MULTI-OBJECTIVE OPTIMIZATION OF ADDITIVE MANUFACTURING PARAMETERS FOR DEFECT AND ENERGY REDUCTION

Abstract

Additive Manufacturing (AM) has emerged as one of the most transformative technologies in modern manufacturing due to its ability to fabricate complex geometries with minimal material waste, rapid prototyping capabilities, and high design flexibility. Technologies such as Selective Laser Melting (SLM), Laser Powder Bed Fusion (LPBF), Direct Metal Laser Sintering (DMLS), Electron Beam Melting (EBM), and Fused Deposition Modeling (FDM) are increasingly being adopted in aerospace, automotive, biomedical, energy, defense, and precision engineering industries. Despite these advantages, additive manufacturing processes often suffer from multiple qualityrelated challenges including porosity, lack of fusion, residual stresses, surface roughness, dimensional inaccuracies, cracking, and excessive energy consumption. These defects significantly influence the mechanical properties, reliability, production cost, and sustainability of manufactured components. Since process parameters such as laser power, scanning speed, hatch spacing, layer thickness, build orientation, nozzle temperature, print speed, and infill density simultaneously affect product quality and energy utilization, optimizing these parameters has become one of the most important research challenges in intelligent manufacturing. This research proposes a comprehensive multi-objective optimization framework for additive manufacturing that simultaneously minimizes manufacturing defects and energy consumption while maximizing part quality and production efficiency. The proposed framework integrates Internet of Things (IoT)-enabled manufacturing systems, machine learning models, real-time process monitoring, and advanced multi-objective optimization algorithms to determine the optimal combination of additive manufacturing process parameters. Manufacturing data are continuously collected from IoT-enabled sensors monitoring laser characteristics, melt pool temperature, chamber conditions, power consumption, build platform status, environmental parameters, and machine operating conditions. These real-time sensor data are integrated with historical production information to develop predictive models capable of estimating defect formation and energy requirements under different process conditions. The proposed framework employs advanced machine learning algorithms including Random Forest, Support Vector Machine (SVM), Artificial Neural Network (ANN), Gradient Boosting, Extreme Gradient Boosting (XGBoost), and Long Short-Term Memory (LSTM) networks to predict manufacturing defects, energy consumption, surface quality, dimensional accuracy, and production performance. The predictive models are coupled with multi-objective optimization techniques such as Non- dominated Sorting Genetic Algorithm II (NSGA-II), Multi-Objective Particle Swarm Optimization (MOPSO), and Pareto-based optimization methods to identify optimal process parameter combinations that satisfy multiple manufacturing objectives simultaneously. Unlike conventional singleobjective optimization approaches, the proposed framework generates Paretooptimal solutions that provide manufacturers with flexible trade-offs between energy efficiency, production quality, productivity, and manufacturing cost. Experimental evaluation demonstrates that the proposed optimization framework significantly reduces porosity, residual stress, surface roughness, and dimensional deviation while simultaneously lowering energy consumption and improving machine utilization compared with traditional parameter selection methods. Continuous feedback from IoT sensors enables adaptive optimization by dynamically updating machine learning models and optimization strategies based on real-time manufacturing conditions. The proposed framework supports intelligent decision-making, sustainable manufacturing, predictive quality control, and autonomous process optimization within Industry 4.0 smart manufacturing environments. Consequently, the research contributes to the development of energy-efficient, high-quality, and sustainable additive manufacturing systems capable of meeting the increasing demands of modern industrial production. Keywords: Additive Manufacturing, MultiObjective Optimization, Process Parameter Optimization, Machine Learning, Internet of Things, Energy Consumption, Manufacturing Defects, Industry 4.0, Smart Manufacturing, Predictive Analytics.