CN118259711A - Atomic gas room temperature control device and method based on Kalman filtering - Google Patents

Abstract

A Kalman filter-based atomic gas temperature control device and method, the device includes: an atomic gas chamber, a heating temperature measurement circuit module, a signal amplification circuit, an analog-to-digital/digital-to-analog conversion module, a power amplification circuit module, a high-frequency inverter circuit module, and a main control module; the temperature signal of the atomic gas chamber at the current moment is collected by a temperature measurement sensor, and the signal is amplified and converted into a digital signal by analog-to-digital conversion, and the optimal temperature estimation value is obtained by extended Kalman filter analysis, and the difference is made with the preset atomic gas chamber target temperature value to be achieved, and the optimal temperature estimation value is input into a PID controller, and the system control value signal is output through PI control, and then converted into an analog signal through digital-to-analog conversion, and converted into a high-frequency alternating current through a high-frequency inverter circuit, and then output to a heating resistor strip after power amplification, so as to adjust the heating power and control the gas chamber temperature. The present invention solves the measurement error caused by the nonlinear relationship between resistance and temperature of thermocouples by combining extended Kalman filtering and PID controller.

Description

Atomic gas room temperature control device and method based on Kalman filtering

Technical Field

The present invention relates to the technical field of heating and temperature control of a rubidium atom gas chamber of a SERF magnetometer, in particular to a device and method for accurately controlling heating power and regulating the gas chamber temperature based on a microprocessor-based PID feedback controller and an extended Kalman filter process.

Background technique

SERF magnetometers are widely used in measuring biological magnetic fields, such as mice, fruit flies, and the magnetic fields of the human brain and heart. Its advantages are that the production cost is much lower than that of SQUID, and it is lighter and its centimeter-level volume makes it highly mobile in space. At the same time, another advantage of SERF magnetometers is that they can achieve vector measurement, which has great potential in solving the problem of tracing the source of brain and nerve-related diseases.

The alkali atom gas cell is an important component of the SERF magnetometer. By heating the alkali atom gas cell to a specific temperature and reaching thermal equilibrium, high-density alkali metal vapor can be obtained. At this time, the spin exchange collision frequency between atoms is much higher than their own Larmor precession frequency. Under this condition, the atoms enter the excited state from the ground state, and the total spin precession of the external field is manifested as the average Larmor frequency, and the spin exchange relaxation is completely eliminated. Therefore, the temperature change of the atomic gas cell directly affects the state of the alkali metal atoms, thereby affecting the sensitivity of the magnetometer. The development engineering of the SERF magnetometer usually has strict requirements on the heating method and temperature control of the gas cell.

The heating method of the conventional magnetometer air chamber is usually to wrap a heating coil around the outer wall of the air chamber, and the temperature sensor collects the surface temperature of the air chamber and feeds it back to the main control module. According to the algorithm, the PID controller uses the difference between the measured temperature and the target temperature as the input value to calculate the output value, and feeds it back to the heating signal generation module in real time to adjust the amplitude, frequency, phase and other parameters of the heating signal, and finally generates heat through the heating coil to heat the air chamber. However, conventional heating methods usually have problems such as uneven heating of the air chamber, resulting in uneven distribution of atomic number density, and PID feedback control lag, which affects the sensitivity of the magnetometer and limits the SERF magnetometer's measurement of biomagnetic fields at the femtosecond (fT) level, such as measuring the magnetic field of the human brain.

Summary of the invention

In order to overcome the above-mentioned deficiencies in the prior art, aiming at the problem of uneven heating of the alkali metal gas chamber and taking into account the miniaturization and integration of the SERF magnetometer, the present invention provides an atomic gas chamber heating and temperature measuring device and a temperature control method based on Kalman filtering.

The present invention is an integrated design scheme for atomic gas chamber heating temperature measurement and its driving circuit and PID temperature control system; based on the hardware module, an implementation method for controlling heating power based on an extended Kalman filter algorithm overcomes the problems of residual error and control lag in traditional PID feedback temperature control technology.

In order to achieve the above-mentioned purpose, the present invention adopts the following technical solutions:

An atomic gas room temperature control device based on Kalman filtering, wherein the hardware module comprises: an atomic gas chamber 1, a heating temperature measurement circuit module 2, a signal amplification circuit 3, an analog-to-digital/digital-to-analog conversion module 4, a power amplification circuit module 5, a high-frequency inverter circuit module 6, and a main control module 7;

The atomic gas chamber 1 is connected to the heating and temperature measuring circuit module 2, and the atomic gas chamber 1 is filled with rubidium atomic gas of a certain density and nitrogen as quenching gas;

The atomic gas chamber 1 is connected to the heating and temperature measuring circuit module 2. When the heating resistor strip of the heating and temperature measuring circuit module 2 is energized, an electric current is generated to heat the atomic gas chamber 1. The temperature of the atomic gas chamber 1 changes, so that a potential difference is generated between the cold end and the hot end of the patch type thermocouple of the heating and temperature measuring circuit module 2. The temperature of the atomic gas chamber 1 is measured by measuring the potential difference.

The heating temperature measurement circuit module 2 is connected to the signal amplifying circuit 3, and the potential difference between the cold end and the hot end of the SMD thermocouple is used as an input analog voltage signal transmitted to the signal amplifying circuit 3, and the signal amplifying circuit 3 amplifies the analog voltage signal and outputs a temperature voltage;

The signal amplifying circuit 3 is connected to the analog-to-digital/digital-to-analog conversion module 4. The signal amplifying circuit 3 outputs the temperature voltage to the analog-to-digital/digital-to-analog conversion module 4, and converts the temperature voltage into a digital signal output through analog-to-digital conversion.

The analog-to-digital/digital-to-analog conversion module 4 is connected to the main control module 7, and the analog-to-digital/digital-to-analog conversion module transmits the digital signal to the main control module 7 as the input signal of the main control module 7; the main control module 7 receives the input signal, and after analysis and processing, outputs the system control quantity to the analog-to-digital/digital-to-analog conversion module 4;

The analysis and processing of the main control module 7 includes:

Step 1: Calculate the input temperature and voltage values according to the thermocouple electromotive force calculation formula to obtain the measured temperature value Z _k . The measured temperature value Z _k is in a floating point data format with 3 decimal places reserved.

Step 2: Input the measured temperature value Z _k into the extended Kalman filter 8 to calculate the posterior estimated temperature value Calculate the a posteriori estimated temperature value The difference T _off from the reference temperature value T _ref ;

Step 3: Using the difference T _off as the input of the PID controller 9, the system control variable U _k is calculated by the PI algorithm as the output value of the PID controller 9;

Step 4: Transmit the output value U _k to the analog-to-digital/digital-to-analog conversion module 4 through the SPI serial port circuit, and output the output value U _k as a low-voltage direct current signal through digital-to-analog conversion;

The analog-to-digital/digital-to-analog conversion module 4 is also connected to the high-frequency inverter circuit module 6. The analog-to-digital/digital-to-analog conversion module 4 outputs a low-voltage DC signal to the high-frequency inverter circuit module 6. The high-frequency inverter circuit module 6 converts the low-voltage DC signal into a high-frequency low-voltage AC signal.

The high-frequency inverter circuit module 6 is connected to the power amplifier circuit module 5. The power amplifier circuit module 5 receives the high-frequency low-voltage AC signal output by the high-frequency inverter circuit module 6, amplifies the power thereof, and outputs a heating signal.

The power amplifier circuit module 5 is connected to the heating and temperature measuring circuit module 2. The heating and temperature measuring circuit module 2 receives the heating signal of the power amplifier circuit module 5. The heating and temperature measuring circuit module 2 controls the heating coil to heat or cool down.

A second aspect of the present invention relates to an atomic gas temperature control method based on Kalman filtering, comprising the following steps:

Step S1: first set the reference temperature value T _ref = 140°C, then convert the measured temperature value Z _k by measuring the voltage value data at time k; determine the nonlinear expression of the prediction model of the temperature control extended Kalman filter 8 as:

T _k =f(T _k-1 ,U _k-1, W _k-1 ) (1)

Z _k =h(T _k ,V _k ) (2)

Among them, formula (1) is the nonlinear expression of the temperature of atomic gas chamber 1 at time k and the input current, formula (2) is the measurement formula of the temperature of atomic gas chamber 1 at time k; T _k is the predicted true value of the temperature of atomic gas chamber 1 at time k, Z _k is the measured true value of the temperature of atomic gas chamber 1 at time k; U _k-1 is the control vector at time k-1, W _k-1 is the process noise at time k-1, and V _k is the measurement noise at time k, both of which conform to the univariate Gaussian distribution:

P(W)～N(0,Q), P(V)～N(0,R)

In actual work, the above process noise and measurement noise are expected to be 0, which conforms to the normal distribution; Q is the covariance matrix of the estimated process noise, and R is the covariance matrix of the measured Gaussian noise, and its value is specifically determined by the noise of the heating temperature measurement circuit module 2;

Step S2: Determine the a priori estimated temperature value and the measured temperature value formula of the temperature control extended Kalman filter 8 prediction model as follows:

in is the a priori estimated temperature value of the temperature of atomic gas chamber 1 at time k, is the prior measurement value at time k;

Step S3: Update the prior state error covariance matrix of the prediction model The update formula is:

Where A is the Jacobian matrix.

Step S4: Update the Kalman gain K _k of the temperature-controlled extended Kalman filter 8 , and the update formula is:

Where H is the transformation matrix;

Step S5: Correct the optimal a posteriori estimated temperature value by information fusion method The posterior correction formula is:

Step S6: Update the predicted value P _k of the posterior state error covariance matrix. The update formula is:

Step S7: Setting the above optimal a posteriori estimated temperature value As the external temperature of the air chamber at time k, calculate the temperature offset from the reference temperature value T _ref As the input of PID controller 9; the calculation equation of general PID control is:

Among them, _Kp is the proportional coefficient of PID controller 9, _Ti is the integral time of PID controller 9, _Td is the differential time of PID controller 9, and u(t) is the system control quantity formed by linear combination of input _Toff through PID controller 9; the above formula cannot be implemented by programming in FPGA, so it is discretized and the integral part is converted into a summation formula, which is:

The above _Ki and _Kd are integral and differential control coefficients respectively, _Uk is the discretized system control quantity, the proportional control _Kp can quickly respond to the system error and reduce the steady-state error. If _Kp is too large, the system will be unstable; the integral control _Ki is the temperature offset that is continuously superimposed to eliminate the steady-state error. If _Ki is too large, the system will overshoot and cause system oscillation; the differential control _Kd is used to reduce the system overshoot and overcome the system oscillation, but the differential term is very sensitive to high-frequency noise, so the differential control is not enabled in the present invention, _Kd = 0;

Step S8: The system control quantity U _k is digitally output from the main control module through programming, and is converted into a heating signal to the heating temperature measurement module after digital-to-analog conversion, high-frequency inversion, and power amplification. The heating power is changed to achieve temperature changes in the gas chamber, and the measured temperature value Z _k+1 at time k+1 is updated after temperature measurement. The extended Kalman filter and PID control are repeated until the temperature of the atomic gas chamber 1 reaches the reference temperature value and stabilizes.

Furthermore, after completing a complete temperature-controlled heating of the atomic gas chamber, a new mathematical model is obtained and the prediction model of the extended Kalman filter 8 is updated.

The working principle of the present invention is as follows: the temperature signal of the atomic gas chamber at the current moment is collected by a temperature measuring sensor, and the signal is amplified and converted into a digital signal through analog-to-digital conversion. The optimal temperature estimation value is obtained through extended Kalman filter analysis, and the difference is made with the preset target temperature value of the atomic gas chamber to be achieved, and the value is input into a PID controller. After PI control, the system control value signal is output, which is then converted into an analog signal through digital-to-analog conversion, converted into high-frequency alternating current through a high-frequency inverter circuit, and output to a heating resistor strip after power amplification to adjust the heating power and further control the gas chamber temperature.

The advantages of the present invention are: the present invention has a simple structure, is economical and practical, overcomes the problems of uneven heating of the atomic gas chamber, instability of PID temperature control, residual error and control lag in the past, and solves the measurement error caused by the nonlinear relationship between resistance and temperature of the thermocouple by combining the extended Kalman filter and the PID controller.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a schematic diagram of the structure of a heating and temperature control system for an atomic gas chamber provided by the invention;

FIG2 is a rectangular double-layer coil structure of a heating and temperature measuring circuit module 2 of the present invention;

FIG3 is a block diagram of an implementation method of the extended Kalman filter provided by the invention;

FIG4 is a flow chart of a PID control algorithm based on extended Kalman filtering provided by the invention;

Detailed ways

The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

In the description of the present invention, it should be noted that the orientations or positional relationships indicated by the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. are based on the orientations or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and "third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected" and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

Embodiment 1

Referring to FIG1 , this embodiment relates to an atomic gas room temperature control device based on Kalman filtering, wherein the hardware modules include: an atomic gas chamber 1, a heating and temperature measurement circuit module 2, a signal amplification circuit 3, an analog-to-digital/digital-to-analog conversion module 4, a power amplification circuit module 5, a high-frequency inverter circuit module 6, and a main control module 7;

The heating and temperature measuring circuit module 2 is connected to the surface of the atomic gas chamber 1. Preferably, the heating and temperature measuring circuit module 2 is fixed to the outer surface of the atomic gas chamber 1 by thermally conductive silicone;

The heating resistor strip of the heating and temperature measuring circuit module 2 is made of pure copper, the temperature measuring sensor is a T-type patch thermocouple, and the hot end of the thermocouple is fixed on the FPC soft board by epoxy resin adhesive;

Referring to FIG. 2 , the heating and temperature measurement circuit module 2 is a rectangular double-layer coil structure, and the thermal temperature measurement circuit module includes a heating resistor bar and a patch-type thermocouple; the heating resistor bar is printed on a flexible printed circuit board (FPC) flexible storage substrate by MEMS technology, and the current passing between the upper and lower layers of the double-layer coil is parallel and reverse, and the magnetic fields generated are offset from each other, further suppressing the three-dimensional magnetic field introduced by the coil heating;

The Seebeck coefficient of the output voltage and temperature of the T-type patch thermocouple in the range of 0°C to 200°C is 38.74μV/°C. In actual work, the slope of voltage/temperature needs to be measured and calibrated.

The output voltage of the T-type patch thermocouple is amplified by the signal amplification circuit 3, and the amplification factor can be adjusted by the active device;

The analog-to-digital/digital-to-analog conversion module 4 sets the sampling frequency to 1 sample/s, a sampling cycle to 1 second, stores the data in the EEPROM, and sends all the data to the main control module 7 for processing after each cycle; preferably, the AD conversion chip of the analog-to-digital/digital-to-analog conversion module 4 is AD7606;

The main control module 7 includes a main control chip and its power supply circuit, crystal oscillator circuit, reset circuit, PCIe interface circuit, and SPI serial port circuit; the main control chip is an FPGA chip XC7A100T;

The function of the high-frequency inverter circuit module 6 is to convert the low-voltage direct current output by the analog-to-digital/digital-to-analog conversion module 4 into high-frequency alternating current, thereby reducing the magnetic field brought by the heating coil and ensuring that the Larmor precession in the atomic gas chamber 1 is not disturbed by the magnetic field;

The algorithm programs of the extended Kalman filter 8 and the PID controller 9 are implemented by programming in the chip of the main control module 7;

Embodiment 2

This embodiment relates to an atomic gas room temperature control method based on Kalman filtering using the atomic gas room temperature control device based on Kalman filtering of the first embodiment. Referring to FIG. 3 , an extended Kalman filter 8 performs optimal estimation on the temperature of the atomic gas chamber 1 and an improved PID controller 9, including the following steps:

Step S1: first set the reference temperature value T _ref = 140°C, then convert the measured temperature value Z _k by measuring the voltage value data at time k; determine the nonlinear expression of the prediction model of the temperature control extended Kalman filter 8 as:

T _k =f(T _k-1 ,U _k-1 ,W _k-1 ) (1)

Z _k =h(T _k ,V _k ) (2)

Among them, formula (1) is the nonlinear expression of the temperature of atomic gas chamber 1 at time k and the input current, formula (2) is the measurement formula of the temperature of atomic gas chamber 1 at time k; T _k is the predicted true value of the temperature of atomic gas chamber 1 at time k, Z _k is the measured true value of the temperature of atomic gas chamber 1 at time k; U _k-1 is the control vector at time k-1, W _k-1 is the process noise at time k-1, and V _k is the measurement noise at time k, both of which conform to the univariate Gaussian distribution:

P(W)～N(0,Q), P(V)～N(0,R)

In actual work, the above process noise and measurement noise are expected to be 0, which conforms to the normal distribution; Q is the covariance matrix of the estimated process noise, and R is the covariance matrix of the measured Gaussian noise, and its value is specifically determined by the noise of the heating temperature measurement circuit module 2;

Step S2: Determine the a priori estimated temperature value and the measured temperature value formula of the temperature control extended Kalman filter 8 prediction model as follows:

in is the a priori estimated temperature value of the temperature of atomic gas chamber 1 at time k, is the prior measurement value at time k;

Step S3: Update the prior state error covariance matrix of the prediction model The update formula is:

Where A is the Jacobian matrix.

Step S4: Update the Kalman gain K _k of the temperature-controlled extended Kalman filter 8. The update formula is:

Where H is the transformation matrix;

Step S5: Correct the optimal a posteriori estimated temperature value by information fusion method

The posterior correction formula is:

Step S6: Update the predicted value P _k of the posterior state error covariance matrix. The update formula is:

Step S7: Referring to FIG. 4 , the optimal a posteriori estimated temperature value is set. As the external temperature of the air chamber at time k, calculate the temperature offset from the reference temperature value T _ref As the input of PID controller 9; the calculation equation of general PID control is:

Among them, _Kp is the proportional coefficient of PID controller 9, _Ti is the integral time of PID controller 9, _Td is the differential time of PID controller 9, and u(t) is the system control quantity formed by linear combination of input _Toff through PID controller 9; the above formula cannot be implemented by programming in FPGA, so it is discretized and the integral part is converted into a summation formula, which is:

The above _Ki and _Kd are integral and differential control coefficients respectively, _Uk is the discretized system control quantity, the proportional control _Kp can quickly respond to the system error and reduce the steady-state error. If _Kp is too large, the system will be unstable; the integral control _Ki is the temperature offset that is continuously superimposed to eliminate the steady-state error. If _Ki is too large, the system will overshoot and cause system oscillation; the differential control _Kd is used to reduce the system overshoot and overcome the system oscillation, but the differential term is very sensitive to high-frequency noise, so the differential control is not enabled in the present invention, _Kd = 0;

Step S8: The system control quantity U _k is digitally output from the main control module through programming, and is converted into a heating signal to the heating temperature measurement module after digital-to-analog conversion, high-frequency inversion, and power amplification, and the heating power is changed to achieve the temperature change in the gas chamber, and then the measured temperature value Z _k+ 1 at time k+1 is updated after temperature measurement, and the extended Kalman filter and PID control are repeated until the temperature of the atomic gas chamber 1 reaches the reference temperature value and stabilizes;

Furthermore, after completing a complete temperature control heating of the atomic gas chamber, a new mathematical model is obtained to update the prediction model of the extended Kalman filter 8;

The contents described in the embodiments of this specification are merely an enumeration of the implementation forms of the inventive concept. The protection scope of the present invention should not be regarded as limited to the specific forms described in the embodiments. The protection scope of the present invention also extends to equivalent technical means that can be conceived by those skilled in the art based on the inventive concept.

Claims (4)

1. An atomic gas room temperature control device based on Kalman filtering, characterized in that it comprises: an atomic gas chamber (1), a heating and temperature measurement circuit module (2), a signal amplification circuit (3), an analog-to-digital/digital-to-analog conversion module (4), a power amplification circuit module (5), a high-frequency inverter circuit module (6), and a main control module (7); The atomic gas chamber (1) is connected to the heating and temperature measuring circuit module (2), and the atomic gas chamber (1) is filled with rubidium atomic gas and nitrogen gas serving as quenching gas; When the heating resistor strip of the heating and temperature measuring circuit module (2) is energized, an electric current is generated to heat the atomic gas chamber (1), and the temperature of the atomic gas chamber (1) changes, so that a potential difference is generated between the cold end and the hot end of the patch type thermocouple of the heating and temperature measuring circuit module (2), and the temperature of the atomic gas chamber (1) is determined by measuring the potential difference; The heating temperature measurement circuit module (2) is connected to the signal amplifying circuit (3), and the potential difference generated between the cold end and the hot end of the patch type thermocouple is transmitted as an input analog voltage signal to the signal amplifying circuit (3), and the signal amplifying circuit (3) amplifies the analog voltage signal and outputs a temperature voltage; The signal amplifying circuit (3) is connected to the analog-to-digital/digital-to-analog conversion module (4), and the signal amplifying circuit (3) outputs a temperature voltage to the analog-to-digital/digital-to-analog conversion module (4), which converts the temperature voltage into a digital signal output through analog-to-digital conversion; The analog-to-digital/digital-to-analog conversion module (4) is connected to the main control module (7). The analog-to-digital/digital-to-analog conversion module (4) transmits the digital signal to the main control module (7). The main control module (7) receives the input signal, and after analysis and processing, outputs the system control quantity to the analog-to-digital/digital-to-analog conversion module (4). The analog-to-digital/digital-to-analog conversion module (4) is also connected to the high-frequency inverter circuit module (6), and the analog-to-digital/digital-to-analog conversion module (4) outputs a low-voltage direct current signal to the high-frequency inverter circuit module (6), and the high-frequency inverter circuit module (6) converts the low-voltage direct current signal into a high-frequency low-voltage alternating current signal; The high-frequency inverter circuit module (6) is connected to the power amplifier circuit module (5), and the power amplifier circuit module (5) receives the high-frequency low-voltage alternating current signal output by the high-frequency inverter circuit module (6), performs power amplification on the signal, and then outputs a heating signal; The power amplifier circuit module (5) is connected to the heating and temperature measuring circuit module (2), the heating and temperature measuring circuit module (2) receives the heating signal of the power amplifier circuit module (5), and the heating and temperature measuring circuit module (2) controls the heating coil to heat or cool. 2. The atomic gas room temperature control device based on Kalman filtering according to claim 1, characterized in that: the main control module (7) includes an extended Kalman filter (8) and a PID controller (9), and the main control module (7) performs the following information processing: Step 1: Calculate the measured temperature value Z _k according to the thermocouple electromotive force calculation formula using the input temperature and voltage values. The measured temperature value Z _k is in a floating point data format with 3 decimal places reserved. Step 2: Input the measured temperature value _Zk into the extended Kalman filter (8) to calculate the posterior estimated temperature value Calculate the a posteriori estimated temperature value The difference T _off from the reference temperature value T _ref ; Step 3: The difference T _off is used as the input of the PID controller (9), and the system control variable U _k is calculated by the PI algorithm as the output value of the PID controller (9); Step 4: The output value U _k is transmitted to the analog-to-digital/digital-to-analog conversion module (4) through the SPI serial port circuit, and the output value U _k is output as a low-voltage direct current signal through digital-to-analog conversion. 3. An atomic gas room temperature control method based on Kalman filtering using the atomic gas room temperature control device based on Kalman filtering as claimed in claim 1, comprising the following steps: Step S1: First, set the reference temperature value T _ref = 140°C, and then convert the measured temperature value Z _k by measuring the voltage value data at time k; determine the nonlinear expression of the prediction model of the temperature control extended Kalman filter (8) as: T _k =f(T _k-1 ,U _k-1 ,W _k-1 ) (1) Z _k =h(T _k ,V _k ) (2) Wherein, formula (1) is the nonlinear expression of the temperature of atomic gas chamber 1 and input current at time k, formula (2) is the measurement formula of the temperature of atomic gas chamber 1 at time k; T _k is the predicted true value of the temperature of atomic gas chamber 1 at time k, Z _k is the measured true value of the temperature of atomic gas chamber 1 at time k; U _k-1 is the control vector at time k-1, W _k-1 is the process noise at time k-1, and V _k is the measurement noise at time k, both of which conform to the univariate Gaussian distribution: P(W)～N(0,Q), P(V)～N(0,R) In actual work, the above process noise and measurement noise are expected to be 0, which conforms to the normal distribution; Q is the covariance matrix of the estimated process noise, and R is the covariance matrix of the measured Gaussian noise, and its value is specifically determined by the noise of the heating temperature measurement circuit module (2); Step S2: Determine the a priori estimated temperature value and measured temperature value formula of the temperature control extended Kalman filter (8) prediction model as follows: in is the a priori estimated temperature value of the temperature of atomic gas chamber 1 at time k, is the prior measurement value at time k; Step S3: Update the prior state error covariance matrix of the prediction model The update formula is: Where A is the Jacobian matrix. Step S4: Update the Kalman gain K _k of the temperature-controlled extended Kalman filter (8), and the update formula is: Where H is the transformation matrix; Step S5: Correct the optimal a posteriori estimated temperature value by information fusion method The posterior correction formula is: Step S6: Update the predicted value P _k of the posterior state error covariance matrix. The update formula is: Step S7: Setting the above optimal a posteriori estimated temperature value As the external temperature of the air chamber at time k, calculate the temperature offset from the reference temperature value T _ref As the input of PID controller (9); the general PID control calculation equation is: Wherein, _Kp is the proportional coefficient of the PID controller (9), _Ti is the integral time of the PID controller (9), _Td is the differential time of the PID controller (9), and u(t) is the system control quantity formed by the linear combination of the input _Toff through the PID controller (9); the above formula cannot be implemented by programming in the FPGA, so it is discretized and the integral part is converted into a summation formula, which is: The above _Ki and _Kd are integral and differential control coefficients respectively, _Uk is the discretized system control quantity, the proportional control _Kp can quickly respond to the system error and reduce the steady-state error. If _Kp is too large, the system will be unstable; the integral control _Ki is the temperature offset that is continuously superimposed to eliminate the steady-state error. If _Ki is too large, the system will overshoot and cause system oscillation; the differential control _Kd is used to reduce the system overshoot and overcome the system oscillation, but the differential term is very sensitive to high-frequency noise, so the differential control is not enabled in the present invention, _Kd = 0; Step S8: The system control quantity U _k is digitally output from the main control module through programming, and is converted into a heating signal to the heating temperature measurement module after digital-to-analog conversion, high-frequency inversion, and power amplification. The heating power is changed to achieve temperature changes in the gas chamber, and the measured temperature value Z _k+1 at time k+1 is updated after temperature measurement. The extended Kalman filter and PID control are repeated until the temperature of the atomic gas chamber 1 reaches the reference temperature value and stabilizes. 4. The method according to claim 1 is characterized in that after completing a complete temperature-controlled heating of the atomic gas chamber, a new mathematical model is obtained to update the prediction model of the extended Kalman filter (8).