E-Book, Englisch, Band 1, 228 Seiten
Dutta / Mandal / Cengiz Experimental Design of Bio-Inspired Algorithms for Optimization Problems in Industry 5.0
1. Auflage 2026
ISBN: 979-8-89881-408-3
Verlag: De Gruyter
Format: EPUB
Kopierschutz: 0 - No protection
E-Book, Englisch, Band 1, 228 Seiten
Reihe: Applied Machine Learning for IoT and Data Analytics
ISBN: 979-8-89881-408-3
Verlag: De Gruyter
Format: EPUB
Kopierschutz: 0 - No protection
Applied Machine Learning for IoT and Data Analytics (Volume 1) is an integrated exploration of nature-inspired optimisation techniques within the emerging Industry 5.0 paradigm- Positioned at the intersection of artificial intelligence, computational intelligence, industrial engineering, and cyber-physical systems, this volume centres on human-centricity, sustainability, resilience, and intelligent automation.
The book comprehensively reviews evolutionary computation, swarm intelligence, neural computation, and hybrid metaheuristics, explaining how these methods can be systematically designed, statistically validated, and benchmarked for real-world deployment. Foundational chapters address Explainable AI (XAI), statistical experimental design, ANOVA-based modelling, parameter tuning strategies, and performance evaluation frameworks.
Through fifteen carefully curated chapters, the book presents practical case studies in wireless sensor networks, smart manufacturing, micro-machining, welding optimisation, renewable energy systems, motor control, wireless communications, banking automation, and advanced antenna design. Emphasis is placed on experimental rigour, benchmarking, and reproducibility-bridging the gap between theoretical advancements and industrial implementation.
Key Features:
-Comprehensive review of classical and hybrid bio-inspired algorithms.
-Integration of optimisation techniques within the Industry 5.0 framework
-Covers Explainable AI for transparent optimisation systems with a strong focus on experimental design, ANOVA modelling, and statistical validation.
-Practical case studies across manufacturing, energy, communications, and automation
-Emphasis on reproducibility and methodological rigour with forward-looking insights into AI-enhanced and explainable optimisation trends.
Autoren/Hrsg.
Weitere Infos & Material
PREFACE
Advances in robots, artificial intelligence, and cyber-physical systems have accelerated the evolution of industrial systems, opening the door for the emergence of Industry 5.0, a paradigm change that prioritizes resilient, sustainable, and human-centric production. Robust optimization strategies become increasingly important as companies shift toward intelligent automation, energy efficiency, and individualized manufacturing. In this regard, bio-inspired algorithms have drawn a lot of interest because of their versatility, capacity for self-organization, and efficiency in resolving challenging optimization problems.
To close the gap between theoretical developments and real-world applications of bio-inspired optimization techniques in the changing industrial environment, this book, "Experimental Design of Bio-Inspired Algorithms for Optimization Problems in Industry 5.0," was developed. It offers a thorough manual for creating, putting into practice, and assessing these algorithms for practical uses in supply chain management, smart manufacturing, human-robot cooperation, predictive maintenance, and other fields.
Leading authorities in the domains of neural computation, swarm intelligence, evolutionary computing, and industrial engineering have contributed to the edited edition. With an emphasis on crucial elements, including parameter tweaking, performance assessment, benchmarking, and scalability, each chapter is thoughtfully written to provide insights into the experimental design features of bio-inspired algorithms. The integration of these algorithms with contemporary computational frameworks is also covered in the book, providing useful answers for both scholars and practitioners.
This book's salient characteristics include:
• A comprehensive examination of bio-inspired algorithms designed for industrial use, such as genetic algorithms, particle swarm optimization, ant colony optimization, and hybrid approaches.
• Recommendations for creating exacting experimental frameworks that guarantee robustness, repeatability, and dependability in optimization solutions.
• Real-world examples and case studies that show how bio-inspired approaches may be used to address Industry 5.0 issues.
• New developments and potential lines of inquiry, such as the application of AI-enhanced optimization in industrial systems.
Chapter 1 explains the way how Artificial Intelligence (AI) has rapidly advanced and changed many industries, ranging from healthcare and finance to manufacturing and entertainment, where people began to interact and process information with machines differently. However, despite the potential offered by AI, some challenges have also emerged. For instance, the black box issue, which is the aspect of AI where many systems, and especially those that rely on complex algorithms such as deep learning, are hard to understand by the human race. The difficulty in understanding how decisions are made and how models work brings worries about trust, fairness, responsibility, and openness in AI systems, especially in situations where they are of critical importance, for instance, in medical diagnosis, autonomous vehicles, or legal matters. This is where Explainable AI (XAI) comes in to remediate this problem and help make decisions more transparently without suppressing the functioning of the AI systems. XAI not only improves the performance of AI systems in healthcare but also fosters trust between patients and providers in Industry 5.0, where human-AI collaboration is crucial.
Chapter 2 covers extending the techniques of ANOVA in optometric research to different experimental designs. In each case, the type of experimental design is described, a statistical model is given, and the advantages and limitations of the appropriate ANOVA are discussed. In addition, the problems of non-conformity to the statistical model and the determination of the number of replications are considered.
Chapter 3 investigates a power series nonlinear mathematical model based on an enhanced Elephant swarm water Search algorithm (ESWSA). The proposed ESWSA algorithm has been developed to predict the average localization error (ALE) of a wireless sensor network system. The performance of the proposed algorithm is evaluated from the effects of anchor ratio, transmission range, node density, and number of iterations. Experimental results demonstrate that higher localization precision with fewer iterations can be achieved. This chapter presents the suggested approach in real-world WSN applications as a foundation for optimizing node location, which is advantageous to lower computational consumption and thereby extend the lifetime of sensor nodes.
Chapter 4 provides a mathematical model of the grinding process based on the input parameters (i.e., infeed, flow rate, and presence of scrapper board) using Regression Equations and ANOVA analysis. The model has been developed from the experimental data of the up-grinding mode with or without a scraperboard. The relationship between response and the grinding parameters has been illustrated graphically by contour plots and surface plots. The output variables are optimized to find the best suitable conditions for grinding using ANOVA. The aim is to find the optimal parametric conditions at which output variables will be minimized.
Chapter 5 analyzes a nonlinear optimization issue for powder injection molding (PIM) of Alumina feedstock, which is addressed by using the artificial neural network (ANN) algorithm. The Jaya algorithm is used to solve the problem of injection variables in a mold-filling optimization issue. The ANN model is implemented with the help of analysis of variance (ANOVA) based on input and response variables. In the last stage, the optimum process parameters are obtained, which show a satisfactory output. Future research may concentrate on creating a posteriori version of the Jaya method, which could then be used to tackle optimization issues for various conventional and contemporary machining processes while taking into account several goals at once.
Chapter 6 demonstrates that a micro-blind hole generation on silica by electrochemical deposition (ECDM) is a very promising method for micro-hole generation on glass. To achieve the target results of the work and to control the process parameters such as machining voltage, pulse frequency (fixed 50Hz), pulse on-time inter-electrode gap (IEG), duty ratio, electrolyte concentration, etc., a series of experiments and analyses are performed to achieve the target results during the micro-ECDM process. The results show that applied voltage leads to a critical voltage that takes action mainly on diametric overcut (DOC) and tool wear rate (TWR).
Chapter 7 presents a micro-drilling of glass by the electrochemical discharge machining (ECDM) process, which is a challenging process due to its high material removal rate and radial overcut during the drilling process. The analysis of variance (ANOVA) is conducted to find out the significant process parameters for the above-mentioned responses of the ECDM process. A mathematical model for Material Removal Rate (MRR) and Radial Overcut (ROC) is developed using the factorial design method. The effects of two process parameters, applied voltage and electrolyte conductivity, on respective machining criteria are justified through ANOVA results and experimental analyses.
Chapter 8 explains Gas Metal Arc (MIG) welding of austenitic stainless steel AISI 304 using ER 308L electrode and 99% pure Argon gas shielding at suitable combinations of input parameters that have been considered for research. A second-order model is developed to represent the responses in terms of the process variables, including welding current (I), gas flow rate (F), and welding speed (S). The relationship between the responses and welding parameters is depicted graphically using contour plots and surface plots. A total of twenty welded specimens are fabricated under varying parametric conditions using the Response Surface Methodology (RSM) with a face-centered central composite design (CCD). ANOVA and ESWSA metaheuristics are applied for the optimization of the welding process, i.e., to find out the optimal welding condition for MIG welding.
Chapter 9 discusses speed control of a 4-phase, 8/6 switched reluctance motor (SRM) powered by photovoltaic. The high-gain converter is controlled by using a Maximum Power Point Tracking (MPPT) technique by implementing an Adaptive Neuro Fuzzy Inference System (ANFIS). The ANFIS-MPPT algorithm generates a control signal and maximizes the efficiency of PV systems using input voltage (VPV) and current (IPV). In addition, a PI and an optimization algorithm for fine-tuning the parameters of the PI result in regulating the speed of SRM.
Chapter 10 explores an experimental framework for bio-inspired algorithms in wireless communication at the physical layer, supporting various components throughout the system. The proposed wireless experimental environment can construct local, small- or medium-sized internal communication systems for factory scenes and provide better understanding and verification in high-definition and high-precision industrial case instruments. The paper focuses on the technical merit and potential of Biological Inheritance (BIA), which not only achieves the same performance as a parallel counterpart scheme for discrete signal data transmission, but also achieves point-to-point signal data communication. BIA's quick operability and the importance of the annealing process in ensuring correct behavior in complex future applications are...
