Showing papers on "Bidding published in 2017"

Cooperation of Wind Power and Battery Storage to Provide Frequency Regulation in Power Markets

Abstract: In the future power system with high penetration of renewables, renewable energy is expected to undertake part of the responsibility for frequency regulation, just as the conventional generators. Wind power and battery storage are complementary in accuracy and durability when providing frequency regulation. Therefore, it would be profitable to combine wind power and battery storage as a physically connected entity or a virtual power plant to provide both energy and frequency regulation in the markets. This paper proposes a real-time cooperation scheme to exploit their complementary characteristics and an optimal bidding strategy for them in joint energy and regulation markets, considering battery cycle life. The proposed cooperation scheme is adopted in a real-time battery operating simulation and then incorporated into the optimal bidding model. The scheme could improve the wind regulation performance score and allow for more regulation bids without affecting the battery life, thus significantly increasing the overall revenue. The validity of the proposed scheme and strategy are proved by the case study.

Real-Time Bidding by Reinforcement Learning in Display Advertising

Abstract: The majority of online display ads are served through real-time bidding (RTB) --- each ad display impression is auctioned off in real-time when it is just being generated from a user visit. To place an ad automatically and optimally, it is critical for advertisers to devise a learning algorithm to cleverly bid an ad impression in real-time. Most previous works consider the bid decision as a static optimization problem of either treating the value of each impression independently or setting a bid price to each segment of ad volume. However, the bidding for a given ad campaign would repeatedly happen during its life span before the budget runs out. As such, each bid is strategically correlated by the constrained budget and the overall effectiveness of the campaign (e.g., the rewards from generated clicks), which is only observed after the campaign has completed. Thus, it is of great interest to devise an optimal bidding strategy sequentially so that the campaign budget can be dynamically allocated across all the available impressions on the basis of both the immediate and future rewards. In this paper, we formulate the bid decision process as a reinforcement learning problem, where the state space is represented by the auction information and the campaign's real-time parameters, while an action is the bid price to set. By modeling the state transition via auction competition, we build a Markov Decision Process framework for learning the optimal bidding policy to optimize the advertising performance in the dynamic real-time bidding environment. Furthermore, the scalability problem from the large real-world auction volume and campaign budget is well handled by state value approximation using neural networks. The empirical study on two large-scale real-world datasets and the live A/B testing on a commercial platform have demonstrated the superior performance and high efficiency compared to state-of-the-art methods.

Risk-Constrained Bidding and Offering Strategy for a Merchant Compressed Air Energy Storage Plant

Abstract: Electricity price forecasts are imperfect. Therefore, a merchant energy storage facility requires a bidding and offering strategy for purchasing and selling the electricity to manage the risk associated with price forecast errors. This paper proposes an information gap decision theory (IGDT)-based risk-constrained bidding/offering strategy for a merchant compressed air energy storage (CAES) plant that participates in the day-ahead energy markets considering price forecasting errors. Price uncertainty is modeled using IGDT. The IGDT-based self-scheduling formulation is then used to construct separate hourly bidding and offering curves. The theoretical approach to develop the proposed strategy is presented and validated using numerical simulations.

Trading on local energy markets: A comparison of market designs and bidding strategies

Abstract: Increasing renewable energy sources and innovative information and communication systems open up new challenges and opportunities to integrate distributed generation into the energy supply system. Formerly centralized energy systems need to be adapted to take full advantage of the immense potential of decentralized energy generation and smart, interconnected energy end users. We introduce a local electricity market on which prosumers and consumers of a community are able to trade electricity directly amongst each other. This local electricity market supports the local integration of renewable energy generation. It facilitates a local balance of energy supply and demand and hence reduces the need for extensive electricity transmission. We introduce, evaluate and compare two local market designs, a direct peer-to-peer market and a closed order book market, as well as two agent behaviors, zero-intelligence agents and intelligently bidding agents. We derive four scenarios by combining each market design with each agent behavior, respectively. All market scenarios offer similar economic advantages for the market participants. However, the peer-to-peer market with intelligent agents appears to be the most advantageous as it results in the lowest average overall electricity price.

Risk-Averse Optimal Bidding Strategy for Demand-Side Resource Aggregators in Day-Ahead Electricity Markets Under Uncertainty

Abstract: This paper first presents a generic model to characterize a variety of flexible demand-side resources (e.g., plug-in electric vehicles and distributed generation). Key sources of uncertainty affecting the modeling results are identified and are characterized via multiple stochastic scenarios. We then propose a risk-averse optimal bidding formulation for the resource aggregator at the demand side based on the conditional value-at-risk (VaR) theory. Specifically, this strategy seeks to minimize the expected regret value over a subset of worst-case scenarios whose collective probability is no more than a threshold value. Our approach ensures the robustness of the day-ahead (DA) bidding strategy while considering the uncertainties associated with the renewable generation, real-time price, and electricity demand. We carry out numerical simulations against three benchmark bidding strategies, including a VaR-based approach and a traditional scenario based stochastic programming approach. We find that the proposed strategy outperforms the benchmark strategies in terms of hedging high regret risks, and results in computational efficiency and DA bidding costs that are comparable to the benchmarks.

Display Advertising with Real-Time Bidding (Rtb) and Behavioural Targeting

Abstract: Online advertising is now one of the fastest advancing areas in the IT industry. In display and mobile advertising, the most significant technical development in recent years is the growth of Real-Time Bidding (RTB), which facilitates a real-time auction for a display opportunity. RTB essentially facilitates buying an individual ad impression in real time while it is still being generated from a users visit. RTB not only scales up the buying process by aggregating a large number of available inventories across publishers but, most importantly, enables direct targeting of individual users. As such, RTB has fundamentally changed the landscape of digital marketing. Scientifically, the demand for automation, integration and optimization in RTB also brings new research opportunities in information retrieval, data mining, machine learning and other related fields. Despite its rapid growth and huge potential, many aspects of RTB remain unknown to the research community for a variety of reasons. This monograph offers insightful knowledge of real-world systems, to bridge the gaps between industry and academia, and to provide an overview of the fundamental infrastructure, algorithms, and technical and research challenges of the new frontier of computational advertising. The topics covered include user response prediction, bid landscape forecasting, bidding algorithms, revenue optimization, statistical arbitrage, dynamic pricing, and ad fraud detection. This is an invaluable text for researchers and practitioners alike. Academic researchers will get a better understanding of the real-time online advertising systems currently deployed in industry. While industry practitioners are introduced to the research challenges, the state of the art algorithms and potential future systems in this field.

Look-Ahead Bidding Strategy for Energy Storage

Abstract: As the cost of battery energy storage continues to decline, we are likely to see the emergence of merchant energy storage operators. These entities will seek to maximize their operating profits through strategic bidding in the day-ahead electricity market. One important parameter in any storage bidding strategy is the state-of-charge at the end of the trading day. Because this final state-of-charge is the initial state-of-charge for the next trading day, it has a strong impact on the profitability of storage for this next day. This paper proposes a look-ahead technique to optimize a merchant energy storage operator's bidding strategy considering both the day-ahead and the following day. Taking into account the discounted profit opportunities that could be achieved during the following day allows us to optimize the state-of-charge at the end of the first day. We formulate this problem as a bilevel optimization. The lower-level problem clears a ramp-constrained multiperiod market and passes the results to the upper-level problem that optimizes the storage bids. Linearization techniques and Karush–Kuhn–Tucker conditions are used to transform the original problem into an equivalent single-level mixed-integer linear program. Numerical results obtained with the IEEE Reliability Test System demonstrate the benefits of the proposed look-ahead bidding strategy and the importance of considering ramping and network constraints.

Learning in Repeated Auctions with Budgets: Regret Minimization and Equilibrium

Abstract: In online advertising markets, advertisers often purchase ad placements through bidding in repeated auctions based on realized viewer information. We study how budget-constrained advertisers may bid in the presence of competition, when there is uncertainty about future bidding opportunities as well as competitors' heterogenous preferences and budgets. We formulate this problem as a sequential game of incomplete information, where bidders know neither their own valuation distribution, nor the budgets and valuation distributions of their competitors. We introduce a family of dynamic bidding strategies we refer to as "adaptive pacing" strategies, in which advertisers adjust their bids throughout the campaign according to the sample path of observed expenditures. We analyze the performance of this class of strategies under different assumptions on competitors' behavior. Under arbitrary competitors' bids, we establish through matching lower and upper bounds the asymptotic optimality of this class of strategies as the number of auctions grows large. When adopted by all the bidders, the dynamics converge to a tractable and meaningful steady state. Moreover, we show that these strategies constitute an approximate Nash equilibrium in dynamic strategies: The benefit of unilaterally deviating to other strategies, including ones with access to complete information, becomes negligible as the number of auctions and competitors grows large. This establishes a connection between regret minimization and market stability, by which advertisers can essentially follow equilibrium bidding strategies that also ensure the best performance that can be guaranteed off-equilibrium.

CVaR-Constrained Optimal Bidding of Electric Vehicle Aggregators in Day-Ahead and Real-Time Markets

Abstract: An electric vehicle aggregator (EVA) that manages geographically dispersed electric vehicles offers an opportunity for the demand side to participate in electricity markets. This paper proposes an optimization model to determine the day-ahead inflexible bidding and real-time flexible bidding under market uncertainties. Based on the relationship between market price and bid price, the proposed optimal bidding model of EVA aims to minimize the conditional expectation of electricity purchase cost in two markets considering price volatility. Moreover, the penalty cost of the deviation between the bidding quantities is included to avoid large power variation and arbitrage. The conditional expectation optimization model is formulated as an expectation minimization problem with the conditional value-at-risk constraints. Based on the price data in the PJM market, simulation results verify that our model is a decision-making tool in electricity markets, which can help market players comprehend the variants of bid price, expected cost and probability of successful bidding.

Risk-Constrained Optimal Bidding Strategy for Pairing of Wind and Demand Response Resources

Abstract: Wind power represents a significant percentage of the Spanish generation mix and this trend will increase due to the commitment of the European Union to the full deployment of Directive 2009/28/EC. The increasing penetration of intermittent renewable energy, and the development of advanced information and communication technologies, give rise to questions on how additional flexibility obtained from loads can be used in order to optimize the use of resources and assets. This paper proposes a comprehensive stochastic decision making model for the coordinated operation of wind power producers and demand response (DR) aggregators participating in the day-ahead market. In order to account for the uncertainty around the true outcomes of day-ahead prices and wind power, a minimum conditional value at risk term has been included in the model formulation. Numerical results illustrate how the proposed bidding strategy for wind and demand response (DR) pairing increases the expected benefit of both resources and reduces the related risk.

A review on Demand-side tools in electricity market

Abstract: With the advent of restructuring in the electricity markets, the Supply-side quickly adapted to the new environment, whereas, the story in the demand side has been different. Demand side dealt with the electric energy as a commodity available to the necessary extent. This caused the Supply-side to realize that the demand side would admit to purchase electric energy at any price, and this resulted in the advent of bidding strategies in the Supply-Side, known as “hockey-stick bidding”. The most important result was transfer of the demand side assets to the Supply-side. After a while, the demand side noticed the self-sloppy condition, therefore looked for tools to deal with these threats. This subject is examined in this paper.

Purchase Bidding Strategy for a Retailer With Flexible Demands in Day-Ahead Electricity Market

Abstract: The paper aims to determine the day-ahead market bidding strategies for retailers with flexible demands to maximize the short-term profit. It proposes a short-term planning framework to forecast the load under dynamic tariffs and construct biding curves. Stochastic programming is applied to manage the uncertainties of spot price, regulating price, consumption behaviors, and responsiveness to dynamic tariffs. A case study based on data from Sweden is carried out. It demonstrates that a real-time selling price can affect the aggregate load of a residential consumer group and lead to load shift toward low-price periods. The optimal bidding curves for specific trading periods are illustrated. Through comparing the bidding strategies under different risk factors, the case study shows that a risk-averse retailer tends to adopt the strategies with larger imbalances. The benefit lies in the reduction of low-profit risk. However, the aversion to risk can only be kept in a certain level. A larger imbalance may lead to a quick reduction of profit in all scenarios.

Alternating Offers Protocols for Multilateral Negotiation

Abstract: This paper presents a general framework for multilateral turn-taking protocols and two fully specified protocols namely Stacked Alternating Offers Protocol (SAOP) and Alternating Multiple Offers Protocol (AMOP). In SAOP, agents can make a bid, accept the most recent bid or walk way (i.e., end the negotiation without an agreement) when it is their turn. AMOP has two different phases: bidding and voting. The agents make their bid in the bidding phase and vote the underlying bids in the voting phase. Unlike SAOP, AMOP does not support walking away option. In both protocols, negotiation ends when the negotiating agents reach a joint agreement or some deadline criterion applies. The protocols have been evaluated empirically, showing that SAOP outperforms AMOP with the same type of conceder agents in a time-based deadline setting. SAOP was used in the ANAC 2015 competition for automated negotiating agents.

Real-Time Bidding by Reinforcement Learning in Display Advertising

Abstract: The majority of online display ads are served through real-time bidding (RTB) --- each ad display impression is auctioned off in real-time when it is just being generated from a user visit. To place an ad automatically and optimally, it is critical for advertisers to devise a learning algorithm to cleverly bid an ad impression in real-time. Most previous works consider the bid decision as a static optimization problem of either treating the value of each impression independently or setting a bid price to each segment of ad volume. However, the bidding for a given ad campaign would repeatedly happen during its life span before the budget runs out. As such, each bid is strategically correlated by the constrained budget and the overall effectiveness of the campaign (e.g., the rewards from generated clicks), which is only observed after the campaign has completed. Thus, it is of great interest to devise an optimal bidding strategy sequentially so that the campaign budget can be dynamically allocated across all the available impressions on the basis of both the immediate and future rewards. In this paper, we formulate the bid decision process as a reinforcement learning problem, where the state space is represented by the auction information and the campaign's real-time parameters, while an action is the bid price to set. By modeling the state transition via auction competition, we build a Markov Decision Process framework for learning the optimal bidding policy to optimize the advertising performance in the dynamic real-time bidding environment. Furthermore, the scalability problem from the large real-world auction volume and campaign budget is well handled by state value approximation using neural networks.

Optimal Demand Response Bidding and Pricing Mechanism With Fuzzy Optimization: Application for a Virtual Power Plant

Abstract: In this paper, a virtual power plant (VPP) that consists of generation, both renewable and conventional, and controllable demand is enabled to participate in the wholesale markets. The VPP makes renewable energy sources (RES) and distributed generations controllable and observable to the system operator. The main objective is to introduce a framework that optimizes the bidding strategies and maximizes the VPP's profit on day-ahead and real-time bases. To achieve this goal, the VPP trades energy externally with a wholesale market, and trades energy and demand response (DR) internally with the consumers in its territory. That is, when generation exceeds demand, the VPP sells the excess energy to the market, and it buys energy from the market when the generation and reduction in demand due to DR scheme are less than the required demand in its territory. Both load curtailment and load shift are modeled. For the day-ahead internal VPP market, fuzzy optimization is proposed to consider the uncertainty in the RES. Comparison results with deterministic and probabilistic optimizations demonstrate the effectiveness of the fuzzy approach in terms of achieving higher realized profits with reasonable computation effort. It is also shown that considering uncertainties in the optimization can result in reduced dependence on the conventional generator.

Real-Time Demand Bidding for Energy Management in Discrete Manufacturing Facilities

Abstract: During periods of power system stress, demand bidding (DB) programs encourage large electricity consumers to submit curtailment capacity bids and carry out load reduction, in return for financial rewards. In this paper, a real-time DB (RT-DB) program with applications in discrete manufacturing facilities is considered. A discrete manufacturing production model is constructed and an automated RT-DB algorithm is designed. An optimization problem, where the objective is to maximize the profits for manufacturers, is formulated. Solving this problem enables the RT-DB algorithm to automatically generate optimal load-reduction bids with adjusted production and energy plans. A case study is described, which shows that the proposed algorithm reduced the load demand during an RT-DB event, increasing the manufacturer's profits. Furthermore, the relationship between the incentive rate and the demand elasticity of the consumer, as well as the production volume and profits is described.

Robust optimisation approach for bidding strategy of renewable generation-based microgrid under demand side management

Abstract: In the restructured electricity market, operator of grid-connected microgrid (MG) tries to supply local load at the lowest cost from alternative energy sources including upstream grid, gas-turbines as local dispatchable units and renewable energy sources (photovoltaic systems and wind-turbines) as well as charge/discharge of energy storage system. In order to purchase power from upstream grid, the bidding curve of MG should be prepared to bid to the market operator. Therefore, this study proposes a robust optimisation approach (ROA) for obtaining optimal bidding strategy of MG. Also, MG operator uses time-of-use rate of demand response program (DRP) to reduce procurement energy cost. For this purpose, ROA is used for uncertainty modelling of upstream grid prices in which the minimum and maximum limits of prices are considered for the uncertainty modelling. The lower and upper bounds of price are consecutively subdivided into sequentially nested subintervals which allow formulating robust mixed-integer linear programming problems. The bidding strategy curves of MG for each time considering DRP are obtained from sufficient data by solving these problems. To show the capability of proposed approach, two cases are studied.

Bidding strategy analysis of virtual power plant considering demand response and uncertainty of renewable energy

Abstract: Due to the gradual exhaustion of petroleum-based energy resources and severe concern for environmental protection, renewable energy (RE) resources and demand response (DR) techniques have been wide deployed in power network. However, the insufficient management as well as technology bottleneck becomes the major obstacle in their further development. Based on the uniform clearing of electricity market, a centralised dispatch model of virtual power plant (VPP) is introduced to improve the competitiveness of distributed energy resources in electricity market. To neutralise the side effect of RE penetration, a bidding strategy optimisation model considering DR and the uncertainty of RE for VPP is proposed and numerical analysis is conducted to prove its applicability. In addition, scenario analysis method is applied to deal with the influence of elastic demand and potential risk, which are associated with utility users’ consumption patterns and VPP's bidding preference, respectively. The application of distributed algorithm into multi-players’ strategy optimisation problem accelerated the convergence of bidding procedure, which verifies the applicability and effectiveness of the proposed models. Furthermore, numerical case studies demonstrate the distinctive superiority of VPP in the integration and management of RE and DR resources, which in turn contribute to its advantage position in electricity market.

Optimal Bidding in Performance-Based Regulation Markets: An MPEC Analysis With System Dynamics

Abstract: In this paper, we address the problem of optimal bidding in performance-based regulation markets for a large price-maker regulation resource. Focusing on the case of the California Independent System Operator (ISO), detailed market components are considered, such as regulation capacity payment, regulation mileage payment, performance accuracy adjustment, automatic generation control dispatch, and participation factor. Our analysis also incorporates system dynamics of the regulation resource for different resource types and technologies. In principle, our problem formulation is a mathematical program with equilibrium constraints (MPEC). However, our fundamentally new formulations introduce several new challenges in solving the MPEC problem in the context of performance-based regulation markets that are not previously addressed. In fact, global optimization techniques fail to solve the original nonlinear program due to its complexity. Therefore, we undergo several innovative steps to transform the problem into a mixed-integer linear program which is solved with accuracy, reliability, and computational efficiency. Insightful case studies are presented using data from a California ISO regulation market project.

Efficient Multi-Attribute Multi-Unit Auctions for B2B E-Commerce Logistics

Abstract: This study is the first proposing allocatively efficient multi‐attribute auctions for the procurement of multiple items. In the B2B e‐commerce logistics problem (ELP), the e‐commerce platform is the shipper generating a large number of online orders between product sellers and buyers, and third‐party logistics (3PL) providers are carriers that can deliver these online orders. This study focuses on the ELP with multiple attributes (ELP‐MA), which is generally the problem of matching the shipper's online orders and 3PL providers given that price and other attributes are jointly evaluated. We develop a one‐sided Vickrey–Clarke–Groves (O‐VCG) auction for the ELP‐MA. The O‐VCG auction leads to incentive compatibility (on the sell side), allocative efficiency, budget balance, and individual rationality. We next introduce the concept of universally unsatisfied set to construct a primal‐dual algorithm, also called the primal‐dual Vickrey (PDV) auction. We prove that the O‐VCG auction can be viewed as a single‐attribute multi‐unit forward Vickrey (SA‐MFV) auction. Both PDV and SA‐MFV auctions realize VCG payments and truthful bidding for general valuations. This result reveals the underlying link not only between single‐attribute and multi‐attribute auctions, but between static and dynamic auctions in a multi‐attribute setting.

First Price Auctions with General Information Structures: Implications for Bidding and Revenue

Abstract: We explore the impact of private information in sealed-bid first-price auctions. For a given symmetric and arbitrarily correlated prior distribution over values, we characterize the lowest winning-bid distribution that can arise across all information structures and equilibria. The information and equilibrium attaining this minimum leave bidders indifferent between their equilibrium bids and all higher bids. Our results provide lower bounds for bids and revenue with asymmetric distributions over values. We also report further characterizations of revenue and bidder surplus including upper bounds on revenue. Our work has implications for the identification of value distributions from data on winning bids and for the informationally robust comparison of alternative auction mechanisms.