Works 3 Simulation Framework 4 Proposed Process 5 Demonstration and Evaluation 6 Conclusions and Perspectives References A SO2 Pollution Concentrations Prediction Approach Using Autoencoders 1 Introduction 2 Database 3 Methodology 4 Results 5 Conclusions References CPU Computation Influence on Energy Consumption Forecasting Activities of a Building 1 Introduction 2 Methodology 3 Case Study and Results 3.1 Case Study 3.2 Results 4 Conclusions References Python-Based Ecosystem for Agent Communities Simulation 1 Introduction 2 Related Works 3 Proposed Solution 3.1 PEAK Multi-agent System Platform 3.2 Management 4 Case Study 5 Conclusion References Deep Learning Approach for the Prediction of the Concentration of Chlorophyll ɑ in Seawater. A Case Study in El Mar Menor (Spain) 1 Introduction 2 Area Description and Datasets 3 Methods 3.1 Artificial Neural Networks 3.2 Bayesian Regularized Neural Networks 3.3 Long Short-Term Memory Neural Networks 3.4 Mutual Information 3.5 Minimum-Redundancy-Maximum-Relevance (mRMR) 4 Experimental Procedure 4.1 Creation of the Lagged Datasets 4.2 Forecasting Models 5 Results and Discussion 6 Conclusions References Evolutionary Computing A Hybrid Discrete Symbiotic Organisms Search Algorithm and List-Based Simulated Annealing Algorithm for Traveling Salesman Problem 1 Introduction 2 A Discrete Symbiotic Organisms Search Algorithm for TSP 2.1 Mutualism Phase 2.2 Commensalism Phase 2.3 Parasitism Phase 3 A List-Based Simulated Annealing Algorithm for TSP 4 A Hybrid DSOS-LBSA Algorithm for TSP 5 Computational Results and Discussion 5.1 Parameter Settings 5.2 Computational Results and Analysis 6 Conclusion and Future Work References Estimation of Distribution Algorithms Applied to the Next Release Problem 1 Introduction 2 Next Release Problem 2.1 Related Work 2.2 Multi-objective Next Release Problem 3 Proposal: Univariate EDAs for the MONRP 3.1 MONRP-UMDA 3.2 MONRP-PBIL 4 Experimental Evaluation 4.1 Algorithms 4.2 Datasets 4.3 Methodology 5 Results and Analysis 5.1 Best Configurations 5.2 Pareto Front Results 5.3 Metrics Results 6 Conclusions and Future Works References An Extremal Optimization Approach to the Pairwise Connectivity Critical Node Detection Problem 1 Introduction 2 Related Work and Problem Formulation 3 Extremal Optimization 4 Numerical Experiments 5 Conclusions References Neural Networks and Data Mining Dimensional Reduction Applied to an Intelligent Model for Boost Converter Switching Operation 1 Introduction 2 Case Study 3 Model Approach 3.1 Dataset 3.2 Methods 3.3 Classification Model 3.4 Experiments Description 4 Results 5 Conclusions and Future Works References Intuitionistic Fuzzy Sets in J-CO-QL+? 1 Introduction 2 Background 2.1 Classical Fuzzy Sets 2.2 Intuitionistic Fuzzy Sets and Relations 2.3 Example: Representing Medical Knowledge 3 Intuitionistic Fuzzy Sets and J-CO-QL+ 3.1 J-CO-QL+ Data Model and Execution Model 3.2 J-CO-QL+ Script 4 Learned Lessons and Conclusions References Assessing the Efficient Market Hypothesis for Cryptocurrencies with High-Frequency Data Using Time Series Classification 1 Introduction 2 Literature Review 3 Methods 4 Experiments and Results 4.1 Datasets Used 4.2 Experimental Settings and Performance Measures 4.3 Results 5 Conclusions References Blockchain for Supply Chain Traceability with Data Validation 1 Introduction 2 Related Work 3 Blockchain-Based GSC Traceability 4 Smart Contract for GSC Traceability 5 Smart Contract Implementation and Performance Evaluation 6 Conclusions and Future Work References Compression of Clustered Ship Trajectories for Context Learning and Anomaly Detection 1 Introduction 2 Background Information 2.1 Data Pre-processing and Data Imbalance 2.2 Trajectory Clustering 2.3 Trajectory Compression 3 Proposed Architecture 3.1 Data Preparation and Cluster Generation 3.2 Compression of Trajectories 3.3 Representative Points Extraction 4 Results Analysis 5 Conclusions and Perspectives References DR Participants’ Actual Response Prediction Using Artificial Neural Networks 1 Introduction 2 Proposed Methodology 3 Case Study 4 Results and Discussion 5 Conclusion References Non-linear Neural Models to Predict HRC Steel Price in Spain 1 Introduction and Previous Work 2 Materials and Methods 2.1 Dataset 2.2 Non-lineal Neural Models 3 Experiments and Results 4 Conclusions and Future Work References Soft Computing Applications First Steps Predicting Execution of Civil Works from Georeferenced Infrastructure Data 1 Introduction 1.1 State of the Art 1.2 Research Proposal 2 Methodology 2.1 Preprocess 2.2 Data Analysis 2.3 Dataset Generation 2.4 Supervised Classification 2.5 Evaluation 2.6 Results 3 Conclusion References Virtual Sensor to Estimate Air Pollution Heavy Metals Using Bioindicators 1 Introduction 2 Database 3 Methodology 4 Results 5 Conclusions References Regression Techniques to Predict the Growth of Potato Tubers 1 Introduction 2 Previous Work 3 Regression Techniques 3.1 Multiple Linear Regression 3.2 Multilayer Perceptron 3.3 Radial-Basis Function Network 3.4 Support Vector Machine 4 Materials and Methods 5 Results and Discussion 6 Conclusions and Future Work References Reliability-Sensitive Optimization for Provision of Ancillary Services by Tempo-Spatial Correlated Distributed Energy Resources 1 Introduction 2 Multivariate Correlation Modeling 2.1 Pair-Copula Construction 2.2 D-Vine Copula Structure 3 Reliability-Sensitive Optimization 3.1 Multivariate Correlation Modeling 3.2 Joint Reliability Evaluation Methodology 4 Simulation Study 5 Conclusion References Special Session on Machine Learning and Computer Vision in Industry 4.0 Predictive Maintenance of ATM Machines by Modelling Remaining Useful Life with Machine Learning Techniques 1 Introduction 2 Materials 3 Methods 3.1 Task Definition 3.2 Feature Extraction and Selection 3.3 Pre-processing 3.4 Machine Learning Model 3.5 Experimental Procedure 4 Results 5 PdM Decision Support System for SIMPLE Project 6 Conclusions References The Impact of Content Deletion on Tabular Data Similarity Using Contextual Word Embeddings 1 Introduction 2 Related Work 3 Research Method 4 Experiments 4.1 Models 4.2 Datasets 4.3 Results 5 Conclusions and Future Work References Deep Learning-Based Dementia Prediction Using Multimodal Data 1 Introduction 2 DementiaBank Dataset 3 Approach 3.1 Audio 3.2 Text 3.3 Multimodal 3.4 Other Approaches 4 Evaluation 5 Conclusion References Lightweight Models in Face Attribute Recognition: Performance Under Oclussions 1 Introduction 2 Related Work 3 Description of the System 3.1 Models 3.2 Datasets 4 Experimental Setup 4.1 Training 4.2 Evaluation 5 Evaluation with Masked Faces 6 Conclusions and Future Work References Small Vessel Detection in Changing Seaborne Environments Using Anchor-Free Detectors on Aerial Images 1 Introduction 2 Related Work 2.1 Vessel Detection 2.2 Datasets 3 Methodology 3.1 Dataset Manipulation 3.2 YOLOX 4 Experimentation 5 Conclusion References Improving Malware Detection with a Novel Dataset Based on API Calls 1 Introduction 2 Related Work 3 Proposal 3.1 Format of the Samples 3.2 Used Datasets 3.3 Used ML Model 4 Results 5 Conclusion References Identifying Places Using Multimodal Social Network Data 1 Introduction 2 Related Work 3 Description of the System 4 Experimental Setup 4.1 Baselines 4.2 Combining Image and Text 4.3 Evaluation 5 Conclusions and Future Work References Monitoring Human Performance Through Deep Learning and Computer Vision in Industry 4.0 1 Introduction 2 Model Proposal 2.1 Manufacturing Description Language 2.2 Manufacturing Process Analyzer 2.3 Fatigue Analyzer 2.4 Recommendations Generator 3 Experimentation 4 Conclusions References Automatic Fish Size Estimation from Uncalibrated Fish Market Images Using Computer Vision and Deep Learning 1 Introduction 2 Fish Size Measuring Methodology 2.1 Inference of Input Data: Instance Segmentation and Classification 2.2 Ground Truth Size Estimation via Visual Metrology 2.3 Fish Size Regression 3 Experimental Evaluation 4 Results and Discussion 5 Conclusions References Vehicle Overtaking Hazard Detection over Onboard Cameras Using Deep Convolutional Networks 1 Introduction 2 Methodology 3 Experiments 3.1 Implementation Details 3.2 Data 3.3 Evaluation 3.4 Results 4 Conclusions References Image Classification Applied to the Problem of Conformity Check in Industry 1 Introduction 2 Related Work 3 Methodology and Results 3.1 Problem Statement - Verifying Presence by Classification 3.2 SVM with HoG 3.3 SVM with BoVW 3.4 Linear SVM with CNN Features 4 Conclusion References A Virtual Sensor Approach to Estimate the Stainless Steel Final Chemical Characterisation 1 Introduction 2 Materials and Methods 2.1 DataSet 2.2 Methods 3 Results 4 Conclusions References Convolutional Neural Networks for Structured Industrial Data 1 Introduction 2 Related Work 3 Materials and Methods 4 Contribution 4.1 Continuous Variable Transmission (CVT) 4.2 Secondary Metallurgy Process 5 Results and Discussion 6 Conclusions and Future Work References Classification of Polymers Based on the Degree of Their Transparency in SWIR Spectrum 1 Introduction 2 Problem Formulation 3 Method Statement 3.1 Polymer Types 3.2 SWIR Camera and Lights 3.3 Background 3.4 Object Detection Method 3.5 CNN for Image Transformation 4 Experiments Procedure 4.1 Hardware Conditions 4.2 Dataset Creation 4.3 ASP U-Net Training 5 Results 6 Conclusions References Deep Learning Based Baynat Foam Classification for Headliners Manufacturing 1 Introduction 2 Deep Learning Methods 2.1 Xception 2.2 Resnet50 2.3 MobilenetV2 2.4 InceptionV3 3 Experimental Results 4 Conclusions and Further Work References Special Session on Time Series Forecasting in Industrial and Environmental Applications A GAN Approach for Anomaly Detection in Spacecraft Telemetries 1 Introduction 2 GANs and Anomaly Detection 3 Related Works 4 Data 5 Method 6 Results 7 Conclusion References Management and Forecasting of the Demand for Caskets in the Funeral Sector. Study Before and During the Covid-19 Pandemic 1 Introduction 2 The Funeral Services Industry 2.1 Funeral Services Sector in Data 3 Dataset Description 4 Methodology 5 Experiments and Results 5.1 Pre-covid 19 Model: Training Data Set 2016-2018 and Test Data Set 2019 5.2 Covid 19 Model: Training Data Set 2016–2018 and Test Data Set 2020 6 Conclusions and Future Work References Explainable Artificial Intelligence for the Electric Vehicle Load Demand Forecasting Problem 1 Introduction 2 Related Works 3 Methodology 3.1 Data Acquisition 3.2 Machine Learning Methods 3.3 Models Explainability 3.4 Quality Parameters 4 Results 4.1 Dataset Description 4.2 Results Discussion 4.3 Explainability 5 Conclusions References A Cluster-Based Deep Learning Model for Energy Consumption Forecasting in Ethiopia 1 Introduction 2 Related Works 3 Methodology 4 Analysis and Result Discussion 4.1 Data Description 4.2 Clustering Process 4.3 Results Achieved 5 Conclusions References Special Session on Optimization, Modeling and Control by Soft Computing Techniques Abnormal Driving Behavior Identification Based on Naturalistic Driving Data Using LSTM Recurrent Neural Networks 1 Introduction 2 Vehicle Dynamics-Based Approach 3 Abnormal Behavior Identification by an LSTM NN Model 3.1 Selected Variables for the Classification Model 3.2 Deep LSTM Recurrent Neural Network Model for Classification 4 Results and Discussion 5 Conclusions and Future Works References Identification of Critical Subgraphs in Drone Airways Graphs by Graph Convolutional Networks 1 Introduction 2 Computational Approach 2.1 Semi-supervised Learning with Graph Convolutional Network 2.2 Enhanced GCN 2.3 Advanced GCN 3 Some Experimental Results 3.1 Basic GCN Dominance Results 3.2 Enhanced GCN Dominance Results 3.3 Advanced GCN Dominance Results 4 Conclusions and Future Work References Robust Velocity Control of an Automated Guided Vehicle Using Artificial Neural Networks 1 Introduction 2 Description of the System 2.1 AGV Model 3 Description of the Control Strategy 3.1 Neural Network Model 3.2 Reference Model 3.3 Neural Network Controller 4 Results 5 Conclusions and Future Works References Studying the Use of ANN to Estimate State-Space Variables for MIMO Systems in a NMPC Strategy 1 Introduction 2 MIMO System: Twin-Rotor 2.1 Mathematical Model Development 2.2 Study of Twin-Rotor Dynamics 3 Experiments and Results 3.1 ANN Model 3.2 NMPC Performance with ANN Model 4 Conclusions References Control of MIMO Systems with iMO-NMPC Strategy 1 Introduction 2 MIMO System Presentation 2.1 SISO System: SNL5 2.2 SISO System: SNL1 2.3 MIMO System: SNL5 and SNL1 3 NN Model 3.1 Topology Analysis 3.2 Prediction Analysis 4 iMO-NMPC Parameters (Control Strategy) 5 Results 5.1 Using Mathematical Model with iMO-NMPC 5.2 Using NN Model with the iMO-NMPC 5.3 Comparison of Prediction Models: Math. vs. NN 6 Conclusions References Optimization of Trajectory Generation for Automatic Guided Vehicles by Genetic Algorithms 1 Introduction 2 Environment Layout Model 3 Description of the Trajectory Optimization Process Based on Genetic Algorithms 3.1 Optimization Methodology 3.2 Genetic Algorithm 4 Results and Discussion 5 Conclusions and Future Works References Special Session on Soft Computing Applied to Renewable Energy Systems Complementing Direct Speed Control with Neural Networks for Wind Turbine MPPT 1 Introduction 2 Mathematical Model of the Wind Turbine 2.1 DFIG Mathematical Description 3 MPPT Control Based on DSC and Neural Networks 4 Discussion of the Results 5 Conclusions and Future Works References A Control Approach on Hybrid Floating Offshore Wind Turbines for Platform and Generated Power Oscillations Reduction at Below-rated Wind Speed 1 Introduction 2 FOWT Model 3 Problem Statement 3.1 RAOs Evaluation 3.2 Control Statement 4 Results 5 Conclusions References Control Tuning by Genetic Algorithm of a Low Scale Model Wind Turbine 1 Introduction 2 Wind Turbine Modelling 2.1 Aerodynamic Model 3 Wind Turbine Low Scale Model 3.1 Blades and Pitch Mechanism 3.2 Electric Generator and Load Actuator 3.3 Measurement System 3.4 Microcontroller 3.5 Load Control System 3.6 Pitch Control System 3.7 Software Architecture 4 Experimental Results 5 Optimal Tuning by Genetic Algorithms 6 Conclusions References Pitch-Based Wind Turbine Tower Vibration Damping Optimized by Simulated Annealing 1 Introduction 2 Pitch-Based Wind Turbine Tower Vibration Damping 3 Tuning of the Controller by Simulated Annealing 3.1 Decision Variables 3.2 Constraints 3.3 Auxiliary Functions 3.4 Cost Function 3.5 Simulated Annealing Algorithm 4 Simulation Results and Discussion 5 Conclusions and Future Works References Neural Networks Techniques for Fault Detection and Offset Prediction on Wind Turbines Sensors 1 Introduction 2 Data Sets 3 Problem Statement 4 LSTM Architectures 5 Methodology 6 Experiments and Results 7 Conclusions and Future Work References Special Session on Pre-processing Big Data in Machine Learning Errors of Identifiers in Anonymous Databases: Impact on Data Quality 1 Introduction 2 Data Collection 2.1 Data Enrichment 2.2 Errors of Identifiers 3 Data Quality 3.1 Quality Metrics 3.2 Aggregated Data Quality 4 Discussion 4.1 Data Cleaning 4.2 Impact of the Anonymization on the Data Quality 5 Conclusions References Feature-Aware Drop Layer (FADL): A Nonparametric Neural Network Layer for Feature Selection 1 Introduction 2 Related Works 3 Methodology 3.1 Description 3.2 Weight Initialization and Regularization 4 Experimentation 4.1 Datasets 4.2 Experimental Settings 4.3 Results and Discussion 5 Conclusions and Future Works References Classification Methods for MOBA Games 1 Introduction 2 Attribute Selection 3 The Proposed Methodology and Experimental Design 4 Results 5 Discussion and Further Research Line References Feature Ranking for Feature Sorting and Feature Selection, and Feature Sorting: FR4(FSoFS)FSo 1 Introduction 2 Feature Selection 3 The Proposed Methodology 4 Experimentation, Results and Discussion 5 Conclusions References Special Session on Tackling Real World Problems with Artificial Intelligence Introducing Intelligence to the Semantic Analysis of Canadian Maritime Case Law: Case Based Reasoning Approach 1 Introduction 2 Problem Statement 3 Related Work 4 Machine Learning and Sentimental Analysis 5 Case-Based Reasoning Information Retrieval System (CBR-IR) 6 Conclusion References Case-Based Reasoning for the Prediction of Flash Flood 1 Introduction 2 Related Work 3 A CBR Proposal for the Flash-Flood Detection 3.1 The Representation of a Past Experience 3.2 The Three Rs CBR Process 3.3 Exploiting the CBR Tool 4 Experimentation and Discussion 5 Conclusion References Weakly Supervised Learning of the Motion Resistance of a Locomotive Powered by Liquefied Natural Gas 1 Introduction 2 Problem Statement 3 Proposed Method 3.1 Speed Model 3.2 Running Resistance Model 3.3 Slope Model 3.4 Numerical Optimization Problem 4 Numerical Results 5 Concluding Remarks and Future Work References Node Location Optimization for Localizing UAVs in Urban Scenarios 1 Introduction 2 Mathematical Model of the NLP 3 Genetic Algorithm Optimization 4 Results 5 Conclusions References Applying Deep Q-learning for Multi-agent Cooperative-Competitive Environments 1 Introduction 2 Cooperative-Competitive Multi-agent Environments 3 Single and Multi-agent Reinforcement Learning 4 Methods and Experiments 4.1 Proposed Approach 4.2 Experimental Setup 4.3 Discussion 5 Conclusions and Future Work References A Comparison of Two Speech Emotion Recognition Algorithms: Pepper Humanoid Versus Bag of Models 1 Introduction and Motivation 2 The SER Subsystems 2.1 Description of the Pepper and Its SER Subsystem 2.2 Descripcion of the SER-BOM Algorithm 3 Materials and Methods 3.1 SER Datasets 3.2 Design of the Testing Workbench 4 Numerical Results 4.1 Results for SER-Pepper 4.2 Results for SER-BoM 4.3 Comparison 5 Conclusion and Future Work References Fine-Tuning of Optimisation Parameters in a Firefly Algorithm in Inventory Management 1 Introduction 2 Related Work 3 Modelling the Inventory Management 4 Experimental Results and Discussion 4.1 Previous Research 4.2 Computational Results and Analysis 5 Conclusion and Future Work References Security Centric Scalable Architecture for Distributed Learning and Knowledge Preservation 1 Introduction 2 Related Work 3 Agricultural Use-Case for Distributed Learning 4 Functional and Non-Functional Requirements for the Distributed Learning Architecture 5 Distributed Learning System Architecture 5.1 Cloud Component Architecture 6 Architecture Validation 7 Discussions 8 Conclusions and Further Research References Author Index