| The technology of strain measurement is very important to modern industry development,such as the areas of mechanical design,construction industry,aerospace industry and so on.It is essential for a strain sensor to be calibrated many times in its service life.No matter when the strain sensor is released from factory,or remains unused for a long time,or a routine check is carried out,calibration is very significant that it guarantees the measurement reliable.However,in many domestic research institutions,problems such as manual manipulation and signal-channel measurement exist,which makes the efficiency of calibration work quite low.Taking this situation under consideration,a auto-calibration system particularly for strain sensor systems is developed to solve the above problems.This paper will introduce the principles of generation and measurement of stain signals and research different calibration theories.Based on these documents,a calibration method is decided for this very auto-calibration system.Besides,this paper will also study the different kinds of strain sensor systems in the market,make clear the measurement means inside them and finally set out a function list as a standard for the system design.In order to automate the calibration of strain gauge,a Cortex-M3 MCU based auto-calibration system was designed and the expected calibration precision is 0.5%.The system performs calibration through comparison between actual and theoretical outputs after auto-loading of reference signal.Finally,test precision and performance parameters will be calculated automatically.The system hardware composes of a multi-channel reference resistance signal generator with temperature compensation.The generator could provide standard resistance signal outputs with high precision through 8 channels based on combination of 3 types of bridge,4 types of base resistor and 5 types of strain simulation through relay control.The channels work together with time division multiplexing.The bridge types include 1/4-bridge,1/2-bridge and full-bridge and the base resistor types include 100?,120?,350? and 1K?.The main design ideology is to change the circuit topology by a group of relays to get the expected signal pattern.The software system insists of the upper software and lower software.System software transmits control commands and sample data through USB protocol.It could control signal generation and estimate the key performance index for the to-be-calibrated gauge.The upper software is coded by Python in Qt to realize parameter setting,data visualization,data processing and result document output.Bayesian statistics and least square method are combined to create a new calibration algorithm.The lower software is coded by C to realize hardware scheduling.Modular design method makes the system easy to modify and extend.In the end of this paper,the experiments of the system,including calibration experiment of strain measurement and function verification experiments,are presented.Experimental results had showed that system hardware could generate reference signal precisely and system software could provide the key performance index efficiently.The calibration system was realized as designed and has high performance. |
PDF Full Text Request