Application of artificial intelligence based on state grid ESG platform in clean energy scheduling optimization

Area of Science:

• * Power Systems Engineering
• * Artificial Intelligence
• * Data Science

Background:

• * The inherent randomness and volatility of clean energy sources complicate grid scheduling.
• * Traditional scheduling methods struggle with multi-objective optimization and real-time adaptation.
• * Existing systems face challenges in efficiently integrating and utilizing diverse clean energy inputs.

Purpose of the Study:

• * To develop an AI-driven method for optimizing clean energy grid scheduling.
• * To enhance clean energy utilization and reduce scheduling costs through advanced algorithms.
• * To address the complexities of grid scheduling in the context of renewable energy integration.

Main Methods:

• * Integration of an Environmental, Social, and Governance (ESG) big data platform for comprehensive data analysis.
• * Application of Particle Swarm Optimization (PSO) for initial scheduling optimization.
• * Implementation of a dual-layer architecture with Deep Q-Network (DQN) for real-time scheduling adjustments based on dynamic grid conditions.

Main Results:

• * Clean energy utilization increased from 62.4% to 87.7% compared to traditional methods.
• * Scheduling costs were reduced by 22% through optimized resource allocation.
• * The method demonstrated adaptability and stability across various load demands and clean energy supply scenarios.

Conclusions:

• * The proposed PSO-DQN method offers a novel and effective solution for clean energy scheduling optimization.
• * The integration with ESG big data platforms enhances decision-making accuracy and efficiency.
• * This research provides a strong foundation for improving clean energy integration in power grids.