A temperature adjustment method for a temperature control instrument

Abstract

The application relates to the technical field of automatic temperature regulation, and discloses a temperature regulation method for a temperature controller, which comprises the following steps: collecting discrete temperature sampling data and smoothing output of a current smoothed temperature value, calculating an algebraic difference value of the current smoothed temperature value and a historical smoothed temperature value to obtain a smoothed temperature difference component, extracting an algebraic polarity of the smoothed temperature difference component, utilizing a discrete polarity counter to time-series accumulate the number of times that the algebraic polarity continuously presents a negative value state in a steady-state temperature holding working condition, and superimposing a fixed-amplitude blocking feedforward power pulse on a power control output end of a temperature regulation loop when a continuous negative polarity count value reaches a thermal penetration constant threshold value. The application realizes early identification of load heat absorption interference through transient algebraic polarity accumulation, compensates for heat loss before the temperature substantially drops, eliminates control discontinuity, overcomes heat lag, and improves the steady-state stability of the regulation process.

Description

Technical Field

[0001] This invention belongs to the field of automatic temperature regulation technology, and in particular relates to a temperature regulation method for a temperature controller. Background Technology

[0002] Temperature controllers are used to regulate the temperature of the controlled object, and their control effect directly affects the process yield. Existing temperature controllers typically use proportional-integral-derivative (PID) control or determine the output duty cycle of the heating mechanism based on fixed feedforward parameters. These control methods generally assume that the heat capacity and heat dissipation coefficient of the controlled system remain basically constant. However, in actual heat transfer, heat needs to pass through the physical interface of the protective outer casing. The thermal resistance and heat capacity of the protective outer casing itself will cause heat transfer lag. When the heating power changes, the temperature detection result cannot immediately reflect the actual temperature change of the controlled object. Although modifying the temperature control hardware or physical form can alleviate the structural limitations at the hardware level to some extent... While existing technologies have limitations, software-level control methods also have shortcomings. For example, Chinese invention patent CN119594622A discloses an adaptive temperature control system and method for a thermoelectric cooler based on transient subcooling. Although it adaptively adjusts by monitoring the heat load, it is highly dependent on the low inertia and fast response characteristics of the thermoelectric cooling components and fails to consider the deep heat transfer lag caused by the physical protection interface. When facing temperature control scenarios with thick-walled outer tubes, existing technologies lack the cumulative calculation and polarity determination of heat penetration delay. As a result, transient adjustments are prone to control loop divergence and power oscillation due to heat transfer time differences, making it difficult to suppress thermal disturbances in a timely manner.

[0003] During the temperature transition phase, or when external disturbances such as the loading of cold materials occur, the temperature difference between the controlled object and the environment will increase, and the heat power lost to the environment will also change nonlinearly. Under such circumstances, fixed parameters are difficult to adapt to the actual heat balance changes in a timely manner. If a large heating power is maintained to shorten the heating time, the heat accumulated in the system will continue to be released after approaching the target temperature due to heat transfer hysteresis, which can easily cause temperature fluctuations or overshoot. If the control gain is reduced to suppress overshoot, the temperature response speed will be slowed down.

[0004] Therefore, existing temperature control methods still need to address the following issues: how to use discrete temperature data to promptly determine heat loss and heat transfer lag, smoothly adjust the control gain and compensate for thermal inertia during the temperature transition phase, and identify the continuous cooling trend caused by external load heat absorption as early as possible during the steady-state temperature holding phase in order to reduce compensation delay. Summary of the Invention

[0005] This invention aims to solve the problems of power surges and temperature oscillations with overshoot in the temperature transition zone, and the inability to compensate for thermal disturbances in a timely manner due to amplitude observation lag in the steady-state temperature holding zone.

[0006] In this technical solution, a temperature regulation method for a temperature controller includes the following steps: Step S1: Obtain discrete temperature sampling data collected by a single temperature detection unit in the temperature regulation loop of the temperature controller within the current sampling period, import the discrete temperature sampling data into the data input interface of the main control microprocessor, smooth the discrete temperature sampling data using a discrete filtering algorithm, and output the current smoothed temperature value through the data input interface. Step S2: Align and differentially calculate the algebraic difference between the current smoothed temperature value and the historical smoothed temperature value output in the previous sampling period within the main control microprocessor, output the smoothed temperature difference component, and separate and extract the positive and negative algebraic signs of the smoothed temperature difference component to determine the transient algebraic polarity of the controlled thermal object in the current sampling period. Step S3: When the controlled thermal object is in a steady-state temperature holding condition, the discrete polarity counter is used to sequentially accumulate the number of discrete cycles in which the transient algebraic polarity continuously presents a negative value, and outputs a continuous negative polarity count value; when the continuous negative polarity count value reaches the heat penetration constant threshold, a fixed amplitude blocking feedforward power pulse is directly superimposed on the power control output terminal of the temperature regulation loop; wherein, the heat penetration constant threshold is determined by the characteristic heat diffusion delay constant determined by the combination of the characteristic geometric thickness of the protective outer jacket of the temperature detection unit and the physical thermal diffusion coefficient of the outer jacket material.

[0007] Preferably, step S1 includes the following sub-steps: Step S11, using the discrete first-order low-pass filter algorithm as the discrete filter algorithm, extracting the discrete temperature sampling data in the current sampling period and the historical smoothed temperature value output in the previous sampling period during smoothing; Step S12, using the discrete first-order low-pass filter algorithm to calculate the weighted algebraic sum of the discrete temperature sampling data and the historical smoothed temperature value, and directly outputting the current smoothed temperature value from the weighted algebraic sum.

[0008] Preferably, the discrete first-order low-pass filtering algorithm satisfies the following formula: ,in, This is the current smoothed temperature value. For discrete temperature sampling data, This is a historical smoothed temperature value. The set first-order low-pass smoothing filter coefficient has a value range of 0.1 to 0.3.

[0009] Preferably, before the steady-state temperature holding condition, a temperature rise control stage is included, which includes the following steps: Step S401, obtaining the current smoothed temperature value and calculating the first-order and second-order time difference components of the current smoothed temperature value; Step S402, calculating the weighted sum of the first-order and second-order time difference components and outputting the phase lead compensation amount; Step S403, superimposing the phase lead compensation amount into the proportional-integral-derivative control, outputting the thermal inertia lead compensation control duty cycle, and using the thermal inertia lead compensation control duty cycle to control the output duty cycle of the temperature regulation loop.

[0010] Preferably, step S3, which involves directly superimposing a fixed amplitude blocking feedforward power pulse at the power control output of the temperature regulation loop, includes the following sub-steps: Step S31, when the continuous negative polarity count value reaches the thermal penetration constant threshold, a fixed blocking feedforward power pulse is superimposed on the proportional-integral-derivative control output, and the output adjusts the control duty cycle; Step S32, in each subsequent sampling period, the blocking feedforward power pulse is attenuated proportionally according to the set attenuation coefficient, and the attenuated feedforward power pulse is output; Step S33, the attenuated feedforward power pulse is added to the proportional-integral-derivative control output, and when the blocking feedforward power pulse attenuates to zero and the algebraic polarity of the smooth temperature difference component is still negative, the proportional attenuation is maintained, and when the algebraic polarity of the smooth temperature difference component changes to a positive value, the blocking feedforward power pulse is forcibly returned to zero and injection is stopped.

[0011] Preferably, after step S3, the following adaptive pulse width modulation steps are further included: Step S601, obtaining the final control duty cycle after superimposed blocking feedforward power pulses, and obtaining the set pulse width modulation period; Step S602, calculating the product of the final control duty cycle and the pulse width modulation period, outputting the high-level conduction time, and calculating the algebraic difference between the pulse width modulation period and the high-level conduction time, and outputting the low-level cutoff time; Step S603, within the pulse width modulation period, outputting a high-level drive signal to the heating drive circuit of the temperature controller according to the high-level conduction time, and outputting a low-level cutoff signal to the heating drive circuit according to the low-level cutoff time.

[0012] Preferably, the method further includes a heat loss power adjustment step, which includes: step S701, when the temperature controller is in a steady-state temperature holding condition, obtaining the reference control duty cycle required to maintain the temperature stability of the controlled thermal object, as a basic heat loss maintenance term; step S702, calculating the convective heat loss correction amount based on the temperature difference between the current smooth temperature value and the set environmental reference temperature value; step S703, adding the basic heat loss maintenance term and the convective heat loss correction amount to calculate the real-time heat loss maintenance duty cycle, and inputting the real-time heat loss maintenance duty cycle as a feedforward quantity into the temperature regulation loop.

[0013] Preferably, the method for determining the heat penetration constant threshold includes: obtaining the wall thickness of the protective outer tube of the temperature detection unit and the thermal diffusivity of the protective outer tube material; calculating the ratio of the square of the wall thickness to the thermal diffusivity to determine the characteristic thermal diffusion delay time of the protective outer tube; obtaining the sampling period of the temperature controller; calculating the integer value of the ratio of the characteristic thermal diffusion delay time to the sampling period; and determining the integer value as the heat penetration constant threshold.

[0014] Preferably, the method for adaptively adjusting the first-order low-pass smoothing filter coefficient includes: obtaining the smoothed temperature difference component and calculating the absolute value of the smoothed temperature difference component; increasing the first-order low-pass smoothing filter coefficient when the absolute value changes with time in an increasing trend; and decreasing the first-order low-pass smoothing filter coefficient when the absolute value changes with time in a stable trend.

[0015] Preferably, when the current smoothed temperature value deviates from the target control temperature value, the steady-state temperature control deviation is adjusted through two-stage feedback control. The two-stage feedback control includes: calculating the algebraic deviation between the current smoothed temperature value and the target control temperature value; calculating the first-stage feedback adjustment gain using a proportional-integral algorithm, and calculating the second-stage feedback adjustment gain using the smooth accumulation of the integral loop; adding the first-stage feedback adjustment gain and the second-stage feedback adjustment gain to obtain the final feedback adjustment amount, and adjusting the output of the temperature control loop based on the final feedback adjustment amount.

[0016] Compared with existing technologies, the temperature regulation method for a temperature controller of the present invention has the following advantages: 1. In the temperature regulation of the temperature controller, by smoothing the discrete temperature data and gradually adjusting the heating power in combination with the rate and trend of temperature change, the temperature rise control can smoothly transition to steady-state temperature holding control, avoiding power abrupt changes and heating interruptions caused by direct switching between different control stages. At the same time, when approaching the target temperature, it can reduce excessive heating in time, reduce temperature fluctuations and overshoot caused by residual thermal inertia, and make the temperature change process more stable.

[0017] 2. By compensating in advance for the heat transfer lag caused by the protective outer casing and other structures based on the change between the actual output power of the temperature controller and the power required to maintain the current temperature, the insufficient heat or heat accumulation caused by the asynchronous response between the heating command and the actual temperature can be reduced. This can maintain a relatively continuous heat supply during the control phase transition and temperature change process, and improve the stability of temperature regulation.

[0018] 3. During steady-state temperature holding, by continuously judging the direction of temperature change, heat absorption disturbances such as cold material loading can be identified before the temperature drops significantly, and heating power can be replenished in time. Compared with the method of waiting for the temperature deviation to reach the set amplitude before compensation, this method can shorten the delay of disturbance identification and compensation, reduce the temperature drop caused by load changes, reduce repeated fluctuations during the temperature recovery process, and improve temperature holding stability. Attached Figure Description

[0019] Figure 1 This is a flowchart of the blocking feedforward temperature regulation method of the present invention; Figure 2 This is a diagram of the control node and signal transmission structure of the temperature controller of the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0021] A temperature regulation method for a temperature controller includes the following steps: Step S1: Obtain discrete temperature sampling data collected by a single temperature detection unit in the temperature regulation loop of the temperature controller within the current sampling period, import the discrete temperature sampling data into the data input interface of the main control microprocessor, smooth the discrete temperature sampling data using a discrete filtering algorithm, and output the current smoothed temperature value through the data input interface. Step S2: Align and differentially calculate the algebraic difference between the current smoothed temperature value and the historical smoothed temperature value output in the previous sampling period within the main control microprocessor, output the smoothed temperature difference component, and separate and extract the positive and negative algebraic signs of the smoothed temperature difference component to determine the transient algebraic polarity of the controlled thermal object in the current sampling period. Step S3: When the controlled thermal object is in a steady-state temperature holding condition, the discrete polarity counter is used to sequentially accumulate the number of discrete cycles in which the transient algebraic polarity continuously presents a negative value, and outputs a continuous negative polarity count value; when the continuous negative polarity count value reaches the heat penetration constant threshold, a fixed amplitude blocking feedforward power pulse is directly superimposed on the power control output terminal of the temperature regulation loop; wherein, the heat penetration constant threshold is determined by the characteristic heat diffusion delay constant determined by the combination of the characteristic geometric thickness of the protective outer jacket of the temperature detection unit and the physical thermal diffusion coefficient of the outer jacket material.

[0022] Preferably, step S1 includes the following sub-steps: Step S11, using the discrete first-order low-pass filter algorithm as the discrete filter algorithm, extracting the discrete temperature sampling data in the current sampling period and the historical smoothed temperature value output in the previous sampling period during smoothing; Step S12, using the discrete first-order low-pass filter algorithm to calculate the weighted algebraic sum of the discrete temperature sampling data and the historical smoothed temperature value, and directly outputting the current smoothed temperature value from the weighted algebraic sum.

[0023] Preferably, the discrete first-order low-pass filtering algorithm satisfies the following formula: ,in, This is the current smoothed temperature value. For discrete temperature sampling data, This is a historical smoothed temperature value. The set first-order low-pass smoothing filter coefficient has a value range of 0.1 to 0.3.

[0024] Preferably, before the steady-state temperature holding condition, a temperature rise control stage is included, which includes the following steps: Step S401, obtaining the current smoothed temperature value and calculating the first-order and second-order time difference components of the current smoothed temperature value; Step S402, calculating the weighted sum of the first-order and second-order time difference components and outputting the phase lead compensation amount; Step S403, superimposing the phase lead compensation amount into the proportional-integral-derivative control, outputting the thermal inertia lead compensation control duty cycle, and using the thermal inertia lead compensation control duty cycle to control the output duty cycle of the temperature regulation loop.

[0025] Preferably, step S3, which involves directly superimposing a fixed amplitude blocking feedforward power pulse at the power control output of the temperature regulation loop, includes the following sub-steps: Step S31, when the continuous negative polarity count value reaches the thermal penetration constant threshold, a fixed blocking feedforward power pulse is superimposed on the proportional-integral-derivative control output, and the output adjusts the control duty cycle; Step S32, in each subsequent sampling period, the blocking feedforward power pulse is attenuated proportionally according to the set attenuation coefficient, and the attenuated feedforward power pulse is output; Step S33, the attenuated feedforward power pulse is added to the proportional-integral-derivative control output, and when the blocking feedforward power pulse attenuates to zero and the algebraic polarity of the smooth temperature difference component is still negative, the proportional attenuation is maintained, and when the algebraic polarity of the smooth temperature difference component changes to a positive value, the blocking feedforward power pulse is forcibly returned to zero and injection is stopped.

[0026] Preferably, after step S3, the following adaptive pulse width modulation steps are further included: Step S601, obtaining the final control duty cycle after superimposed blocking feedforward power pulses, and obtaining the set pulse width modulation period; Step S602, calculating the product of the final control duty cycle and the pulse width modulation period, outputting the high-level conduction time, and calculating the algebraic difference between the pulse width modulation period and the high-level conduction time, and outputting the low-level cutoff time; Step S603, within the pulse width modulation period, outputting a high-level drive signal to the heating drive circuit of the temperature controller according to the high-level conduction time, and outputting a low-level cutoff signal to the heating drive circuit according to the low-level cutoff time.

[0027] Preferably, the method further includes a heat loss power adjustment step, which includes: step S701, when the temperature controller is in a steady-state temperature holding condition, obtaining the reference control duty cycle required to maintain the temperature stability of the controlled thermal object, as a basic heat loss maintenance term; step S702, calculating the convective heat loss correction amount based on the temperature difference between the current smooth temperature value and the set environmental reference temperature value; step S703, adding the basic heat loss maintenance term and the convective heat loss correction amount to calculate the real-time heat loss maintenance duty cycle, and inputting the real-time heat loss maintenance duty cycle as a feedforward quantity into the temperature regulation loop.

[0028] Preferably, the method for determining the heat penetration constant threshold includes: obtaining the wall thickness of the protective outer tube of the temperature detection unit and the thermal diffusivity of the protective outer tube material; calculating the ratio of the square of the wall thickness to the thermal diffusivity to determine the characteristic thermal diffusion delay time of the protective outer tube; obtaining the sampling period of the temperature controller; calculating the integer value of the ratio of the characteristic thermal diffusion delay time to the sampling period; and determining the integer value as the heat penetration constant threshold.

[0029] Preferably, the method for adaptively adjusting the first-order low-pass smoothing filter coefficient includes: obtaining the smoothed temperature difference component and calculating the absolute value of the smoothed temperature difference component; increasing the first-order low-pass smoothing filter coefficient when the absolute value changes with time in an increasing trend; and decreasing the first-order low-pass smoothing filter coefficient when the absolute value changes with time in a stable trend.

[0030] Preferably, when the current smoothed temperature value deviates from the target control temperature value, the steady-state temperature control deviation is adjusted through two-stage feedback control. The two-stage feedback control includes: calculating the algebraic deviation between the current smoothed temperature value and the target control temperature value; calculating the first-stage feedback adjustment gain using a proportional-integral algorithm, and calculating the second-stage feedback adjustment gain using the smooth accumulation of the integral loop; adding the first-stage feedback adjustment gain and the second-stage feedback adjustment gain to obtain the final feedback adjustment amount, and adjusting the output of the temperature control loop based on the final feedback adjustment amount.

[0031] Example 1: This example applies a temperature regulation method for a temperature controller to a semiconductor wafer rapid thermal annealing furnace. The furnace's sealed cavity contains a single temperature detection unit. A quartz protective sleeve is fitted around the sensing end to isolate it from the measured medium. The temperature regulation loop adjusts the output duty cycle of the heating mechanism based on the temperature data output by the temperature detection unit. When the wafer temperature rises from room temperature to the target control temperature of 600°C, the thermal resistance and capacity of the quartz protective sleeve cause a phase lag in heat transfer. The radiative heat loss from the controlled thermal object to the outside of the furnace increases with the temperature difference. When the measured temperature approaches the control switching boundary, segmented control is prone to power jumps and energy breaks, resulting in discontinuous energy supply. This leads to oscillations in the temperature regulation loop, causing localized temperature fluctuations and temperature overshoot on the wafer surface.

[0032] The processing unit uses a fixed sampling period of 10ms. During operation, discrete temperature sampling data collected by a single temperature detection unit is imported through the data input interface of the main control microprocessor. The current smoothed temperature value is calculated using a discrete first-order low-pass filter algorithm. The discrete first-order low-pass filter algorithm satisfies: ,in, This is the historical smoothed temperature value output from the previous sampling period. The values ​​of the first-order low-pass smoothing filter coefficients are maintained within the range of 0.1 to 0.3. The processing unit calculates the candidate values ​​of the first-order low-pass smoothing filter coefficients using the following formula during initial assignment: ,in, The noise threshold parameter is 0.05℃. Before the candidate value is written into the filter, it is limited to the range of 0.1 to 0.3. During operation, the processing unit calculates the absolute value of the smoothed temperature difference component. When the absolute value shows an increasing trend for 5 consecutive sampling periods, the current first-order low-pass smoothing filter coefficient is increased by 0.02 until it reaches 0.3. When the absolute value remains within the noise threshold for 10 consecutive sampling periods and shows a stable trend, the current first-order low-pass smoothing filter coefficient is decreased by 0.01 until it drops to 0.1.

[0033] The processing unit will display the current smoothing temperature value. With the target control temperature value The difference is calculated to obtain the current deviation. When the current deviation decreases to the critical threshold of 2°C... When the threshold is reached, the critical deceleration control mode is activated, and the variable gain attenuation factor is calculated. : The processing unit calculates the conventional feedback adjustment term based on the current deviation and the slope of temperature change. : ,in, This is the proportional gain coefficient. The differential damping coefficient is used during the temperature control phase. The processing unit calculates the first-order and second-order time difference components sequentially based on the current sampling period and the current smoothed temperature values ​​from the previous two sampling periods. The first-order time difference component is multiplied by 0.6, and the second-order time difference component is multiplied by 0.4. The two products are then added together and the result is calculated using the heat capacity compensation coefficient. The duty cycle scale is converted to obtain the phase lead compensation amount. The phase advance compensation is superimposed on the proportional-integral-derivative control output to form the thermal inertia advance compensation control duty cycle.

[0034] During the steady-state temperature holding calibration of the temperature controller, the processing unit records the reference control duty cycle required to maintain the temperature stability of the controlled thermal object. This is used as the basic heat dissipation maintenance term. During the current sampling period, the processing unit adjusts the current smoothed temperature value accordingly. Compared with ambient reference temperature value The temperature difference between them, multiplied by the thermal resistance offset coefficient The convective heat loss correction is obtained; then, the basic heat loss maintenance term is added to the convective heat loss correction to obtain the real-time heat loss maintenance duty cycle. : Real-time heat dissipation maintains duty cycle As a feedforward input to the temperature control loop, the processing unit also calculates the final control duty cycle of the previous sampling period. Maintaining duty cycle with current real-time heat dissipation The algebraic difference between them is used to obtain the time delay verification quantity according to the following formula. : ,in, As a dimensional balancing factor, For a single-layer delay constant calibrated to 8.5s, the delay check quantity is... It is used to characterize the degree of deviation between the feedback regulation output and the heat dissipation feedforward quantity, and is not used as the basis for generating the phase lead compensation quantity.

[0035] The processing unit utilizes a variable gain attenuation factor The final control duty cycle for the current sampling period is obtained by continuously weighting the thermal inertia lead compensation control duty cycle and the real-time heat dissipation maintenance duty cycle. : During the temperature rise control phase, the blocking feedforward power pulse is in an untriggered state; therefore, the above-mentioned synthesized value is the final control duty cycle of the current sampling period, which increases with the current smoothed temperature value. Approaching the target control temperature value Variable gain attenuation factor The temperature regulation loop gradually decreases, switching from deviation feedback regulation to heat dissipation feedforward regulation. When the processing unit determines the current smoothed temperature value... The critical temperature point for retracting the advanced control is reached, and the critical temperature point for force withdrawal satisfies the following: At this time, the variable gain attenuation factor The value has been reduced to below 0.01, and the control unit sets it to zero, thus reducing the final control duty cycle. Converging to real-time heat dissipation to maintain duty cycle The operation only eliminates the residual quantization value of the variable gain attenuation factor and does not create a significant power step in the heating drive channel. The control unit will ultimately control the duty cycle. Multiplying the pulse width modulation period yields the high-level on-time, and subtracting the high-level on-time from the pulse width modulation period yields the low-level off-time. The heating actuator outputs power according to these two time parameters.

[0036] Under the aforementioned control commands, the temperature of the rapid thermal annealing furnace rises to the target temperature range of 600℃. The measured temperature curve on the wafer surface gradually converges as it approaches the target control temperature value. The local control temperature difference is maintained within ±0.05℃. The temperature regulation loop maintains continuous power output between the heating control stage and the steady-state holding temperature condition. The wafer completes the lattice rearrangement annealing process in a stable thermal field.

[0037] Example 2: In the current multi-zone thermal response test bench, a cross-linked heating wire is arranged in the closed reaction chamber. A quartz protective outer tube is set on the outside of the temperature sensing end of the single temperature detection unit, so that the heat between the heating source and the measured medium is transferred through the protective outer tube. The test bench records the temperature regulation process under rapid heating and steady-state cold material loading disturbance conditions. The core computing chip of the temperature regulation loop adopts a 32-bit single-core microprocessor with a hardware floating-point arithmetic unit, and is configured with a 12-bit successive approximation analog-to-digital converter interface. The minimum voltage recognition resolution of the interface is 0.02℃ when converted to temperature change. The discrete temperature sampling data output by the interface is used to calculate the first-order and second-order time difference components. The processing unit multiplies the first-order time difference component by 0.6 and the second-order time difference component in the same sampling period by 0.4, and then adds the two products to obtain the phase lead compensation amount. The phase lead compensation is superimposed on the proportional-integral-derivative control output to form the thermal inertia lead compensation control duty cycle, and participates in the final control duty cycle. The generation of .

[0038] Sampling period Based on the highest thermal response cutoff frequency of the controlled thermal object during the heating phase It is confirmed that the relationship is as follows: ,in, The unit is s. The unit is Hz, measured during the pre-run of the test bench. The frequency is 20Hz. After substituting, we get... The sampling period is 10ms, and the sampling period is fixed by the microprocessor's timer interrupt.

[0039] The interference injection module superimposes Gaussian white noise with a signal-to-noise ratio of 20dB onto the discrete temperature sampling data output from the analog-to-digital conversion interface, and adds power frequency harmonic interference at a frequency of 50Hz. The temperature rise test setup includes the sample group of this invention, a partially missing control group, and an out-of-range control group. The sample group of this invention employs the aforementioned phase lead compensation method and uses a variable gain attenuation factor. With the current smoothing temperature value The temperature changes close to the target control temperature; for the partially missing control group, other parameters remain unchanged, and the phase lead compensation is adjusted. Set to zero; for out-of-range control groups, retain the compensation output interface, replace its generation method with an alternative compensation channel involving a single-layer delay constant, and set the single-layer delay constant... The time was set to 45.0 s, deviating from the actual thermal diffusion time determined by the wall thickness and thermal diffusivity of the protective outer casing material. All sample groups were started from the ambient reference temperature. Heating begins at 25.4℃. As the current smoothing temperature approaches the target control temperature of 600.0℃, the sampling point numbers are recorded sequentially. Current smoothing temperature value Variable gain attenuation factor Compensation amount and final control duty cycle .

[0040] The sample group of this invention runs to the sampling point sequence number When the value is 5800, the current smoothing temperature is 598.12℃. The processing unit obtains a variable gain attenuation factor of 0.9400 from the temperature state, and obtains a phase lead compensation of 0.0415 based on the weighted sum of the first-order and second-order timing difference components. The synthesized final control duty cycle is 0.3412, and the sampling point number is... When the speed reaches 5900, the current smoothing temperature rises to 599.05℃, the variable gain attenuation factor decreases to 0.4750, the phase lead compensation decreases to 0.0218, and the final control duty cycle drops to 0.1824. (Sampling point sequence number...) When the value is 6000, the current smoothed temperature value reaches 599.96℃, the variable gain attenuation factor further decreases to 0.0200, the phase lead compensation is 0.0052, and the final control duty cycle is 0.0815. As the current smoothed temperature value approaches the target control temperature value, the variable gain attenuation factor, the phase lead compensation, and the final control duty cycle all gradually decrease.

[0041] The partial missing control group was sampled at the following serial number: When the value is 5800, the current smoothing temperature is 597.23℃, the variable gain attenuation factor is 1.0000, the phase lead compensation remains at 0.0000, and the final control duty cycle is 0.4852; sampling point sequence number. When the value is 5900, the current smoothing temperature is 598.92℃, the variable gain attenuation factor is 0.5400, the phase lead compensation is still 0.0000, and the final control duty cycle is 0.3125; sampling point sequence number. When the value is 6000, the current smoothed temperature rises to 600.84℃. The variable gain attenuation factor and phase lead compensation are both 0.0000, and the final control duty cycle is 0.0214. Compared to the target control temperature of 600.0℃, the partial missing control group at sampling point number... The temperature overshoot at 6000 is 0.84℃.

[0042] The out-of-range control group was located at the sampling point number. When the value is 5800, the current smoothed temperature is 596.45℃, the variable gain attenuation factor is 1.0000, the compensation amount of the replacement compensation channel output is 0.1245, and the final control duty cycle is 0.6128; sampling point sequence number. When the value is 5900, the current smoothing temperature is 598.11℃, the variable gain attenuation factor is 0.9450, the compensation is 0.1012, and the final control duty cycle is 0.5214; sampling point sequence number. When the value is 6000, the current smoothed temperature value rises to 601.56℃, the variable gain attenuation factor is 0.0000, the compensation amount is 0.0624, and the final control duty cycle is 0.1245. The compensation amount of the out-of-range control group continues to be input into the temperature regulation loop at the end of the heating process, and the current smoothed temperature value exceeds the target control temperature value.

[0043] After the temperature rise test was completed, 10g, 20g, and 30g of cold simulated material were sequentially added to the test bench under a steady-state temperature of 600.00℃. When Gaussian white noise and power frequency harmonic interference were present, the cold simulated material absorbed heat, causing the smooth temperature difference component at the single temperature detection unit to continuously show a negative value. The discrete polarity counter accumulated the number of consecutive negative transient algebraic polarities cycle by cycle and output a continuous negative polarity count value. In order to conduct rapid anti-interference verification, the test bench replaced the protective outer tube test section of the single temperature detection unit with an ultra-thin high thermal conductivity protective outer tube test section with a wall thickness of 0.05mm. The processing unit calculated the ratio of the square of the wall thickness value to the thermal diffusivity coefficient of the protective outer tube material to obtain the characteristic thermal diffusivity delay time. Then, the time was divided by the 10ms sampling period and truncated with integer data. The resulting thermal penetration constant threshold was 5 cycles. When the continuous negative polarity count value reached the threshold, the processing unit directly superimposed a fixed amplitude blocking feedforward power pulse at the power control output of the temperature regulation loop.

[0044] Under the aforementioned material disturbance conditions, the maximum temperature drop values ​​measured by the sample group of this invention were 0.04℃, 0.07℃, and 0.11℃, respectively. The control curves all recovered to a steady state of 600.00℃ within 15.0s. The maximum temperature drop values ​​of the conventional control group under the same material disturbance were 0.45℃, 1.12℃, and 2.58℃, respectively. The adjustment swing time exceeded 120.0s. The discrete polarity counter triggered a blocking feedforward power pulse before the temperature amplitude dropped. The steady-state temperature fluctuation within the target temperature range of the test bench remained within ±0.05℃. The temperature regulation loop adjusted the driving pulse width of the cross-linked heating wire according to the final control duty cycle.

[0045] Example 3: This example deploys a temperature regulation method for a temperature controller in a multi-zone thermal response test bench with a high-temperature gas scouring environment. A cross-linked heating wire is installed within the closed reaction chamber of the test bench. A quartz protective outer tube is fitted around the sensing end of a single temperature detection unit to isolate the sensing end from the test medium. The protective outer tube has a characteristic geometric thickness. The thermal diffusivity of the outer tube material is 0.003m. for In the convective heat transfer process, the thermal resistance and thermal capacity of the protective outer jacket cause a thermal diffusion delay when the electrothermal energy is transferred from the external heat source to the interior of the reaction chamber. Consequently, the discrete temperature sampling data collected by the temperature detection unit is phase-lagging. As the temperature difference increases, the radiative heat power lost by the reaction chamber to the outside is prone to cause output duty cycle jumps and heating circuit power oscillations when the temperature approaches the control switching boundary.

[0046] The main control microprocessor's data input interface reads the static parameters of the protective outer sleeve and determines the characteristic geometric thickness. Set to 0.003m, and set the physical thermal diffusivity of the outer tube material to 0.003m. Set as The characteristic thermal diffusion delay time is obtained by calculating the ratio of the square of the characteristic geometric thickness to the physical thermal diffusivity of the outer casing material, using m² / s. : ,in, The characteristic thermal diffusion delay time, For characteristic geometric thickness, Given the physical thermal diffusivity of the outer tube material, the characteristic thermal diffusion delay time calculated based on the above parameters is 3s. The main control microprocessor then modifies this time with a 10ms sampling period. The ratio was calculated, and the integer value of 300 was obtained by truncating the integer data. This value was then determined as the threshold for the heat penetration constant. .

[0047] After the multi-zone thermal response test bench enters the steady-state temperature holding condition, the main control microprocessor continuously calculates the current smoothed temperature value of adjacent sampling periods. Compared with historical smooth temperature values The algebraic difference between the components is used to output a smooth temperature difference component, and the transient algebraic polarity is determined based on its positive or negative algebraic sign. When the external flowing medium absorbs heat and causes the local temperature to continuously decrease in one direction, the smooth temperature difference component continuously exhibits a negative value. The discrete polarity counter accumulates the number of consecutive occurrences of the negative value state cycle by cycle and outputs a continuous negative polarity count value.

[0048] When the continuous negative polarity count reaches the thermal penetration constant threshold of 300, the power determination state machine superimposes a fixed-amplitude blocking feedforward power pulse into the proportional-integral-derivative control output. The initial amplitude of the pulse corresponds to a duty cycle increment of 0.15. After superposition, the adjusted control duty cycle is obtained. In each subsequent sampling period, the attenuation control module adjusts the attenuation according to a geometrical attenuation coefficient of 0.95. The blocking feedforward power pulse is decayed cycle by cycle to obtain a decayed feedforward power pulse, which is then added to the proportional-integral-derivative control output. When the blocking feedforward power pulse decays to zero and the transient algebraic polarity is still negative, the decay control module continues to update according to the proportional decay coefficient of 0.95, and the pulse value remains zero. When the transient algebraic polarity changes to a positive value, the interrupt judgment logic forces the blocking feedforward power pulse to zero and stops injection.

[0049] During the continuous accumulation of the discrete polarity counter for 300 sampling cycles, the proportional-integral-derivative feedback control loop operates independently to regulate random temperature fluctuations within the reaction chamber. At this time, the discrete polarity counter only records the number of times the transient algebraic polarity is continuously negative, without rewriting the proportional-integral-derivative control output. Only when the continuous negative polarity count value reaches 300 without interruption will the power determination state machine superimpose the blocking feedforward power pulse onto the power control output.

[0050] The power determination state machine generates the final control duty cycle based on the proportional-integral-derivative control output and the attenuated feedforward power pulse. Based on this, the output power of the cross-linked heating wire is adjusted, and the duty cycle is continuously changed to reduce the temperature fluctuation caused by the heat conduction delay at the interface of the protective outer tube. Under the condition of material heat absorption, the temperature field data of the multi-zone thermal response test bench is maintained within ±0.05℃ of the target control temperature value of 600.0℃. No temperature overshoot exceeding the process index occurs, and the output power of the heating actuator does not oscillate due to sudden changes in control commands.

[0051] Example 4: This example combines Figures 1 to 2 A description of a temperature regulation method for a temperature controller, such as... Figure 1As shown, in step S1, discrete data smoothing is performed by collecting discrete temperature sampling data from a single temperature detection unit and smoothing it using a discrete filtering algorithm to output the current smoothed temperature value. Next, in step S2, transient algebraic polarity is determined by calculating the algebraic difference between the current smoothed temperature value and the historical smoothed temperature value to obtain the smoothed temperature difference component, and separating and extracting the positive and negative algebraic signs to determine the transient algebraic polarity. Then, in step S3, steady-state polarity accumulation is performed. When the controlled thermal object is in a steady-state holding temperature condition, the discrete polarity counter is used to sequentially accumulate the number of discrete cycles in which the transient algebraic polarity continuously exhibits a negative value. Based on this, it is determined whether the continuous negative polarity count value reaches the heat penetration constant threshold. If it does not reach the heat penetration constant threshold, the process returns to step S3 to continue using the discrete polarity counter to sequentially accumulate the number of discrete cycles in which the transient algebraic polarity continuously exhibits a negative value. If it reaches the heat penetration constant threshold, a superimposed blocking feedforward power pulse is executed, superimposing a fixed amplitude blocking feedforward power pulse at the power control output of the temperature regulation loop.

[0052] like Figure 2 As shown, the main control microcomputer node of the temperature controller receives discrete temperature sampling data transmitted by the single temperature detection unit in the physical acquisition and sensing node through the discrete temperature sampling data transmission channel via the data input interface. The physical acquisition and sensing node is equipped with a quartz protective outer tube covering the single temperature detection unit. The data input interface transmits the received discrete temperature sampling data to the 32-bit single-core microprocessor. The 32-bit single-core microprocessor reads the solidified configuration in its connected non-volatile memory to perform filtering and control quantity calculation. The final control duty cycle obtained by the calculation is output to the heating actuator in the thermal execution and cavity node via the power control output terminal and along the final control duty cycle signal path. The heating actuator performs temperature control heating on the controlled thermal object, and convective heat transfer and physical heat conduction occur between the controlled thermal object and the quartz protective outer tube in the physical acquisition and sensing node.

[0053] Example 5: The control chip acquires the initial temperature value of the heater when it is in a zero-power output state during the pre-calibration stage of the multi-zone thermal response test bench. Subsequently, the control energy field generator continuously inputs 50W of thermal power into the reaction chamber, and the data buffer records 10 consecutive sampling cycles. The temperature change rate sequence within the unit is processed by the processing unit to calculate the arithmetic mean of the sequence. And based on the initial temperature value with arithmetic mean Calculate the thermal resistance offset coefficient : ,in, This is the thermal resistance offset coefficient. This is the arithmetic mean of the temperature change rate sequence. The initial temperature value is the thermal resistance offset coefficient calculated by the processing unit. The value is 0.12. The control chip writes the value as an update parameter into the fixed configuration sector of the non-volatile memory. When calculating the real-time heat dissipation maintenance duty cycle in the closed loop, the processing unit multiplies the thermal resistance offset coefficient with the full-scale power calibration coefficient pre-stored in the chip, so that the ratio of the temperature change rate to the initial temperature value is mapped to a scalar contribution term in the range of 0 to 1, and it is used to calculate the output duty cycle.

[0054] When the multi-zone thermal response test bench experiences time-varying drift in heat transfer impedance due to scaling interference, the control unit initiates correction logic within the self-test cycle, enabling the test bench to continuously run for 3000 sampling cycles under steady-state temperature conditions. In variable gain attenuation factor After converging to zero, the processing unit calculates the final control duty cycle cycle by cycle. Maintaining duty cycle with real-time heat dissipation The transient difference between the two samples; the difference over 50 consecutive sampling periods. When the value is greater than the threshold of 0.02, an incremental correction with a step of 0.001 is added to the initial constant in the solidification configuration sector, and the reference control parameters in the current control register are updated accordingly. The heating drive channel adjusts the output duty cycle according to the updated reference control parameters, so that the control temperature difference in the reaction chamber is kept within the preset tolerance range of ±0.05℃ of 600.0℃.

[0055] Example 6: When the multi-zone thermal response test bench operates under a steady-state temperature of 600.0℃, the thermal conduction coupling between adjacent temperature control channels causes continuous small fluctuations in the local temperature of the controlled thermal object. The temperature regulation loop uses two-stage feedback control to adjust the steady-state temperature control deviation. During the data calibration stage, the main control microprocessor calculates the current smoothed temperature value. With the target control temperature value Algebraic deviation between contemporary number bias The absolute value of 100 consecutive sampling periods When all values ​​are less than the zero-point parameter by 0.5℃, the chip calibration unit uses a proportional-integral algorithm to calculate the first-stage feedback adjustment gain. The calculation uses a scaling factor. and integration time constant Among them, the scaling factor The integral time constant is 1.2. It lasts for 40 seconds.

[0056] Meanwhile, the smoothing accumulator register inside the main control microprocessor uses an integration circuit to adjust for algebraic deviations. Smooth accumulation is performed to obtain the second-stage feedback adjustment gain. Within the current sampling period, the second-stage feedback adjustment gain satisfies: ,in, The gain is adjusted by the second-stage feedback. This is the historical value of the second-stage feedback adjustment gain from the previous sampling period. This is the integral gain coefficient, with a value of 0.05. For algebraic deviation, The sampling period is 10ms, and the main microprocessor will adjust the gain using the first-stage feedback. With second-stage feedback adjustment gain The values ​​are added together to obtain the final feedback adjustment amount, which is then used as a feedback correction term input to the power modulation module to update the heating duty cycle.

[0057] First-stage feedback adjustment gain With second-stage feedback adjustment gain After jointly adjusting the heating duty cycle, the test bench compensates for the thermal conduction coupling between adjacent temperature control channels at the heating source end. The temperature data fluctuation amplitude caused by cross heat transfer in the reaction chamber is reduced from 0.42℃ and stabilized within the range of ±0.05℃. The residual steady-state error of the controlled thermal object converges to zero under the effect of continuous smoothing accumulation of the integral loop. Under the condition that the temperature field of adjacent loops continues to change, the temperature regulation loop maintains a convergent state during the steady-state long-cycle operation, so that the temperature in the multi-loop coupling environment remains constant.

[0058] The embodiments of this application have been described above with reference to the accompanying drawings. Unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other. This application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit of this application and the scope of protection of this invention, and all of these forms are within the protection scope of this application.

Claims

1. A temperature regulation method for a temperature controller, characterized in that, Includes the following steps: Step S1: Obtain discrete temperature sampling data collected by a single temperature detection unit in the temperature regulation loop of the temperature controller within the current sampling period, import the discrete temperature sampling data into the data input interface of the main control microprocessor, smooth the discrete temperature sampling data using a discrete filtering algorithm, and output the current smoothed temperature value through the data input interface. Step S2: Align and differentially calculate the algebraic difference between the current smoothed temperature value and the historical smoothed temperature value output in the previous sampling period within the main control microprocessor, output the smoothed temperature difference component, and separate and extract the positive and negative algebraic signs of the smoothed temperature difference component to determine the transient algebraic polarity of the controlled thermal object in the current sampling period. Step S3: When the controlled thermal object is in a steady-state holding temperature condition, the discrete polarity counter is used to accumulate the number of discrete cycles in which the transient algebraic polarity continuously presents a negative value, and the continuous negative polarity count value is output. When the continuous negative polarity count value reaches the thermal penetration constant threshold, a fixed amplitude blocking feedforward power pulse is directly superimposed on the power control output terminal of the temperature regulation circuit; wherein, the thermal penetration constant threshold is determined by the characteristic thermal diffusion delay constant determined by the combination of the characteristic geometric thickness of the protective outer jacket of the temperature detection unit and the physical thermal diffusion coefficient of the outer jacket material.

2. The temperature adjustment method for a temperature controller according to claim 1, characterized in that, Step S1 includes the following sub-steps: Step S11, using the discrete first-order low-pass filter algorithm as the discrete filter algorithm, extracting the discrete temperature sampling data in the current sampling period and the historical smoothed temperature value output in the previous sampling period during smoothing; Step S12, using the discrete first-order low-pass filter algorithm to calculate the weighted algebraic sum of the discrete temperature sampling data and the historical smoothed temperature value, and directly outputting the current smoothed temperature value from the weighted algebraic sum.

3. The temperature adjustment method for a temperature controller according to claim 2, characterized in that, The discrete first-order low-pass filter algorithm satisfies the following formula: ,in, This is the current smoothed temperature value. For discrete temperature sampling data, The historical smoothed temperature value, The set first-order low-pass smoothing filter coefficient has a value range of 0.1 to 0.

3.

4. The temperature adjustment method for a temperature controller according to claim 1, characterized in that, Before the steady-state temperature holding condition, there is also a temperature rise control stage, which includes the following steps: Step S401, obtain the current smoothed temperature value, and calculate the first-order and second-order time difference components of the current smoothed temperature value; Step S402, calculate the weighted sum of the first-order and second-order time difference components, and output the phase lead compensation amount; Step S403, superimpose the phase lead compensation amount into the proportional-integral-derivative control, output the thermal inertia lead compensation control duty cycle, and use the thermal inertia lead compensation control duty cycle to control the output duty cycle of the temperature regulation loop.

5. A temperature regulation method for a temperature controller according to claim 1, characterized in that, Step S3, which involves directly superimposing a fixed amplitude blocking feedforward power pulse at the power control output of the temperature regulation loop, includes the following sub-steps: Step S31, when the continuous negative polarity count value reaches the thermal penetration constant threshold, a fixed blocking feedforward power pulse is superimposed on the proportional-integral-derivative control output, and the output adjusts the control duty cycle; Step S32, in each subsequent sampling period, the blocking feedforward power pulse is attenuated proportionally according to the set attenuation coefficient, and the attenuated feedforward power pulse is output; Step S33, the attenuated feedforward power pulse is added to the proportional-integral-derivative control output, and when the blocking feedforward power pulse attenuates to zero and the algebraic polarity of the smooth temperature difference component is still negative, the proportional attenuation is maintained, and when the algebraic polarity of the smooth temperature difference component changes to a positive value, the blocking feedforward power pulse is forced to zero and injection is stopped.

6. The temperature adjustment method for a temperature controller according to claim 1, characterized in that, Following step S3, the following adaptive pulse width modulation steps are also included: Step S601, obtaining the final control duty cycle after superimposed blocking feedforward power pulses, and obtaining the set pulse width modulation period; Step S602, calculating the product of the final control duty cycle and the pulse width modulation period, outputting the high-level conduction time, and calculating the algebraic difference between the pulse width modulation period and the high-level conduction time, and outputting the low-level cutoff time; Step S603, within the pulse width modulation period, outputting a high-level drive signal to the heating drive circuit of the temperature controller according to the high-level conduction time, and outputting a low-level cutoff signal to the heating drive circuit according to the low-level cutoff time.

7. A temperature regulation method for a temperature controller according to claim 1, characterized in that, It also includes a heat dissipation power adjustment step, which includes: step S701, when the temperature controller is in a steady-state temperature holding condition, obtaining the reference control duty cycle required to maintain the temperature stability of the controlled thermal object, as the basic heat dissipation maintenance term; step S702, calculating the convective heat dissipation correction amount based on the temperature difference between the current smooth temperature value and the set environmental reference temperature value; step S703, adding the basic heat dissipation maintenance term and the convective heat dissipation correction amount to calculate the real-time heat dissipation maintenance duty cycle, and inputting the real-time heat dissipation maintenance duty cycle as a feedforward quantity into the temperature regulation loop.

8. A temperature regulation method for a temperature controller according to claim 1, characterized in that, The method for determining the heat penetration constant threshold includes: obtaining the wall thickness of the protective outer tube of the temperature detection unit and the thermal diffusivity of the protective outer tube material; calculating the ratio of the square of the wall thickness to the thermal diffusivity to determine the characteristic thermal diffusion delay time of the protective outer tube; obtaining the sampling period of the temperature controller; calculating the integer value of the ratio of the characteristic thermal diffusion delay time to the sampling period; and determining the integer value as the heat penetration constant threshold.

9. A temperature regulation method for a temperature controller according to claim 3, characterized in that, The method for adaptively adjusting the first-order low-pass smoothing filter coefficient includes: obtaining the smoothed temperature difference component and calculating the absolute value of the smoothed temperature difference component; increasing the first-order low-pass smoothing filter coefficient when the absolute value changes with time in an increasing trend; and decreasing the first-order low-pass smoothing filter coefficient when the absolute value changes with time in a stable trend.

10. A temperature regulation method for a temperature controller according to claim 1, characterized in that, When the current smoothed temperature value deviates from the target control temperature value, the steady-state temperature control deviation is adjusted through two-stage feedback control. The two-stage feedback control includes: calculating the algebraic deviation between the current smoothed temperature value and the target control temperature value; calculating the first-stage feedback adjustment gain using a proportional-integral algorithm, and calculating the second-stage feedback adjustment gain using the smooth accumulation of the integral loop; adding the first-stage feedback adjustment gain and the second-stage feedback adjustment gain to obtain the final feedback adjustment amount, and adjusting the output of the temperature control loop based on the final feedback adjustment amount.

Images