Aiming at the problem that the current weighing calibration method for weighing sensors ignores the impact of abnormal water loss or gain

which can easily lead to larger uncertainties

a calibration function of weighing sensor with lower uncertainty is obtained by estimating the actual water weight injected from the standard vessel into the weighing tank.This estimation is achieved by fusing dualsensor signals based on the Kalman filtering algorithm

which incorporates a water loss prediction.A calibration system is established

and comparisons between the original and new methods are carried out in terms of uncertainty and performance of the calibration function used for in-situ flowmeter weighing method calibration with or without abnormal water loss/gain.Experimental results demonstrate that regardless of the presence of abnormal water loss/gain factors

the proposed method can effectively reduce uncertainty of the calibration function and significantly decrease flow measurement deviation during in-situ calibrations of flowmeter weighing method.