A bottom-up building energy modelling at the urban scale based on Geographic Information System and semantic 3D city models can provide quantitative insights to tackle critical urban energy challenges. Nevertheless, incomplete information is a common obstacle to produce reliable modelling results. The residential building heating demand simulation performance gap caused by input uncertainties is discussed in this study. We present a data-driven urban scale energy modelling framework from open-source data harmonization, sensitivity analysis, heating demand simulation at the postcode level to Bayesian calibration with six years of training data and two years of validation data. Comparing the baseline and the calibrated simulation results, the averaged absolute percentage errors of energy use intensity in the study area have significantly improved from 25.0% to 8.3% and from 19.9% to 7.7% in two validation years, while ${\mathrm{C}\mathrm{V}\mathrm{R}\mathrm{M}\mathrm{S}\mathrm{E}}_{2016}=11.5\mathrm{\%}$ and ${\mathrm{C}\mathrm{V}\mathrm{R}\mathrm{M}\mathrm{S}\mathrm{E}}_{2017}=13.2\mathrm{\%}$. The overall methodology is extendable to other urban contexts.

基于地理信息系统和语义3D城市模型的自下而上的城市规模建筑能源模型可以提供定量的见解，以解决关键的城市能源挑战。但是，信息不完整是产生可靠建模结果的常见障碍。本研究讨论了由输入不确定性引起的住宅建筑供热需求模拟性能差距。我们提供了一个数据驱动的城市规模能源建模框架，从开源数据协调，敏感性分析，邮政编码水平的供热需求模拟到贝叶斯校准，其中包含六年的培训数据和两年的验证数据。比较基准线和校准后的模拟结果，${\mathrm{C}\mathrm{V}\mathrm{[R}\mathrm{中号}\mathrm{小号}\mathrm{Ë}}_{\mathrm{2016年}}=11.5\mathrm{％}$ 和 ${\mathrm{C}\mathrm{V}\mathrm{[R}\mathrm{中号}\mathrm{小号}\mathrm{Ë}}_{\mathrm{2017年}}=13.2\mathrm{％}$。总体方法可扩展到其他城市环境。