第五届机器人、自动化与智能控制国际会议| Workshop21征稿中

人工智能算法把建模与仿真从“写方程、求解析”变成“喂数据、学规律”：先以深度神经网络、图网络、扩散模型等从海量量测、图像、文本中自动提取设备与系统的低维流形，生成可毫秒级响应的“数字替身”，取代传统高阶微分方程；(1) 论文须是全英文原创稿件，非纯综述类，整体须符合科技论文要求，摘要、关键词和结论部分须体现会议主题，须有方法、图表、实验数据和结果，应具有学术或实用推广价值，且未在国内外学术期刊或会

Joyelsp

703人浏览 · 2025-08-27 09:21:22

Joyelsp · 2025-08-27 09:21:22 发布

The 5th International Conference on Robotics, Automation and Intelligent Control
（ICRAIC 2025）

2025年10月31日-11月2日中国 | 成都

二轮截稿时间：9月25日

第五届机器人、自动化与智能控制国际会议（ICRAIC 2025）由成都理工大学主办，成都大学、爱迩思出版社（ELSP）、ESBK国际学术交流中心协办，将于2025年10月31日至11月2日在中国成都举行。大会邀请香港中文大学任洪亮教授、日本东京国立大学邓明聪教授、电子科技大学程洪教授进行主旨报告(主讲持续更新中,欢迎专家学者推荐交流）。

华北电力大学张悦副教授受邀参与组建专题论坛二十一，并担任论坛主席。张老师将于会议期间就人工智能方法在新型电力系统建模与仿真中的应用主题进行交流与探讨，欢迎加入专题论坛讨论交流。投稿参与该论坛请选择【Workshop 21】

【Workshop 21】专属投稿链接：

https://ocs.academicenter.com/submission/stepone?conf_id=1911663315256549376&workshop_id=1959819291708125184

WorkShop 21 简介

Title:

Applications of artificial intelligence methods in modeling and simulation of new-type power systems.

Keywords:

New power systems, artificial intelligence, modeling, simulation

新型电力系统，人工智能，建模，仿真

Summary：

建模与仿真是新型电力系统的“数字试航”，没有它就无法在源-网-荷-储深度融合、高比例新能源与电力电子设备并存的复杂环境中预判风险、优化运行、保障安全。

人工智能算法把建模与仿真从“写方程、求解析”变成“喂数据、学规律”：先以深度神经网络、图网络、扩散模型等从海量量测、图像、文本中自动提取设备与系统的低维流形，生成可毫秒级响应的“数字替身”，取代传统高阶微分方程；再借强化学习、迁移学习、物理信息神经网络把机理约束嵌入黑箱，实现跨尺度、跨场景的快速外推与可信度自评；

最后经云端并行与边缘推理，将原本需数小时的潮流、稳定、可靠性仿真压缩到秒级甚至毫秒级，并可在线滚动更新模型，使复杂系统“边运行、边学习、边演化”。

Modeling and simulation are the “digital sea trials” of new-type power systems: without them, it would be impossible to foresee risks, optimize operation, and ensure security in an environment where generation, network, load, and storage are deeply fused and high shares of renewables and power-electronic devices coexist.

Artificial-intelligence algorithms transform modeling and simulation from “writing equations and seeking analytical solutions” into “feeding data and learning patterns.” First, deep neural networks, graph networks, diffusion models, and similar techniques automatically extract low-dimensional manifolds of equipment and systems from vast streams of measurements, images, and text, creating millisecond-response “digital twins” that replace traditional high-order differential equations. Next, reinforcement learning, transfer learning, and physics-informed neural networks embed mechanistic constraints into the black box, enabling rapid cross-scale and cross-scenario extrapolation with built-in credibility assessment.

Finally, cloud-side parallelism and edge inference compress power-flow, stability, and reliability simulations that once took hours into seconds—or even milliseconds—while models are updated online in a rolling fashion, allowing the complex system to “run, learn, and evolve” in real time.

Chair

张悦

华北电力大学

He earned a B.Eng. (Sep 1999–Jul 2003) and a Ph.D. in Engineering (Sep 2003–Jun 2008) from North China Electric Power University.

Research interests include:

• Modeling and optimization control of power-generation processes

• Intelligent control theory and applications

• Application of artificial-intelligence algorithms in power systems

Over the past five years, he has led or participated in more than 20 government- and industry-sponsored projects, received two provincial/ministerial-level awards, published ten first-author papers, and authored the monograph Intelligent Control Theory and Applications.

Workshop参与形式

1. 研究论文投稿（每篇注册稿件均附赠一名免费线下参会的名额，不参会无退费）；

2. 作相关研究成果的口头报告；

3. 海报展示参会；

4. 仅参会交流。

投稿指南

1. 论文出版检索信息：IEEE出版&提交EI、Scopus数据库；

2. 论文模板：ICRAIC

3. Workshop21专属投稿链接：

https://ocs.academicenter.com/submission/stepone?conf_id=1911663315256549376&workshop_id=1959819291708125184

4. 请根据以下几点准备您的论文：

(1) 论文须是全英文原创稿件，非纯综述类，整体须符合科技论文要求，摘要、关键词和结论部分须体现会议主题，须有方法、图表、实验数据和结果，应具有学术或实用推广价值，且未在国内外学术期刊或会议发表过；