Luca Rebuffi, Runyu Zhang, Xianbo Shi et al. · 2025
We present an integrated system that combines advanced wavefront diagnostics with artificial intelligence (AI) to automate and optimize X-ray optics at synchrotron beamlines. This system couples real-time wavefront sensing with AI-driven control algorithms to achieve precise beam alignment, stabilization, and performance optimization. A key feature is the use of multi-fidelity transfer learning, which enables knowledge gained from both real-world beamline optimizations and ultra-realistic digital twin simulations to be effectively applied to in situ optimization. By leveraging Multi-Objective Bayesian Optimization, the system continuously refines its performance, reducing optimization time and minimizing the need for manual adjustments. Designed for seamless deployment, it operates with existing beamline hardware and provides an intuitive graphical interface. Initial deployments at the Advanced Photon Source beamlines have demonstrated its ability to enhance beam stability, improve reproducibility, and significantly streamline alignment procedures. This AI-enhanced control framework represents a significant step toward fully autonomous beamline operation in next-generation synchrotron facilities.
Cited by 0
Ahmed Elkaseer, Islam H. Abdelgaliel, Jon Lambarri et al. · Scientific Reports · 2024
This study investigates the ablation performance of Inconel 718, a nickel-based superalloy, and metal matrix polycrystalline diamond (MMPCD), a super composite, using a nano-second (ns) pulsed laser across a range of ablation conditions. Single trenches varying in energy fluence and scanning speeds were created, analyzing the experimental responses in terms of ablation rate and surface roughness. Using regression techniques, models were developed to understand these relationships. Four multi-objective optimization algorithms, weighted value grey wolf optimizer (WVGWO), multi-objective Pareto search (MOPS), multi-objective genetic algorithm (MOGA), and multi-objective sunflower optimization (MOSFO), were employed to optimize these models. Key findings include MMPCD achieving the highest ablation rates at maximum fluence and lower speeds with negligible recast, resulting in smoother surfaces, whereas Inconel 718 reached its peak rates at similar conditions but exhibited significant surface recast. This research provides valuable insights into ns-pulsed laser machining for advanced materials, emphasizing the impact of fluence and scanning speed on achieving high ablation rates and minimal surface roughness.
Cited by 0Open access
Rebuffi, Luca, Zhang, Runyu, Shi, Xianbo et al. · Figshare · 2025
We present an integrated system that combines advanced wavefront diagnostics with artificial intelligence (AI) to automate and optimize X-ray optics at synchrotron beamlines. This system couples real-time wavefront sensing with AI-driven control algorithms to achieve precise beam alignment, stabilization, and performance optimization. A key feature is the use of multi-fidelity transfer learning, which enables knowledge gained from both real-world beamline optimizations and ultra-realistic digital twin simulations to be effectively applied to in situ optimization. By leveraging Multi-Objective Bayesian Optimization, the system continuously refines its performance, reducing optimization time and minimizing the need for manual adjustments. Designed for seamless deployment, it operates with existing beamline hardware and provides an intuitive graphical interface. Initial deployments at the Advanced Photon Source beamlines have demonstrated its ability to enhance beam stability, improve reproducibility, and significantly streamline alignment procedures. This AI-enhanced control framework represents a significant step toward fully autonomous beamline operation in next-generation synchrotron facilities.
Cited by 0Open access
Rebuffi, Luca, Zhang, Runyu, Shi, Xianbo et al. · Figshare · 2025
We present an integrated system that combines advanced wavefront diagnostics with artificial intelligence (AI) to automate and optimize X-ray optics at synchrotron beamlines. This system couples real-time wavefront sensing with AI-driven control algorithms to achieve precise beam alignment, stabilization, and performance optimization. A key feature is the use of multi-fidelity transfer learning, which enables knowledge gained from both real-world beamline optimizations and ultra-realistic digital twin simulations to be effectively applied to in situ optimization. By leveraging Multi-Objective Bayesian Optimization, the system continuously refines its performance, reducing optimization time and minimizing the need for manual adjustments. Designed for seamless deployment, it operates with existing beamline hardware and provides an intuitive graphical interface. Initial deployments at the Advanced Photon Source beamlines have demonstrated its ability to enhance beam stability, improve reproducibility, and significantly streamline alignment procedures. This AI-enhanced control framework represents a significant step toward fully autonomous beamline operation in next-generation synchrotron facilities.
Cited by 0Open access
Luca Rebuffi, Runyu Zhang, Xianbo Shi et al. · 2025
We present an integrated system that combines advanced wavefront diagnostics with artificial intelligence (AI) to automate and optimize X-ray optics at synchrotron beamlines. This system couples real-time wavefront sensing with AI-driven control algorithms to achieve precise beam alignment, stabilization, and performance optimization. A key feature is the use of multi-fidelity transfer learning, which enables knowledge gained from both real-world beamline optimizations and ultra-realistic digital twin simulations to be effectively applied to in situ optimization. By leveraging Multi-Objective Bayesian Optimization, the system continuously refines its performance, reducing optimization time and minimizing the need for manual adjustments. Designed for seamless deployment, it operates with existing beamline hardware and provides an intuitive graphical interface. Initial deployments at the Advanced Photon Source beamlines have demonstrated its ability to enhance beam stability, improve reproducibility, and significantly streamline alignment procedures. This AI-enhanced control framework represents a significant step toward fully autonomous beamline operation in next-generation synchrotron facilities.
Cited by 0
Luca Rebuffi, Runyu Zhang, Xianbo Shi et al. · 2025
We present an integrated system that combines advanced wavefront diagnostics with artificial intelligence (AI) to automate and optimize X-ray optics at synchrotron beamlines. This system couples real-time wavefront sensing with AI-driven control algorithms to achieve precise beam alignment, stabilization, and performance optimization. A key feature is the use of multi-fidelity transfer learning, which enables knowledge gained from both real-world beamline optimizations and ultra-realistic digital twin simulations to be effectively applied to in situ optimization. By leveraging Multi-Objective Bayesian Optimization, the system continuously refines its performance, reducing optimization time and minimizing the need for manual adjustments. Designed for seamless deployment, it operates with existing beamline hardware and provides an intuitive graphical interface. Initial deployments at the Advanced Photon Source beamlines have demonstrated its ability to enhance beam stability, improve reproducibility, and significantly streamline alignment procedures. This AI-enhanced control framework represents a significant step toward fully autonomous beamline operation in next-generation synchrotron facilities.
Cited by 0
Luca Rebuffi, Runyu Zhang, Xianbo Shi et al. · 2025
We present an integrated system that combines advanced wavefront diagnostics with artificial intelligence (AI) to automate and optimize X-ray optics at synchrotron beamlines. This system couples real-time wavefront sensing with AI-driven control algorithms to achieve precise beam alignment, stabilization, and performance optimization. A key feature is the use of multi-fidelity transfer learning, which enables knowledge gained from both real-world beamline optimizations and ultra-realistic digital twin simulations to be effectively applied to in situ optimization. By leveraging Multi-Objective Bayesian Optimization, the system continuously refines its performance, reducing optimization time and minimizing the need for manual adjustments. Designed for seamless deployment, it operates with existing beamline hardware and provides an intuitive graphical interface. Initial deployments at the Advanced Photon Source beamlines have demonstrated its ability to enhance beam stability, improve reproducibility, and significantly streamline alignment procedures. This AI-enhanced control framework represents a significant step toward fully autonomous beamline operation in next-generation synchrotron facilities.
Cited by 0
Luca Rebuffi, Runyu Zhang, Xianbo Shi et al. · 2025
We present an integrated system that combines advanced wavefront diagnostics with artificial intelligence (AI) to automate and optimize X-ray optics at synchrotron beamlines. This system couples real-time wavefront sensing with AI-driven control algorithms to achieve precise beam alignment, stabilization, and performance optimization. A key feature is the use of multi-fidelity transfer learning, which enables knowledge gained from both real-world beamline optimizations and ultra-realistic digital twin simulations to be effectively applied to in situ optimization. By leveraging Multi-Objective Bayesian Optimization, the system continuously refines its performance, reducing optimization time and minimizing the need for manual adjustments. Designed for seamless deployment, it operates with existing beamline hardware and provides an intuitive graphical interface. Initial deployments at the Advanced Photon Source beamlines have demonstrated its ability to enhance beam stability, improve reproducibility, and significantly streamline alignment procedures. This AI-enhanced control framework represents a significant step toward fully autonomous beamline operation in next-generation synchrotron facilities.
Cited by 0
Rebuffi, Luca, Zhang, Runyu, Shi, Xianbo et al. · Figshare · 2025
We present an integrated system that combines advanced wavefront diagnostics with artificial intelligence (AI) to automate and optimize X-ray optics at synchrotron beamlines. This system couples real-time wavefront sensing with AI-driven control algorithms to achieve precise beam alignment, stabilization, and performance optimization. A key feature is the use of multi-fidelity transfer learning, which enables knowledge gained from both real-world beamline optimizations and ultra-realistic digital twin simulations to be effectively applied to in situ optimization. By leveraging Multi-Objective Bayesian Optimization, the system continuously refines its performance, reducing optimization time and minimizing the need for manual adjustments. Designed for seamless deployment, it operates with existing beamline hardware and provides an intuitive graphical interface. Initial deployments at the Advanced Photon Source beamlines have demonstrated its ability to enhance beam stability, improve reproducibility, and significantly streamline alignment procedures. This AI-enhanced control framework represents a significant step toward fully autonomous beamline operation in next-generation synchrotron facilities.
Cited by 0Open access
Kumar Mahesh, Perumal Nallagownden, Irraivan Elamvazuthi · RePEc: Research Papers in Economics
This paper proposes an advanced Pareto-front non-dominated sorting multi-objective particle swarm optimization (Advanced-PFNDMOPSO) method for optimal configuration (placement and sizing) of distributed generation (DG) in the radial distribution system. The distributed generation consists of single and multiple numbers of active power DG, reactive power DG and simultaneous placement of active-reactive power DG. The optimization problem considers two multi-objective functions, i.e., power loss reduction and voltage stability improvements with voltage profile and power balance as constraints. First, the numerical output results of objective functions are obtained in the Pareto-optimal set. Later, fuzzy decision model is engendered for final selection of the compromised solution. The proposed method is employed and tested on standard IEEE 33 bus systems. Moreover, the results of proposed method are validated with other optimization algorithms as reported by others in the literature. The overall outcome shows that the proposed method for optimal placement and sizing gives higher capability and effectiveness to the final solution. The study also reveals that simultaneous placement of active-reactive power DG reduces more power losses, increases voltage stability and voltage profile of the system.
Cited by 0
Ashok D. Belegundu, Tirupathi R. Chandrupatla · Cambridge University Press eBooks · 2019
Organizations and businesses strive toward excellence, and solutions to problems are based mostly on judgment and experience. However, increased competition and consumer demands require that the solutions be optimum and not just feasible. Theory leads to algorithms. Algorithms need to be translated into computer codes. Engineering problems need to be modeled. Optimum solutions are obtained using theory and computers, and then interpreted. Revised and expanded in its third edition, this textbook integrates theory, modeling, development of numerical methods, and problem solving, thus preparing students to apply optimization to real-world problems. This text covers a broad variety of optimization problems using: unconstrained, constrained, gradient, and non-gradient techniques; duality concepts; multi-objective optimization; linear, integer, geometric, and dynamic programming with applications; and finite element-based optimization. It is ideal for advanced undergraduate or graduate courses in optimization design and for practicing engineers.
Cited by 0
Ashok D. Belegundu, Tirupathi R. Chandrupatla · Cambridge University Press eBooks · 2019
Organizations and businesses strive toward excellence, and solutions to problems are based mostly on judgment and experience. However, increased competition and consumer demands require that the solutions be optimum and not just feasible. Theory leads to algorithms. Algorithms need to be translated into computer codes. Engineering problems need to be modeled. Optimum solutions are obtained using theory and computers, and then interpreted. Revised and expanded in its third edition, this textbook integrates theory, modeling, development of numerical methods, and problem solving, thus preparing students to apply optimization to real-world problems. This text covers a broad variety of optimization problems using: unconstrained, constrained, gradient, and non-gradient techniques; duality concepts; multi-objective optimization; linear, integer, geometric, and dynamic programming with applications; and finite element-based optimization. It is ideal for advanced undergraduate or graduate courses in optimization design and for practicing engineers.
Cited by 1
Juhai Wang · Journal of Electrical Systems · 2024
Tactical optimization of soccer games using motion trajectory analysis algorithms involves the utilization of advanced computational techniques to analyze player movements and optimize team strategies. By applying motion trajectory analysis algorithms to tracking data obtained from sensors or video feeds, coaches and analysts can gain insights into player positioning, ball movement patterns, and team interactions during matches. This analysis helps identify tactical strengths and weaknesses, assess opponent strategies, and inform adjustments to game plans in real-time or during post-match analysis. By leveraging these algorithms, teams can refine their tactics, improve performance, and gain a competitive edge on the field. This paper introduces the Motion Trajectory Trigonometric Multi-Objective (MTTM-O) framework as a comprehensive approach to optimizing tactical decisions and predicting match outcomes in soccer. With trigonometric functions to model player trajectories and multi-objective optimization techniques, MTTM-O offers insights into the impact of tactical formations, expected goals, ball possession, and defensive solidity on match results. Through scenario analysis and multi-objective optimization, we demonstrate the effectiveness of MTTM-O in enhancing strategic decision-making and performance analysis in soccer. Findings underscore the importance of tactical flexibility, adaptation to opponent strategies, and the balance between offensive and defensive aspects of gameplay for success on the field. Through multi-objective optimization, MTTM-O evaluates various tactical formations and their impact on match results. For instance, in scenario analysis, employing a 3-5-2 formation resulted in an expected goal (xG) of 2.2 and a ball possession percentage of 62%, leading to a decisive win. Conversely, a 4-4-2 formation yielded lower xG (0.8) and ball possession (47%), resulting in a draw. These findings highlight the importance of tactical flexibility and adaptation to opponent strategies for success on the field.
Cited by 2Open access
Murali Krishna Pasupuleti · 2024
Abstract: This chapter explores how AI-powered innovation is accelerating the design and discovery of advanced materials, transforming traditional approaches in materials science. By leveraging machine learning algorithms and data-driven models, AI enables researchers to predict material properties, optimize synthesis processes, and expedite the discovery of novel materials with enhanced performance. The chapter delves into AI-driven techniques for high-throughput screening, inverse design, and multi-objective optimization in fields such as energy storage, aerospace, and healthcare. It also discusses the integration of AI with existing computational methods like molecular dynamics and density functional theory (DFT) to further advance materials research. Ethical considerations, challenges in data quality, and the future potential of AI in revolutionizing materials science are also addressed. Keywords: AI-powered materials innovation, machine learning, materials discovery, advanced materials, high-throughput screening, inverse design, multi-objective optimization, energy storage, aerospace, healthcare, molecular dynamics, density functional theory, AI in materials science, predictive modeling, material synthesis.
Cited by 5Open access
Mingshan Chen, Weichao Ding, Mengyang Zhu et al. · Processes · 2024
Container technology has gained a widespread application in cloud computing environments due to its low resource overhead and high flexibility. However, as the number of containers grows, it becomes increasingly challenging to achieve the rapid and coordinated optimization of multiple objectives for container scheduling, while maintaining system stability and security. This paper aims to overcome these challenges and provides the optimal allocation for a large number of containers. First, a large-scale multi-objective container scheduling optimization model is constructed, which involves the task completion time, resource cost, and load balancing. Second, a novel optimization algorithm called LSMOF-AD (large-scale multi-objective optimization framework with muti-stage and adaptive differential strategies) is proposed to effectively handle large-scale container scheduling problems. The experimental results show that the proposed algorithm has a better performance in multiple benchmark problems compared to other advanced algorithms and can effectively reduce the task processing delay, while achieving a high resource utilization and load balancing compared to other scheduling strategies.
Cited by 1Open access
Cajander, Niclas · Aaltodoc (Aalto University) · 2025
Multi-objective reinforcement learning (MORL) addresses decision-making problems where multiple, possibly conflicting objectives must be optimized simultaneously. Unlike single-objective reinforcement learning, MORL employs vector-valued rewards to capture the trade-offs between objectives, necessitating advanced theoretical frameworks and algorithms. This thesis explores the gaps in MORL theory and algorithms for learning problems, focusing on fully observable, single-task, single agent Multi-objective Markov Decision Processes (MOMDPs) and the challenges posed by different utility functions and optimality criteria. The thesis examines key theoretical aspects, including solution sets such as the Pareto Front (PF) and Convex Coverage Set (CCS), and their dependence on utility functions and policy types. The thesis discusses single-policy algorithms and highlights their limitations, particularly for the understudiedESRcriterion, whichis crucialforscenarios requiring optimization over single episodes, such as medical treatments. The thesis also reviews multi-policy algorithms, categorizing them into outer-loop and inner-loop methods, and discusses their applicability to real-world problems. Crucial aspects include sample-efficiency, scalability with respect to the number of objectives, and scalability of state-action spaces. It is clear that the ESR criterion is still understudied, as very few algorithms have been proposed. Moreover, as no solution set has been defined for ESR, there only exist single-policy algorithms for the criterion. There exist many multi-policy methods with clear strengths, but they are often limited by some other factor.
Cited by 0Open access
Takumi Sonoda, Masaya Nakata · 2021
Surrogate-assisted multi-objective evolutionary algorithms have advanced the field of computationally expensive optimization, but their progress is often restricted to low-dimensional problems. This manuscript presents a multiple classifiers-assisted evolutionary algorithm based on decomposition, which is adapted for high-dimensional expensive problems in terms of the following two insights. Compared to approximation-based surrogates, the accuracy of classification-based surrogates is robust for few high-dimensional training samples. Further, multiple local classifiers can hedge the risk of over-fitting issues. Accordingly, the proposed algorithm builds multiple classifiers with support vector machines on a decomposition-based multi-objective algorithm, wherein each local classifier is trained for a corresponding scalarization function. Experimental results statistically confirm that the proposed algorithm is competitive to the state-of-the-art algorithms and computationally efficient as well.
Cited by 3Open access
Takumi Sonoda, Masaya Nakata · 2021
Surrogate-assisted multi-objective evolutionary algorithms have advanced the field of computationally expensive optimization, but their progress is often restricted to low-dimensional problems. This manuscript presents a multiple classifiers-assisted evolutionary algorithm based on decomposition, which is adapted for high-dimensional expensive problems in terms of the following two insights. Compared to approximation-based surrogates, the accuracy of classification-based surrogates is robust for few high-dimensional training samples. Further, multiple local classifiers can hedge the risk of over-fitting issues. Accordingly, the proposed algorithm builds multiple classifiers with support vector machines on a decomposition-based multi-objective algorithm, wherein each local classifier is trained for a corresponding scalarization function. Experimental results statistically confirm that the proposed algorithm is competitive to the state-of-the-art algorithms and computationally efficient as well.
Cited by 16Open access
DongSeop Lee, Jacques Périaux, Felipé Gonzalez · 2009
This paper presents the application of advanced optimization techniques to unmanned aerial system mission path planning system (MPPS) using multi-objective evolutionary algorithms (MOEAs). Two types of multi-objective optimizers are compared; the MOEA nondominated sorting genetic algorithm II and a hybrid-game strategy are implemented to produce a set of optimal collision-free trajectories in a three-dimensional environment. The resulting trajectories on a three-dimensional terrain are collision-free and are represented by using Bezier spline curves from start position to target and then target to start position or different positions with altitude constraints. The efficiency of the two optimization methods is compared in terms of computational cost and design quality. Numerical results show the benefits of adding a hybrid-game strategy to a MOEA and for a MPPS.
Cited by 0
António Gaspar‐Cunha, João Melo, Tomás Marques et al. · Polymers · 2025
Plastic injection molding is a fundamental manufacturing process used in various industries, accounting for approximately 30% of the global plastic product market. A significant challenge of this process lies in the need to employ sophisticated computational techniques to optimize the various phases. This review examines the optimization methodologies in injection molding, with a focus on integrating advanced modeling, surrogate models, and multi-objective optimization techniques to enhance efficiency, quality, and sustainability. Key phases such as plasticizing, filling, packing, cooling, and ejection are analyzed, each presenting unique optimization challenges. The review emphasizes the importance of cooling, which accounts for 50-80% of the cycle time, and examines innovative strategies, such as conformal cooling channels (CCCs), to enhance uniformity and minimize defects. Various computational tools, including Moldex3D and Autodesk Moldflow, are discussed due to their role in process simulation and optimization. Additionally, optimization algorithms such as evolutionary algorithms, simulated annealing, and multi-objective optimization methods are explored. The integration of surrogate models, such as Kriging, response surface methodology, and artificial neural networks, has shown promise in addressing computational cost challenges. Future directions emphasize the need for adaptive machine learning and artificial intelligence techniques to optimize molds in real time, offering more innovative and sustainable manufacturing solutions. This review is a comprehensive guide for researchers and practitioners, bridging theoretical advancements with practical implementation in injection molding optimization.
Cited by 9Open access