Abstract In European households, 79% of the energy is consumed for space heating and cooling. The remote detection of possible retrofitting targets can help to increase the renovation rate and hence contribute to the realization of the 2000 W society. Here, a new method to characterize buildings based on smart meter monitoring data and a simplified physical simulation model is presented. The aim of this method is to estimate the time dependent demand of heating energy based on weather data applying these simplified building models. The method has been successfully applied on simulation and real-world smart meter monitoring data. The annual space energy demand was excellently reproduced with a deviation of less than 1% and 8% for simulation and real-world buildings, respectively. The recovered relevant building parameters deviate less than 1% for the reference case. The successful application of the algorithm on in-silico and real-world data monitoring data indicates the vast potential of this automated modelling technique on heat load prediction and energy-efficient operation of buildings. Furthermore, the derived heat demand profile may help utilities and facility managers in the future to identify better operation schedules of small areas and districts.

中文翻译：

---

基于智能电表测量数据的远程楼宇表征初步结果

摘要 在欧洲家庭中，79% 的能源用于空间供暖和制冷。远程检测可能的改造目标有助于提高改造率，从而有助于实现 2000 W 社会。在这里，提出了一种基于智能电表监控数据和简化的物理仿真模型来表征建筑物的新方法。该方法的目的是根据天气数据应用这些简化的建筑模型来估计热能的时间相关需求。该方法已成功应用于模拟和现实世界的智能电表监测数据。每年的空间能源需求得到了很好的再现，模拟和现实世界建筑的偏差分别小于 1% 和 8%。对于参考案例，恢复的相关建筑参数偏差小于 1%。该算法在计算机和真实世界数据监测数据上的成功应用表明这种自动化建模技术在热负荷预测和建筑物节能运行方面的巨大潜力。此外，派生的热量需求概况可以帮助公用事业和设施管理者在未来确定更好的小区域和地区的运营时间表。