The Online Diagnosis And Real-time Correction Of Sensor Faults In Building Energy Systems

With the continuous improvement of people's requirements for the comfort of the indoor environment,the application of HVAC systems in the construction field is becoming more and more extensive,but it also brings very serious energy problems.The existence of sensor error not only affects the precise control of the building system,resulting in a decline in the performance of the building system,but also has a very serious negative impact on building energy consumption,causing a lot of energy waste.Therefore,the accuracy of data collected by sensors is of great significance for reducing energy consumption in buildings.However,the traditional sensor calibration method is time-consuming and laborious,and it can't do anything for hundreds of sensors in large buildings.In order to solve the practical problems in building energy system,the following work is carried out:Firstly,a new sensor method,virtual in situ calibration(VIC),is proposed.The basic mathematical formula and calibration principle of the method are studied,and the calibration process of the method in building energy system is determined.Based on Bayes theory,this method can repair the systematic error of multiple working sensors in the system at the same time and greatly reduce the uncertainty brought by random error.Moreover,it does not need to remove the working sensors or add additional standard sensors like the traditional calibration method,which solves the limitations of existing sensor calibration methods in building energy system.Secondly,the virtual in-situ calibration method is applied to the actual primary air return reheating system with available boundary conditions,and the accuracy of this method is studied by six normal working conditions and four extreme working conditions.The results show that under ten different working conditions,the systematic error and random error of all working sensors in the system can be accurately identified,and the correction error of all sensors can be controlled within ± 5%.Finally,the virtual in-situ calibration method is applied to the actual freezing station system with unavailable boundary conditions,and the same design conditions as the air system are used to further verify the effectiveness of the method.The results show that the systematic errors of all sensors in the system have been accurately identified and repaired,and there is no significant difference between the calibration results of water system and air system.