Peter Luh | University of Connecticut (original) (raw)
Papers by Peter Luh
In current U.S. deregulated wholesale electricity markets, an auction mechanism that minimizes th... more In current U.S. deregulated wholesale electricity markets, an auction mechanism that minimizes the total bid cost is used to select bids and their output levels. Energy market clearing prices (MCPs) are derived from the shadow prices associated with the system demand constraints in an economic dispatch process with fixed unit commitment status. Therefore they do not reflect no-load and start-up costs, resulting in the uplift payments. To reduce such side payments and improve the market transparency, the ldquoconvex hull pricing modelrdquo is adopted, and ldquoonline capacity constraintsrdquo are introduced to the energy market, requiring the total online capacities to be greater than or equal to the system demand. With the associated multipliers serving as uniform ldquono-load clearing prices,rdquo the total uplift payment is proved to be reduced based on optimality conditions for multipliers. Numerical examples support the analytical results and shed insights on different types of uplift payments.
More and more wind generation is introduced to electricity markets. Yet, it is not clear how to c... more More and more wind generation is introduced to electricity markets. Yet, it is not clear how to conduct auctions with significant stochastic wind generation while avoiding dramatically increasing computational complexity. In this paper, the Markov process concept is innovatively used to model wind generation with transition matrices established based on the historical data. Wind generation is then integrated into system demand, and the day-ahead auction problem considering system demand, unit maximal and minimal generations and ramp rates is formulated. The goal is to optimize the commitment and dispatch of normal generators to minimize the total expected cost while satisfying all possible loads. Lagrangian relaxation is used to demonstrate the capability of the formulation. Since there are many states at each hour, a novel enhanced version of Dynamic Programming is developed to solve the subproblems. Numerical testing results obtained by CPLEX demonstrate the effectiveness and scalability of the method compared with the standard stochastic programming formulation based on scenarios, and the feasibility compared with the deterministic formulation.
In the current US day-ahead electricity markets, auction is complicated because of the existence ... more In the current US day-ahead electricity markets, auction is complicated because of the existence of both discrete variables and continuous variables, which implies that there may be no linear prices to support equilibrium. To meet the system demand, the Independent System Operator (ISO) uses lump sum “uplift payments” to make generators follow its schedule. However, these uplift payments are opaque and difficult to hedge in terms of price signals. In addition, there are incentives for suppliers to bid other than their true costs. Untruthful bidding is undesirable, since in this case minimizing the total bid cost does not imply minimizing the total production cost, or maximizing the social welfare. As a result, the market may not be efficient. In this paper, system-wide redundant constraints on weighted generation levels are introduced to the day-ahead energy market to achieve equilibrium. With multipliers relaxing system demand constraints and redundant constraints serving as multipart prices, equilibrium can be achieved under the general condition that the optimal primal solution is among subproblem solutions for all units. Although incentive compatibility may not be achievable, the novel “local incentive compatibility” concept is introduced and quantified as the second best alternative to incentive compatibility.
ABSTRACT for the past several years, Kleinman and his associates . . . have collaborated on an in... more ABSTRACT for the past several years, Kleinman and his associates . . . have collaborated on an interdisciplinary program of experimental and theoretical research involving applied mathematics, experimental psychology, systems and human engineering, and computer science to develop a normative-descriptive theory to address problems of team-distributed decision making / they have developed empirically validated normative-descriptive models that capture the complexity, dynamicity, and uncertainty of the task environment and, in turn, quantify the resulting team performance, coordination, and decision strategies in specific situations objectives of this chapter are to present several of these models, to introduce the mathematical tools used to support the modeling activity, to demonstrate the process by which the normative-descriptive theory is applied to generate predictions of actual team performance, and to offer some novel hypotheses on team decision-making behavior and performance / point out the characteristics of the distributed dynamic decision-making problems that have been addressed, to define the normative-descriptive modeling approach, and to describe a unique research paradigm for distributed decision making that has been used to generate and collect the data to develop and validate the mathematical models (PsycINFO Database Record (c) 2012 APA, all rights reserved)
IEEE Control Systems Magazine, 1990
Journal of Optimization Theory and Applications, 1987
Journal of Vascular Surgery, 1983
Pricing problems are formulated as non-nested, stochastic Stackelberg games, and studied by using... more Pricing problems are formulated as non-nested, stochastic Stackelberg games, and studied by using the inducible region concept. It is shown that the inducible region can indeed be delineated, and the optimal pricing scheme can be constructed.
2015 IEEE International Conference on Automation Science and Engineering (CASE), 2015
Journal of Intelligent Manufacturing, 2015
1990 American Control Conference, May 23, 1990
... More recently, we harve used the tedmi to obtain near optimal solutions for the parallel mach... more ... More recently, we harve used the tedmi to obtain near optimal solutions for the parallel machine Sch ling of multi-operaion jobs with simple fork ... If a capacity consaint is violated at t k, a greedy hIuristic determines which new operations should beg -t that tinn slot and which ones ...
1982 American Control Conference, 1982
ABSTRACT
International Journal of Systems Science, Apr 1, 1989
IEEE Control Systems Magazine, 1990
Mathematical and Computer Modelling, 1996
[1993] Proceedings IEEE International Conference on Robotics and Automation, 1993
ABSTRACT
1990 IEEE International Conference on Systems, Man, and Cybernetics Conference Proceedings, 1990
Proceedings of the 10th World Congress on Intelligent Control and Automation, 2012
ABSTRACT Many optimization problems that frequently arise and have been extensively used in pract... more ABSTRACT Many optimization problems that frequently arise and have been extensively used in practice are modeled as linear mixed-integer programming problems. Among the many of such problems are transportation and assignment problems. In such problems, constraints that couple decision variables can be viewed as hyperplanes in their respective spaces. These hyperplanes frequently intersect at different angles thus making the feasible set of the problems complex and leading to difficulties defining the convex hull when using the cutting planes method. The efficiency of branch-and-cut is therefore low, since the branching tree grows quickly due to the combinatorial nature of these problems. This paper overcomes these difficulties by using additional cuts that can better define the convex hull. Thus when the Lagrangian relaxation and surrogate optimization method is used, the computational burden of obtaining the multiplier updating directions is significantly reduced. In the surrogate optimization framework, the constraints of the original problem are relaxed by introducing the Lagrange multipliers. After a good dual solution is found, it is then improved to obtain a good feasible solution, while the dual value provides a lower bound on the feasible cost. Numerical examples indicate that the surrogate optimization can obtain feasible solutions when the standard branch-and-cut method cannot. Additional cuts help improve the feasible solutions and tighten the lower bound.
In current U.S. deregulated wholesale electricity markets, an auction mechanism that minimizes th... more In current U.S. deregulated wholesale electricity markets, an auction mechanism that minimizes the total bid cost is used to select bids and their output levels. Energy market clearing prices (MCPs) are derived from the shadow prices associated with the system demand constraints in an economic dispatch process with fixed unit commitment status. Therefore they do not reflect no-load and start-up costs, resulting in the uplift payments. To reduce such side payments and improve the market transparency, the ldquoconvex hull pricing modelrdquo is adopted, and ldquoonline capacity constraintsrdquo are introduced to the energy market, requiring the total online capacities to be greater than or equal to the system demand. With the associated multipliers serving as uniform ldquono-load clearing prices,rdquo the total uplift payment is proved to be reduced based on optimality conditions for multipliers. Numerical examples support the analytical results and shed insights on different types of uplift payments.
More and more wind generation is introduced to electricity markets. Yet, it is not clear how to c... more More and more wind generation is introduced to electricity markets. Yet, it is not clear how to conduct auctions with significant stochastic wind generation while avoiding dramatically increasing computational complexity. In this paper, the Markov process concept is innovatively used to model wind generation with transition matrices established based on the historical data. Wind generation is then integrated into system demand, and the day-ahead auction problem considering system demand, unit maximal and minimal generations and ramp rates is formulated. The goal is to optimize the commitment and dispatch of normal generators to minimize the total expected cost while satisfying all possible loads. Lagrangian relaxation is used to demonstrate the capability of the formulation. Since there are many states at each hour, a novel enhanced version of Dynamic Programming is developed to solve the subproblems. Numerical testing results obtained by CPLEX demonstrate the effectiveness and scalability of the method compared with the standard stochastic programming formulation based on scenarios, and the feasibility compared with the deterministic formulation.
In the current US day-ahead electricity markets, auction is complicated because of the existence ... more In the current US day-ahead electricity markets, auction is complicated because of the existence of both discrete variables and continuous variables, which implies that there may be no linear prices to support equilibrium. To meet the system demand, the Independent System Operator (ISO) uses lump sum “uplift payments” to make generators follow its schedule. However, these uplift payments are opaque and difficult to hedge in terms of price signals. In addition, there are incentives for suppliers to bid other than their true costs. Untruthful bidding is undesirable, since in this case minimizing the total bid cost does not imply minimizing the total production cost, or maximizing the social welfare. As a result, the market may not be efficient. In this paper, system-wide redundant constraints on weighted generation levels are introduced to the day-ahead energy market to achieve equilibrium. With multipliers relaxing system demand constraints and redundant constraints serving as multipart prices, equilibrium can be achieved under the general condition that the optimal primal solution is among subproblem solutions for all units. Although incentive compatibility may not be achievable, the novel “local incentive compatibility” concept is introduced and quantified as the second best alternative to incentive compatibility.
ABSTRACT for the past several years, Kleinman and his associates . . . have collaborated on an in... more ABSTRACT for the past several years, Kleinman and his associates . . . have collaborated on an interdisciplinary program of experimental and theoretical research involving applied mathematics, experimental psychology, systems and human engineering, and computer science to develop a normative-descriptive theory to address problems of team-distributed decision making / they have developed empirically validated normative-descriptive models that capture the complexity, dynamicity, and uncertainty of the task environment and, in turn, quantify the resulting team performance, coordination, and decision strategies in specific situations objectives of this chapter are to present several of these models, to introduce the mathematical tools used to support the modeling activity, to demonstrate the process by which the normative-descriptive theory is applied to generate predictions of actual team performance, and to offer some novel hypotheses on team decision-making behavior and performance / point out the characteristics of the distributed dynamic decision-making problems that have been addressed, to define the normative-descriptive modeling approach, and to describe a unique research paradigm for distributed decision making that has been used to generate and collect the data to develop and validate the mathematical models (PsycINFO Database Record (c) 2012 APA, all rights reserved)
IEEE Control Systems Magazine, 1990
Journal of Optimization Theory and Applications, 1987
Journal of Vascular Surgery, 1983
Pricing problems are formulated as non-nested, stochastic Stackelberg games, and studied by using... more Pricing problems are formulated as non-nested, stochastic Stackelberg games, and studied by using the inducible region concept. It is shown that the inducible region can indeed be delineated, and the optimal pricing scheme can be constructed.
2015 IEEE International Conference on Automation Science and Engineering (CASE), 2015
Journal of Intelligent Manufacturing, 2015
1990 American Control Conference, May 23, 1990
... More recently, we harve used the tedmi to obtain near optimal solutions for the parallel mach... more ... More recently, we harve used the tedmi to obtain near optimal solutions for the parallel machine Sch ling of multi-operaion jobs with simple fork ... If a capacity consaint is violated at t k, a greedy hIuristic determines which new operations should beg -t that tinn slot and which ones ...
1982 American Control Conference, 1982
ABSTRACT
International Journal of Systems Science, Apr 1, 1989
IEEE Control Systems Magazine, 1990
Mathematical and Computer Modelling, 1996
[1993] Proceedings IEEE International Conference on Robotics and Automation, 1993
ABSTRACT
1990 IEEE International Conference on Systems, Man, and Cybernetics Conference Proceedings, 1990
Proceedings of the 10th World Congress on Intelligent Control and Automation, 2012
ABSTRACT Many optimization problems that frequently arise and have been extensively used in pract... more ABSTRACT Many optimization problems that frequently arise and have been extensively used in practice are modeled as linear mixed-integer programming problems. Among the many of such problems are transportation and assignment problems. In such problems, constraints that couple decision variables can be viewed as hyperplanes in their respective spaces. These hyperplanes frequently intersect at different angles thus making the feasible set of the problems complex and leading to difficulties defining the convex hull when using the cutting planes method. The efficiency of branch-and-cut is therefore low, since the branching tree grows quickly due to the combinatorial nature of these problems. This paper overcomes these difficulties by using additional cuts that can better define the convex hull. Thus when the Lagrangian relaxation and surrogate optimization method is used, the computational burden of obtaining the multiplier updating directions is significantly reduced. In the surrogate optimization framework, the constraints of the original problem are relaxed by introducing the Lagrange multipliers. After a good dual solution is found, it is then improved to obtain a good feasible solution, while the dual value provides a lower bound on the feasible cost. Numerical examples indicate that the surrogate optimization can obtain feasible solutions when the standard branch-and-cut method cannot. Additional cuts help improve the feasible solutions and tighten the lower bound.