Research on fast and intelligent calibration method based on automatic test system

Abstract

The paper studies a new calibration method based on the framework of automatic test system. The method uses IVI (Interchangeable Virtual Instruments) as the hardware basis of the system to discuss the fast and intelligent automatic calibration of the test equipment. It intelligently analyzes the correlation of system parameters for highly complex calibrated instruments using a stepwise regression, and accurately selects high correlation factors for the calibration of each parameter. And it performs accurate multistage segmentation of the calibrated parameters using the M5 model tree algorithm to ensure calculating the linear regression equation under the least mean square and realizing the fitting and calibration of each parameter; and calibration results are then configured as a file to the folder set by the calibrated instrument to replace the original calibration file. It does not require highly experienced engineers to perform a large number of analyses, but automatically tests, autonomously analyzes, automatically calculates, generates and gives calibration files. The system is fast, intelligent, and efficient. In the paper, the self-made pulse generator is taken as an example, and the amplitude of the pulse generator is calibrated. The calibrated system error is less than 1.5%.