当前位置: X-MOL 学术Sci. Tech. Built Environ. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
The virtual in-situ calibration of various physical sensors in air handling units
Science and Technology for the Built Environment ( IF 1.9 ) Pub Date : 2020-08-13 , DOI: 10.1080/23744731.2020.1798175
Peng Wang 1, 2 , Kaihong Han 1 , Ruobing Liang 1 , Liangdong MA 1 , Sungmin Yoon 3, 4
Affiliation  

Systematic and random errors for working sensors in building systems can seriously affect the operation performance and have a significant negative impact on energy efficiency and indoor comfort. However, the traditional method for sensor calibration is very time-consuming and labor-intensive. Therefore, a virtual in-situ calibration (VIC) method was proposed based on Bayesian inference that can simultaneously correct systematic errors for multiple working and greatly reduce random errors. Using an actual system with reheating of the primary return air, we introduced the principle of the VIC method and its application. The robustness and applicability of the method were verified under six normal and four extreme operating conditions. The effects of sensor system errors on the energy consumption of the building system and on indoor thermal comfort were also studied. The results showed that under various working conditions, systematic and random errors for all the working sensors in the system were accurately identified. On average, the systematic and random error after calibration were reduced by approximately 95% and 60%, respectively. Compared to operating conditions with systematic errors, the average energy consumption after VIC was reduced by more than 50%, and predicted mean vote values after calibration were within the comfort zone.



中文翻译:

空气处理机组中各种物理传感器的虚拟原位校准

建筑系统中工作传感器的系统性和随机性误差会严重影响运行性能,并对能源效率和室内舒适度产生显着的负面影响。然而,传统的传感器校准方法非常耗时和劳动密集型。因此,提出了一种基于贝叶斯推理的虚拟原位校准(VIC)方法,该方法可以同时纠正多个工作的系统误差并大大减少随机误差。以一次回风再加热的实际系统为例,介绍了VIC法的原理及其应用。在六种正常和四种极端操作条件下验证了该方法的稳健性和适用性。还研究了传感器系统误差对建筑系统能耗和室内热舒适度的影响。结果表明,在各种工作条件下,系统中所有工作传感器的系统误差和随机误差都得到了准确的识别。平均而言,校准后的系统误差和随机误差分别减少了约 95% 和 60%。与有系统误差的工况相比,VIC后的平均能耗降低了50%以上,校准后的预测平均投票值在舒适区之内。校准后的系统误差和随机误差分别降低了约 95% 和 60%。与有系统误差的工况相比,VIC后的平均能耗降低了50%以上,校准后的预测平均投票值在舒适区之内。校准后的系统误差和随机误差分别降低了约 95% 和 60%。与有系统误差的工况相比,VIC后的平均能耗降低了50%以上,校准后的预测平均投票值在舒适区之内。

更新日期:2020-08-13
down
wechat
bug