Multi-industry zero-carbon park energy and multi-agent scheduling system

By combining a multi-agent scheduling system and a large language model (LLM) with a Transformer model and a mixed integer programming algorithm, the complexity problem in microgrid scheduling is solved, achieving efficient and reliable power supply scheduling and improving the autonomous operation capability and environmental adaptability of the microgrid.

CN122175220APending Publication Date: 2026-06-09BEIJING INST OF TECH +1
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
BEIJING INST OF TECH
Filing Date
2026-02-11
Publication Date
2026-06-09

AI Technical Summary

Technical Problem

Existing microgrid dispatching methods struggle to achieve efficient and reliable power dispatching when faced with the volatility of renewable energy, the dynamic complexity of loads, insufficient battery degradation management, and the challenges of heterogeneous data fusion and decision-making. In particular, they lack autonomous operation capabilities in high-uncertainty environments.

Method used

A multi-agent scheduling system is adopted, which combines an environmental perception module, a demand prediction module, a battery health management module, an optimization scheduling module, an execution and feedback module, and an LLM control module. Through the Transformer model and mixed integer linear programming algorithm, multi-objective optimization and anomaly response are achieved, generating high-precision scheduling decisions.

Benefits of technology

It significantly improves the operational reliability, economy, and environmental adaptability of microgrids under high renewable energy penetration and abnormal scenarios, enables autonomous dispatch and efficient data fusion, reduces battery degradation, and enhances system transparency and user trust.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122175220A_ABST
    Figure CN122175220A_ABST
Patent Text Reader

Abstract

This invention discloses a multi-industry zero-carbon park energy and multi-agent scheduling system. By embedding a microgrid operating system and adopting a modular multi-agent architecture, it combines real-time environmental perception, high-precision load demand prediction, battery health status optimization, dynamic multi-objective scheduling algorithms, and anomaly response mechanisms to achieve real-time adaptive energy management without human intervention. The system uses LLM as the core control module, which is responsible for integrating complex heterogeneous data (environmental parameters, prediction results, equipment status, user behavior, etc.), dynamically optimizing key decision parameters (such as trigger thresholds, optimization weights, and constraint boundaries), and achieving intelligent trade-offs among multiple objectives such as cost minimization, battery degradation minimization, supply and demand balance, and maximizing environmental benefits.
Need to check novelty before this filing date? Find Prior Art