Description

Virtual In-situ sensor Calibration (VIC) can calibrate systematic and random errors of sensors in building energy systems. It is conducted in-situ to handle the working sensors of operational systems, using Bayesian MCMC. It is important to understand hidden facors and their negative impact on the accuracy of VIC to solve the various errors at different working conditions. If the hidden factors can be defined and their negative impacts can be handled in the VIC process, the VIC can provide the good calibration results no matter what the sensor errors happen. Accordingly, this study investigates a sensitivity effect (as a hidden factor) and its impact on the VIC accuracy through a case study. It shows how the sensitivity affects the calibration results. In addition to that, the handling strategy to overcome the negative sensitivity effect is proposed and then evaluated in the second case study. The VIC method is applied to a single-effect LiBr-H2O refrigeration system for the two case studies.

Citation: 2018 Winter Conference, Chicago, IL, Conference Papers

Product Details

Published:

2018

Number of Pages:

8

Units of Measure:

Dual

File Size:

1 file , 2.6 MB

Product Code(s):

D-CH-18-C043