| Title | Publication Details | Authors | Abstract |
| Distribution system reliability assessment due to lightning storms |
IEEE Transactions on Power Delivery Publication Date: July 2005 Volume: 20, Issue: 3 On page(s): 2153- 2159 ISSN: 0885-8977 INSPEC Accession Number: 8526451 Digital Object Identifier: 10.1109/TPWRD.2005.848724 Current Version Published: 2005-06-27 |
|
Abstract: Lightning is a significant cause of faults and outages in many electric power systems and is one of the major causes of poor system reliability. Predictive assessment of distribution reliability indices can be used to identify areas that have poor reliability so that appropriate changes in system design can be implemented. The assessment of distribution system performance under lightning conditions requires modeling of storm characteristics and system response. In this paper, a Monte Carlo simulation for evaluating distribution system reliability under lightning storm conditions is presented. The results from a practical distribution system show the importance of detailed modeling of storm characteristics and simulation of the system response in assessing distribution system reliability during lightning storms. |
| An ISO market settlement design validation approach using probabilistic market simulation study techniques |
2004 International Conference on Probabilistic Methods Applied to Power Systems Publication Date: 12-16 Sept. 2004 On page(s): 338- 342 ISBN: 0-9761319-1-9 INSPEC Accession Number: 8331609 Current Version Published: 2005-01-17 |
|
Abstract: In the independent system operator organizations, settlement activity is an integral part of the market operation. One of the key components for measuring independent system operator market operational design correctness is to analyze how well market price signals create incentives and/or penalties for creating an efficient market to achieve the design goals under uncertainties due to unpredictable market conditions. The market settlement rule is an important vehicle to implement such price signals. This paper presents an approach for a probabilistic method to validate settlement systems design using probabilistic techniques for independent system operator market simulation. In addition, the paper presents critical comments on various differences in settlement design approach. |
| Improving Market Participant Strategies with FTR Options |
PICA 2001 Sydney, Australia |
|
Abstract: Strategies that reduce risk and maximize profit are of great interest to electricity market participants. Commodities and financial instruments that can be hedged to reduce uncertainty and manage risk are important components of such a strategy. One such vehicle is the fixed transmission right (FTR). The holder of an FTR is entitled to the transmission congestion rents collected between specified points of delivery and receipt. However, if congestion occurs in an unanticipated direction, the holder of an FTR is obligated to pay. An interesting compromise derivative [1,2] is the FTR option (FTRO). The FTRO gives its holder the right to the congestion rents if they are positive, without the obligation to pay if they are negative. In this paper the authors provide examples of how FTR options can be utilized as part of a comprehensive market strategy. Profitability with and without the FTRO is determined under various market and network scenarios. Methods and assumptions required for valuing the options are examined, including equations similar to Black-Scholes. |
| Gaining the Competitive Edge: Enhancing Market Demand Forecasts with Hartley Transforms | Submitted |
|
Abstract: Succeeding in the competitive electricity market requires well-planned operational and bidding strategies. Choosing the appropriate strategy is highly dependent on recognizing in what state the system is operating. Much data (including time series) is available, and a proper analysis of this data can be very time consuming. Developing techniques to quickly identify the system state should help the savvy market participant in reacting intelligently to market changes before its competitors. Additionally, data analysis techniques can reveal certain patterns in the data that would not be discovered with other techniques. Identification of these patterns may be very helpful in forecasting demand or price. This paper compares techniques that identify useful patterns in relevant time series data. These patterns are keys or leading indicators of future electric utility price or demand of electricity. The importance of improving methods of price analysis will increase in the deregulated environment as competition increases. Techniques being investigated are Fourier Transforms, Hartley Transforms, Line Spectrum analysis due to both Fourier Transforms and Hartley Transforms. |
| Gaming Strategies on Electric Power Markets Driven by Agents | Submitted |
|
Abstract: The deregulation of the electric industry in the United States opened some sectors of the power market to competition. Buyers and sellers of electric power are competing for limited resources. When large amounts of money are at stake, the participants have incentives to engage in gaming, although regulations attempting to limit such activity may exist. The goal of this study is to model an agent driven bilateral power market auction where some of the players attempt to benefit from causing economic instabilities by applying different gaming strategies. The market structure is similar to the California power market. One of the companies engages in gaming behavior, using the congestion management policies combined with pricing schemes to cut off one or more of the generators of the other company. Considering the shutdown and start-up costs, minimum up and downtimes, and generator limits allows participants even more opportunities for gaming under certain pricing schemes. It is also that in some cases the power delivered to a customer is reduced since this is sometimes the best solution to the total congestion management optimization problem. |
| Price Signal Analysis for Competitive Electric Generation Companies | DRPT2000 Conference |
|
Abstract: Successful operation and bidding in the competitive electricity marketplace requires well-planned strategies. The appropriate strategy is dependent on the state of the system. Much data (including time series) is available, and a proper analysis of this data can provide insight in choosing the right strategies. Traditional data analysis techniques can be time consuming. Techniques that quickly analyze the data can assist in forecasting price and demand and identifying the present state of the market, which should help the savvy trader in reacting intelligently to the market before its competitors. Advanced data analysis techniques may reveal patterns in the data that may be very helpful in forecasting demand or price. This paper compares several techniques that may help in identifying useful patterns in relevant time series data. These patterns are keys or leading indicators of future electric utility price or demand of electricity. The importance of quickly identifying these signals will increase as competition increases. The techniques being investigated are Fourier and Hartley Transforms, Line Spectrum analysis using both Fourier Transforms and Hartley Transforms. |
| Modeling and Evaluating Electricity Options Markets with Intelligent Agents | DRPT2000 Conference |
|
Abstract: Under deregulation, the formation of electricity markets is a
topic of great interest in the power industry and in financial
institutions worldwide. Using derivative financial instruments
(including options) becomes important for hedging against uncertainty
and managing risk--limiting exposure to adverse market conditions. Black
and Scholes' equation is often used to value options, but its validity
is questionable due to assumptions that may not hold for electricity,
most notably the assumption of log-normally distributed prices for the
underlying commodity. In this research, a put options market for
electricity is modeled. Adaptive agents trade in this market to
maximize profit. They are not forced to use an explicit economic or
financial model (e.g., Black-Scholes) in their valuation. A genetic
algorithm (GA) is used to find alternate valuations that are used to
generate buy and sell signals. The results show that it is possible to
evolve profitable valuations for use with buying and selling options in
this simple model. Reasons for and implications of this finding (e.g.,
that Black-Scholes may not be a good method for pricing electricity
derivatives) are discussed.
Keywords: Black-Scholes, options pricing, adaptive agents, agent-based economics, risk management, power system deregulation. |
| Predatory Gaming Strategies for Electric Power Markets | DRPT2000 Conference |
|
Abstract: The recent deregulation of the electric industry in the United
States opened some sectors of the power market to competition.
Buyers and sellers of electric power are competing for limited
resources. Although regulations exist attempting to limit such activity,
when large amounts of money are at stake, the participants have
incentives to engage in predatory behavior. The goal of this study is
to model an agent driven bilateral power market auction where some
of the players attempt to benefit from causing instabilities like
brownouts and blackouts, as well as economic instabilities by applying
different gaming strategies. The market structure is similar to the
California power market. The network considered consists of six
generators in three zones and two loads connected by a six bus power
network. An independent entity takes care of the congestion management,
as well of allocation of the available resources. One of the
companies engages in predatory behavior, using the congestion management
policies combined with carefully chosen bids to cut off one or more of
the generators of the other company. Vulnerabilities associated with
shutdown and startup costs, minimum up and downtimes, ramp rate and
generator limits for each generator, are utilized to achieve market
destabilization. Customers may be negatively impacted by the predatory
behavior, since reducing the power delivered to a customer is sometimes
the best solution to the total congestion management optimization
problem.
Keywords: Power systems deregulation, congestion management, bidding strategies, auction gaming, market collusion. |
| Modeling Dynamic Electricity Markets with Intelligent Agent Based Economics | North American Power Systems, 1999 |
|
Abstract: In recent years, numerous competitive electricity markets have emerged around the world. Regulators need tools to help them study the impacts of their policy decisions on the physical operation and security of the power system, as well as the financial impact to the various market participants, and the economic aspects of the market itself (e.g., is it stable?). Alvarado [1] investigates the issue of market stability for simple market scenarios by setting up a set of dynamic market equations and computing the eigenvalues. This research begins by reproducing the results obtained from the dynamic market model, and suggests several extensions to make it more realistic. In addition, the authors propose that an agent-based approach to studying this market might reduce the number of assumptions required. This paper presents the steps in modeling agents that supply and demand electricity. This work discusses some of the benefits of using agent-based computational economics. The authors suggest some agent-based experiments to explore how changing the underlying assumptions, varying the pricing rules, and modifying the representation of the agent can produce changes to the results. |
| Economically Destabilizing Electricity Markets for Profit | North American Power Systems, 1999 |
|
Abstract: The deregulation of the electric industry in the United States opens the power market to competition. Buyers and sellers of electric power will be competing for limited resources. Although regulators will attempt to limit such activity, when large amounts of money are at stake, the participants have incentives to engage in destabilizing behavior. The goal of this study is to model agent driven power market auctions where some of the players attempt to benefit from causing economic instabilities and intentionally driving market prices by applying different strategies. |
| Optimal Control Applied to the Transmission Investment Strategy Problem | North American Power Systems, 1999 |
|
Abstract: In the deregulated competitive electric power marketplace, the transaction price and quantity are decided by the supply and demand. With the growth of power demand in recent years, the limit of transmission capacity becomes an obstacle to competitive transactions. This moves the market solution away from the ideal competitive solution, and brings deadweight losses that must be carried by society. In order to maximize total social welfare and achieve economic efficiency, the market regulator must build some new transmission facilities, e.g., new lines or power flow controller. If the economic cost is less than the increase of trading surplus, (i.e., the benefit), then the investment is worthy. This paper discusses investment strategies for market regulators deciding to invest in new transmission facilities in a growing competitive power market. We can see in such a market, the economic profit and cost from one period to another can be represented by a series of discrete-time dynamic equations, and the objective function is in quadratic form of state and control variables. After constructing models, the decision making can be viewed as a LQR (Linear Quadratic Regulator) problem and the discrete-time Ricatti equation can be applied to solve it. |
| Determining Electricity Market Share with Chaos | North American Power Systems, 1999 |
|
Abstract: Experimental economics and computerized market simulations can be important tools to explore the consequences of choosing various market rules of strategies used in a market. Much attention has been focused on the economic stability of a given set of market rules following the extreme prices in emerging markets seen in recent years. This research seeks to discover economically stable points of system operation generation and pricing. The role of randomness and chaos in these simulations will be explained and then utilized in a positive feedback model. This model is explained in detail, and is used to simulate a simplified deregulated power economy. With the random events recorded, the probability of certain solutions can be calculated. For each of these models, the triggering events can be studied and used to initiate actions proper for the conditions likely to follow. The effects of dynamic market equilibrium conditions are also investigated. |
| A Profit-Based Unit Commitment GA for Uncertain Price and Demand Forecasts | North American Power Systems, 1998 |
|
Abstract: As the electrical industry restructures, many of the traditional algorithms for controlling generating units need modification or replacement. Previously utilized to schedule generation units in a manner that minimizes costs while meeting all demand, the unit commitment (UC) algorithm must be updated. A UC algorithm that maximizes profit will play an essential role in developing successful bidding strategies for the competitive generator. Simply bidding to win contracts is insufficient; bidding strategies must result in contracts that, on average, cover the total generation costs. No longer guaranteed to be the only electricity supplier, a generation company'?s share of the demand will be more difficult to predict than in the past. Removing the obligation to serve softens the demand constraint giving the generator additional flexibility in scheduling units. In this paper the authors provide a price/profit-based UC formulation which considers the softer demand constraint and allocates fixed and transitional costs to the scheduled hours. The authors describe a genetic algorithm solution to this new UC problem. The algorithm uses hourly demand and price forecasts as inputs. We recognize that these forecasts may be uncertain, and so we describe extensions to the UC GA needed to handle these ?fuzzy? demand and price forecasts. |
| A Profit-Based Unit Commitment GA for the Competitive Environment | Accepted for publication in IEEE Transactions on Power Systems, 1999 |
|
Abstract: As the electrical industry restructures, many of the traditional algorithms for controlling generating units need modification or replacement. Previously utilized to schedule generation units in a manner that minimizes costs while meeting all demand, the unit commitment (UC) algorithm must be updated. A UC algorithm that maximizes profit will play an essential role in developing successful bidding strategies for the competitive generator. Simply bidding to win contracts is insufficient; bidding strategies must result in contracts that, on average, cover the total generation costs. No longer guaranteed to be the only electricity supplier, a generation company'?s share of the demand will be more difficult to predict than in the past. Removing the obligation to serve softens the demand constraint. In this paper the authors provide a price/profit-based UC formulation which considers the softer demand constraint and allocates fixed and transitional costs to the scheduled hours. The authors describe a genetic algorithm solution to this new UC problem and present results for an illustrative example. |
| Comprehensive Bidding Strategies with Genetic Programming/Finite State Automata | Accepted to the IEEE Transactions on Power Systems, 1998 |
|
Abstract: This research is an extension of the
authors' previous work in double auctions aimed
at developing bidding strategies for electric
utilities which trade electricity competitively.
The improvements detailed in this paper come from
using data structures which combine genetic
programming and finite state automata termed
GP-Automata. The strategies developed by the
method described here are adaptive--reacting to
inputs--whereas the previously developed strategies
were only suitable in the particular scenario for
which they had been designed. The strategies
encoded in the GP-Automata are tested in an
auction simulator. The simulator pits them
against other distribution companies (distcos)
and generation companies (gencos), buying
and selling power via double auctions implemented
in regional commodity exchanges. The GP-Automata
are evolved with a genetic algorithm so that they
possess certain characteristics. In addition to
designing successful bidding strategies (whose
usage would result in higher profits) the resulting
strategies can also be designed to imitate certain
types of trading behaviors. The resulting strategies
can be implemented directly in on-line trading,
or can be used as realistic competitors in an
off-line trading simulator.
Keywords: Competitive auction markets, genetic algorithms, bidding strategies, deregulation, energy broker, power systems, genetic programming, GP-Automata. |
| Effects of Tree Size and State Number on GP-Automata Bidding Strategies | GP-98 Conference |
|
Abstract: The impending deregulation of the electrical industry in the USA promises to open a multi- billion dollar industry to competition. Current research indicates that the double auction will be at the heart of several regional electrical commodity exchanges. The authors are attempting to design comprehensive profitable bidding strategies for traders. The advantages of the strategies detailed here come from using data structures which combine genetic programming and finite state automata termed GP-Automata. Adaptive strategies encoded by two populations of GP-Automata are tested in an auction simulator modeling distribution companies and generation companies buying and selling power via a double auction. In addition to evolving profitable bidding strategies, the resulting strategies can also be designed to imitate certain types of trading behaviors. These strategies can be used directly in on-line trading, or as realistic competitors in an off-line trading simulator. In this paper we report the results of specific experiments which test the effect of changing the size of the GP trees, and the effect of changing the number of states. |
| Using Adaptive Agents to Study Bilateral Contracts and Trade Networks | GP-98 Conference |
|
Abstract: This research is an extension of research done by Charles Richter, Gerald Sheble' and Dan Ashlock (1997, 1998) on double auction bidding strategies for electric utilities which trade competitively. This research considers the network topology and whether a successful bid transaction can occur given the flow constraints on the network. The ATC (Available Transmission Capacity) of the network is a flow constraint indicator that is used to provide feedback to agents attempting to engage in bilateral contracts. The aim is to develop adaptive agents that are able to recognize with whom they can enter a profitable bilateral contract. In other words, the agents develop preferential partner selection lists and bidding strategies in a simulated electric market. The idea of evolving preferred trading partner lists comes from the Trade Network Game (TNG) developed by Tesfatsion, Ashlock and Stanley (1995). The strategies being developed by the method described here are adaptive. The strategies are encoded in GP-Automata, a technique which combines genetic programming and finite state automata. |
| Bidding Strategies that Minimize Risk with Options and Futures Contracts | American Power Conference, '98 |
|
Abstract: This research builds on earlier research in developing bidding
strategies through the inclusion of options and futures contracts. In
the competitive environment, electric traders's profits depend on the
implementation of a successful bidding strategy. Bidding strategies are
studied in an environment in which distribution companies (distcos) and
generation companies (gencos), buy and sell power via double auctions in
regional commodity exchanges. The market framework being used was
proposed by Kumar and Sheble'[1] and allows participants to trade in
the spot, future, planning and swap markets, and also gives rise to the
use of option contracts. Bid-strategy research previously published by
the authors focused on increasing electric generators' profit as they
participated in a spot/cash market. Here we incorporate techniques such
as game theory and decision analysis to minimize the risk to the
electric utility or energy trader. The goal is to ensure reliable power
system operation while also ensuring that contracts are fulfilled and
traders adopting the strategies remain profitable. The developed
strategies are tested in our electric market trading simulator which can
be used off-line to predict whether bid strategies will be profitable
and successful.
Keywords: Competitive auction markets, optimization, genetic algorithms, bidding strategies, deregulation, energy broker, power systems, options and futures contracts, risk management. [1] Jayant Kumar and Gerald Sheble'. "Framework for Energy Brokerage System with Reserve Margin and Transmission Losses." IEEE PES/96 WM 190-9 PWRS, 1996. |
| Genetic Algorithm Evolution of Utility Bidding Strategies for the Competitive Marketplace | IEEE Transactions on Power Systems, PWRS |
|
Abstract: This paper describes an environment in which distribution
companies (discos) and generation companies (gencos), buy and sell power
via double auctions implemented in a regional commodity exchange. The
electric utilities' profits depend on the implementation of a
successful bidding strategy. In this research, a genetic algorithm
evolves bidding strategies as gencos and discos trade power. A
framework in which bidding strategies may be tested and modified is
presented. This simulated electric commodity exchange can be used
off-line to predict whether bid strategies will be profitable and
successful. It can also be used to experimentally verify how bidding
behavior affects the competitive electric marketplace.
Keywords: Competitive auction markets, optimization, genetic algorithms, bidding strategies, deregulation, energy broker, power systems, operations. |
| The Effect of Splitting Populations on Bidding Strategies | GP-97 |
|
Abstract: In this paper we explore the effects of splitting a single population of artificial agents engaging in a simple double auction game into two competing populations by modifying experiments reported in [Ashlock, 1997]. The original paper used a new genetic programming tool, termed GP-Automata, to induce bidding strategies with a genectic algorithm for Nash's game divide the dollar. The motivation for performing the research is the biological notion of inclusive fitness and kinship theory. The a priori hypothesis of the authors was that behavior of the agents in the simulated market would change substantially when they were no longer forced to be similar to one another by the genetic mechanism used to induce new bidding strategies. While breeding takes place only within each poulation, all bidding is between agents from different populations. The agents in the original (single population) paper strongly favored "fair" Nash equilibria of the divide the dollar game, at odds with the economic theory for egoistic agents. When controls for kinship effects are implemented by splitting the population a substantial effect is observed. When agents doing the bidding are not close genetic kin to one another the "unfair" Nash equilbria regain a great deal of their former prominence. This result is of importance to any sort of evolutionary algorithm creating artificial agents, as kinship theory can confound game-theoretic predictions that assume egoistic agents. The current research also arguably increases the level of realism in the simulation of a double auction market. |
| Building Fuzzy Bidding Strategies for the Competitive Generator | North American Power Symposium, '97 |
|
Abstract: In this paper, the authors build on previous research that
they have done in the area of building bidding strategies for electric
utilities in the competitive environment. The previous research is
briefly reviewed. The deregulated market-place is defined and modeled.
Fuzzy logic is included to make bidding strategies adaptive. Four
methods for building bidding strategies which use fuzzy logic and/or
genetic algorithms are discussed and outlined. Economical inputs are
fuzzified for use in determining a generator's bid. Methods of tuning
and searching for the optimal rule are discussed. We discuss how an
agent using the bidding strategies can compare them based on
profitability.
Keywords: Bidding Strategies, Auctions, Trading, Fuzzy Economics, Expert-System Bidding, Genetic Algorithms. |
| Automatic Generation Control with a Fuzzy Logic Controller | North American Power Symposium, '96 |
|
Abstract: In this paper, fuzzy rules are developed for automatic
generation control of an electric power system. Interchange error, time
error and area control error (ACE) are used in conjunction with fuzzy
rules to control generation levels in a two area system. The fuzzy
controller described in this paper not only outperforms a traditional
controller connected to a system with the same generator parameters and
same load changes, but also performs better than a fuzzy AGC controller
that uses only one input as previously presented.
Keywords: Automatic Generation Control, AGC, Fuzzy Logic Control, FLC. |
| Genetic Algorithm Development of a Healthcare Expert System | Midwest Electro-Technologies Conference |
|
Abstract: Many industries are currently seeking to take advantage of computer-based expert systems to assist in making decisions when many inputs must be considered. Designing the expert system has been a difficult process, typically requiring human experts to construct complicated lists of rules. This is often a lengthy process sometimes requiring several expert person-years. The method described in this paper could minimize or eliminate the need for the expert's input in the development process. If records of the expert's past decisions exist, it is possible to use this data to develop expert systems rules. The method proposed in this paper uses a genetic algorithm to deduce the expert system rules from existing records. The experimental results outlined in this paper demonstrate the proposed algorithms ability to learn from a given set of decision's inputs and outputs. |