Abstract: Based on the thermocouple calibration method and combined with virtual instrument technology, an automatic thermocouple calibration system was developed. The hardware and software design of the device are presented, and the automatic analysis and processing of the acquired experimental data is highlighted. The developed system performs well in online automatic calibration of various working thermocouples. Keywords : Thermocouple; Calibration; Data analysis ; LABVIEW When the thermoelectric properties change beyond the specified range, the temperature indicated by the thermocouple will become distorted, and the temperature measurement error will increase. Therefore, thermocouples not only need to be calibrated before use, but also need to be verified after a period of use to determine the magnitude of their error. Verification refers to all the work done to evaluate whether the thermocouple performance is up to standard. However, most university laboratories still use traditional thermocouple simulation verification devices, directly comparing second-class standard platinum-rhodium-platinum (S-type) thermocouples with calibration thermocouples to verify industrial thermocouples. Data measurement is done by reading pointer dials and recording manually. This method is cumbersome, slow in response, long in testing cycle, difficult in data storage, cannot directly obtain verification conclusions, and the experimental results are unsatisfactory. Since this traditional thermocouple simulation verification device is difficult to adapt to the needs of modern production, an automatic thermocouple verification system has been developed based on thermocouple verification methods and the LabVIEW development platform combined with virtual instrument technology. 2. Verification Principle of Traditional Thermocouple Simulation Verification Device The verification principle and process of traditional thermocouples can be described by the following example: Assume the set measurement temperature is 600℃. After the furnace temperature stabilizes, the average thermoelectric potential of the standard thermocouple is VS = 5.252mV. On the S-type calibration table, when the working end is 600℃ and the free end is 0℃, the thermoelectric potential is VS_standard = 5.222mV. The S-type calibration certificate of this standard thermocouple shows that the thermoelectric potential when the working end is 600℃ and the free end is 0℃ is VS_certified = 5.242mV. The deviation value Δ = 5.222 - 5.242 = -0.02mV. The actual potential value measured by the standard thermocouple: VS_standard/actual = 5.252 + (-0.02) = 5.232mV. By reverse-checking the S-table, TS_standard = 599.5℃, and obtaining VS_tested = 5.230mV for the tested S(K ) type thermocouple. By reverse-checking the S-table, TS_tested = 599.2℃. The error of the tested thermocouple is directly calculated to be 0.3℃. The error is then analyzed to determine its temperature range and level, leading to a conclusion. 3. Design of the Thermocouple Self-Verification System Based on the thermocouple calibration principle diagram, we can use LabVIEW to modularize and integrate different parts of the simulation device. The PID controller can automatically control the temperature. A computer screen replaces the operating screen, allowing us to observe the real-time furnace temperature and historical control curves, which is both intuitive and accurate. This is much simpler and more direct than manual reading, and avoids human error. Simultaneously, the virtual instrument can automatically process data, allowing direct viewing of the calibration results from the panel. Calibration records can also be stored and printed for future retrieval. [align=center]Figure 1 System Structure Diagram[/align] In the specific software and hardware implementation, we use a standard thermocouple to perform two functions: both a standard thermocouple and a temperature control thermocouple. This reduces the difficulty of implementing temperature control and data acquisition in the software, and also avoids the phenomenon that the data acquisition program cannot be executed when the heating stable temperature differs greatly from the set temperature. The plug-and-play data acquisition device allows us to better integrate the software and hardware. 3.1 Self-tuning Fuzzy-PID Dual-Mode Composite Control of Tubular Calibration Furnace This paper proposes a composite self-tuning FUZZY-PID dual-mode control algorithm. Its basic idea is to use a parallel connection of a PID controller and a fuzzy controller. Fuzzy control is used for large deviations, while self-tuning PID control is used for small deviations. This improves the control accuracy and eliminates limit cycle oscillations, thus giving full play to the advantages of both and achieving optimal control. The schematic diagram of the self-tuning Fuzzy-PID dual-mode hybrid controller is shown in Figure 2. The system incorporates a Bang-Bang switch. The program switches control modes based on the magnitude of the deviation E. When the deviation E exceeds a certain threshold X, the system switches to a conventional fuzzy controller, which accelerates the response speed. When the deviation E is less than the threshold X, the system switches to a conventional PID controller, which significantly improves the steady-state accuracy. During control, a PID parameter relay self-tuning program can be initiated as needed, allowing the PID parameter tuning to be completed automatically by the control system without manual intervention. This results in a dual-mode controller characterized by fast response, high steady-state accuracy, and ease of use. Furthermore, the fuzzy and PID controllers can be designed separately. If a conventional fuzzy controller already exists, the original program does not need to be modified; only the PID algorithm and PID parameter self-tuning module need to be embedded, along with a threshold judgment statement. When E>X, K1=0, K2=1. When E≤X, K1=1, K2=0. [align=center] Figure 2 Temperature control scheme for tubular calibration furnace[/align] Where X is determined according to the characteristics of the object and the control requirements, or it can be determined during on-site debugging. Through this variable structure control, when the system error is large, the system response speed can be improved and the response process can be accelerated; when the error is small, the system damping can be improved, giving the process a small overshoot characteristic. It is worth mentioning that the design of the self-tuning Fuzzy-PID dual-mode controller does not require high accuracy of the mathematical model of the controlled object, and is not sensitive to changes in system parameters caused by changes in external conditions such as ambient temperature, indicating that the system has a certain robustness. In addition, compared with the traditional PID controller, the self-tuning Fuzzy-PID dual-mode controller reduces the tuning requirements of PID parameters due to the participation of fuzzy algorithms and parameter self-adjustment. 3.2 Signal Conditioning and Acquisition The hardware structure of the test system based on virtual instruments is: sensor → signal conditioner → data acquisition equipment → computer. Sensors convert various physical quantities such as temperature, pressure, and displacement into electrical quantities; signal conditioners perform preprocessing such as amplification, filtering, and isolation on the electrical signals; the main function of data acquisition equipment is to convert analog signals into digital signals, and it generally also has functions such as amplification, sample-and-hold, and multiplexing/resetting. [align=center]Figure 3 Signal Conditioning Circuit Design[/align] Signal conditioning systems can filter out unwanted components and noise from the tested signal. For slowly changing weak signals such as thermocouple temperature measurements, low-pass filters are often needed to reduce high-frequency components and improve the accuracy of digital-to-analog conversion. Low-pass filters can filter out all signal frequency components above the cutoff frequency. Many signal conditioning devices have 4Hz low-pass filters, which are well-suited for filtering out 50Hz AC noise from low-frequency sampled signals. When the detected signal contains high voltage peaks, it may damage the computer or injure the operator. In this case, it is necessary to isolate the computer from the sensor for safety reasons. Another reason for isolation is to ensure that the measurements of the data acquisition equipment are not affected by terrain differences. 3.3 Online Automatic Data Analysis and Processing The relationship between thermocouple potential and temperature is a high-order function, and the conversion relationship of temperature electromotive force is different in different temperature ranges. Therefore, segmental judgment is required. The conversion relationship between thermoelectric potential and temperature of commonly used S-type thermocouples is shown below. [align=center] Figure 4 Potential-temperature curve of S-type thermocouple[/align] In the range of -50 to 630.74℃, the thermoelectric potential of the S-type thermocouple has a 6th power polynomial relationship with temperature: E=ai ti (1) t is the temperature a0=0 a1 =5.3995782346 a2=1.251977E-2 a3=-2.2448217997E-5 a4=2.8452164949E-8 a5=-2.2440584544E-11 a6=8.5054166936E-15 In the range of 630.74 to 1064.43℃, the thermoelectric potential of the S-type thermocouple has a polynomial relationship with temperature: E=bi ti (2) b0=-2.9824481615E+2 b1=8.2375528221 b2=1.6453909942E-3 At 1064.43～1665℃, the polynomial relationship between the thermoelectric potential and temperature of the S-type thermocouple is: E=ci ti (3) c0=1.2766292175E+3 c1=3.4970908041 c2=6.3824648666E-3 c3=-1.5722424599E-6 The online automatic analysis and processing of data contains several important modules, which are: 1. Temperature-Electromotive Force Conversion Module. 2. Electromotive Force-Temperature Conversion Module. 3. Reading, Acquisition and Judgment Module. 4. Reading and Acquisition Module. 5. Δ Calculation Module (potential difference Δ between the calibration table and the standard certificate). 6. Mean Value Processing Module. 7. Verification Conclusion Processing Module. 8. Data Storage Processing Module. [align=center]Figure 5 Front Panel Design for Automatic Thermocouple Calibration[/align] The operator sets the calibration temperature point. After inputting the temperature point, this information is sent in three ways: one through the Δ module to display the potential difference Δ between the certificate and the standard calibration table; another through the temperature-to-electromotive force conversion module to display the standard calibration table potential at the set point; and the third through the reading and acquisition judgment module, waiting to compare it with the temperature signal from the standard thermocouple channel after electromotive force-to-temperature conversion. The reading and acquisition module controls its own on/off state. If the temperature from the standard channel does not reach the set temperature or differs significantly, the reading and acquisition judgment module will make a corresponding judgment—sending a shutdown command. Since the initial requirements for reading and acquisition are not met, the indicator light remains off, waiting for the temperature to rise. Only when the temperature from the standard thermocouple channel is not significantly different from the set temperature will the reading and acquisition judgment module control the opening of the reading and acquisition switch, allowing the signal from the channel to enter the data pre-acquisition state, and the acquisition preparation indicator light will illuminate. When the data acquisition conditions are fully met, the acquisition program will complete data acquisition within the specified interval (initially set to acquire data from 5 points within 3 minutes, the same number as the simulation device). The acquired potential signals of the standard thermocouple and the measured thermocouple are sent to the calibration result processing module and the data storage module respectively through the averaging module. The calibration result processing section compares the processed data with the set thermocouple standard to determine which level it meets. The processed information can be stored in a specified file path for viewing and printing. Simultaneously, the calibration results are displayed in real-time on the front panel for on-site observation by the operator. The entire data processing process is fully automated, avoiding operator reading errors and minimizing manual operations. Compared to traditional simulation devices, data acquisition time is significantly reduced, improving both efficiency and accuracy. We use the ice-point bath method for cold junction compensation, which provides high measurement accuracy and small error, facilitating data processing. [align=center] Figure 6 Automatic calibration flowchart of thermocouple[/align] 4 Conclusion Virtual instruments are the product of the integration of computer technology and modern measurement and control technology. They follow the concept of "software as instrument" and effectively combine computer resources, instrument measurement/control hardware and software for data analysis, process communication and graphical user interface, thereby greatly reducing the hardware resources of the instrument. They can also define the instrument function and structure according to the user's needs and design the user's own instrument. Therefore, in the calibration and division of thermocouples, the application of virtual instrument technology can improve work efficiency, save costs and improve the accuracy of calibration and division. References: [1] Lei Lin. Microcomputer Automatic Detection and System Design. Beijing. Electronic Industry Press, 2003. [2] National Instruments Corporation. Labview User Manual, April 2003 Edition [3] State Bureau of Standards. National Standard of the People's Republic of China GB3773-83 Platinum-Rhodium 10-Platinum Thermocouple Wire and Calibration Table. Beijing. China Standards Press, 1983.