The design of Human Occupied Vehicle (HOV) is a typical multidisciplinary problem, but heavily dependent on the experience of naval architects at present engineering design. In order to relieve the experience depend...The design of Human Occupied Vehicle (HOV) is a typical multidisciplinary problem, but heavily dependent on the experience of naval architects at present engineering design. In order to relieve the experience dependence and improve the design, a new Multidisciplinary Design Optimization (MDO) method "Bi-Level Integrated System Collaborative Optimization (BLISCO)" is applied to the conceptual design of an HOV, which consists of hull module, resistance module, energy module, structure module, weight module, and the stability module. This design problem is defined by 21 design variables and 23 constraints, and its objective is to maximize the ratio of payload to weight. The results show that the general performance of the HOV can be greatly improved by BLISCO.展开更多
Since COVID-19 was declared as a pandemic in March 2020,the world’s major preoccupation has been to curb it while preserving the economy and reducing unemployment.This paper uses a novel Bi-Level Dynamic Optimal Cont...Since COVID-19 was declared as a pandemic in March 2020,the world’s major preoccupation has been to curb it while preserving the economy and reducing unemployment.This paper uses a novel Bi-Level Dynamic Optimal Control model(BLDOC)to coordinate control between COVID-19 and unemployment.The COVID-19 model is the upper level while the unemployment model is the lower level of the bi-level dynamic optimal control model.The BLDOC model’s main objectives are to minimize the number of individuals infected with COVID-19 and to minimize the unemployed individuals,and at the same time minimizing the cost of the containment strategies.We use the modified approximation Karush–Kuhn–Tucker(KKT)conditions with the Hamiltonian function to handle the bi-level dynamic optimal control model.We consider three control variables:The first control variable relates to government measures to curb the COVID-19 pandemic,i.e.,quarantine,social distancing,and personal protection;and the other two control variables relate to government interventions to reduce the unemployment rate,i.e.,employment,making individuals qualified,creating new jobs reviving the economy,reducing taxes.We investigate four different cases to verify the effect of control variables.Our results indicate that rather than focusing exclusively on only one problem,we need a balanced trade-off between controlling each.展开更多
In this work we propose a solution method based on Lagrange relaxation for discrete-continuous bi-level problems, with binary variables in the leading problem, considering the optimistic approach in bi-level programmi...In this work we propose a solution method based on Lagrange relaxation for discrete-continuous bi-level problems, with binary variables in the leading problem, considering the optimistic approach in bi-level programming. For the application of the method, the two-level problem is reformulated using the Karush-Kuhn-Tucker conditions. The resulting model is linearized taking advantage of the structure of the leading problem. Using a Lagrange relaxation algorithm, it is possible to find a global solution efficiently. The algorithm was tested to show how it performs.展开更多
An algorithm is proposed in this paper for solving two-dimensional bi-level linear programming problems without making a graph. Based on the classification of constraints, algorithm removes all redundant constraints, ...An algorithm is proposed in this paper for solving two-dimensional bi-level linear programming problems without making a graph. Based on the classification of constraints, algorithm removes all redundant constraints, which eliminate the possibility of cycling and the solution of the problem is reached in a finite number of steps. Example to illustrate the method is also included in the paper.展开更多
The urban transit fare structure and level can largely affect passengers’travel behavior and route choices.The commonly used transit fare policies in the present transit network would lead to the unbalanced transit a...The urban transit fare structure and level can largely affect passengers’travel behavior and route choices.The commonly used transit fare policies in the present transit network would lead to the unbalanced transit assignment and improper transit resources distribution.In order to distribute transit passenger flow evenly and efficiently,this paper introduces a new distance-based fare pattern with Euclidean distance.A bi-level programming model is developed for determining the optimal distance-based fare pattern,with the path-based stochastic transit assignment(STA)problem with elastic demand being proposed at the lower level.The upper-level intends to address a principal-agent game between transport authorities and transit enterprises pursing maximization of social welfare and financial interest,respectively.A genetic algorithm(GA)is implemented to solve the bi-level model,which is verified by a numerical example to illustrate that the proposed nonlinear distance-based fare pattern presents a better financial performance and distribution effect than other fare structures.展开更多
As the proportion of renewable energy power generation continues to increase,the number of grid-connected microgrids is gradually increasing,and geographically adjacent microgrids can be interconnected to form a Micro...As the proportion of renewable energy power generation continues to increase,the number of grid-connected microgrids is gradually increasing,and geographically adjacent microgrids can be interconnected to form a Micro-Grid Community(MGC).In order to reduce the operation and maintenance costs of a single micro grid and reduce the adverse effects caused by unnecessary energy interaction between the micro grid and the main grid while improving the overall economic benefits of the micro grid community,this paper proposes a bi-level energy management model with the optimization goal of maximizing the social welfare of the micro grid community and minimizing the total electricity cost of a single micro grid.The lower-level model optimizes the output of each equipment unit in the system and the exchange power between the system and the external grid with the goal of minimizing the operating cost of each microgrid.The upper-level model optimizes the goal ofmaximizing the socialwelfare of themicrogrid.Taking amicrogrid community with four microgrids as an example,the simulation analysis shows that the proposed optimization model is beneficial to reduce the operating cost of a single microgrid,improve the overall revenue of the microgrid community,and reduce the power interaction pressure on the main grid.展开更多
Objective:To analyze the clinical efficacy of early application of bi-level positive airway pressure ventilation in the treatment of COPD with type II respiratory failure.Method:A total of 58 patients with COPD and ty...Objective:To analyze the clinical efficacy of early application of bi-level positive airway pressure ventilation in the treatment of COPD with type II respiratory failure.Method:A total of 58 patients with COPD and type II respiratory failure admitted to our hospital from January 2017 to January 2019 were randomly divided into observation group and control group,with 29 cases in each group.Among them,the control group was received routine treatment while the observation group was treated with bi-level positive pressure airway ventilation in addition of conventional treatment.The arterial blood gas analysis,mortality rate and hospitalization time of these two groups before and after treatment were compared.Result:The blood pH,partial pressure of oxygen(PaO2)and arterial oxygen saturation(SaO2)of these two groups were significantly higher after the treatment while PaO2 alone was decreased.The difference was statistically significant(P<0.05).The results of arterial blood gas analysis in the observation group were significantly improved compared with those before treatment.The mortality rate and hospitalization time were significantly less than the control group,and the difference was statistically significant(P<0.05).Conclusion:Early clinical application of bi-level positive airway pressure ventilation in the treatment of COPD with type II respiratory failure has a significant clinical effect in reducing the mortality rate and hospitalization time of patients,and thus it is worthy of clinical application.展开更多
Features in educational data are ambiguous which leads to noisy features and curse of dimensionality problems.These problems are solved via feature selection.There are existing models for features selection.These mode...Features in educational data are ambiguous which leads to noisy features and curse of dimensionality problems.These problems are solved via feature selection.There are existing models for features selection.These models were created using either a single-level embedded,wrapper-based or filter-based methods.However single-level filter-based methods ignore feature dependencies and ignore the interaction with the classifier.The embedded and wrapper based feature selection methods interact with the classifier,but they can only select the optimal subset for a particular classifier.So their selected features may be worse for other classifiers.Hence this research proposes a robust Cascade Bi-Level(CBL)feature selection technique for student performance prediction that will minimize the limitations of using a single-level technique.The proposed CBL feature selection technique consists of the Relief technique at first-level and the Particle Swarm Optimization(PSO)at the second-level.The proposed technique was evaluated using the UCI student performance dataset.In comparison with the performance of the single-level feature selection technique the proposed technique achieved an accuracy of 94.94%which was better than the values achieved by the single-level PSO with an accuracy of 93.67%for the binary classification task.These results show that CBL can effectively predict student performance.展开更多
In silico approaches for metabolites optimization have been derived from the flood of sequenced and annotated genomes. However, there exist still numerous degrees of freedom in terms of optimization algorithm approach...In silico approaches for metabolites optimization have been derived from the flood of sequenced and annotated genomes. However, there exist still numerous degrees of freedom in terms of optimization algorithm approaches that can be exploited in order to enhance yield of processes which are based on biological reactions. Here, we propose an evolutionary approach aiming to suggest different mutant for augmenting ethanol yield using glycerol as substrate in Escherichia coli. We found that this algorithm, even though is far from providing the global optimum, is able to uncover genes that a global optimizer would be incapable of. By over-expressing accB, eno, dapE, and accA mutants in ethanol production was augmented up to 2 fold compared to its counterpart E. coli BW25113.展开更多
Considering the decision-making variables of the capacities of branch roads and the optimization targets of lowering the saturation of arterial roads and the reconstruction expense of branch roads, the bi-level progra...Considering the decision-making variables of the capacities of branch roads and the optimization targets of lowering the saturation of arterial roads and the reconstruction expense of branch roads, the bi-level programming model for reconstructing the branch roads was set up. The upper level model was for determining the enlarged capacities of the branch roads, and the lower level model was for calculating the flows of road sections via the user equilibrium traffic assignment method. The genetic algorithm for solving the bi-level model was designed to obtain the reconstruction capacities of the branch roads. The results show that by the bi-level model and its algorithm, the optimum scheme of urban branch roads reconstruction can be gained, which reduces the saturation of arterial roads apparently, and alleviates traffic congestion. In the data analysis the arterial saturation decreases from 1.100 to 0.996, which verifies the micro-circulation transportation's function of urban branch road network.展开更多
Based on genetic algorithms, a solution algorithm is presented for the bi-level decision making problem with continuous variables in the upper level in accordance with the bi-level decision making principle. The algor...Based on genetic algorithms, a solution algorithm is presented for the bi-level decision making problem with continuous variables in the upper level in accordance with the bi-level decision making principle. The algorithm is compared with Monte Carlo simulated annealing algorithm, and its feasibility and effectiveness are verified with two calculating examples.展开更多
Multi-objective bi-level optimization(MOBLO)addresses nested multi-objective optimization problems common in a range of applications.However,its multi-objective and hierarchical bi-level nature makes it notably comple...Multi-objective bi-level optimization(MOBLO)addresses nested multi-objective optimization problems common in a range of applications.However,its multi-objective and hierarchical bi-level nature makes it notably complex.Gradient-based MOBLO algorithms have recently grown in popularity,as they effectively solve crucial machine learning problems like meta-learning,neural architecture search,and reinforcement learning.Unfortunately,these algorithms depend on solving a sequence of approximation subproblems with high accuracy,resulting in adverse time and memory complexity that lowers their numerical efficiency.To address this issue,we propose a gradient-based algorithm for MOBLO,called gMOBA,which has fewer hyperparameters to tune,making it both simple and efficient.Additionally,we demonstrate the theoretical validity by accomplishing the desirable Pareto stationarity.Numerical experiments confirm the practical efficiency of the proposed method and verify the theoretical results.To accelerate the convergence of gMOBA,we introduce a beneficial L2O(learning to optimize)neural network(called L2O-gMOBA)implemented as the initialization phase of our gMOBA algorithm.Comparative results of numerical experiments are presented to illustrate the performance of L2O-gMOBA.展开更多
How to effectively use the multi-energy demand elasticity of users to bid in the multi-energy market and formulate multi-energy retail packages is an urgent problem which needs to be solved by integrated energy servic...How to effectively use the multi-energy demand elasticity of users to bid in the multi-energy market and formulate multi-energy retail packages is an urgent problem which needs to be solved by integrated energy service providers(IESPs)to attract more users and reduce operating costs.This paper presents a unified clearing of electricity and natural gas based on a bi-level bidding and multi-energy retail price formulation method for IESPs considering multi-energy demand elasticity.First,we propose an operating structure of IESPs in the wholesale and retail energy markets.The multi-energy demand elasticity model of retail-side users and a retail price model for electricity,gas,heat and cooling are established.Secondly,a bi-level bidding model for IESPs considering multi-energy demand elasticity is established to provide IESPs with wholesale-side bidding decisions and retail-side energy retail price decisions.Finally,an example is given to verify the proposed method.The results show that the method improves the total social welfare of the electricity and natural gas markets by 7.99%and the profit of IESPs by 1.40%.It can reduce the variance of the electricity,gas,and cooling load curves,especially the reduction of the variance of the electricity load curve can which reach 79.90%.It can be seen that the research in this paper has a positive effect on repairing the limitations of integrated energy trading research and improving the economics of the operation of IESPs.展开更多
This paper focuses on optimal voltage regulator(VR)planning to maximize the photovoltaic(PV)energy integration in distribution grids.To describe the amount of dynamic PV energy that can be integrated into the power sy...This paper focuses on optimal voltage regulator(VR)planning to maximize the photovoltaic(PV)energy integration in distribution grids.To describe the amount of dynamic PV energy that can be integrated into the power system,the concept of PV accommodation capability(PVAC)is introduced and modeled with optimization.Our proposed planning model is formulated as a Benders decomposition based bi-level stochastic optimization problem.In the upper-level problem,VR planning decisions and PVAC are determined via mixed integer linear programming(MILP)before considering uncertainty.Then in the lower-level problem,the feasibility of first-level results is checked by critical network constraints(e.g.voltage magnitude constraints and line capacity constraints)under uncertainties considered by time-varying loads and PV generations.In this paper,these uncertainties are represented in the form of operational scenarios,which are generated by the Gaussian copula theory and reduced by a well-studied backward-reduction algorithm.The modified IEEE 33-node distribution grid is utilized to verify the effectiveness of the proposed model.The results demonstrate that a PV energy integration can be significantly enhanced after optimal voltage regulator planning.展开更多
This paper proposes a robust and computationally efficient control method for damping ultra-low frequency oscillations(ULFOs) in hydropower-dominated systems. Unlike the existing robust optimization based control form...This paper proposes a robust and computationally efficient control method for damping ultra-low frequency oscillations(ULFOs) in hydropower-dominated systems. Unlike the existing robust optimization based control formulation that can only deal with a limited number of operating conditions, the proposed method reformulates the control problem into a bi-level robust parameter optimization model. This allows us to consider a wide range of system operating conditions. To speed up the bi-level optimization process, the deep deterministic policy gradient(DDPG) based deep reinforcement learning algorithm is developed to train an intelligent agent. This agent can provide very fast lower-level decision variables for the upper-level model, significantly enhancing its computational efficiency. Simulation results demonstrate that the proposed method can achieve much better damping control performance than other alternatives with slightly degraded dynamic response performance of the governor under various types of operating conditions.展开更多
基金financially supported by the National Natural Science Foundation of China(Grant No.51109132)the Specialized Research Fund for the Doctoral Program of Higher Education of China(Grant No.20110073120015)
文摘The design of Human Occupied Vehicle (HOV) is a typical multidisciplinary problem, but heavily dependent on the experience of naval architects at present engineering design. In order to relieve the experience dependence and improve the design, a new Multidisciplinary Design Optimization (MDO) method "Bi-Level Integrated System Collaborative Optimization (BLISCO)" is applied to the conceptual design of an HOV, which consists of hull module, resistance module, energy module, structure module, weight module, and the stability module. This design problem is defined by 21 design variables and 23 constraints, and its objective is to maximize the ratio of payload to weight. The results show that the general performance of the HOV can be greatly improved by BLISCO.
基金The authors extend their appreciation to the Deanship of Scientific Research at King Saud University for funding this work through research Group No.RG-1441-309.
文摘Since COVID-19 was declared as a pandemic in March 2020,the world’s major preoccupation has been to curb it while preserving the economy and reducing unemployment.This paper uses a novel Bi-Level Dynamic Optimal Control model(BLDOC)to coordinate control between COVID-19 and unemployment.The COVID-19 model is the upper level while the unemployment model is the lower level of the bi-level dynamic optimal control model.The BLDOC model’s main objectives are to minimize the number of individuals infected with COVID-19 and to minimize the unemployed individuals,and at the same time minimizing the cost of the containment strategies.We use the modified approximation Karush–Kuhn–Tucker(KKT)conditions with the Hamiltonian function to handle the bi-level dynamic optimal control model.We consider three control variables:The first control variable relates to government measures to curb the COVID-19 pandemic,i.e.,quarantine,social distancing,and personal protection;and the other two control variables relate to government interventions to reduce the unemployment rate,i.e.,employment,making individuals qualified,creating new jobs reviving the economy,reducing taxes.We investigate four different cases to verify the effect of control variables.Our results indicate that rather than focusing exclusively on only one problem,we need a balanced trade-off between controlling each.
文摘In this work we propose a solution method based on Lagrange relaxation for discrete-continuous bi-level problems, with binary variables in the leading problem, considering the optimistic approach in bi-level programming. For the application of the method, the two-level problem is reformulated using the Karush-Kuhn-Tucker conditions. The resulting model is linearized taking advantage of the structure of the leading problem. Using a Lagrange relaxation algorithm, it is possible to find a global solution efficiently. The algorithm was tested to show how it performs.
文摘An algorithm is proposed in this paper for solving two-dimensional bi-level linear programming problems without making a graph. Based on the classification of constraints, algorithm removes all redundant constraints, which eliminate the possibility of cycling and the solution of the problem is reached in a finite number of steps. Example to illustrate the method is also included in the paper.
基金the Humanities and Social Science Foundation of the Ministry of Education of China(Grant No.20YJCZH121).
文摘The urban transit fare structure and level can largely affect passengers’travel behavior and route choices.The commonly used transit fare policies in the present transit network would lead to the unbalanced transit assignment and improper transit resources distribution.In order to distribute transit passenger flow evenly and efficiently,this paper introduces a new distance-based fare pattern with Euclidean distance.A bi-level programming model is developed for determining the optimal distance-based fare pattern,with the path-based stochastic transit assignment(STA)problem with elastic demand being proposed at the lower level.The upper-level intends to address a principal-agent game between transport authorities and transit enterprises pursing maximization of social welfare and financial interest,respectively.A genetic algorithm(GA)is implemented to solve the bi-level model,which is verified by a numerical example to illustrate that the proposed nonlinear distance-based fare pattern presents a better financial performance and distribution effect than other fare structures.
基金This paper is supported by Science and Technology Project of State Grid(The construction of provincial energy big data ecosystem and the application practice research of data value-added service for the park,5400-202012224A-0-0-00).
文摘As the proportion of renewable energy power generation continues to increase,the number of grid-connected microgrids is gradually increasing,and geographically adjacent microgrids can be interconnected to form a Micro-Grid Community(MGC).In order to reduce the operation and maintenance costs of a single micro grid and reduce the adverse effects caused by unnecessary energy interaction between the micro grid and the main grid while improving the overall economic benefits of the micro grid community,this paper proposes a bi-level energy management model with the optimization goal of maximizing the social welfare of the micro grid community and minimizing the total electricity cost of a single micro grid.The lower-level model optimizes the output of each equipment unit in the system and the exchange power between the system and the external grid with the goal of minimizing the operating cost of each microgrid.The upper-level model optimizes the goal ofmaximizing the socialwelfare of themicrogrid.Taking amicrogrid community with four microgrids as an example,the simulation analysis shows that the proposed optimization model is beneficial to reduce the operating cost of a single microgrid,improve the overall revenue of the microgrid community,and reduce the power interaction pressure on the main grid.
文摘Objective:To analyze the clinical efficacy of early application of bi-level positive airway pressure ventilation in the treatment of COPD with type II respiratory failure.Method:A total of 58 patients with COPD and type II respiratory failure admitted to our hospital from January 2017 to January 2019 were randomly divided into observation group and control group,with 29 cases in each group.Among them,the control group was received routine treatment while the observation group was treated with bi-level positive pressure airway ventilation in addition of conventional treatment.The arterial blood gas analysis,mortality rate and hospitalization time of these two groups before and after treatment were compared.Result:The blood pH,partial pressure of oxygen(PaO2)and arterial oxygen saturation(SaO2)of these two groups were significantly higher after the treatment while PaO2 alone was decreased.The difference was statistically significant(P<0.05).The results of arterial blood gas analysis in the observation group were significantly improved compared with those before treatment.The mortality rate and hospitalization time were significantly less than the control group,and the difference was statistically significant(P<0.05).Conclusion:Early clinical application of bi-level positive airway pressure ventilation in the treatment of COPD with type II respiratory failure has a significant clinical effect in reducing the mortality rate and hospitalization time of patients,and thus it is worthy of clinical application.
文摘Features in educational data are ambiguous which leads to noisy features and curse of dimensionality problems.These problems are solved via feature selection.There are existing models for features selection.These models were created using either a single-level embedded,wrapper-based or filter-based methods.However single-level filter-based methods ignore feature dependencies and ignore the interaction with the classifier.The embedded and wrapper based feature selection methods interact with the classifier,but they can only select the optimal subset for a particular classifier.So their selected features may be worse for other classifiers.Hence this research proposes a robust Cascade Bi-Level(CBL)feature selection technique for student performance prediction that will minimize the limitations of using a single-level technique.The proposed CBL feature selection technique consists of the Relief technique at first-level and the Particle Swarm Optimization(PSO)at the second-level.The proposed technique was evaluated using the UCI student performance dataset.In comparison with the performance of the single-level feature selection technique the proposed technique achieved an accuracy of 94.94%which was better than the values achieved by the single-level PSO with an accuracy of 93.67%for the binary classification task.These results show that CBL can effectively predict student performance.
基金the support of the National BioResource Project(NIG,Japan):E.coli Strain for kindly providing us with the Keio Collection using for our experimental sectionAlso this work is funded by Vicerrectoria de investigaciones at Universidad de los Andes.
文摘In silico approaches for metabolites optimization have been derived from the flood of sequenced and annotated genomes. However, there exist still numerous degrees of freedom in terms of optimization algorithm approaches that can be exploited in order to enhance yield of processes which are based on biological reactions. Here, we propose an evolutionary approach aiming to suggest different mutant for augmenting ethanol yield using glycerol as substrate in Escherichia coli. We found that this algorithm, even though is far from providing the global optimum, is able to uncover genes that a global optimizer would be incapable of. By over-expressing accB, eno, dapE, and accA mutants in ethanol production was augmented up to 2 fold compared to its counterpart E. coli BW25113.
基金Project(2006CB705507) supported by the National Basic Research and Development Program of ChinaProject(20060533036) supported by the Specialized Research Foundation for the Doctoral Program of Higher Education of China
文摘Considering the decision-making variables of the capacities of branch roads and the optimization targets of lowering the saturation of arterial roads and the reconstruction expense of branch roads, the bi-level programming model for reconstructing the branch roads was set up. The upper level model was for determining the enlarged capacities of the branch roads, and the lower level model was for calculating the flows of road sections via the user equilibrium traffic assignment method. The genetic algorithm for solving the bi-level model was designed to obtain the reconstruction capacities of the branch roads. The results show that by the bi-level model and its algorithm, the optimum scheme of urban branch roads reconstruction can be gained, which reduces the saturation of arterial roads apparently, and alleviates traffic congestion. In the data analysis the arterial saturation decreases from 1.100 to 0.996, which verifies the micro-circulation transportation's function of urban branch road network.
文摘Based on genetic algorithms, a solution algorithm is presented for the bi-level decision making problem with continuous variables in the upper level in accordance with the bi-level decision making principle. The algorithm is compared with Monte Carlo simulated annealing algorithm, and its feasibility and effectiveness are verified with two calculating examples.
文摘目的:评价经鼻持续气道正压通气(NCPAP)和双水平NCPAP(Bi-level NCPAP)对早产儿中度呼吸窘迫综合征(RDS)的治疗效果及对炎症反应的影响。方法:42例符合标准中度RDS的早产儿,胎龄28~34周,随机分为A、B两组,分别采用NCPAP治疗(压力6 cmH2O)和Bi-level NCPAP治疗(低压4.0 cm H2O,高压7.5 cm H2O)。在出生后第1、7日检测早产儿血清细胞因子(IL-6、IL-8、TNF-α)水平,记录患儿需要呼吸支持和氧依赖的时间以及出院时的胎龄,比较两组上述指标的差异。结果:两组早产儿均存活,无支气管肺发育不良或中枢神经系统疾病的发生。出生后第1、7日B组血清IL-6、IL-8、TNF-α水平均明显低于NCPAP组(P均<0.05)。两组早产儿组内不同时间血清三种细胞因子水平比较差异无统计学意义(P均>0.05)。A组需要呼吸支持的时间、氧依赖时间均长于B组(P均<0.05)、出院时胎龄大于B组(P<0.05)。结论:与NCPAP相比,Bi-level NCPAP能更好地改善通气、缩短呼吸支持和氧依赖的时间,缩短早产儿住院时间,所引起炎症反应程度也较NCPAP低,因此Bi-level NC-PAP对早产儿具有更好的耐受性和安全性。
基金supported by the Major Program of National Natural Science Foundation of China(Grant Nos.11991020 and 11991024)supported by National Natural Science Foundation of China(Grant No.12371305)+2 种基金supported by National Natural Science Foundation of China(Grant No.12222106)Guangdong Basic and Applied Basic Research Foundation(Grant No.2022B1515020082)Shenzhen Science and Technology Program(Grant No.RCYX20200714114700072)。
文摘Multi-objective bi-level optimization(MOBLO)addresses nested multi-objective optimization problems common in a range of applications.However,its multi-objective and hierarchical bi-level nature makes it notably complex.Gradient-based MOBLO algorithms have recently grown in popularity,as they effectively solve crucial machine learning problems like meta-learning,neural architecture search,and reinforcement learning.Unfortunately,these algorithms depend on solving a sequence of approximation subproblems with high accuracy,resulting in adverse time and memory complexity that lowers their numerical efficiency.To address this issue,we propose a gradient-based algorithm for MOBLO,called gMOBA,which has fewer hyperparameters to tune,making it both simple and efficient.Additionally,we demonstrate the theoretical validity by accomplishing the desirable Pareto stationarity.Numerical experiments confirm the practical efficiency of the proposed method and verify the theoretical results.To accelerate the convergence of gMOBA,we introduce a beneficial L2O(learning to optimize)neural network(called L2O-gMOBA)implemented as the initialization phase of our gMOBA algorithm.Comparative results of numerical experiments are presented to illustrate the performance of L2O-gMOBA.
基金supported in part by the National Key R&D Program of China(2018YFB0905000)the Science and Technology Project of the State Grid Corporation of China(SGTJDK 00DWJS1800232)。
文摘How to effectively use the multi-energy demand elasticity of users to bid in the multi-energy market and formulate multi-energy retail packages is an urgent problem which needs to be solved by integrated energy service providers(IESPs)to attract more users and reduce operating costs.This paper presents a unified clearing of electricity and natural gas based on a bi-level bidding and multi-energy retail price formulation method for IESPs considering multi-energy demand elasticity.First,we propose an operating structure of IESPs in the wholesale and retail energy markets.The multi-energy demand elasticity model of retail-side users and a retail price model for electricity,gas,heat and cooling are established.Secondly,a bi-level bidding model for IESPs considering multi-energy demand elasticity is established to provide IESPs with wholesale-side bidding decisions and retail-side energy retail price decisions.Finally,an example is given to verify the proposed method.The results show that the method improves the total social welfare of the electricity and natural gas markets by 7.99%and the profit of IESPs by 1.40%.It can reduce the variance of the electricity,gas,and cooling load curves,especially the reduction of the variance of the electricity load curve can which reach 79.90%.It can be seen that the research in this paper has a positive effect on repairing the limitations of integrated energy trading research and improving the economics of the operation of IESPs.
基金Natural Science Foundation of Guangdong(2019A1515111173)Young Talent Program(Dept of Education of Guangdong)(2018KQNCX223)+2 种基金High-level University Fund,G02236002National Natural Science Foundation of China(71971183)Hong Kong UGC PolyU Grant under Project P0038972.
文摘This paper focuses on optimal voltage regulator(VR)planning to maximize the photovoltaic(PV)energy integration in distribution grids.To describe the amount of dynamic PV energy that can be integrated into the power system,the concept of PV accommodation capability(PVAC)is introduced and modeled with optimization.Our proposed planning model is formulated as a Benders decomposition based bi-level stochastic optimization problem.In the upper-level problem,VR planning decisions and PVAC are determined via mixed integer linear programming(MILP)before considering uncertainty.Then in the lower-level problem,the feasibility of first-level results is checked by critical network constraints(e.g.voltage magnitude constraints and line capacity constraints)under uncertainties considered by time-varying loads and PV generations.In this paper,these uncertainties are represented in the form of operational scenarios,which are generated by the Gaussian copula theory and reduced by a well-studied backward-reduction algorithm.The modified IEEE 33-node distribution grid is utilized to verify the effectiveness of the proposed model.The results demonstrate that a PV energy integration can be significantly enhanced after optimal voltage regulator planning.
基金supported by the National Natural Science Foundation of China (No.52277083)。
文摘This paper proposes a robust and computationally efficient control method for damping ultra-low frequency oscillations(ULFOs) in hydropower-dominated systems. Unlike the existing robust optimization based control formulation that can only deal with a limited number of operating conditions, the proposed method reformulates the control problem into a bi-level robust parameter optimization model. This allows us to consider a wide range of system operating conditions. To speed up the bi-level optimization process, the deep deterministic policy gradient(DDPG) based deep reinforcement learning algorithm is developed to train an intelligent agent. This agent can provide very fast lower-level decision variables for the upper-level model, significantly enhancing its computational efficiency. Simulation results demonstrate that the proposed method can achieve much better damping control performance than other alternatives with slightly degraded dynamic response performance of the governor under various types of operating conditions.