User-centric Vehicle-to-Grid Optimization with an Input Convex Neural Network-based Battery Degradation Model

TL;DR

This work addresses how to balance V2G revenue with battery health by introducing a data-driven, user-centric framework. It leverages Partially Input Convex Neural Networks to model battery degradation as a function of temperature and time while maintaining convexity with respect to charging rate, enabling a convex optimization problem for personalized charging profiles. The approach combines offline PICNN training on extensive degradation data with a convex, multi-objective optimization that yields globally optimal, user-tailored strategies; the degradation predictor achieves $R^2=0.974$ on unseen data and a trade-off curve guides user decisions. Practically, the method offers reliable degradation-aware V2G optimization, adaptable to realistic price signals and scenarios, with future work focusing on experimental validation and forecasting prosumer preferences.

Abstract

We propose a data-driven, user-centric vehicle-to-grid (V2G) methodology based on multi-objective optimization to balance battery degradation and V2G revenue according to EV user preference. Given the lack of accurate and generalizable battery degradation models, we leverage input convex neural networks (ICNNs) to develop a data-driven degradation model trained on extensive experimental datasets. This approach enables our model to capture nonconvex dependencies on battery temperature and time while maintaining convexity with respect to the charging rate. Such a partial convexity property ensures that the second stage of our methodology remains computationally efficient. In the second stage, we integrate our data-driven degradation model into a multi-objective optimization framework to generate an optimal smart charging profile for each EV. This profile effectively balances the trade-off between financial benefits from V2G participation and battery degradation, controlled by a hyperparameter reflecting the user prioritization of battery health. Numerical simulations show the high accuracy of the ICNN model in predicting battery degradation for unseen data. Finally, we present a trade-off curve illustrating financial benefits from V2G versus losses from battery health degradation based on user preferences and showcase smart charging strategies under realistic scenarios.

Figures (6)

• Figure 1: Pictorial illustration of the proposed data-driven user-centric V2G system.
• Figure 2: Voltage, current, and temperature plots of RW9 battery cell during randomized battery cycling. The data for the above plot is taken from Bole2014.
• Figure 3: PICNN training loss with validation loss.
• Figure 4: Battery capacity prediction vs true capacity comparison.
• Figure 5: Trade-off between charging and battery health degradation cost.

Theorems & Definitions (3)

• Definition 1
• Definition 2
• Remark 1