LAaDT-DSS:面向非专家的土耳其屋顶光伏决策支持系统
LAaDT-DSS: decision support for non-experts on rooftop PV, incentive and grid simulations under Turkish tariff rules
这套系统让不懂建模的人也能算屋顶光伏该不该装,语言模型只负责调工具不瞎算,参数还标注来源。±3.6% 的预测偏差和 2,048 次抽样验证说得过去,做能源决策工具的可以看看它的架构。
土耳其研究团队发布 LAaDT-DSS,一套针对屋顶光伏投资、激励成本与电网馈线限制三类问题的决策支持软件,全部计算遵循土耳其电价规则,由单一仿真内核完成。用户通过在 3D 地图上点击房屋或用自然语言提问即可使用;语言模型只负责调用仿真工具,不参与数值计算。系统维护一个 125 条目的参数注册表,每条参数标注为实测来源或估算值。
验证方面,其光伏发电量预测与两个独立工具相比偏差在 ±3.6% 以内(同等损耗条件下)。不确定性测试中,对估算输入做 2,048 次抽样,仅光伏方案的结论除一次外全部保持稳定;加入电池储能后,31%-41% 的抽样中结论发生变化。四个脚本化对话场景端到端跑通,但尚未开展用户研究。
该系统的价值在于把投资分析、政策成本核算与电网约束检验放进同一工具,并通过参数溯源标注提升可信度。对发展中国家的电网规划和户用光伏推广具有参考意义。
LAaDT-DSS: decision support for non-experts on rooftop PV, incentive and grid simulations under Turkish tariff rules
Rooftop photovoltaics raise three questions: should a household invest, which incentive costs the state least, and does the feeder stay within limits? Those who ask are rarely modellers. We present LAaDT-DSS, decision-support software that answers all three under Turkish tariff rules from one simulation core. Non-experts reach it by clicking a house on a 3D map or by asking in plain words; a language model is asked to call the simulation tools, not to compute. Each entry of its 125-entry parameter registry is marked as sourced or estimated. Case studies test the system. Its PV yield lies within $\pm$3.6% of two independent tools at equal losses. Across 2,048 draws of its estimated inputs, the answer for PV alone held in all but one; with a battery, it changed in 31-41% of the draws. Four scripted chat scenarios ran end to end; no user study was run.
本条情报归属 AITOP「论文」核心追踪栏目。在跨区域产业观察坐标中,该动态直接关联「AI 开发者工具链、Coding Agent 与工程化实践」研究主线,反映出当前技术演进与商业落地的最新风向。涉及核心实体与技术议题包括:LAaDT-DSS、光伏、决策支持系统、语言模型,读者可通过下方标签进入对应专题页查阅历史脉络与关联对比。
信源核验说明:本文原始线索采编自「arXiv cs.AI」(@Baran Kaynak),经 AITOP 多维评分管线结构化提炼。