Search

This paper introduces a bilevel programming model for optimizing transit network departure frequency. In the upper-level model, user satisfaction is reflected by considering congestion effects in the cost function. The lower-level assignment model simulates passenger travel behavior more realistically by incorporating congestion effects. This problem is solved by a heuristic gradient descent algorithm, where an approximation of the gradient is obtained at each iteration by using sensitivity analysis for transit equilibrium problems. The effectiveness of the proposed model and algorithm is demonstrated through two test examples, one of which involves a real-world scenario comprising over 130 transit lines. Numerical results conclusively indicate that the incorporation of congestion effects in the proposed model leads to improved transit system performance and enhanced user satisfaction. [ABSTRACT FROM AUTHOR]

Detection of real-world factors in digital photos and videos is one of the most important challenges in computer recognition for object detection. The main goal of generic object detection is to identify and locate specific objects. Despite abundant benefits, there exist some problems, such as the accuracy and extraction of lower and higher-level features. In order to obtain the low and high level characteristics of an image for precise classification using Bi-directional Feature Pyramid Network (Bi-FPN) and Mask Regional based Convolutional Neural Network (Mask RCNN) with Support Vector Machine (SVM) are utilised. In this proposed model, the features are extracted with the Bi-FPN model and are pooled with the adaptive feature polling technique. These features are aligned with the ROI alignment and separated into convoluted and fully connected layers. The SVM replaces the fully connected layer for classification and the convoluted layer is used for masking and bounding the object with the box. This combined model with Mask RCNN and SVM represents the Hybrid Mask RCNN-SVM. The proposed model has been implemented in Python for calculating and comparing performance metrics such as the accuracy, error, precision and recall etc., for the proposed and existing model. The resultant values for accuracy, recall, error and precision for real-time object detection utilizing hybrid Mask RCNN-SVM are 0.98, 0.96, 0.027 and 0.98. Thus, the evaluation of performance metrics results that the values of the proposed model being better compared to the existing techniques. As a result, the proposed hybrid Mask RCNN-SVM model effectively segments the object from the real-time video. [ABSTRACT FROM AUTHOR]

The growing deployment of distributed resources significantly affects the distribution grid performance in most countries. The optimal sizing and placement of these resources have become increasingly crucial to mitigating grid issues and reducing costs. Particle Swarm Optimization (PSO) is widely used to address such problems but faces computational inefficiency due to its numerical convergence behavior. This limits its effectiveness, especially for power system problems, because the numerical distance between two nodes in power systems might be different from the actual electrical distance. In this paper, a scalable bilevel optimization problem with two novel algorithms enhances PSO's computational efficiency. While the resistivity-driven algorithm strategically targets low-resistivity regions and guides PSO toward areas with lower losses, the connectivity-driven algorithm aligns solution spaces with the grid's physical topology. It prioritizes actual physical neighbors during the search to prevent local optima traps. The tests of the algorithms on the IEEE 33-bus and the 69-bus and Norwegian networks show significant reductions in power losses (up to 74% for PV, wind, and storage) and improved voltage stability (a 21% reduction in mean voltage deviation index) with respect to the results of classical PSO. The proposed network-oriented PSO outperforms classical PSO by achieving a 2.84% reduction in the average fitness value for the IEEE 69-bus case with PV, wind, and storage deployment. The Norwegian case study affirms the effectiveness of the proposed approach in real-world applications through significant improvements in loss reduction and voltage stability. [ABSTRACT FROM AUTHOR]

This paper presents and analyzes a new viscosity approximation method with the inertial technique for finding a common fixed point of a countable family of nonexpansive mappings and then its strong convergence theorem is established under some suitable conditions. As a consequence, we employ our proposed algorithm for solving some convex bilevel optimization problems and then apply it for solving regression of a graph of cosine function and classification of some noncommunicable diseases by using the extreme learning machine model. We perform a comparative analysis with other algorithms to demonstrate the performance of our approach. Our numerical experiments confirm that our proposed algorithm outperforms other methods in the literature. [ABSTRACT FROM AUTHOR]

In this paper, we study a generalized bilevel optimal control problem that has a variational inequality parametrized by the final state on the follower and pure state constraints on the leader. After reducing the problem with a gap function to an analogous single-level optimal control problem, we focus on the development of a necessary optimality condition of the Pontryagin type. We highlight some significant issues originating from the generalized bilevel structure and its pure state constraints on the leader, which give rise to a degenerated maximum principle in the absence of constraint qualifications. To ensure the nondegeneracy of the derived maximum principle, we employ a partial penalization strategy and a well-known regularity criterion for optimal control problems with pure state constraints. [ABSTRACT FROM AUTHOR]

In this paper, we develop and analyze the concepts of gH-Dini Hadamard ε-subdifferential and Hε-subgradient for interval-valued functions (IVFs). Some important characteristics of gH-Dini Hadamard ε-subdifferential such as closedness, convexity, and monotonicity are studied. The interrelations between gH-subgradient and gH-Dini Hadamard ε-subgradient, and between gH-Fréchet derivative and gH-Dini Hadamard ε-subdifferential are investigated. To define the concept of Hε-subgradient, the notions of the sponge of a set around a point and gH-calm IVF at a point are studied. A variational description of gH-Dini Hadamard ε-subgradient with Hε-subgradient is proposed. Various necessary and sufficient conditions for obtaining an ε-efficient solution to an interval optimization problem (IOP) with the help of gH-Dini Hadamard ε-subgradient of an IVF are derived. Lastly, an application of proposed results is discussed in the sparsity regularizer for IOPs. [ABSTRACT FROM AUTHOR]

The purpose of this paper is to derive primal and dual second-order necessary optimality conditions for a standard bilevel optimization problem with both smooth and nonsmooth data. The approach involves utilizing two different reformulations of the hierarchical model as a single-level problem under a partial calmness assumption. The first reformulation consists on the use of the value function of the lower-level problem, which is then tackled by using second-order directional derivatives. However, for the dual conditions, this approach is not suitable except for cases that the value function is smooth. Therefore, we adopt a second approach that relies on the Ψ-reformulation. In both cases, the obtained necessary optimality conditions can be expressed according to the problem data. Finally, some examples are given to illustrate the proven results. [ABSTRACT FROM AUTHOR]

This article studies the problem of semi-supervised learning on graphs, which aims to incorporate ubiquitous unlabeled knowledge (e.g., graph topology, node attributes) with few-available labeled knowledge (e.g., node class) to alleviate the scarcity issue of supervised information on node classification. While promising results are achieved, existing works for this problem usually suffer from the poor balance of generalization and fitting ability due to the heavy reliance on labels or task-agnostic unsupervised information. To address the challenge, we propose a dual-channel framework for semi-supervised learning on Graphs via Knowledge Transfer between independent supervised and unsupervised embedding spaces, namely, GKT. Specifically, we devise a dual-channel framework including a supervised model for learning the label probability of nodes and an unsupervised model for extracting information from massive unlabeled graph data. A knowledge transfer head is proposed to bridge the gap between the generalization and fitting capability of the two models. We use the unsupervised information to reconstruct batch-graphs to smooth the label probability distribution on the graphs to improve the generalization of prediction. We also adaptively adjust the reconstructed graphs by encouraging the label-related connections to solidify the fitting ability. Since the optimization of the supervised channel with knowledge transfer contains that of the unsupervised channel as a constraint and vice versa, we then propose a meta-learning-based method to solve the bi-level optimization problem, which avoids the negative transfer and further improves the model's performance. Finally, extensive experiments validate the effectiveness of our proposed framework by comparing state-of-the-art algorithms. [ABSTRACT FROM AUTHOR]

Three-dimensional soft tissue simulation has become a popular tool in the process of virtual orthognathic surgery planning and patient–surgeon communication. To apply 3D soft tissue simulation software in routine clinical practice, both qualitative and quantitative validation of its accuracy are required. The objective of this study was to systematically review the literature on the accuracy of 3D soft tissue simulation in orthognathic surgery. The Web of Science, PubMed, Cochrane, and Embase databases were consulted for the literature search. The systematic review (SR) was conducted according to the PRISMA statement, and 40 articles fulfilled the inclusion and exclusion criteria. The Quadas-2 tool was used for the risk of bias assessment for selected studies. A mean error varying from 0.27 mm to 2.9 mm for 3D soft tissue simulations for the whole face was reported. In the studies evaluating 3D soft tissue simulation accuracy after a Le Fort I osteotomy only, the upper lip and paranasal regions were reported to have the largest error, while after an isolated bilateral sagittal split osteotomy, the largest error was reported for the lower lip and chin regions. In the studies evaluating simulation after bimaxillary osteotomy with or without genioplasty, the highest inaccuracy was reported at the level of the lips, predominantly the lower lip, chin, and, sometimes, the paranasal regions. Due to the variability in the study designs and analysis methods, a direct comparison was not possible. Therefore, based on the results of this SR, guidelines to systematize the workflow for evaluating the accuracy of 3D soft tissue simulations in orthognathic surgery in future studies are proposed. [ABSTRACT FROM AUTHOR]

The separation section of the methanol-to-olefin (MTO) process is energy-intensive, and the optimization and heat integration can enhance energy efficiency and reduce costs. A bi-level optimization model framework is proposed to optimize the separation process and simultaneously integrate the heat exchanger network (HEN). The upper level employs a data-driven BP neural network proxy model instead of the mechanism model for the separation process, while the lower level adopts a stage-wise superstructure for the HEN without stream splits. The interaction between the two systems is realized effectively through information exchange. A bi-level particle swarm algorithm is employed to optimize complex problems and determine the optimal operational parameters for the distillation system and HEN. Compared with the typical sequential synthesis method, the optimization by the proposed approach reduces the total annual cost by 1.4293 × 10 6 USD/y, accounting for 4.76%. [ABSTRACT FROM AUTHOR]

This paper presents the generation and transmission expansion planning (GTEP) in electricity and gas networks by considering their resilience against floods and earthquakes. These networks supply electricity, heat, and gas consumption energies as a multi‐carrier microgrid. The scheme is expressed in the form of bi‐level optimization, the upper level of which is the minimization of generation and transmission planning cost (total investment cost and expected operating cost) in the mentioned networks constrained to the investment budget and the planning model of the mentioned elements. Lower‐level formulation minimizes the total expected annual operating cost of these networks and the expected outage cost of electricity, heat, and gas consumers in the event of floods and earthquakes. This formulation is bound by the power flow equations of electricity and gas networks, the operation and resilience constraints of the networks, and the limitation on generation capability. In this problem, the expected energy not‐supplied and the outage cost during natural disasters are considered resilience indicators. Next, a single‐level model for the proposed design is extracted from the Karush–Kuhn–Tucker (KKT) method. The basic requirement of this method is the convexity of the lower‐level constraints. For this purpose, first, a linear approximation model is obtained for the lower‐level constraints of the problem. Furthermore, stochastic optimization is adopted to model the uncertainty of load, renewable power, and network equipment availability during floods and earthquakes. Finally, the extracted numerical results confirm the capability of the proposed scheme in improving the operation and resilience of the mentioned networks using optimal generation and transmission planning. [ABSTRACT FROM AUTHOR]

Microgrid (MG) integrated with renewable energy sources (RES) has become increasingly popular, especially when the lack of resources and environmental pollution are serious. However, the uncertainty of RES is the major problem when operating MG. The hydrogen energy storage system (HESS) is a prominent solution for the RES uncertainty since it can store excess energy, supply when in shortage, help to regulate power consumption, and improve MG's flexibility. In this research, a bi-level optimization model for planning the microgrid is proposed. The upper layer of the optimization model aims to minimize the total operation cost of the MG, while the lower level focuses on minimizing the operation cost of the hydrogen energy storage system. The results demonstrate that the proposed bi-level model can help to reduce the operation cost of MG by 42.75% in comparison with typical single level optimization model. Additionally, the electricity generated by diesel generators has also decreased by 52.43%, resulting in a significant reduction in CO 2 emissions and contributing to environmental protection. • Green hydrogen can be effectively stored from surplus renewable energy. • HESS operation in Day-Ahead economic optimization scheduling model for MG with high penetration of RES. • A novel bi-level optimization model for optimal scheduling of MG with 42.75% reduction in operation cost. • The interaction between the distributed sources and the behavior of HESS are incorporated into the optimal scheduling. • The electricity generated by diesel generators decreases by 52.43%, which leads to the CO 2 emission reduction. [ABSTRACT FROM AUTHOR]

In this paper, we consider a shortest path network interdiction problem with incomplete information and multiple levels of interdiction intensity. The evader knows the attacker's decision on the network arcs that have been interdicted. However, the extent of damage on each arc depends on the interdiction intensity and the amount of budget spent for interdiction. We consider two cases in which the evader has incomplete information about both the intensity of attack on the interdicted arcs and the additional cost imposed for traversing those arcs. In the first case, the evader's perception of this cost falls in an interval of uncertainty. In the second case, it is assumed that the evader estimates a relative frequency for each level of interdiction intensity. This gives rise to multiple uncertainty sets for the evader's estimates of the additional cost. To handle the uncertainty that arises in both cases, a robust optimization approach is employed to derive the mathematical formulation of underlying bilevel optimization problem. For each case, we first take the well-known duality-based approach to reformulate the problem as a single-level model. We show that this method does not always end up with an integer solution or fails in achieving a solution within the time limit. Therefore, we develop an alternative algorithm based on the decomposition approach. Computational results show that the proposed algorithm outperforms the duality-based method to obtain the optimal solution. Last, a real case study is presented to show the applicability of the studied problem. [ABSTRACT FROM AUTHOR]

This paper proposes a pricing model to ensure the budget balance of a company responsible for collecting and treating solid household waste in a city. The model used is of the Principal-Agent type which considers two actors at play, namely the leader (the service provider company) and the follower (the households and establishments that are the beneficiaries of the waste collection service). The model is a constrained bi-level optimization program. The model was applied in the case of Benin to the Grand Nokoué waste management and healthiness Company (SGDS-GN) whose intervention area, in the beginning, covers five communes in the southern part of the country. The solving of the optimization program was made with three (03) scenarios: the company spends two (02) times a week for waste collection (scenario 1), the company spends three (03) times a week for waste collection (scenario 2) and the company spends four (04) times a week for waste collection (scenario 3). The simulation results show that scenario 2 with three (03) passages per week appears more optimal for both the company (the operating result is higher) and the beneficiaries of waste collection services (the rate to be paid is lower). [ABSTRACT FROM AUTHOR]

This study optimized the design of an aeration device for pond engineered recirculating aquaculture systems (RASs) whose application is aimed at increasing dissolved oxygen (DO) levels in RAS aquaculture practice. DO is a key factor in aquaculture productivity, and oxygenators are the power devices used for regulating its levels in aquaculture ponds. In this study, grass carp (Ctenopharyngodon idellus) aquaculture trials were conducted in a self-built RAS by using the new aeration device (NAD); the microporous and impeller aeration components were individually tested in terms of performance, and then combined for the orthogonal testing of their operating parameters in order to assess the NAD's oxygenation capacity. The test results show that the device effectively increased the dissolved oxygen levels in the RAS tank, enhanced the upper–lower water layer exchange and directional flow, and met the design and parameter selection requirements. Compared with the existing RAS oxygenation equipment, the NAD operated with the optimal parameters and increased the oxygen transfer rate in the pond water tank by 122%. [ABSTRACT FROM AUTHOR]

In this paper, we study a class of unconstrained and constrained bilevel optimization problems in which the lower level is a possibly nonsmooth convex optimization problem, while the upper level is a possibly nonconvex optimization problem. We introduce a notion of varepsilon -KKT solution for them and show that an varepsilon -KKT solution leads to an...-hypergradient--based stationary point under suitable assumptions. We also propose first-order penalty methods for finding an varepsilon -KKT solution of them, whose subproblems turn out to be a structured minimax problem and can be suitably solved by a first-order method recently developed by the authors. Under suitable assumptions, an operation complexity of... 1), measured by their fundamental operations, is established for the proposed penalty methods for finding an varepsilon-KKT solution of the unconstrained and constrained bilevel optimization problems, respectively. Preliminary numerical results are presented to illustrate the performance of our proposed methods. To the best of our knowledge, this paper is the first work to demonstrate that bilevel optimization can be approximately solved as minimax optimization, and moreover, it provides the first implementable method with complexity guarantees for such sophisticated bilevel optimization. [ABSTRACT FROM AUTHOR]

In a multi-objective optimization problem, a decision maker has more than one objective to optimize. In a bilevel optimization problem, there are the following two decision-makers in a hierarchy: a leader who makes the first decision and a follower who reacts, each aiming to optimize their own objective. Many real-world decision-making processes have various objectives to optimize at the same time while considering how the decision-makers affect each other. When both features are combined, we have a multi-objective bilevel optimization problem, which arises in manufacturing, logistics, environmental economics, defence applications and many other areas. Many exact and approximation-based techniques have been proposed, but because of the intrinsic nonconvexity and conflicting multiple objectives, their computational cost is high. We propose a hybrid algorithm based on batch Bayesian optimization to approximate the upper-level Pareto-optimal solution set. We also extend our approach to handle uncertainty in the leader's objectives via a hypervolume improvement-based acquisition function. Experiments show that our algorithm is more efficient than other current methods while successfully approximating Pareto-fronts. [ABSTRACT FROM AUTHOR]

In this paper, we consider the problem of learning variational models in the context of supervised learning via risk minimization. Our goal is to provide a deeper understanding of the two approaches of learning of variational models via bilevel optimization and via algorithm unrolling. The former considers the variational model as a lower level optimization problem below the risk minimization problem, while the latter replaces the lower level optimization problem by an algorithm that solves said problem approximately. Both approaches are used in practice, but unrolling is much simpler from a computational point of view. To analyze and compare the two approaches, we consider a simple toy model, and compute all risks and the respective estimators explicitly. We show that unrolling can be better than the bilevel optimization approach, but also that the performance of unrolling can depend significantly on further parameters, sometimes in unexpected ways: While the stepsize of the unrolled algorithm matters a lot (and learning the stepsize gives a significant improvement), the number of unrolled iterations plays a minor role. [ABSTRACT FROM AUTHOR]

Since the outbreak of COVID-19, diverse changes have been made in schools. The decision was made for students to be back in school online, and synchronous bi-directional online classes were launched as a means of 'blended learning'. Against this background, we aim to explore how synchronous bi-directional online classroom space is understood and materialised through a lens of socio-material assemblage. This article builds on existing literature on the potential of liminal space, which encourages a focus on the limits of the space and the simultaneous possibility of students' agency. Online ethnography was conducted to analyse the dynamics of spatiality in an online classroom. Participant observation was conducted in online classes in one 6th grade (elementary school) class in Seoul, South Korea, followed by semi-structured interviews with the teacher and students. We explore themes of students being controlled and (simultaneously) exercising their agency, thus opening up understanding of the agency of children as active actors who construct the space. [ABSTRACT FROM AUTHOR]

The integration of renewable energy units into power systems brings a huge challenge to the flexible regulation ability. As an efficient and convenient flexible resource, energy storage systems (ESSs) have the advantages of fast-response characteristics and bi-directional power conversion, which can provide flexible support for the power system. This paper establishes an optimization model for the ESS based on a bi-level programming model. The upper-level model optimizes the decision strategy of ESS configuration planning. The lower-level model is based on scenario analysis theory to simulate the operation of typical daily scenarios. Flexibility requirement constraints are added to characterize the required flexibility resources of the power system. In addition, the conditional value-at-risk (CVaR) is applied to characterize the risk of wind curtailment and load shedding during operation. To simplify the model, a set of association constraints is introduced to convert the original bi-level programming model into a direct-solvable single-level mixed-integer linear programming (MILP) model. Finally, the effectiveness of the proposed model is verified through case studies. [ABSTRACT FROM AUTHOR]

This paper considers an extension of the shortest path network interdiction problem that incorporates robustness to account for parameter uncertainty. The shortest path interdiction problem is a game of two players with conflicting agendas and capabilities: an evader, who traverses the arcs of a network from a source node to a sink node using a path of shortest length, and an interdictor, who maximizes the length of the evader's shortest path by interdicting arcs on the network. It is usually assumed that the parameters defining the network are known exactly by both players. We consider the situation where the evader assumes the nominal parameter values while the interdictor uses robust optimization techniques to account for parameter uncertainty or sensor degradation. We formulate this problem as a nonlinear mixed‐integer semi‐infinite bilevel program and show that it can be converted into a mixed‐integer linear program with a second order cone constraint. We use random geometric networks and transportation networks to perform computational studies and demonstrate the unique decision strategies that our variant produces. Solving the shortest path interdiction problem with asymmetric uncertainty protects the interdictor from investing in a strategy that hinges on key interdictions performing as promised. It also provides an alternate strategy that mitigates the risk of these worst‐case possibilities. [ABSTRACT FROM AUTHOR]

Bilevel programming is characterized by the existence of two optimization problems in which the constraint region of the upper level problem is implicitly determined by the lower level optimization problem. This hierarchical design of optimization is suitable to model a large number of real-life applications. However, when dealing with a non linear multi-objective optimization context, new complexities arise due to conflicting objectives. In this paper, an exact method is described to solve an integer indefinite quadratic bilevel maximization problem with multiple objectives at the upper level, where the objective functions at both levels are the product of two linear functions. The algorithm suggested aims to produce a set of efficient solutions by employing a branch and cut approach. It optimizes the indefinite quadratic problem of the upper level within the feasible region of the original problem in an iterative manner. Then, it introduces the Dantzig cut technique to identify the optimal solution for the integer indefinite quadratic bilevel programming problem. Additionally, the algorithm utilizes an efficient cut that reduces the search process for obtaining the set of efficient solutions of the main problem, along with a branching constraint for the integer decision variables. The algorithm was implemented and tested on instances generated randomly, yielding positive outcomes. [ABSTRACT FROM AUTHOR]

New energy storage methods based on electrochemistry can not only participate in peak shaving of the power grid but also provide inertia and emergency power support. It is necessary to analyze the planning problem of energy storage from multiple application scenarios, such as peak shaving and emergency frequency regulation. This article proposes an energy storage capacity configuration planning method that considers both peak shaving and emergency frequency regulation scenarios. A frequency response model based on emergency frequency regulation combined with low-frequency load shedding is established, taking into account the frequency safety constraints of the system and the principle of idle time reuse, to establish a bi-level programming model. In the upper-level model, the optimization objective is to minimize the annual operating cost of the system during the planning period, combined with the constraints of power grid operation to plan the energy storage capacity. The lower-level model embeds frequency safety constraints with the optimization objective of minimizing the cost of fault losses. To solve the bi-level optimization problem, the Karush–Kuhn–Tucher (KKT) condition and Big-M method were used to transform the bi-level model into a single-layer linear model. Finally, an improved IEEE RTS-24 system was used for numerical verification. The results show that the method proposed in this article can reasonably plan the capacity of energy storage, improve frequency safety during system operation, and reduce the operating cost of the power grid. [ABSTRACT FROM AUTHOR]

Domain generalization studies how to generalize a machine learning model to unseen distributions. Learning invariant representation across different source distributions has been shown high effectiveness for domain generalization. However, the intrinsic possibility of overfitting in source domains can limit the generalization of invariance when faced with a target domain with large discrepancy to the source domains. To address this problem, we propose a meta-learning algorithm via bilevel optimization for domain generalization, where the inner-loop objective aims to minimize the discrepancy across different source domains while the outer-loop objective aims to minimize the discrepancy between source domains and a potential target domain. We show from a geometric perspective that the proposed algorithm can improve out-of-domain robustness for invariance learning. Empirically, we evaluate on five datasets and achieve the best results among a range of strong domain generalization baselines. [ABSTRACT FROM AUTHOR]

In this paper, we propose a new accelerated algorithm for solving convex bilevel optimization problems using some fixed point and two-step inertial techniques. Our focus is on analyzing the convergence behavior of the proposed algorithm. We establish a strong convergence theorem for our algorithm under some control conditions. To demonstrate the effectiveness of our algorithm, we utilize it as a machine learning algorithm to solve data classification problems of some noncommunicable diseases, and compare its efficacy with BiG-SAM and iBiG-SAM. [ABSTRACT FROM AUTHOR]

Grid operation and inspection a key links to ensure the safe operation of the power system, which requires efficient task allocation and resource scheduling. To address this problem, this paper proposes a resource scheduling model for grid operation and inspection based on bi-level programming. Firstly, the O&I process is analyzed and defined as a combined optimization problem of the multiple traveling salesman problem (MTSP) and the job-shop scheduling problem (JSP). Secondly, a bi-level programming model of MTSP and JSP is established according to the characteristics of the problem. Finally, an adaptive genetic algorithm is used to solve the problem. The feasibility of the model and the advancement of the algorithm are verified through the simulation of real scenarios and a large number of tests, which provide strong support for the sustainable development of the power system. [ABSTRACT FROM AUTHOR]

As consumers and governments prioritize cost-effectiveness and ecological sustainability, the limitations of traditional manufacturing paradigms become apparent in the context of constrained resources. The adverse effects of these paradigms on the environment and society hinder the achievement of a sustainable product life cycle. Intelligent manufacturing processes offer a solution by efficiently gathering meaningful data, such as usage and product recycling information, from previous product generations to enhance product design and subsequent sustainable manufacturing processes (SMPs). Modular product family architecture (PFA) design holds promise in promoting product sustainability and addressing diverse consumer needs. PFA design and SMPs are inherently interconnected within intelligent manufacturing frameworks. This paper aims to integrate the decision-making processes underlying PFA with SMPs. We model integrated PFA and SMP decisions as a Stackelberg game, which involves hierarchical joint optimization (HJO) for assessing product modularity and sustainable manufacturing fulfillment. We develop a bilevel 0–1 integer nonlinear programming model to represent the HJO decision-making process and propose a nested genetic algorithm (NGA) to solve the HJO problem. A case study with a laptop is conducted to validate the feasibility and potential of the proposed HJO model for joint optimization problems in PFA design and SMPs. [ABSTRACT FROM AUTHOR]

In September 2020, the European Union (EU) decided to include the maritime industry into the EU emissions trading system to effectively control carbon emissions from ships, namely the maritime emissions trading system (METS). Various allocation and trading schemes are adopted in practice. This paper investigates a comparison of allocation and trading schemes of the METS from the perspective of liner shipping carriers and regulators, respectively. We explore how to determine the optimal allocation and trading scheme of the METS and how the liner shipping route (LSR) choice decision (following or avoiding the METS) and carbon emissions will be affected. For liner shipping carriers, we use a mixed-integer linear programming model that aims to reduce carriers' costs to formulate our problem. For regulators, we present a bi-level programming model that aims to reduce carbon emissions. In numerical experiments, we further analyze the impact of joining the METS on the scheme choice, LSR choice and emissions by considering the European Economic Area and the Red Sea being covered by the METS. Numerical experiments are provided to show the effectiveness of our proposed models. • Investigate allocation and trading schemes of the maritime emission trading system. • Propose a liner shipping route choice problem. • Develop a mixed-integer linear programming model. [ABSTRACT FROM AUTHOR]

In this paper, we present BilevelJuMP.jl, a new Julia package to support bilevel optimization within the JuMP framework. The package is a Julia library that enables the user to describe both upper and lower-level optimization problems using the JuMP algebraic syntax. Because of the generality and flexibility that our library inherits from JuMP's syntax, our package allows users to model bilevel optimization problems with conic constraints in the lower level and all constraints supported by JuMP in the upper level including conic, quadratic, and nonlinear constraints. Moreover, the models defined with the syntax from BilevelJuMP.jl can be solved by multiple techniques that are based on reformulations as mathematical programs with equilibrium constraints (MPEC). Manipulations on the original problem data are possible due to MathOptInterface.jl's structures and Dualization.jl features. Hence, the proposed package allows quick modeling, deployment, and thereby experimenting with bilevel models based on off-the-shelf mixed-integer linear programming and nonlinear solvers. History: Accepted by Ted Ralphs, Area Editor for Software Tools. Funding: The authors were partially supported by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - Brasil (CAPES) - Finance Code 001. The work of A. Street was also partially supported by Fundação Carlos Chagas Filho de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ) and Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq). The work was partially funded by the project P&D ANEEL PD-00403-0050/2020 sponsored by ENGIE BRASIL ENERGIA S.A. Supplemental Material: The software that supports the findings of this study is available within the paper and its Supplemental Information (https://pubsonline.informs.org/doi/suppl/10.1287/ijoc.2022.0135) as well as from the IJOC GitHub software repository (https://github.com/INFORMSJoC/2022.0135). The complete IJOC Software and Data Repository is available at https://informsjoc.github.io/. [ABSTRACT FROM AUTHOR]

This work aims to investigate a multimodal feeder plan for a metro terminal station in the suburbs, where fixed-route transit, flex-route feeder transit (FRFT), shared bike, online car-hailing, and private cars are included. A slack arrival strategy, which relaxes the schedule constraints of a checkpoint but does not affect the bus operating cycle, is proposed to deal with the real-time demand insertion (RDI) problem in FRFT. A dynamic optimization model that is based on slack arrival strategy, which considers penalty time costs, is established to solve rejection decisions and route modification problems in RDI. Then, a method for a multimodal feeder plan is proposed, where FRFT is set for the growing number of commuters and factories, and fixed-route transit is set for mature communities with large passenger demand. A bilevel programming model is built, in which the upper-level multimodal feeder plan model is constructed by comprehensively considering the total travel cost of passengers and the bus profit (ticket income minus operational cost). The lower level is a multimode traffic assignment model. A genetic algorithm (GA) is adopted in the concrete optimization solution. The proposed method is validated through a sample application to Tianjin City, China. Compared with the method without public transit (PT) routes, the method with fixed routes, and the method with conventional flex-route transit in growing communities and factories, the total travel cost is reduced by 9.14%, 5.65%, and 3.86%, respectively. The bus profit increased by 34.67%, 56.89%, and 38.68%, respectively. In addition, the effectiveness of policies for FRFT is verified by evaluating the performance of a multimodal feeder plan under various advanced reservation rates. The feeder problem in a metro terminal station in the suburbs could be solved, and the development of the increasing number of commuters and factories could be ensured with the proposed method. [ABSTRACT FROM AUTHOR]

Sustainable aviation fuel (SAF) will significantly impact global warming in the aviation sector, and important SAF targets are emerging. Isoprenol is a precursor for a promising SAF compound DMCO (1,4-dimethylcyclooctane) and has been produced in several engineered microorganisms. Recently, Pseudomonas putida has gained interest as a future host for isoprenol bioproduction as it can utilize carbon sources from inexpensive plant biomass. Here, we engineer metabolically versatile host P. putida for isoprenol production. We employ two computational modeling approaches (Bilevel optimization and Constrained Minimal Cut Sets) to predict gene knockout targets and optimize the "IPP-bypass" pathway in P. putida to maximize isoprenol production. Altogether, the highest isoprenol production titer from P. putida was achieved at 3.5 g/L under fed-batch conditions. This combination of computational modeling and strain engineering on P. putida for an advanced biofuels production has vital significance in enabling a bioproduction process that can use renewable carbon streams. • Biosynthetic isoprenol is an emerging precursor for sustainable aviation fuel DMCO. • Pseudomonas putida was engineered for isoprenol production via the IPP-bypass pathway. • Genome scale metabolic modeling-aided gene knockouts improves isoprenol production. • Frequency scoring of two computational methods enables to prioritize knockout targets. • Combination of genome and pathway engineering results in isoprenol titer improvement. [ABSTRACT FROM AUTHOR]

In order to improve the efficiency of the last-mile delivery system when customers are possibly absent for deliveries, we propose the idea of employing the crowd to work as keepers and to provide storage services for their neighbors. Crowd keepers have extra flexibility, more availability, and lower costs than fixed storage options such as automated lockers, and this leads to a more efficient and a more profitable system for last-mile deliveries. We present a bilevel program that jointly determines the assignment, routing, and pricing decisions while considering customer preferences, keeper behaviors, and platform operations. We develop an equivalent single-level program, a mixed-integer linear program with subtour elimination constraints, that can be solved to optimality using a row generation algorithm. To improve the efficiency of the solution procedure, we further derive exact best response sets for both customers and keepers, and approximate optimal travel times using linear regression. We present a numerical study using a real-world data set from Amazon. The fixed-storage and no-storage systems are used as benchmarks to assess the performance of the crowdkeeping system. The results show that the crowdkeeping delivery system has the potential to generate higher profits because of its ability to consolidate deliveries and to eliminate failed deliveries. Funding: Funding provided by the Natural Sciences and Engineering Research Council of Canada [Grants 2022-04979 and 2022-05261], the Canada Research Chair program [Grant CRC-2018-00105], and the China Scholarship Council [Grant 202006190051] is gratefully acknowledged. Supplemental Material: The online appendix is available at https://doi.org/10.1287/trsc.2022.0323. [ABSTRACT FROM AUTHOR]

In many subsurface studies, facies and petrophysical properties are two important reservoir parameters that are closely correlated. They are routinely used in well interpretation, hydrocarbon reserve calculation and production profile prediction. These two parameters have commonly been determined in two separate tasks because of their mathematical differences (facies are discrete, and petrophysical properties are continuous). However, this is incorrect because facies and petrophysical properties are often strongly correlated. Therefore, we propose a new joint inversion method of facies and petrophysical properties based on a bi‐level optimization model. In the bi‐level optimization model, the upper‐level problem is the petrophysical property inversion while the lower‐level problem can identify the facies and add a facies‐related constraint for the upper‐level optimization. We also develop a new genetic algorithm for the discrete‐continuous inversion problem based on the bi‐level optimization model because the inversion problem usually has multiple local solutions. In addition, rock physics and statistics are combined in the inversion process. A rock physics model is used to establish the basic relationships between the petrophysical and elastic parameters, and the statistical approach is used to describe the intrinsic connection among the multiple reservoir parameters based on well log data. The numerical experiments indicate that the traditional separate prediction method and the new joint inversion method can quickly obtain more accurate results. In the application examples of real data, the inversion results can be matched to the well log data within the limits of the input data resolution, which further verifies the reliability and application potential of this new method. [ABSTRACT FROM AUTHOR]

To accurately calculate the carbon emission of integrated energy system (IES), and fully explore the potential for load side on carbon emission reduction, this paper proposes a method of guiding load to participate in demand response based on node carbon emission intensity, and constructs a bi‐layer low‐carbon scheduling model for source‐load collaborative optimization. First, the carbon flow calculation model of IES in the total process is established, such as source, network, load, and storage. It can depict the carbon emission characteristics of energy conversion equipment and energy storage devices. The principle of proportional sharing is used to track carbon emission flows, the changes in carbon emission intensity at each node is perceived from a spatiotemporal perspective. Second, carbon flow analysis is incorporated into the load demand response mechanism, and a carbon emission model for load aggregator (LA) after demand response is established based on the node carbon emission intensity. Third, a bi‐layer low‐carbon scheduling model is constructed, which considers the source‐load collaborative optimization. The upper layer is the optimal economic dispatch of IES operators, while the lower layer is the optimal economic dispatch of LAs. Finally, the effectiveness of the proposed method is verified using the system as an example, such as improved 14‐node power grid, 6‐node heating network, and 6‐node nature gas network. [ABSTRACT FROM AUTHOR]

This article considers resilience of service networks that are composed of service and control nodes to node‐targeted attacks. Two complementary problems of selecting attacked nodes and placing control nodes reflect the interaction between the network operator and the network attacker. This interaction can be analyzed within the framework of game theory. Considering the limited performance of the previously introduced iterative solution algorithms based on non‐compact problem models, new compact integer programming formulations of the node attack optimization problem are proposed, which are based on the notion of pseudo‐components and on a bilevel model. The efficiency of the new formulations is illustrated by the numerical study that uses two reference networks (medium‐size and large‐size), and a wide range of the sizes of attacks and controllers placements. [ABSTRACT FROM AUTHOR]

Methods for finding pure Nash equilibria have been dominated by variational inequalities and complementarity problems. Since these approaches fundamentally rely on the sufficiency of first-order optimality conditions for the players' decision problems, they only apply as heuristic methods when the players are modeled by nonconvex optimization problems. In contrast, this work approaches Nash equilibrium using theory and methods for the global optimization of nonconvex bilevel programs. Through this perspective, we draw precise connections between Nash equilibria, feasibility for bilevel programming, the Nikaido–Isoda function, and classic arguments involving Lagrangian duality and spatial price equilibrium. Significantly, this is all in a general setting without the assumption of convexity. Along the way, we introduce the idea of minimum disequilibrium as a solution concept that reduces to traditional equilibrium when an equilibrium exists. The connections with bilevel programming and related semi-infinite programming permit us to adapt global optimization methods for those classes of problems, such as constraint generation or cutting plane methods, to the problem of finding a minimum disequilibrium solution. We propose a specific algorithm and show that this method can find a pure Nash equilibrium even when the players are modeled by mixed-integer programs. Our computational examples include practical applications like unit commitment in electricity markets. [ABSTRACT FROM AUTHOR]

We propose a new approach to solving bilevel optimization problems, intermediate between solving full-system optimality conditions with a Newton-type approach, and treating the inner problem as an implicit function. The overall idea is to solve the full-system optimality conditions, but to precondition them to alternate between taking steps of simple conventional methods for the inner problem, the adjoint equation, and the outer problem. While the inner objective has to be smooth, the outer objective may be nonsmooth subject to a prox-contractivity condition. We prove linear convergence of the approach for combinations of gradient descent and forward-backward splitting with exact and inexact solution of the adjoint equation. We demonstrate good performance on learning the regularization parameter for anisotropic total variation image denoising, and the convolution kernel for image deconvolution. [ABSTRACT FROM AUTHOR]

An innovative strategy for an optimal design of planar frames able to resist seismic excitations is proposed. The optimal design is performed considering the cross sections of beams and columns as design variables. The procedure is based on genetic algorithms (GA) that are performed according to a nested structure suitable to be implemented in parallel on several computing devices. In particular, this bi-level optimization involves two nested genetic algorithms. The first external one seeks the size of the structural elements of the frame which corresponds to the most performing solution associated with the highest value of an appropriate fitness function. The latter function takes into account, among other considerations, the seismic safety factor and the failure mode that are calculated by means of the second internal algorithm. The proposed procedure aims at representing a prompt performance-based design procedure which observes earthquake engineering principles, that is displacement capacity and energy dissipation, although based on a limit analysis, thus avoiding the need of performing cumbersome nonlinear analyses. The details of the proposed procedure are provided and applications to the seismic design of two frames of different size are described. [ABSTRACT FROM AUTHOR]

In this paper, an approach is suggested to solve nonlinear bilevel programming (NBLP) problems. In the suggested method, we convert the NBLP problem into a standard nonlinear programming problem with complementary constraints by applying the Karush-Kuhn-Tucker condition to the lower-level problem. By using the Chen-Harker-Kanzow-Smale (CHKS) smoothing function, the nonlinear programming problem is successively smoothed. A nonmonton active interior-point trust-region algorithm is introduced to solve the smoothed nonlinear programming problem to obtain an approximately optimal solution to the NBLP problem. Results from simulations on several benchmark problems and a real-world case about a watershed trading decision-making problem show how the effectiveness of the suggested approach in NBLP solution development. [ABSTRACT FROM AUTHOR]

By classifying loads as elastic or inelastic and restructuring the load demand model, demand side management (DSM) may help bring down the distribution system's operational costs. One way to accomplish this goal is to shift the time of day that the flexible loads are used to one that has a lower utility cost per unit of use. Using renewable energy sources (RES), and fossil fuel-powered generators in grid-connected mode, this article implements a bi-level optimization technique to reduce operational costs. First, the load model is reorganized in accordance with the DSM involvement level. The second stage involves considering the revised load demand model and percolating an ideal allocation for distributed generators (DGs) to reduce the microgrid's generation costs. This study made use of a novel hybrid swarm intelligence algorithm developed for the purpose of optimizing power systems, which has proven effective in the past for a wide range of optimization problems. All types of grid involvement and price methods were explored for producing cost without considering DSM. When 10–30% DSM involvement was considered, the numerical findings showed a noticeable decrease in generating cost. A detailed statistical and sensitivity analysis of the proposed hybrid optimization tool also corroborated about the robustness and preciseness of the same. [ABSTRACT FROM AUTHOR]

Green Supply Chain Management requires coordinated decisions between the strategic and operational organization layers to address strict green goals. Furthermore, linking CO2 emissions to supply chain operations is not always easy. This study proposes a new mathematical model to minimize CO2 emissions in a three-layered supply chain. The model foresees using a financial budget to mitigate emissions contributions and optimize supply chain operations planning. The three-stage supply chain analyzed has inbound logistics and handling operations at the intermediate level. We assume that these operations contribute to emissions quadratically. The resulting bilevel programming problem is solved by transforming it into a nonlinear mixed-integer program by applying the Karush-Kuhn-Tucker conditions. We show, on different sets of synthetic data and on a case study, how our proposal produces solutions with a different flow of goods than a modified linear model version. This results in lower CO2 emissions and more efficient budget expenditure.

Mass customization enables the integration of traditional flow line production with product platforms to accommodate abundant product-process varieties. These platform-based flow lines explore common process routes while highlighting rebatching scheduling with selectivity banks (RBS) to handle large process varieties across production stages at minimum setup cost. Given the inherent coupling between decision making in job diverging and retrieval quality, an interactive optimization approach is necessary for the RBS problem. This study proposes a bilevel interactive optimization (BIO) model for RBS to accommodate high variety flow line production. The model addresses the conflicting goals of lane occupancy cost, process setup cost, and job divergence and retrieval efficiency. Regarding job divergence at the leader-level, a vehicle routeing problem with precedence constraints is formulated and solved by a constructed genetic algorithm (GA). Concerning job retrieval at the follower-level and the ongoing characteristic of selectivity banks, a dispatching problem with various batch size preference and dynamic time window is established and dealt with a restricted dynamic programming (RDP) algorithm after balancing search efficiency and accuracy. Thus, to solve the BIO, a hybrid GA-RDP is developed and implemented. A practical application to an automotive painting shop illustrates the operational benefits of the BIO model for the RBS problem. [ABSTRACT FROM AUTHOR]

The volatility of renewable energy poses challenges to the stability and economic benefits of the power grid, as the unstable output of wind and solar energy increases the difficulty of supply-demand balance. To address this issue, optimizing the scheduling strategy of wind-photovoltaic hybrid power generation systems to deal with the uncertainty of renewable energy has become an urgent problem to be addressed. This optimization can not only improve the adaptability of the power grid to fluctuations in renewable energy, but also enhance economic efficiency by reducing reliance on expensive energy storage and backup power sources. The study adopted the methods of bilevel programming and sparse optimization, in which system operators optimize operating costs and system efficiency by adjusting the output ratio of wind-photovoltaic power, energy storage system operation, and grid interaction. The power grid operator adjusted the scheduling plan based on upper level decisions to ensure the stability of the power grid. Sparse optimization techniques were applied to improve the sparsity of solutions and the generalization ability of models. The research results showed that the proposed bilevel programming and sparse optimization strategies performed well in simulation experiments. The SOP-MLP model achieved a recall rate of 0.98 and an precision rate that quickly stabilized at over 90% after 400 training cycles, outperforming traditional MLP, Transformer, SVM, and Extra-Trees models. In the case analysis, the SOP-MLP model effectively reduced abandoned electricity and optimized power resource allocation. In the S1 scenario, the comprehensive dispatch response rates of wind-photovoltaic power reached 94% and 91%, while the expected operating cost of the system was 705600 yuan. The cost-benefit ratio considering scenario probability was 79400 yuan. The study provides a new optimization strategy for power system scheduling, which can effectively handle the uncertainty and complexity of wind-photovoltaic grid connection, and has important theoretical value and practical application prospects.

Much attention has been paid to how limited in-orbit satellite resources can be better utilized to meet the increasingly heavy demand for space observation. A variety of single-stage optimization problems have been discussed, with few considerations regarding the effect of the two stages in multi-satellite cooperative observation mission planning. In this study, bilevel programming is applied to simultaneously consider both mission assignment and satellite scheduling. The purpose of the bilevel programming model is to optimize the planning scheme of multi-satellite cooperative observation mission and maximize the comprehensive benefit from the perspective of the system as a whole. The upper level of the model formulates the mission assignment scheme, and the lower level determines the optimal resource scheduling scheme by a mathematical method on the basis of the upper level, and then feeds the results back to the upper level. The upper level and lower level affect each other, and the optimal solution is obtained through an iterative process under the solution framework of genetic algorithm. Extensive experiments are simulated to demonstrate the feasibility and efficiency of the proposed bilevel programming model.

The pricing problem for demand-responsive feeder transit (RFT) in competitive relationships with public bicycles in bicycle-sharing systems is investigated from a game-theoretic perspective to establish a feeder program for passengers traveling to and from transport hubs. First, a revenue model including the base fare and passenger travel delay penalty is constructed. In this model, the base fares of responsive feeder buses and public bicycles are functions of the distance from the demand point to the metro station and the rental period, respectively. Then, the pricing game process under mutual competition between responsive feeder buses and public bicycles is analyzed. Moreover, a nested two-layer planning pricing model is constructed. Finally, a solution algorithm based on sensitivity analyses and a genetic algorithm is designed. The results show that this method can be used to effectively determine the unit fares and passenger flow sharing rates of RFT and public bicycles. Compared with that under the single-vehicle RFT operation mode, the RFT fare under the two-vehicle RFT operation mode is lower, and the passenger flow sharing rate is higher, which can significantly increase the social benefits. However, the total revenue of the operation is significantly reduced and thus requires greater financial input from the government. Under the dual-vehicle operation mode, the fares of public bicycles are greater than those of RFT, and the passenger flow sharing rate is low, which significantly reduces both the social benefits and the total benefits of the operation. With the improvement of the operation mode of the RFT system, the competitiveness of using public bicycles decreases.

The increasing proportion of distributed photovoltaics (DPVs) and electric vehicle charging stations in low-voltage distribution networks (LVDNs) has resulted in challenges such as distribution transformer overloads and voltage violations. To address these problems, we propose a coordinated planning method for flexible interconnections and energy storage systems (ESSs) to improve the accommodation capacity of DPVs. First, the power-transfer characteristics of flexible interconnection and ESSs are analyzed. The equipment costs of the voltage source converters (VSCs) and ESSs are also analyzed comprehensively, considering the differences in installation and maintenance costs for different installation locations. Second, a bilevel programming model is established to minimize the annual comprehensive cost and yearly total PV curtailment capacity. Within this framework, the upper-level model optimizes the installation locations and capacities of the VSCs and ESSs, whereas the lower-level model optimizes the operating power of the VSCs and ESSs. The proposed model is solved using a non-dominated sorting genetic algorithm with an elite strategy (NSGA-II). The effectiveness of the proposed planning method is validated through an actual LVDN scenario, which demonstrates its advantages in enhancing PV accommodation capacity. In addition, the economic benefits of various planning schemes with different flexible interconnection topologies and different PV grid-connected forms are quantitatively analyzed, demonstrating the adaptability of the proposed coordinated planning method.