CN-121804713-B - Dilution refrigerator thermometer in-situ calibration system based on natural cooling dynamics

CN121804713BCN 121804713 BCN121804713 BCN 121804713BCN-121804713-B

Abstract

The invention relates to the technical field of temperature calibration of dilution refrigerators and discloses an in-situ calibration system of a dilution refrigerator thermometer based on natural cooling dynamics, which comprises the steps of calculating temperature curvature according to a dynamic equation of cooling power and temperature, substituting the temperature into the temperature curvature according to the temperature corresponding to a resistance value of the thermometer to obtain a calibration function of the resistance value and the temperature; the method comprises the steps of identifying a multi-value region of temperature, evolving a calibration function into a plurality of branch functions in the multi-value region, calculating the deviation metric of each branch function and the calibration function, calculating the support degree of each branch function on the calibration function in the multi-value region, adjusting calculation load to obtain real-time data so as to minimize the experiment times of verifying the calibration function, dynamically adjusting the verification strategy of a non-multi-value region according to the branch verification result of the multi-value region, and obviously reducing the repeated experiment cost of low-temperature calibration of a dilution refrigerator through dynamic branch and load guidance on the premise of not relaxing the calibration precision.

Inventors

WANG YANFEI
TAO JING
YIN HAIFENG
XU CHIDONG
LI TIANYU

Assignees

合肥科光量子科技有限公司

Dates

Publication Date: 20260508
Application Date: 20260310

Claims (8)

1. The in-situ calibration system for the dilution refrigerator thermometer based on natural cooling dynamics is characterized by comprising the following components: The temperature and cooling power mapping module is used for calculating temperature curvature according to a dynamic equation of cooling power and temperature, substituting the temperature into the temperature curvature according to the temperature corresponding to the resistance value of the thermometer, and obtaining a calibration function of the resistance value and the temperature; The temperature evolution branch judging module is used for identifying a multi-value area of the temperature, evolving a calibration function into a plurality of branch functions in the multi-value area, and calculating the deviation measurement of each branch function and the calibration function; The branch verification and load adjustment module is used for calculating the support degree of each branch function on the calibration function in the multi-value area, and adjusting the calculation load to acquire real-time data so as to minimize the experiment times for verifying the calibration function; And the non-multi-value area verification strategy optimization module is used for dynamically adjusting the verification strategy of the non-multi-value area according to the branch verification result of the multi-value area so as to reduce the repeated verification times.
2. The dilution refrigerator thermometer in-situ calibration system based on natural cooling dynamics as set forth in claim 1, wherein said temperature and cooling power mapping module is configured to calculate a temperature curvature according to a kinetic equation of cooling power and temperature, comprising: An equation that the cooling power and the temperature of the dilution refrigerator show nonlinear characteristics in a low temperature area; The low temperature zone represents The following are set forth; , wherein, In order to dilute the cooling power of the refrigerator, In order to be able to determine the temperature, Is a constant related to the dilution refrigerator and is used for representing the characteristics of the dilution refrigerator; calculating the curvature of the temperature change with time: , And Respectively a thermal response constant and a thermal capacity; substituting the cooling power into the curvature gives: ; Calculating the derivative of the curvature gives: ; For a pair of And (3) deriving to obtain: ; substituting the curvature formula to obtain: For indicating the acceleration degree of the temperature change with time, Is time.
3. The dilution refrigerator thermometer in-situ calibration system based on natural cooling dynamics as set forth in claim 2, wherein the substituting the temperature into the temperature curvature according to the temperature corresponding to the resistance value of the thermometer to obtain a calibration function of the resistance value and the temperature comprises: Acquiring the relation between resistance and temperature of the thermometer without exogenous heating: , wherein, Is the temperature The corresponding resistance value of the resistor is obtained, For the resistance value at the reference temperature, Is the temperature coefficient of the resistor; temperature was calculated from the resistance value: ; The calibration function is used for describing the constraint of resistance and temperature in the cooling process of the dilution refrigerator; substituting the temperature into curvature to obtain a calibration function of temperature and resistance , And The first and second derivatives of temperature, respectively.
4. The dilution refrigerator thermometer in-situ calibration system based on natural cooling dynamics as set forth in claim 3, wherein the temperature evolution branch discrimination module is configured to identify a multi-valued area of temperature, evolve a calibration function into a plurality of branch functions in the multi-valued area, and calculate a deviation metric of each branch function and the calibration function, and the method comprises: real-time recording of dilution refrigerator during operation Resistance value of time thermometer ; Real-time recording Temperature of time thermometer ; If there is a resistance value Corresponding to a plurality of temperatures and meeting a calibration function Constraint, determining the resistance value The time interval is a multi-value area; obtaining resistance values at the start of a multi-value region And a corresponding plurality of temperature values , wherein, For the ith temperature, according to The multiple branch functions of the evolution multivalue area are specifically: Setting arbitrary candidate first derivative ; Constructing a set of equations: , , solving the equation set in the multi-value region to obtain the temperature conforming to the calibration function An ith branch function evolving as a calibration function; Calculation of Inverse mapping of (a) Obtain the resistance value Calculating a resistance deviation term The method is used for describing the deviation degree of the resistance of the branch function and the measured resistance value; Calculating the first and second derivatives of the branching function as And ; Calculating a trend deviation term: , ; calculating a deviation measure from the resistance deviation term and the trend deviation term 。
5. The dilution refrigerator thermometer in-situ calibration system based on natural cooling dynamics as set forth in claim 4, wherein the branch verification and load adjustment module is configured to calculate a support of each branch function to the calibration function in a multi-valued area, adjust a calculation load to obtain real-time data, and minimize a number of experiments to verify the calibration function, and comprises: setting a deviation metric threshold ; Calculation of Support of the branching function of (2) on the calibration function , wherein, ; Estimating cost of adjusting computational load The method specifically comprises the following steps: Setting a plurality of load adjusting modules, and calculating the adjusting amplitude as Cost of behavior at time , wherein, As the amplitude weight coefficient, the higher the amplitude is used for quantization, the higher the cost is; Setting a risk of branch confusion caused by regulating a load ; , For the number of branching functions, Respectively predicting the s time before load adjustment The resistance values of the two branching functions, Prediction of load after adjustment The resistance values of the two branching functions, A positive number to prevent zero denominator; , wherein, Cost for branch confusion risk Is a weight of influence of (1); setting a cost threshold If (if) Disabling the adjustment of the computational load; If it is The computational load is adjusted in the multi-valued area to minimize the number of times the calibration function is repeatedly verified.
6. The dilution refrigerator thermometer in-situ calibration system based on natural cooling dynamics as set forth in claim 5, wherein said adjusting the computational load to obtain real-time data to minimize the number of experiments to verify the calibration function comprises: Acquiring data of a plurality of sampling points recorded in real time in a multi-value area, wherein each data comprises temperature and resistance value, and acquiring the highest value of the temperature And minimum value ; The branching function has a temperature interval in a multi-value region, and the temperature interval is between And The branch function of (2) is marked as a candidate function; setting a binary function , The representation candidate function is used to verify the calibration function, otherwise, ; Calculating for the candidate function: , ; Setting probability threshold ; Defining candidate function coverage for validating calibration functions A percentage of sampling points, wherein the number of sampling points is fixed in each verification experiment; solving to obtain an objective function, wherein the objective function is used for verifying a calibration function; The adjustment quantity of the load corresponding to the two objective functions is calculated and switched, specifically: Obtaining two objective functions with intersection points And Adjusting the calculated load from the objective function via the intersection point Switching to an objective function The method specifically comprises the following steps: Setting up The load adjustment amount corresponding to the intersection point is Calculating a first derivative ; Objective function at intersection point And Is equal to the temperature change rate: 。
7. The dilution refrigerator thermometer in-situ calibration system based on natural cooling dynamics as set forth in claim 6, wherein said adjusting the computational load obtains real-time data to minimize the number of experiments verifying the calibration function, further comprising: According to Solving the condition of switching is , Is the temperature at the intersection point, wherein, The constants related to the dilution refrigerator when two objective functions are calculated respectively; By adjusting load control For completing slave objective functions Switching to an objective function , , Is that Sensitivity to load, fitted by experiment; Obtaining 。
8. The dilution refrigerator thermometer in-situ calibration system based on natural cooling dynamics as set forth in claim 5, wherein the non-multi-valued area verification policy optimization module is configured to dynamically adjust a verification policy of the non-multi-valued area according to a branch verification result of the multi-valued area, so as to reduce a number of repeated verifications, and comprises: in the multi-value region, if the calibration function constraint is satisfied If the number of the branch functions is 1, performing non-multivalued area verification; Calculating the credibility of the target function to the verification of the calibration function , , For the sum of the support of the objective function at this time, Is a scale parameter; Adjusting verification frequency of non-multi-value area according to credibility of multi-value area Sum error threshold The method specifically comprises the following steps: , , is the default verification frequency for non-multi-valued areas, As the weight coefficient of the light-emitting diode, As a basis for the error threshold value, For adjusting the coefficients.

Description

Dilution refrigerator thermometer in-situ calibration system based on natural cooling dynamics Technical Field The invention relates to the technical field of dilution refrigerator temperature calibration, in particular to a dilution refrigerator thermometer in-situ calibration system based on natural cooling dynamics. Background The dilution refrigerator is widely used for achieving millikelvin-level refrigeration in low-temperature experiments, a thermometer of the dilution refrigerator is usually measured through resistance change, a cooling process in a low-temperature area has obvious dynamic characteristics, temperature change along with time has a nonlinear relation, and the same resistance value possibly corresponds to different temperature evolution paths, so that challenges are brought to in-situ calibration and real-time verification. The existing dilution refrigerator thermometer calibration technology generally relies on limited discrete calibration points or experience curves to establish a resistance-temperature mapping relation, and is difficult to continuously describe temperature, resistance and dynamic evolution relation thereof in a natural cooling process, so that calibration accuracy in a low-temperature area with rapid cooling rate change or obvious thermal inertia is reduced. In the prior art, nonlinear dynamics characteristics between cooling power below 100mK and temperature are generally ignored, and the first derivative and the second derivative of the temperature are not modeled, so that temperature curvature change cannot be effectively restrained, and the calibration accuracy of a low-temperature region is affected. In the prior art, the verification process of a multi-value region has the difficulty that the same resistance value can correspond to different temperatures under different cooling paths or different experimental conditions, and the single-point verification result is easy to have larger fluctuation and difficult to reproduce. The prior art lacks a recognition means for the temperature evolution state of the multi-value area, cannot determine a specific temperature branch corresponding to a certain measurement point, is difficult to take the observation data of the multi-value area as a reliable verification basis, can only verify the calibration function by repeated experiments or prolonged stabilization time, has high verification cost and long time consumption, is greatly influenced by a cooling path and initial conditions, and is difficult to meet the requirement of in-situ rapid calibration of a low-temperature area. The scheme provides a dilution refrigerator thermometer in-situ calibration system based on natural cooling dynamics. Disclosure of Invention The invention provides an in-situ calibration system of a dilution refrigerator thermometer based on natural cooling dynamics, which is used for promoting solving of the problems in the background art. The invention provides a dilution refrigerator thermometer in-situ calibration system based on natural cooling dynamics, which comprises the following steps: The temperature and cooling power mapping module is used for calculating temperature curvature according to a dynamic equation of cooling power and temperature, substituting the temperature into the temperature curvature according to the temperature corresponding to the resistance value of the thermometer, and obtaining a calibration function of the resistance value and the temperature; The temperature evolution branch judging module is used for identifying a multi-value area of the temperature, evolving a calibration function into a plurality of branch functions in the multi-value area, and calculating the deviation measurement of each branch function and the calibration function; The branch verification and load adjustment module is used for calculating the support degree of each branch function on the calibration function in the multi-value area, and adjusting the calculation load to acquire real-time data so as to minimize the experiment times for verifying the calibration function; And the non-multi-value area verification strategy optimization module is used for dynamically adjusting the verification strategy of the non-multi-value area according to the branch verification result of the multi-value area so as to reduce the repeated verification times. Optionally, the temperature and cooling power mapping module is configured to calculate a temperature curvature according to a dynamic equation of cooling power and temperature, and includes: An equation that the cooling power and the temperature of the dilution refrigerator show nonlinear characteristics in a low temperature area; The low temperature zone represents The following are set forth; , wherein, In order to dilute the cooling power of the refrigerator,In order to be able to determine the temperature,Is a constant related to the dilution refrigerator and is used for representing the characte