New method combines FMEA and Bayesian Network for root cause analysis in lithium-ion battery production.
problem Complex cause-effect relationships in lithium-ion battery production.
method Combining FMEA with Bayesian Network to detect and resolve inconsistencies.
result Holistic method builds large-scale cross-process Bayesian Failure Network for root cause analysis.
Survival models predict component failures using neural networks and resampled data.
problem Accurately predicting component failure times for maintenance planning.
method Neural network-based survival models trained on non-independent, homogeneously sampled data.
result Random resampling during training reduces dataset size and improves efficiency.
Study examines barriers to grid-connected battery systems in Spain, finding high cycle cost remains main obstacle.
problem Barriers to grid-connected battery systems in Spain's deregulated electricity market.
method Utilization analysis and concept of 'potentially profitable utilization time' introduced.
result High cycle cost remains the main barrier for grid-connected battery systems in Spain.
A latent function decomposition method is proposed for forecasting the capacity of lithium-ion battery cells. The method uses the Multi-Output Gaussian Process, a generative machine learning framework for multi-task and transfer learning. The MCGP decomposes the available capacity trends from multiple battery cells int…
Modeling intraday dispatch for wind-battery assets to meet grid targets.
problem Meeting grid targets for wind-battery assets in an intraday context.
method Developed a mathematical model with closed-form solutions and a novel algorithm for stochastic control.
result Calibrated model to 140+ wind-battery assets in Texas, demonstrating economic benefits.
This paper proposes a TL method for SOH estimation of lithium-ion batteries.
problem Accurately estimating battery SOH to prevent unexpected failures.
method Temporal synchronization and distribution similarity analysis for transfer learning.
result The proposed method achieves a root mean squared error of 0.0034, improving accuracy by 77%.
Method detects lithium-ion battery knee onset for early warning.
problem Early warning of lithium-ion battery performance changes.
method Temporal dynamics extraction and Matrix Profile anomaly detection.
result Knee onset detected for categorizing battery health.
The paper models battery valuation in intraday electricity markets, incorporating liquidity costs.
problem Valuing batteries in intraday electricity markets considering liquidity costs.
method Stochastic model for mid-prices combined with a deterministic model for liquidity costs, using dynamic programming for optimization.
result Liquidity costs significantly impact battery valuation, especially with multiple batteries.
Recent studies have shown that the aggregated dynamic flexibility of an ensemble of thermostatic loads can be modeled in the form of a virtual battery. The existing methods for computing the virtual battery parameters require the knowledge of the first-principle models and parameter values of the loads in the ensemble.…
New features from early battery cycles predict lifetime with high accuracy.
problem Accurately predicting battery lifetime under varying conditions is challenging due to manufacturing variability and usage-dependent degradation.
method Extracted features from regularly scheduled reference performance tests and used them to predict battery lifetime using a hierarchical Bayesian regression model.
result Demonstrated a lifetime prediction of in-distribution cells with 15.1% mean absolute percentage error using only the first 15% of data.
A new learning framework reduces PV-Battery system costs by 3.6%.
problem Optimizing PV-Battery systems for cost savings with accurate forecasts.
method Decision-focused learning framework integrating optimization and prediction.
result Decision-focused method reduces average electricity costs by 3.6%.
Grid-scale batteries' bid patterns in price uncertainty markets
problem Interpreting bids from grid-scale batteries in wholesale electricity markets under price uncertainty
method Developing an asset-level model of a price-taking battery
result Empirical results deliver insights into withholding behavior, uncertainty effects, and risk management reshaping bid curves
The study analyzes how probabilistic forecasts improve battery trading strategies in electricity markets.
problem Improvements in statistical forecast quality do not directly translate to economic value in battery trading strategies.
method The study frames battery optimization as a stochastic program based on fully probabilistic forecasts and examines decision quality under different uncertainty models.
result The study identifies two critical flaws in quantile-based trading strategies and provides theoretical justification and empirical evidence.
Accurately predicting the future capacity and remaining useful life of batteries is necessary to ensure reliable system operation and to minimise maintenance costs. The complex nature of battery degradation has meant that mechanistic modelling of capacity fade has thus far remained intractable; however, with the advent…
Framework optimizes battery storage for markets by separating long-term degradation from short-term market dynamics.
problem Intractable computation due to timescale mismatch between battery degradation and market dynamics.
method Approximate dynamic programming with value function approximation and pseudo-time encoding.
result Policy outperforms benchmarks in real-time market scenarios.
High-frequency trading strategy boosts battery storage profits.
problem Maximizing revenue for battery energy storage systems in intraday markets.
method Adapted dynamic programming for continuous intraday markets, considering limit order book dynamics.
result Dynamic programming strategy outperforms standard re-optimization methods, increasing profits by 58% and 14% respectively.
Graph neural network improves SOH estimation of lithium-ion batteries.
problem Accurate SOH estimation requires alignment of statistical distributions between training and testing datasets.
method Graph convolutional networks (GCNs) with anomaly detection for selecting discharge voltage segments.
result Achieves precise SOH estimation with a root mean squared error of less than 1%.
Accurately predicting the future health of batteries is necessary to ensure reliable operation, minimise maintenance costs, and calculate the value of energy storage investments. The complex nature of degradation renders data-driven approaches a promising alternative to mechanistic modelling. This study predicts the ch…
We use data on 124 batteries released by Stanford University to first try to solve the binary classification problem of determining if a battery is "good" or "bad" given only the first 5 cycles of data (i.e., will it last longer than a certain threshold of cycles), as well as the prediction problem of determining the e…
BESS shows potential in European markets for frequency support, but not for energy arbitrage.
problem Lack of profitability for BESS in energy arbitrage in most European markets.
method Proposed a general payoff model for BESS operation and calculated utilization factors for common applications.
result BESS shows higher potential in providing frequency support services, especially in Central Western and Northern Europe.
Paper optimizes battery storage in multiple energy markets for better profits.
problem Optimizing battery storage participation in multiple energy markets to balance supply and demand.
method Developed a joint bidding strategy combining intraday and frequency markets using mixed integer linear programming and a learned classifier strategy.
result The LCS increases overall profits by over 4% compared to static strategies and by more than 3% over a naive dynamic benchmark.
Combines multi-fidelity and asynchronous batch methods for faster experimental design.
problem Designing optimal experimental setups for battery performance.
method Algorithm combining multi-fidelity and asynchronous batch Bayesian Optimization.
result Algorithm outperforms single-fidelity batch and multi-fidelity sequential methods.
Recent changes to greenhouse gas emission policies are catalyzing the electric vehicle (EV) market making it readily accessible to consumers. While there are challenges that arise with dense deployment of EVs, one of the major future concerns is cyber security threat. In this paper, cyber security threats in the form o…
Generative model calibrates 3D battery cathode morphologies from 2D images.
problem Calibrate 3D morphologies of all-solid-state battery cathodes from 2D microscopy images.
method Combining GANs with excursion sets of Gaussian random fields.
result Calibrated digital twins enable systematic exploration of morphological scenarios.
The paper analyzes profitable bidding strategies for BESS in day-ahead and intraday markets.
problem Optimizing profitability of Battery Energy Storage Systems (BESS) in day-ahead and intraday markets.
method Employing the rolling intrinsic approach to model continuous intraday markets, accounting for bid-ask spreads and liquidity constraints.
result Multi-market bidding strategies outperform single-market participation, and relaxing daily cycling constraints can unlock additional value.
We simplify complex regression coefficients using linearization and feature comparison.
problem Interpreting high-dimensional regression coefficients from nonlinear responses.
method Developed a linearization method to derive feature coefficients and compare them with regression coefficients.
result Shows how regression coefficients relate to linearized feature coefficients and how they change under regularization.
New findings challenge the importance of forecast accuracy in battery storage optimization, highlighting the role of rank correlation instead.
problem The challenge of optimizing battery storage dispatch decisions in multi-market electricity trading using forecast accuracy metrics.
method A hierarchical three-layer optimization system trading in multiple markets (FCR, aFRR, day-ahead, intraday) with real market data.
result Rank correlation (Kendall tau) is a better predictor of intraday dispatch value than forecast accuracy (MAE), with a threshold of tau around 0.85-0.95 capturing up to 97-100% of perfect-foresight revenue.
CoExBO optimizes lithium-ion batteries with user input, enhancing trust and efficiency.
problem User distrust in Bayesian optimization due to opacity and lack of user input.
method Preference learning and iterative explanation to integrate user insights.
result Algorithm converges to optimal solution even with adversarial user inputs.
Model predicts BESS interactions and price impacts in energy markets.
problem Understanding BESS interactions and price formation in energy markets.
method Stochastic game-theoretic model with linear-quadratic differential game.
result Equilibrium controls and prices derived for BESSs in both heterogeneous and homogeneous settings.
Strategic valuation of efficient and well-timed network investments under uncertain electricity market environment has become increasingly challenging, because there generally exist multiple interacting options in these investments, and failing to systematically consider these options can lead to decisions that underva…
Wireless sensor networks are composed of distributed sensors that can be used for signal detection or classification. The likelihood functions of the hypotheses are often not known in advance, and decision rules have to be learned via supervised learning. A specific such algorithm is Fisher discriminant analysis (FDA),…
Optimizes BESS for cross-market energy arbitrage to boost revenues.
problem Charging and discharging BESS at optimal times to maximize profits.
method Developed a generic framework, backtest engine, and optimization strategy.
result Boosted revenues by 10% through strategic BESS operation.
This study examines how financial tick data becomes more random with time aggregation.
problem Investigating the randomness of financial tick data over time.
method Applied statistical randomness tests from NIST and TestU01 batteries to ultra-high frequency financial data.
result Financial tick data becomes increasingly random as the aggregation level of transaction time increases.
Proposes a new model for simulating electricity prices and their correlation structure.
problem Simulating and understanding the complex dynamics of intraday electricity prices.
method Develops a multidimensional statistical model based on Poisson measures, estimating three key parameters.
result Demonstrates the model's effectiveness in battery valuation through dynamic programming.
Extracts important peaks from XRD spectra using Attention mechanism.
problem Identifying significant peaks in XRD patterns for material properties.
method Convolutional neural network with Attention mechanism to analyze deep features.
result Selected lattice constant predicts cathodic material cell voltage.
The nullspace and regularization impact high-dimensional linear regression interpretability.
problem Interpreting high-dimensional linear regression coefficients in complex data.
method Optimization formulation to compare coefficients and physical knowledge.
result Regularization and z-scoring choices affect interpretability and true coefficient closeness.
Many applications require recovering a matrix of minimal rank within an affine constraint set, with matrix completion a notable special case. Because the problem is NP-hard in general, it is common to replace the matrix rank with the nuclear norm, which acts as a convenient convex surrogate. While elegant theoretical c…
A large consensus now seems to take for granted that the distributions of empirical returns of financial time series are regularly varying, with a tail exponent close to 3. We revisit this results and use standard tests as well as develop a battery of new non-parametric and parametric tests (in particular with stretche…
SQR Averaging improves probabilistic electricity price forecasting.
problem Accurate short-term price forecasting in electricity markets.
method Smoothing Quantile Regression Averaging.
result SQR Averaging leads to profit increases of up to 3.5% in day-ahead power trading.
This paper improves context-aware recommender systems by selecting and incorporating relevant low-dimensional contextual information.
problem Generating accurate recommendations is not enough; contextual information can cause issues like battery drain and privacy.
method Developed a feature-selection algorithm based on genetic algorithms to reduce context dimensions while maintaining explainability.
result The approach improves accuracy and transparency in recommendations, outperforming state-of-the-art models.
Economics tool predicts failure times in reliability systems.
problem Predicting optimal failure times in weighted k-out-of-n reliability systems with heterogeneous component failure.
method Using rational expectations to analyze and predict failure times in reliability systems with heterogeneous component failure.
result Different measures are optimal for predicting system failure depending on component failure distributions.
SEFR is a fast, energy-efficient classifier for ultra-low power devices.
problem Running machine learning on battery-powered devices is challenging due to time and energy constraints.
method SEFR is an ultra-low power classifier with linear time complexity for training and testing.
result SEFR is 63 times faster and 70 times more energy efficient than state-of-the-art classifiers.
Unified model predicts multi-mode failure with multi-sensor data.
problem Independent failure mode and RUL prediction ignores inherent relationship.
method Hierarchical Bayesian framework with Cox model, Gaussian process, and multinomial distributions.
result Robust uncertainty quantification and accurate prediction of multi-mode failure.
PAGER detects failures in deep regression models using a new framework.
problem Detecting failures in deep regression models.
method PAGER uses a combination of epistemic uncertainty and manifold non-conformity scores.
result PAGER accurately characterizes and detects failures in deep regressors.
Factor analysis aims to determine latent factors, or traits, which summarize a given data set. Inter-battery factor analysis extends this notion to multiple views of the data. In this paper we show how a nonlinear, nonparametric version of these models can be recovered through the Gaussian process latent variable model…
In this paper we investigate the adaptive market efficiency of the agricultural commodity futures market, using a sample of eight futures contracts. Using a battery of nonlinear tests, we uncover the nonlinear serial dependence in the returns series. We run the Hinich portmanteau bicorrelation test to uncover the momen…
Risk Advisor predicts and mitigates ML deployment failures.
problem Predicting and mitigating test-time failure risks of ML systems.
method Post-hoc meta-learner for estimating failure risks and uncertainties.
result Reliably predicts deployment-time failure risks across various ML models.
CalNF models rare failures with limited data, improving safety in autonomous systems.
problem Challenges in modeling and debugging rare safety-critical failures due to limited data.
method CalNF, a self-regularized framework for posterior learning from limited data.
result Achieves state-of-the-art performance on data-limited failure modeling and inverse problems.