项目 | 彭立群

Portfolio

Models, systems, and the decisions they support

Each case study states the problem, what I built, what the analysis found, and where its conclusions stop.

Theme 01

How can planners evaluate large new loads more quickly without hiding engineering and regulatory constraints?

四类专家智能体、混合检索，以及一套针对并网与大型负荷问题的盲评框架。

一个 4300 万变量的选址模型，把风资源、储能、电网备用和网络时延耦合起来，评估吉瓦级清洁算力。

Theme 02

Where can large loads and industrial facilities obtain reliable low-carbon energy at a workable cost?

548 家钢铁、合成氨和甲醇工厂逐一匹配逐小时可再生能源资源：66% 可完全离网运行，允许 8% 电网供电后升至 79%。

决定项目落地的是省级优先事项，而非全国成本最优。到 2050 年海上风电可提供中国 3% 至 18% 的电力。

Theme 03

Which deployment and policy choices reduce cost and emissions, and where do those objectives diverge?

让各省自选储能策略总是最省钱的，但成本最低和排放最低并不是同一种策略。

在以煤电为主的电力系统里，补贴电网电解氢会让排放高于煤制氢。