Multi-objective Optimization Method Research Articles

Multi-objective fuzzy green scheduling optimization method of special vehicle body-in-white prototype shop considering equipment preventive maintenance

Based on the advantages of non-contact, flexibility, and small heat-affected zone, laser processing technology is widely used in material removal, addition, modification and other manufacturing fields. However, the emission of large amounts of smoke during laser processing of metal materials will result in environmental pollution in the job shop. Additionally, the processing time of laser equipment becomes uncertain as the power decays with prolonged use. Therefore, this paper is mainly focus on a multi-objective flexible job shop fuzzy green scheduling problem to minimize the fuzzy makespan, fuzzy total energy consumption and fuzzy total smoke emission. To solve it, an improved multi-objective particle swarm optimization algorithm has been designed. Firstly, a preventive maintenance strategy is proposed to reduce the makespan and equipment failure frequency by considering periodic power attenuation of laser equipment. Next, different particle updating strategies are employed to enhance particles’ exploration capabilities. Subsequently, five neighborhood structures are introduced to improve the performance of local search. Finally, comparison experiments are conducted with an expanded common benchmark, and a production case study in a special vehicle body-in-white prototype job shop. The results effectively demonstrate the feasibility and effectiveness of the developed model and improved multi-objective particle swarm optimization algorithm.

An efficient parallel multi-fidelity multi-objective Bayesian optimization method and application to 3-stage axial compressor with 144 variables

Parallel infill sampling is a promising approach to improve the efficiency of multi-fidelity multi-objective Bayesian optimization (MOBO). In existing literature, the number of infill samples per iteration is typically limited to 10. Additionally, the application of the multi-fidelity MOBO method in engineering optimization designs with over 100 variables is rare. To that end, a novel generalized expected improvement matrix (GEIM) criterion is proposed by using generalized reference values for the element in expected improvement matrix. Parallel infill sampling strategy based on GEIM is developed and incorporated into the multi-fidelity MOBO framework. The distinct feature of the proposed method is that the number of infill samples at each iteration can be significantly larger than existing methods (i.e. as much as 100), so as to further enhance the optimization efficiency. Empirical experiment results on analytic problems show that the proposed parallel multi-fidelity MOBO method is highly competitive in comparison with existing methods. To address the curse of dimensionality in optimizing multi-stage axial compressor, an efficient modeling method for the Hierarchical Kriging (HK) is incorporated. The HK predictor for the efficiency of the 3-stage compressor with 144 variables is built in 16.71 s with acceptable accuracy. Remarkable improvements over the two objectives of the 3-stage compressor with 144 variables are achieved simultaneously within 870 high-fidelity CFD simulations.

Dynamic multi-layer walkability model for transit-oriented movement: Nodes and Routes optimization method

This paper presents a novel approach to optimizing movement in urban areas through a dynamic multi-layer walkability model. This research uncovers new facets of walkability modeling within transit-oriented movement, aligning pedestrian pathways (Routes) with urban architecture, public spaces, and green spaces (Nodes). The dynamic multi-layer (DML) approach involves optimizing both the Nodes and Routes of the transit system using a multi-objective optimization method. This method improves accessibility and connectivity by aggregating the results of agent-based modeling for route simulations and considers multiple criteria, including greenness, distance to transit, and destination accessibility. Experimentation with a case study produced several findings that underscore the value of multi-layered models for transit movement and the power of computational methods in optimizing both Nodes and Routes. This discovery offers valuable insights into the DML process and its potential applications in the field of urban design and architecture.

Evaluation of cutting parameters during dry turning of AISI 304 stainless steel and optimization through a modified WPCA approach

The main objective of this study is to analyze how input parameters such as cutting speed ( Vc), depth of cut ( ap), feed per revolution ( f), and tool nose radius ( r) influence the performance of dry turning AISI 304 stainless steel. Performance evaluation includes cutting force components ( Fx, Fy, and Fz), surface roughness (SR), material removal rate (MRR), cutting power ( Cp), and cutting temperature ( T). Experiments are planned using the Taguchi L18 experimental design, and a coated carbide tool (GC 2025) is employed. ANOVA revealed that the most influential factors on SR are f and r, contributing 52.40% and 26.54%, respectively. Axial and tangential cutting force components are influenced by ap and f, with contributions of 72.99% and 5.93% for Fx and 81.57% and 13.38% for Fz, respectively. The radial component is influenced by ap (85.74%). Vc, f, and ap all contribute to cutting power, with contributions of 55.13%, 24.35%, and 10.74%, respectively. Finally, ap and the interaction f*ap contribute 45.18% and 8.81%, respectively, to the temperature in the cutting zone. This article introduces a novel multi-objective optimization method based on weighted principal component analysis. This method distinguishes itself by its ability to address challenges related to data matrix normalization, the ambiguity of eigenvector signs, and the issue of negative signs in Taguchi’s philosophy. The results of the new method, the desirability function, and the classical WPCA method are compared. The results show that the new method suggests a compromise solution that is similar to the desirability function and better than the classical WPCA’s suggestion.

A Probability-based Fuzzy Multi-objective Optimization for Material Selection

Abstract: In the present paper, a rational fuzzy multi-objective optimization for material selection is developed in respect of probabilistic method for multi-objective optimization, which agrees with the viewpoint of system theory for the whole optimization of a system and is the novelty of this work. The basic ideas and algorithms of fuzzy theory together with probability theory are taken as the cornerstone to perform the formulation. In the treatment, the intersection of the membership function of fuzzy numbers of alternative material performance and the membership function of fuzzy numbers of desired material performance is used as the utility of the material performance index. Thereafter, the utility of each material performance index is further used to conduct the assessment of its partial preferable probability and formulate the multi-objective optimization by means of probability theory. Moreover, a typical example is presented to provide the rational process of the probabilistic fuzzy multi-objective optimization for material selection.

A deep learning-aided multi-objective optimization of a downstream process for production of monoclonal antibody products

Monoclonal antibodies (mAbs) have emerged as the dominant class of therapeutic moieties due to their superior therapeutic potential and safety profile. Downstream processing plays a pivotal role in ensuring the purity, yield, and quality of these therapeutic products. The biopharmaceutical industry generates a significant amount of data, making it an attractive target for creating machine learning (ML) and deep learning (DL) applications. This paper reports a DL-based 2D-convolutional neural network (2D-CNN) to predict: (i) Protein A mAb elute concentration, (ii) Cation exchange (CEX) mAb elute concentration, (iii) % Acidic variant, (iv) Aggregate (%), and (v) % Basic variant from process data that is routinely collected during biopharma manufacturing, including Protein A elution UV, CEX elution UV, CEX elution conductivity, and post viral inactivation pH profiles. Further, interpretation of predicted outputs using the SHAPLEY technique and optimization through multi-objective Bayesian optimization methods have been attempted so as to match the biosimilar to the reference molecule with respect to the critical quality attributes. It is observed that the proposed model outperforms other published approaches in prediction. The optimum results were experimentally validated by performing three consecutive runs, and a mean percentage deviation of less than 3 % was reported. The proposed approach demonstrates the feasibility of real-time prediction and optimization in biopharmaceutical product manufacturing using CNN-aided modeling and multi-objective weighted sum optimization.

Optimization of a wind farm layout to mitigate the wind power intermittency

A multi-objective optimization method utilizing genetic algorithms is developed to optimize wind farm layout design with the dual objectives of enhancing the production of wind power and reducing hour-level intermittency in the generated power. This method introduces a novel metric for annual wind power intermittency based on the transition probability of wind conditions (i.e., speed and direction). The present multi-objective optimization method ensures that Pareto optimal layouts are evenly distributed according to multiple cost function values. To mitigate wind power intermittency, the spatial distribution of wake fields is strategically manipulated to counteract hourly fluctuations in wind conditions. In wind conditions favoring high power generation, it is found that increasing the overlap between wakes and turbines helps to minimize disparities in generated power compared to other wind conditions.

A variable stiffness design method for soft robotic fingers based on grasping force compensation and linearization

Abstract Soft fingers play an increasingly important role in robotic grippers to achieve adaptive grasping with variable stiffness features. Previous studies of soft finger design have primarily focused on the optimization of the structural parameters of existing finger structures, but limited efforts have been put into the design methodology from fundamental grasping mechanisms to finger structures with desired grasping force features. To this aim, a fundamental architecture of soft fingers is proposed for analyzing common soft finger features and the influence of the internal structures on the overall grasping performance. In addition, three general performance metrics are introduced to evaluate the grasping performance of soft finger designs. Then, a novel method is proposed to combine the variable stiffness structure with the fundamental architecture to compensate for the grasping force of the finger and linearization. Subsequently, an embodiment design is proposed with a cantilever spring-based variable stiffness (CSVS) mechanism based on the method, and a multi-objective optimization method is employed to optimize the design. Besides, the CSVS features are analyzed through finite element analysis (FEA), and by comparing the grasping performance between the V-shape finger and the CSVS finger, it is demonstrated that the design method can effectively shorten the pre-grasp stage and linearize the grasping force in the post-grasp stage while reducing the likelihood of sliding friction between the finger and the grasped object. Finally, experiments are conducted to validate the accuracy of the FEA model, the effectiveness of the design methodology, and the adaptability of the CSVS finger.

A DYNAMIC PERFORMANCE-BASED PAYMENT MECHANISM FOR PUBLIC-PRIVATE PARTNERSHIP PROJECTS: AN INTEGRATED MODEL FOR PRINCIPAL-AGENT AND MULTI-OBJECTIVE OPTIMIZATION MODELS

Performance-based payment mechanism is one of the key issues to ensure all stakeholders’ benefits in infrastructure Public-Private Partnership (PPP) projects. However, most existing research on performance-based payment with a fixed incentive coefficient can’t play a good incentive role. This study aims to the intrinsic mechanism between the performance appraisal score and the performance-based payment structure, so as to design the dynamic performance-based payment mechanism for infrastructure PPP projects. Firstly, the multi-objective optimal method is used to calculate the unit-payment. Second, principal-agent theory is used to construct the performance-based payment model with a changeable incentive coefficient. The findings of this study show that, the performance-based payment mechanism can effectively motivate participants to provide high-quality and efficient services, because their remuneration directly depends on their performance. When the outcome does not meet expectations, the amount paid can be adjusted accordingly, thus ensuring the maximum protection of public resources as well as the private sector’s profits. They serve a dual purpose, on one hand, they offer insights to rectify the shortcomings in the current unsatisfactory payment structure. On the other hand, the study provides a theoretical reference for the public sector to effectively incentivize the private sector in enhancing project performance.

Research on Multi-Objective Process Parameter Optimization Method in Hard Turning Based on an Improved NSGA-II Algorithm

To address the issue of local optima encountered during the multi-objective optimization process with the Non-dominated Sorting Genetic Algorithm II (NSGA-II) algorithm, this paper introduces an enhanced version of the NSGA-II. This improved NSGA-II incorporates polynomial and simulated binary crossover operators into the genetic algorithm’s crossover phase to refine its performance. For evaluation purposes, the classic ZDT benchmark functions are employed. The findings reveal that the enhanced NSGA-II algorithm achieves higher convergence accuracy and surpasses the performance of the original NSGA-II algorithm. When applied to the machining of the high-hardness material 20MnCrTi, four algorithms were utilized: the improved NSGA-II, the conventional NSGA-II, NSGA-III, and MOEA/D. The experimental outcomes show that the improved NSGA-II algorithm delivers a more optimal combination of process parameters, effectively enhancing the workpiece’s surface roughness and material removal rate. This leads to a significant improvement in the machining quality of the workpiece surface, demonstrating the superiority of the improved algorithm in optimizing machining processes.

Streamlining Computational Fragment-Based Drug Discovery through Evolutionary Optimization Informed by Ligand-Based Virtual Prescreening.

Recent advances in computational methods provide the promise of dramatically accelerating drug discovery. While mathematical modeling and machine learning have become vital in predicting drug-target interactions and properties, there is untapped potential in computational drug discovery due to the vast and complex chemical space. This paper builds on our recently published computational fragment-based drug discovery (FBDD) method called fragment databases from screened ligand drug discovery (FDSL-DD). FDSL-DD uses in silico screening to identify ligands from a vast library, fragmenting them while attaching specific attributes based on predicted binding affinity and interaction with the target subdomain. In this paper, we further propose a two-stage optimization method that utilizes the information from prescreening to optimize computational ligand synthesis. We hypothesize that using prescreening information for optimization shrinks the search space and focuses on promising regions, thereby improving the optimization for candidate ligands. The first optimization stage assembles these fragments into larger compounds using genetic algorithms, followed by a second stage of iterative refinement to produce compounds with enhanced bioactivity. To demonstrate broad applicability, the methodology is demonstrated on three diverse protein targets found in human solid cancers, bacterial antimicrobial resistance, and the SARS-CoV-2 virus. Combined, the proposed FDSL-DD and a two-stage optimization approach yield high-affinity ligand candidates more efficiently than other state-of-the-art computational FBDD methods. We further show that a multiobjective optimization method accounting for drug-likeness can still produce potential candidate ligands with a high binding affinity. Overall, the results demonstrate that integrating detailed chemical information with a constrained search framework can markedly optimize the initial drug discovery process, offering a more precise and efficient route to developing new therapeutics.

Optimizing electricity generation through the utilization of wasted heat from process streams in a LURGI methanol plant

In the realm of petrochemical operations, persistent efforts have been made to curtail specific energy consumption; however, certain heat sources continue to be underutilized. This paper investigates the condition of a Lurgi methanol plant to identify process streams with enough wasted heat that have the potential to produce electricity. After a thorough analysis, one particular zone is identified where an air cooler package and seawater exchanger are used to decrease the temperature of crude methanol vapor. To harness this untapped heat for electricity generation, an organic Rankine cycle (ORC) is proposed. In the quest for efficient electricity production, three distinct working fluids-dry (R600a), wet (R134a), and isentropic (R11)- are scrutinized to compare their performance in generating electricity within the ORC system. In addition, the maximum amount of electricity that can be generated from this waste heat recovery (WHR) project is determined and optimized using a multi-objective optimization method. By leveraging genetic algorithms, the system's exergy is enhanced while minimizing costs. A comprehensive economic comparison is conducted using probability analysis to evaluate each system's financial viability. The results show that dry and isentropic working fluids yield the best results for electricity production, generating approximately 9–10 MW. The return on investment (ROI) for these working fluids is also nearly similar, at 16 %. Among the chosen working fluids, R600a is selected as the superior option due to its lower outlet temperature for crude methanol. Additionally, because the cooling water flow for the condenser of the ORC is identified as a limitation in the current petrochemical plant, a new optimization is conducted with a constraint on the flow of cooling water, aiming for approximately 3000 tons/h. The results show that the proposed WHR-ORC system can generate a maximum of approximately 7.5 MW net electricity in the studied petrochemical plant, while also reducing CO2 emissions by up to 40,000 tons per year. The innovative methodology showcased in this research underscores its economic viability, paving the way for heightened energy efficiency and environmentally conscious methanol production practices.

Convergence of a new nonmonotone memory gradient method for unconstrained multiobjective optimization via robust approach

Convergence of a new nonmonotone memory gradient method for unconstrained multiobjective optimization via robust approach

An inexact nonmonotone projected gradient method for constrained multiobjective optimization

An inexact nonmonotone projected gradient method for constrained multiobjective optimization

Design and multi-criteria optimization and financial assessment of an innovative combined power plant and desalination process

Geothermal power generation programs are crucial due to their long-term sustainability and environmentally friendly characteristics, particularly for potential energy-based applications in the future. Innovative designs can yield improve efficien by mitigating irreversibility, resulting in enhanced power production. Hence, this study investigates an innovative cascade thermal design model that employs geothermal resources to enhance the efficiency and output of low-temperature power generation systems, specifically through the integration of organic flash and organic Rankine cycles. Additionally, this model introduces a novel cogeneration approach by incorporating a desalination unit based on humidification-dehumidification processes, aiming to concurrently produce electricity and fresh water. Employing a combination of thermodynamic analysis and financial assessment, the system's performance was simulated and optimized in MATLAB, using the Multi-Objective Particle Swarm Optimization (MOPSO) method. A sensitivity analysis preceded the optimization, leading to the development of three distinct optimization scenarios focused on balancing power and freshwater production, maximizing exergetic efficiency and net present value, and optimizing fixed capital investment for maximum financial viability. The results highlight the second scenario as particularly effective, achieving a power generation of 254.3 kW, an exergetic efficiency of 44.84 %, and a net present value of $405,099. Conversely, the third scenario offers the best balance between freshwater production capacity (0.504 kg/s), fixed capital investment ($820,822), and a payback period of 7.12 years. This research demonstrates the potential of integrating advanced thermal models with geothermal resources for sustainable and efficient energy and freshwater production, marking a significant step forward in the development of eco-friendly cogeneration systems.

Integrated design optimization of ducted lift fan for lift enhancement and noise reduction based on nested NFFD control body

In order to enhance the aerodynamic performance and reduce the aerodynamic noise of the coaxial dual-rotor ducted fan, this paper proposes a stepwise optimization strategy based on nested NFFD (Nurbs-Based Free-Form Deformations) control body parameterization. This strategy takes into account the optimization of aerodynamic performance, which involves significant deformations, and noise reduction, which requires fine deformations. By doing so, it effectively addresses the issue of drastically increased computational complexity and time consumption associated with conventional multi-objective optimization methods. By designing the aerodynamic shape of the blades, along with aerodynamic performance parameter calculation and aerodynamic noise calculation modules, it is possible to achieve lift enhancement and noise reduction for the ducted fan. The numerical simulation results demonstrate that the flow field over the optimized blade surface is improved significantly. The lift has increased from 3.45N to 3.54N, resulting in a 2.6% enhancement. Additionally, the noise level has decreased from 72.33dB to 69.52dB, achieving a noise reduction of 2.81dB.

Network-level pavement maintenance and rehabilitation planning considering uncertainties using chance-constrained programming

ABSTRACT In practice, given limited funds, to consider multiple strategic goals/objectives that different stakeholders concern, pavement network-level maintenance and rehabilitation (M&R) planning becomes a multi-objective optimisation (MOO) based project selection and budget allocation problem. In an attempt to solve this problem, most agencies established MOO models under the deterministic situation without appropriate consideration of uncertainties. However, ignoring performance uncertainties often leads to unreasonable decisions. To provide more convincing and reliable pavement M&R decisions, this paper proposes a Chance-Constrained Programming (CCP) based MOO method to incorporate performance uncertainties in network-level single period pavement M&R planning. First, a general deterministic MOO model with budget and network performance constraints is established. Then, three commonly-used statistical forms of network-level performance measures are introduced. To incorporate uncertainties, the probability distribution of each form of performance measure is derived. Based on the CCP method, the MOO model is transformed to an equivalent deterministic formulation as a mixed non-linear integer programming (MNLIP) problem. To demonstrate the proposed method, a case study using real data is conducted. The results show that the proposed method can effectively help decision-makers to appropriately incorporate performance uncertainties in conducting network-level pavement M&R planning.

Multi-objective optimization for gymnasium layout in early design stage: Based on genetic algorithm and neural network

The layout of gymnasium directly affects environment performance. The current methods are insufficient to provide quantified decision support for gymnasium layouts in the early design stages (EDS). This study proposes a framework for optimizing the layout of gymnasiums using a multi-objective optimization (MOO) method based on genetic algorithms (GA) and neural networks. The study tested the framework using a community sports arena as an example, and the results indicate: the final optimized solutions achieved a maximum reduction of 11.1 % in cooling energy consumption (CE) and 3.3 % in solar radiation (SR) compared to the earlier generations, along with a 0.9 % improvement in thermal comfort percentage (TCP). This framework promotes the development of algorithm-driven methods for stadium layout design, while the prediction model based on RBF neural networks can simultaneously provide effective performance predictions for similar design outcomes.

Multi-objective optimization method for steel frame with top-and-seat angles considering yield control

Based on the yield mechanism control principle of “strong columns and weak beams, strong joints and weak components” in steel frames connected by top-and-seat steel angles, this paper proposes a method to define the relationship between the bending capacity of various components through inequality iteration and deduction. A number of component parameters has been identified that enables control of the yielding sequence in structural components. The study also developed an objective function and used genetic algorithms to create a multi-objective optimization technique for semi-rigid steel frame structures. The method considers the non-dominated cost interval, resulting in a more comprehensive approach. From the optimization case studies against the published experimental results and finite element simulations, it demonstrates that the proposed approach efficiently manages the yielding sequence and minimizes the decision space. The optimization model provides the coordination optimization between structural cost and lateral base shear capacity.

Multi-objective optimization of variable-stiffness composite for CFRP-winding buckle arrestor by using NSGA-Ⅲ

This study introduces a multi-objective optimization method utilizing the NSGA-III algorithm to tailor the fiber shape of variable-stiffness CFRP (Carbon Fiber Reinforced Plastics) shell structures. Combined with offshore oil and gas engineering, this method is employed to perform fiber shape optimization design for CFRP-winding buckle arrestors. Based on cubic polynomial path functions' contour lines, curved fibers are represented and fitted using the Curvature Minimum Bi-Arc (CMBiArc) interpolation NURBS curve method. Employing the NSGA-III algorithm, the fiber shapes of a CFRP laminate are optimized under manufacturing constraints, targeting mean compliance, fundamental frequency, and fiber curvature. With the study of numerical modeling, buckling crossover mode and arresting efficiency, multi-objective optimization is carried out for fiber shape of CFRP-winding buckle arrestors designed for subsea pipelines. With the goal of maximizing arresting efficiency, structural fundamental frequency, and minimizing fiber curvature, the method yields an enhanced CFRP-winding arrestor design, which demonstrates a high practicability through a layered design approach. By parametric study of optimization result and comparison with slip-on arrestor, the feasibility of the optimization method and the practical value of the novel improved arrestor are substantiated.