arXiv subjects

Lang Tong

Publications and source records attributed to Lang Tong.

At least 37 records · Page 2Linked to original sources

arXiv2024

Co-Optimizing Distributed Energy Resources in Linear Complexity under Net Energy Metering

The co-optimization of behind-the-meter distributed energy resources is considered for prosumers under the net energy metering tariff. The distributed energy resources considered include renewable generations, flexible demands, and battery energy storage systems. An energy management system co-optimizes the consumptions and battery storage based on locally available stochastic renewables by solving a stochastic dynamic program that maximizes the expected operation surplus. To circumvent the exponential complexity of the dynamic program solution, we propose a closed-form and linear computation complexity co-optimization algorithm based on a relaxation-projection approach to a constrained stochastic dynamic program. Sufficient conditions for optimality for the proposed solution are obtained. Numerical studies demonstrate orders of magnitude reduction of computation costs and significantly reduced optimality gap.

eess.SY↗

arXiv2024

On the Optimality of Procrastination Policy for EV charging under Net Energy Metering

We consider the problem of behind-the-meter EV charging by a prosumer, co-optimized with rooftop solar, electric battery, and flexible consumptions such as water heaters and HVAC. Under the time-of-use net energy metering tariff with the stochastic solar production and random EV charging demand, a finite-horizon surplus-maximization problem is formulated. We show that a procrastination threshold policy that delays EV charging to the last possible moment is optimal when EV charging is co-optimized with flexible demand, and the policy thresholds can be computed easily offline. When battery storage is part of the co-optimization, it is shown that the net consumption of the prosumer is a two-threshold piecewise linear function of the behind-the-meter renewable generation under the optimal policy, and the procrastination threshold policy remains optimal, although the thresholds cannot be computed easily. We propose a simple myopic solution and demonstrate in simulations that the performance gap between the myopic policy and an oracle upper bound appears to be 0.5-7.5%.

eess.SY↗

arXiv2024

Generative Probabilistic Time Series Forecasting and Applications in Grid Operations

Generative probabilistic forecasting produces future time series samples according to the conditional probability distribution given past time series observations. Such techniques are essential in risk-based decision-making and planning under uncertainty with broad applications in grid operations, including electricity price forecasting, risk-based economic dispatch, and stochastic optimizations. Inspired by Wiener and Kallianpur's innovation representation, we propose a weak innovation autoencoder architecture and a learning algorithm to extract independent and identically distributed innovation sequences from nonparametric stationary time series. We show that the weak innovation sequence is Bayesian sufficient, which makes the proposed weak innovation autoencoder a canonical architecture for generative probabilistic forecasting. The proposed technique is applied to forecasting highly volatile real-time electricity prices, demonstrating superior performance across multiple forecasting measures over leading probabilistic and point forecasting techniques.

cs.LG↗

arXiv2024

Dynamic Net Metering for Energy Communities

We propose a social welfare maximizing market mechanism for an energy community that aggregates individual and community-shared energy resources under a general net energy metering (NEM) policy. Referred to as Dynamic NEM (D-NEM), the proposed mechanism dynamically sets the community NEM prices based on aggregated community resources, including flexible consumption, storage, and renewable generation. D-NEM guarantees a higher benefit to each community member than possible outside the community, and no sub-communities would be better off departing from its parent community. D-NEM aligns each member's incentive with that of the community such that each member maximizing individual surplus under D-NEM results in maximum community social welfare. Empirical studies compare the proposed mechanism with existing benchmarks, demonstrating its welfare benefits, operational characteristics, and responsiveness to NEM rates.

eess.SY↗

arXiv2024

Wholesale Market Participation of DERA: DSO-DERA-ISO Coordination

Distributed energy resource aggregators (DERAs) must share the distribution network together with the distribution utility in order to participate in the wholesale electricity markets that are operated by independent system operators (ISOs). We propose a forward auction that a distribution system operator (DSO) can utilize to allocate distribution network access limits to DERAs. As long as the DERAs operate within their acquired limits, these limits define operating envelopes that guarantee distribution network security, thus defining a mechanism that requires no real-time intervention from the DSOs for DERAs to participate in the wholesale markets. Our auctions take the form of robust and risk-sensitive markets with bids/offers from DERAs and utility's operational costs. Properties of the proposed auction, e.g., resulting surpluses of DSO and the DERAs, and the auction prices, along with empirical performance studies, are presented.

eess.SY↗

arXiv2023

Resource Sharing in Energy Communities: A Cooperative Game Approach

We analyze the overall benefits of an energy community cooperative game under which distributed energy resources (DER) are shared behind a regulated distribution utility meter under a general net energy metering (NEM) tariff. Two community DER scheduling algorithms are examined. The first is a community with centrally controlled DER, whereas the second is decentralized letting its members schedule their own DER locally. For both communities, we prove that the cooperative game's value function is superadditive, hence the grand coalition achieves the highest welfare. We also prove the balancedness of the cooperative game under the two DER scheduling algorithms, which means that there is a welfare re-distribution scheme that de-incentivizes players from leaving the grand coalition to form smaller ones. Lastly, we present five ex-post and an ex-ante welfare re-distribution mechanisms and evaluate them in simulation, in addition to investigating the performance of various community sizes under the two DER scheduling algorithms.

cs.GT↗

arXiv2023

Enhancing Microgrid Resilience with Green Hydrogen Storage

We consider the problem of hydrogen storage integration in microgrids to improve the electricity supply resilience. Nonlinear effects from electrochemical models of electrolyzers and fuel cells for hydrogen storage are considered, making scheduling under the nonlinear model intractable and the conventional linear approximation infeasible. A piecewise linear model approximation with feasibility projection is proposed, resulting in a computationally efficient model predictive control for hydrogen storage operation. Several resilience performance measures, such as loss-of-load, duration-of-outage, and system cost, are used in performance evaluation. Simulations for the proposed optimization demonstrated a 13%-48% reduction in duration-of-outage, a 6.4%-21.7% reduction in system cost, and a 95% reduction in loss-of-load for critical loads compared to the scheduling algorithm involving linear model approximation. The performance gap of the proposed optimization to the benchmark involving the accurate nonlinear electrochemical model is less than 1% in most metrics.

eess.SY↗

arXiv2023

Convexifying Regulation Market Clearing of State-of-Charge Dependent Bid

We consider the problem of merchant storage participating in the regulation market with state-of-charge (SoC) dependent bids. Because storage can simultaneously provide regulation up and regulation down capacities, the market-clearing engine faces the computation challenge of evaluating storage costs under different regulation scenarios. One approach is to employ a bilevel optimization that minimizes the worst-case storage cost among all potential regulation events. However, subproblems of such a bilevel optimization are nonconvex, resulting in prohibitive computation challenges for the real-time clearing of the regulation market. We show that the complex nonconvex market clearing problem can be convexified by a simple restriction on the SoC-dependent bid, rendering the intractable market clearing computation to standard linear programs. Numerical simulations demonstrate that SoC-dependent bids satisfying the convexification conditions increase the profits of merchant storage owners by 12.32-77.38% compared with SoC-independent bids.

eess.SY↗

arXiv2023

Operating-Envelopes-Aware Decentralized Welfare Maximization for Energy Communities

We propose an operating-envelope-aware, prosumer-centric, and efficient energy community that aggregates individual and shared community distributed energy resources and transacts with a regulated distribution system operator (DSO) under a generalized net energy metering tariff design. To ensure safe network operation, the DSO imposes dynamic export and import limits, known as dynamic operating envelopes, on end-users' revenue meters. Given the operating envelopes, we propose an incentive-aligned community pricing mechanism under which the decentralized optimization of community members' benefit implies the optimization of overall community welfare. The proposed pricing mechanism satisfies the cost-causation principle and ensures the stability of the energy community in a coalition game setting. Numerical examples provide insights into the characteristics of the proposed pricing mechanism and quantitative measures of its performance.

eess.SY↗

arXiv2023

Co-optimizing Consumption and EV Charging under Net Energy Metering

We consider the co-optimization of flexible household consumption, electric vehicle charging, and behind-the-meter distributed energy resources under the net energy metering tariff. Using a stochastic dynamic programming formulation, we show that the solution to the dynamic programming co-optimization is a procrastination threshold policy that delays and minimizes electricity purchasing for EV charging in each time interval. The policy thresholds can be computed off-line, simplifying the continuous action space dynamic optimization to decoupled closed-form charging and consumption decisions. Empirical studies using renewable, consumption, and EV data demonstrate the benefits of co-optimization.

eess.SY↗

arXiv2023

Non-parametric Probabilistic Time Series Forecasting via Innovations Representation

Probabilistic time series forecasting predicts the conditional probability distributions of the time series at a future time given past realizations. Such techniques are critical in risk-based decision-making and planning under uncertainties. Existing approaches are primarily based on parametric or semi-parametric time-series models that are restrictive, difficult to validate, and challenging to adapt to varying conditions. This paper proposes a nonparametric method based on the classic notion of {\em innovations} pioneered by Norbert Wiener and Gopinath Kallianpur that causally transforms a nonparametric random process to an independent and identical uniformly distributed {\em innovations process}. We present a machine-learning architecture and a learning algorithm that circumvent two limitations of the original Wiener-Kallianpur innovations representation: (i) the need for known probability distributions of the time series and (ii) the existence of a causal decoder that reproduces the original time series from the innovations representation. We develop a deep-learning approach and a Monte Carlo sampling technique to obtain a generative model for the predicted conditional probability distribution of the time series based on a weak notion of Wiener-Kallianpur innovations representation. The efficacy of the proposed probabilistic forecasting technique is demonstrated on a variety of electricity price datasets, showing marked improvement over leading benchmarks of probabilistic forecasting techniques.

cs.LG↗

arXiv2023

Convexifying Market Clearing of SoC-Dependent Bids from Merchant Storage Participants

State-of-charge (SoC) dependent bidding allows merchant storage participants to incorporate SoC-dependent operation and opportunity costs in a bid-based market clearing process. However, such a bid results in a non-convex cost function in the multi-interval economic dispatch and market clearing, limiting its implementation in practice. We show that a simple restriction on the bidding format removes the non-convexity, making the multi-interval dispatch of SoC-dependent bids a standard convex piece-wise linear program.

eess.SY↗

arXiv2022

Co-optimizing Behind-The-Meter Resources under Net Metering

We consider the problem of co-optimizing behind-the-meter (BTM) storage and flexible demands with BTM stochastic renewable generation. Under a generalized net energy metering (NEM) policy-NEM X, we show that the optimal co-optimization policy schedules the flexible demands based on a load priority list that defers less prioritized loads to times when the BTM generation is abundant. This gives rise to the notion of a net-zero zone, which we quantify under different distributed energy resources (DER) compositions. We highlight the special case of inflexible demands that results in a storage policy that minimizes the imports and exports from and to the grid. Comparative statics are provided on the optimal co-optimization policy. Simulations using real residential data show the surplus gains of various customers under different DER compositions.

eess.SY↗

arXiv2022

DSO-DERA Coordination for the Wholesale Market Participation of Distributed Energy Resources

We design a coordination mechanism between a distribution system operator (DSO) and distributed energy resource aggregators (DERAs) participating directly in the wholesale electricity market. Aimed at ensuring system reliability while providing open access to DERAs, the coordination mechanism includes a forward auction that allocates access limits to aggregators based on aggregators' bids that represent their benefits of aggregation. The proposed coordination mechanism results in decoupled DSO and DERAs operations that satisfy the network constraints, independent of the stochasticity of the renewables, wholesale real-time locational marginal prices, and individual DERA's aggregation policy. Optimal bidding strategies by competitive aggregators are also derived. The efficiency of the coordination mechanism and the locational price distribution at buses of a radial distribution grid are demonstrated for a 141-bus radial network.

eess.SY↗

arXiv2022

Achieving Social Optimality for Energy Communities via Dynamic NEM Pricing

We propose a social welfare maximizing mechanism for an energy community that aggregates individual and shared community resources under a general net energy metering (NEM) policy. Referred to as Dynamic NEM, the proposed mechanism adopts the standard NEM tariff model and sets NEM prices dynamically based on the total shared renewables within the community. We show that Dynamic NEM guarantees a higher benefit to each community member than possible outside the community. We further show that Dynamic NEM aligns the individual member's incentive with that of the overall community; each member optimizing individual surplus under Dynamic NEM results in maximum community's social welfare. Dynamic NEM is also shown to satisfy the cost-causation principle. Empirical studies using real data on a hypothetical energy community demonstrate the benefits to community members and grid operators.

eess.SY↗

arXiv2022

Novelty Detection in Time Series via Weak Innovations Representation: A Deep Learning Approach

We consider novelty detection in time series with unknown and nonparametric probability structures. A deep learning approach is proposed to causally extract an innovations sequence consisting of novelty samples statistically independent of all past samples of the time series. A novelty detection algorithm is developed for the online detection of novel changes in the probability structure in the innovations sequence. A minimax optimality under a Bayes risk measure is established for the proposed novelty detection method, and its robustness and efficacy are demonstrated in experiments using real and synthetic datasets.

cs.LG↗

arXiv2022

Pricing Real-time Stochastic Storage Operations

Pricing storage operation in the real-time market under demand and generation stochasticities is considered. A scenario-based stochastic rolling-window dispatch model is formulated for the real-time market, consisting of conventional generators, utility-scale storage, and distributed energy resource aggregators. We show that uniform pricing mechanisms require discriminative out-of-the-market uplifts, making settlements under locational marginal pricing (LMP) discriminative. It is shown that the temporal locational marginal pricing (TLMP) that adds nonuniform shadow prices of ramping and state-of-charge to LMP removes the need for out-of-the-market uplifts. Truthful bidding incentives are also established for price-taking participants under TLMP. Revenue adequacy and uplifts are evaluated in numerical simulations.

eess.SY↗

arXiv2022

Wholesale Market Participation of Storage with State-of-Charge Dependent Bids

Wholesale market participation of storage with state-of-charge (SoC) dependent bids results in a non-convex cost in a multi-interval economic dispatch, which requires a mixed-integer linear program in the market clearing. We show that the economic dispatch can be convexified to the standard linear program when the SoC-dependent bid satisfies the equal decremental-cost ratio (EDCR) condition. Such EDCR bids are shown to support individual rationalities of all market participants in both the day-ahead multi-interval economic dispatch under locational marginal pricing and the rolling-window look-ahead dispatch under temporal-locational marginal pricing in the real-time market. A numerical example is presented to demonstrate a higher profit margin with an SoC-dependent bid over that from an SoC-independent bid.

eess.SY↗