Large Language Model Assisted Optimal Bidding of BESS in FCAS Market: An AI-agent based Approach

TL;DR

The paper tackles profitable BESS bidding in the FCAS market under Australia's NEM by modeling a two-stage, multi-band bidding process and addressing both aleatoric and epistemic uncertainties. It introduces a CVaR-DRL framework based on PPO to manage downside risk while maintaining exploration, and couples it with a novel LLMs-assisted AI-agent framework for online analysis, hybrid decision-making, and interpretable feedback. The key contributions are (i) a market-aware CVaR-DRL bidding strategy, (ii) a bi-level optimization framework that respects actual FCAS procedures, and (iii) an interactive DRL-LLMs loop that improves decision timeliness and robustness in unseen scenarios. Empirical results show higher profits and reduced risk compared to traditional methods, with LLMs further enhancing performance in unfamiliar market conditions, supporting practical deployment for BESS in FCAS operations.

Abstract

To incentivize flexible resources such as Battery Energy Storage Systems (BESSs) to offer Frequency Control Ancillary Services (FCAS), Australia's National Electricity Market (NEM) has implemented changes in recent years towards shorter-term bidding rules and faster service requirements. However, firstly, existing bidding optimization methods often overlook or oversimplify the key aspects of FCAS market procedures, resulting in an inaccurate depiction of the market bidding process. Thus, the BESS bidding problem is modeled based on the actual bidding records and the latest market specifications and then formulated as a deep reinforcement learning (DRL) problem. Secondly, the erratic decisions of the DRL agent caused by imperfectly predicted market information increases the risk of profit loss. Hence, a Conditional Value at Risk (CVaR)-based DRL algorithm is developed to enhance the risk resilience of bidding strategies. Thirdly, well-trained DRL models still face performance decline in uncommon scenarios during online operations. Therefore, a Large Language Models (LLMs)-assisted artificial intelligence (AI)-agent interactive decision-making framework is proposed to improve the strategy timeliness, reliability and interpretability in uncertain new scenarios, where conditional hybrid decision and self-reflection mechanisms are designed to address LLMs' hallucination challenge. The experiment results demonstrate that our proposed framework has higher bidding profitability compared to the baseline methods by effectively mitigating the profit loss caused by various uncertainties.

Figures (10)

• Figure 1: LLMs-based BESS Ancillary Service Autonomous Intelligent Bidding System.
• Figure 2: Optimization framework of the proposed models.
• Figure 3: Clearing results of the market clearing model under different BESS bidding capacities.
• Figure 4: BESS bidding results of the mathematical model.
• Figure 5: BESS bidding results of the DRL model.