A smart expansion generation method and system for extreme operation scenarios of a new power system

CN122948333APending Publication Date: 2026-10-20NORTH CHINA ELECTRIC POWER UNIV +2
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611106301.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-07-24
Publication Date
2026-10-20

AI Technical Summary

Technical Problem

[0010]鉴于上述的分析，本发明实施例旨在提供一种面向新型电力系统极端运行场景的智能扩展生成方法及系统，用以解决现有新型电力系统极端运行场景因历史样本稀缺、物理约束难以满足，以及风险特征不可控而导致的极端运行场景生成困难的技术问题

Benefits of technology

1.本发明破解极端样本稀缺困境，实现场景样本的智能扩展与风险分布可控迁移；针对现有技术因极端运行场景历史样本稀少而难以覆盖高风险运行状态的缺陷，本发明通过构建多维极端性度量指标，对历史样本进行量化评估与分级，并采用加权组合与时间序列重构两种方式对有限样本进行智能扩展，显著丰富了可用样本的数量与多样性。在此基础上，本发明进一步引入基于幂权重公式的分布调节机制，通过重采样主动提高高风险样本的占比，使样本分布向极端区间迁移，从而解决了极端场景“样本少、覆盖难”的技术问题，为后续生成模型提供了高质量、风险可控的训练数据基础；

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure FT_1
    Figure FT_1
  • Figure FT_2
    Figure FT_2
  • Figure SMS_149
    Figure SMS_149
Patent Text Reader

Abstract

This invention relates to an intelligent extended generation method and system for novel extreme operating scenarios of power systems. The method includes: acquiring historical operating data of the power system and preprocessing it to obtain an initial operating scenario sample set; calculating the extremity metric of each initial operating scenario sample in the initial operating scenario sample set; reconstructing the samples in the initial operating scenario sample set based on the extremity metric of each initial operating scenario sample to generate extended operating scenario samples; calculating the extremity metric of the extended operating scenario samples and adjusting the distribution of the extended operating scenario samples based on the extremity metric of the extended operating scenario samples to obtain a reconstructed scenario sample set; determining the probability level label of each sample in the reconstructed scenario sample set based on the mapping relationship between the extremity metric and the probability of occurrence of a preset target; and generating multi-type extreme operating scenario sequences using a CGAN network based on the reconstructed scenario samples, their corresponding probability level labels, and given conditions.
Need to check novelty before this filing date? Find Prior Art