Temperature control system and method

Abstract

The application provides a temperature control system and method, which can be applied to the technical field of temperature control equipment. The temperature control system comprises the following connected elements: a temperature measurement circuit, which is used to generate an Nth temperature measurement signal for a current environment at an Nth time; N is an integer greater than 1; a signal processing circuit, which is used to analyze the Nth temperature measurement signal and determine an Nth temperature of the current environment at the Nth time; a controller, which is used to process N temperatures of the current environment at N times by using a trained neural network to obtain an N+1th control coefficient, correct the N+1th control coefficient based on an Nth temperature change rate from an N-1th temperature to the Nth temperature, and generate a corresponding pulse width modulation signal based on the corrected control coefficient; and a heating circuit, which is used to generate corresponding heat energy under the control of the pulse width modulation signal.

Description

Technical Field

[0001] This invention relates to the field of temperature control equipment technology, and more specifically to a temperature control system and method. Background Technology

[0002] With the development of technologies such as low-temperature physics experiments, superconductivity, quantum computing, aerospace, and biomedicine, the demand for precise temperature control in ultra-low temperature environments has increased significantly. However, temperature control devices still have shortcomings in terms of temperature control range, accuracy, and the universality of temperature control algorithms in ultra-low temperature environments. Summary of the Invention

[0003] In view of the above problems, the present invention provides a temperature control system and method.

[0004] According to one aspect of the present invention, a temperature control system is provided, comprising: a temperature measuring circuit for generating an Nth temperature measuring signal for the current environment at time N; where N is an integer greater than 1; a signal processing circuit for parsing the Nth temperature measuring signal to determine the Nth temperature of the current environment at time N; a controller for processing N temperatures of the current environment at N times using a trained neural network to obtain an (N+1)th control coefficient, correcting the (N+1)th control coefficient based on the rate of change of the Nth temperature from the (N-1)th temperature to the Nth temperature, and generating a corresponding pulse width modulation signal based on the corrected control coefficient; and a heating circuit for generating corresponding heat energy under the control of the pulse width modulation signal.

[0005] According to another aspect of the present invention, a temperature control method is provided, comprising: generating an Nth temperature measurement signal for the current environment at a Nth time; where N is an integer greater than 1; parsing the Nth temperature measurement signal to determine the Nth temperature of the current environment at a Nth time; processing the N temperatures of the current environment at N times using a trained neural network to obtain an (N+1)th control coefficient; correcting the (N+1)th control coefficient based on the rate of change of the Nth temperature from the (N-1)th temperature to the Nth temperature; and generating a corresponding pulse width modulation signal based on the corrected control coefficient; and controlling the temperature of a thermally conductive material in the current environment under the control of the pulse width modulation signal.

[0006] According to an embodiment of the present invention, the temperature measuring circuit can generate the Nth temperature signal of the current environment. Subsequently, the signal processing circuit can determine the Nth temperature of the current environment based on the temperature measuring signal. Then, the neural network deployed by the processor can synthesize the N temperatures at N times to calculate the preliminary N+1th control coefficient. Furthermore, the processor can calculate the rate of change of the Nth temperature from time N-1 to time N based on the Nth temperature and the N-1th temperature at the previous time.

[0007] Furthermore, the processor can correct the (N+1)th control coefficient based on the Nth temperature change rate to adjust the heating power of the heating circuit, thereby at least partially eliminating the influence of the specific heat capacity of the heated object on the temperature control, and accurately performing high-precision temperature control. Moreover, since the controller calculates N temperatures based on a trained neural network, it has a faster response speed compared to PID (Proportional Integral Derivative) controllers in some solutions. Thus, accurate real-time temperature control is possible in ultra-low temperature environments. Attached Figure Description

[0008] The above-described features, other objects, and advantages of the present invention will become clearer from the following description of embodiments of the invention with reference to the accompanying drawings, in which:

[0009] Figure 1 A schematic diagram of a temperature control system according to an embodiment of the present invention is shown.

[0010] Figure 2 A schematic diagram of a temperature control system according to another embodiment of the present invention is shown.

[0011] Figure 3 A schematic diagram of a temperature measuring circuit and a signal processing circuit according to an embodiment of the present invention is shown.

[0012] Figure 4 A schematic diagram of a trained neural network according to an embodiment of the present invention is shown.

[0013] Figure 5 A temperature calibration curve for a second temperature range according to an embodiment of the present invention is shown.

[0014] Figure 6 A temperature calibration curve for a first temperature range according to an embodiment of the present invention is shown.

[0015] Figure 7 A comparison graph is shown showing the temperature control curves of some PID controllers at 60K and the temperature control curves of the controller of this embodiment at 60K and 62K.

[0016] Figure 8 A comparison graph of the temperature control curves of some PID controller schemes and the controller of the present invention embodiment at 295K is shown.

[0017] Figure 9 A schematic diagram of a heating circuit according to an embodiment of the present invention is shown.

[0018] Figure 10 A schematic diagram of a temperature control method according to an embodiment of the present invention is shown. Detailed Implementation

[0019] Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the invention. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of the invention for ease of explanation. However, it will be apparent that one or more embodiments may be practiced without these specific details. Furthermore, descriptions of well-known structures and techniques are omitted in the following description to avoid unnecessarily obscuring the concept of the invention.

[0020] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the invention. The terms “comprising,” “including,” etc., as used herein indicate the presence of the stated features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0021] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0022] When using expressions such as "at least one of A, B and C", they should generally be interpreted in accordance with the meaning that is commonly understood by those skilled in the art (e.g., "a system having at least one of A, B and C" should include, but is not limited to, a system having A alone, a system having B alone, a system having C alone, a system having A and B, a system having A and C, a system having B and C, and / or a system having A, B and C, etc.).

[0023] With the development of technologies such as low-temperature physics experiments, superconductivity, quantum computing, aerospace, and biomedicine, the demand for precise temperature control in ultra-low temperature environments has increased significantly. However, temperature control devices have shortcomings in terms of control range, accuracy, and the universality of control algorithms in ultra-low temperature environments. Regarding the control range, temperature control devices are mostly used in room temperature and high-temperature environments, and they struggle to provide accurate heat compensation in ultra-low temperature environments, thus limiting the control range.

[0024] In terms of temperature control, devices capable of controlling temperatures in ultra-low temperature environments often rely on large, precision instruments. This results in issues such as high resource consumption, poor portability, and limited applicability when controlling temperatures in ultra-low temperature environments.

[0025] In view of this, embodiments of the present invention provide a temperature control system based on an improved PID neural network algorithm, suitable for precise temperature control in ultra-low temperature environments. The following description is in conjunction with the accompanying drawings.

[0026] Figure 1 A schematic diagram of a temperature control system according to an embodiment of the present invention is shown.

[0027] like Figure 1 As shown, the temperature control system of this embodiment includes a connected temperature measuring circuit, a signal processing circuit, a controller, and a heating circuit.

[0028] In this embodiment of the invention, the temperature measuring circuit may include components such as a temperature-sensitive diode. Based on this, the temperature measuring circuit can detect the current ambient temperature and generate a corresponding temperature measurement signal. Subsequently, the temperature measuring circuit can provide the Nth temperature measurement signal to the signal processing circuit. For example, the temperature measuring circuit can provide the corresponding Nth temperature measurement signal to the signal processing circuit based on the ambient temperature at time N. Here, N is an integer greater than 1.

[0029] The signal processing circuit may include units such as a processor. Based on this, the processor can then parse the digital signal to determine the specific value of the Nth temperature of the current environment at time N. Subsequently, the processor can provide a signal indicating the Nth temperature to the controller. It should be understood that this example only uses the ambient temperature at time N; however, in this embodiment of the invention, the temperature measuring circuit and the signal processing circuit can provide corresponding signals to the controller at N times, respectively, based on the N temperatures of the current environment at N times. Thus, the controller will store the N temperatures at time N.

[0030] The controller can be deployed with a feedback network based on a trained neural network. This neural network can be a PID neural network. Further, in this embodiment of the invention, the controller can utilize the trained neural network to process N temperatures at N times in the current environment to obtain the (N+1)th control coefficient. Specifically, the controller can input the N temperatures at N times into the trained PID neural network and output the (N+1)th control coefficient. The (N+1)th control coefficient includes a proportional gain, an integral gain, and a derivative gain. Specifically, the proportional gain can be the proportional gain coefficient, the integral gain coefficient can be the integral gain coefficient, and the derivative gain coefficient can be the integral gain coefficient.

[0031] Furthermore, the controller can calculate the rate of change of the Nth temperature based on the time difference (e.g., per unit time) between the (N-1)th and Nth temperatures, and between the (N-1)th and Nth temperatures. It can then correct the (N+1)th control coefficient based on this rate of change, generating a corresponding pulse-width modulation (PWM) signal. The controller can then use this PWM signal to control the heating circuit to generate the corresponding heat. Specifically, the heating circuit can heat an object (e.g., a material to be heated) in an ultra-low temperature environment (temperatures from 10 K to 300 K). It should be understood that the heating circuit generates heat in the current environment, thus heating the object in that environment.

[0032] In this embodiment of the invention, while keeping the heating power of the heating circuit constant, the rate of temperature change of the heated object may differ at different temperatures in an ultra-low temperature environment. Furthermore, even with the same heating time, the greater the absolute value of the temperature change rate, the smaller the specific heat capacity of the heated object, thus affecting the accurate control of the heated object's temperature.

[0033] To address this, the temperature measurement circuit of this embodiment can generate the Nth temperature signal of the current environment. Subsequently, the signal processing circuit can determine the Nth temperature of the current environment based on this temperature signal. Then, the neural network deployed by the processor can synthesize the N temperatures at N times to calculate a preliminary N+1th control coefficient. Furthermore, the processor can calculate the rate of change of the Nth temperature from time N-1 to time N based on the Nth temperature and the N-1th temperature at the previous time.

[0034] Furthermore, the processor can correct the (N+1)th control coefficient based on the Nth temperature change rate to adjust the heating power of the heating circuit, thereby at least partially eliminating the influence of the specific heat capacity of the heated object on the temperature control, and achieving accurate high-precision temperature control. Moreover, since the controller calculates N temperatures based on a trained neural network, it has a faster response speed compared to PID controllers in some other solutions. Thus, accurate real-time temperature control is possible in ultra-low temperature environments.

[0035] Figure 2 A schematic diagram of a temperature control system according to another embodiment of the present invention is shown.

[0036] like Figure 2 As shown, in this embodiment of the invention, the temperature measuring circuit may include a constant current source circuit, a temperature sensor, and a sampling circuit. The constant current source circuit can provide a fixed current to the temperature sensor so that the temperature sensor operates based on this current. The sampling circuit can sample the voltage generated by the temperature sensor based on temperature to generate the Nth temperature measurement signal.

[0037] The signal processing circuit may include a signal processing module and a processor. For example, the Nth temperature measurement signal may be an analog signal. An analog-to-digital converter can perform analog-to-digital conversion on the Nth temperature measurement signal to obtain a corresponding digital signal. Subsequently, the processor can parse the digital signal, determine the Nth temperature of the current environment at time N, and indicate the Nth temperature to the controller.

[0038] In this embodiment of the invention, the temperature control system may further include a host computer and a display. The host computer can be used to configure the neural network of the controller, and is not limited thereto. The controller can display the Nth temperature through the display.

[0039] Furthermore, the controller can process the Nth temperature based on a trained neural network to control the heating circuit to generate heat, thereby affecting the ambient temperature. The ambient temperature, in turn, affects the voltage generated by the temperature sensor, thus forming a closed loop. Further, the following combines... Figure 3 The temperature measurement circuit and signal processing circuit of the embodiments of the present invention will be further described.

[0040] Figure 3 A schematic diagram of a temperature measuring circuit and a signal processing circuit according to an embodiment of the present invention is shown.

[0041] like Figure 3 As shown, the temperature measurement circuit may include a connected constant current source circuit, a silicon diode, and a sampling circuit.

[0042] In this embodiment of the invention, the constant current source circuit can provide a fixed current to the silicon diode. When the current remains constant, the voltage across the silicon diode is only related to temperature, thereby reducing measurement errors caused by current fluctuations. For example, the temperature measurement range of the silicon diode can be 10 K to 300 K.

[0043] Based on this, the voltage drop of a silicon diode under forward bias changes systematically with junction temperature, exhibiting a good temperature-voltage characteristic curve. Therefore, under the influence of the ambient temperature, and based on the constant current of the constant current source circuit, the voltage drop between the first and second terminals of the silicon diode can be adjusted at time N, so that the sampling circuit can generate the Nth temperature measurement signal based on the voltage drop. Specifically, the sampling circuit can sample the voltage containing temperature information from the silicon diode and transmit the sampled Nth temperature measurement signal to the signal processing circuit for further processing.

[0044] Furthermore, the signal processing module of the signal processing circuit may include a differential amplifier, a voltage follower, and a filter circuit connected in sequence.

[0045] Because the forward voltage drop of silicon diodes is typically in the hundreds of millivolts range, the signal amplitude is relatively small. Direct analog-to-digital conversion is susceptible to noise and insufficient resolution, making high-precision temperature measurement difficult. Therefore, to meet high-precision requirements, the sampled Nth temperature measurement signal can be input to a differential amplifier. The differential amplifier amplifies the Nth temperature measurement signal from the temperature measurement circuit. After amplification, the differential amplifier provides the amplified Nth temperature measurement signal to the filter circuit via a voltage follower. The voltage follower can be based on an operational amplifier. In this way, the voltage follower achieves near-identical input and output voltages, possessing extremely high input impedance and extremely low output impedance, enabling impedance matching between the preceding and following stages and reducing signal attenuation or distortion.

[0046] The filtering circuit can filter the amplified Nth temperature measurement signal to remove high-frequency noise and electromagnetic interference, thereby obtaining a more stable and reliable filtered Nth temperature measurement signal. The above has already been discussed. Figure 3 The structure shown is illustrated, but it should be understood that the embodiments of the present invention are not limited thereto. For example, in an embodiment of the present invention, the signal processing module may further include components not shown in the diagram. Figure 3 The analog-to-digital converter is shown in the figure. The filtering circuit can provide the filtered Nth temperature measurement signal to the analog-to-digital converter, so that the analog-to-digital converter converts the type of the filtered Nth temperature measurement signal from an analog signal to a digital signal, and provides the Nth temperature measurement signal as a digital signal to the processor described above.

[0047] The processor may include a microcontroller unit (MCU), etc. For example, the microcontroller unit may be implemented based on a single-chip microcomputer. In this embodiment of the invention, a temperature-voltage characteristic curve of the silicon diode can be pre-fitted based on the voltage drop across the silicon diode at different temperatures and the corresponding temperatures. Thus, the processor can use this pre-set temperature-voltage characteristic curve of the silicon diode to determine the corresponding Nth temperature based on the voltage value corresponding to the received Nth temperature measurement signal. In this way, the voltage drop across the silicon diode can be restored to the accurate temperature value of the Nth temperature.

[0048] Furthermore, regarding temperature control algorithms and accuracy, some temperature control devices can employ PID algorithms. However, because the proportional, integral, and derivative coefficients of PID algorithms are fixed, they lack adaptive adjustment capabilities, making them unsuitable for high-precision temperature control in ultra-low temperature environments. Specifically, in complex, nonlinear, and dynamically changing ultra-low temperature environments (e.g., below 100 K), the specific heat capacity of materials decreases significantly, and even a very small heat input can lead to large temperature fluctuations. Consequently, temperature control devices may suffer from overshoot, hysteresis, or even instability, making it difficult to meet the requirements for high-precision, wide-temperature-range temperature control.

[0049] In response, the PID neural network deployed in the controller in this embodiment of the invention has been improved. The controller outputs the (N+1)th control coefficient based on N temperature error values ​​between N temperatures and the target temperature. Specifically, the controller can calculate the N temperature error values ​​with respect to the target temperature. Then, the N temperature error values ​​and the Nth temperature can be input into the trained neural network to output the (N+1)th control coefficient.

[0050] Specifically, the trained neural network deployed in the controller includes an input layer. The input layer can determine the temperature error values ​​between each of the N temperatures and the target temperature, thus obtaining the aforementioned N temperature error values. Furthermore, it can determine the sum of the N temperature error values, and the difference between the Nth temperature error value and the (N-1)th temperature error value. Subsequently, the controller can further process the N temperature error values, the sum, the difference, and the Nth temperature using the trained neural network, outputting the (N+1)th control coefficient. The following combines... Figure 4 Please provide a detailed explanation.

[0051] Figure 4 A schematic diagram of a trained neural network according to an embodiment of the present invention is shown.

[0052] like Figure 4 As shown, a trained neural network can include at least an input layer, hidden layers, and an output layer. The input layer has 4 neurons, the hidden layers have 5 neurons, and the output layer has 3 neurons. It should be understood that these numbers are merely illustrative.

[0053] The input layer can extract features based on the following formula:

[0054] (1)

[0055] (2)

[0056] (3)

[0057] (4)

[0058] Where X1 represents the Nth temperature error value out of N temperature error values, which can be used to represent the error between the Nth temperature at time N and the target temperature. e(n) represents the nth error value. n is a positive integer less than N. X2 represents the sum of the above values, that is, the cumulative value of the temperature error values ​​from time 1 to time N. X3 represents the difference between the Nth temperature error value at time N and the Nth temperature error value at time N-1. X4, i.e., y(N), represents the Nth temperature at time N. r(N) is the predetermined target temperature at time N. It should be noted that the predetermined target temperatures at N times can be the same, but can also be set to be different according to requirements, which is not limited here. Based on this, by using the four indicators X1~X4 extracted above as input features, the four dimensions that need to be paid attention to in the temperature control process, namely error, cumulative value, trend of change and current operating temperature, can be covered, thereby completely describing the temperature control status of the controller. Since the temperature control in this embodiment of the invention is based on the proportional, integral, and derivative actions of temperature error, using the above four indicators as input features during model training allows the neural network to learn directly in the feature space corresponding to the PID structure. This makes it easier to calculate accurate proportional, derivative, and integral coefficients, resulting in faster convergence and lower training difficulty. Furthermore, it at least partially avoids a large number of redundant sampling points, improving the training effect of the neural network and thus enhancing the performance of the trained neural network.

[0059] The five neurons in the hidden layer provide moderate computational power, capable of handling complex temperature scenarios without becoming computationally overloaded. Each node in the hidden layer has specific weights and biases, and the Sigmoid activation function can be used to introduce nonlinear computations into the hidden layer's calculation process.

[0060] The output layer has three nodes: a non-negative KP (i.e., proportional coefficient) node, a KI (i.e., integral coefficient) node, and a KD (i.e., differential coefficient) node.

[0061] Based on this, in the adaptive adjustment process of PID parameters in this embodiment of the invention, at time N, the i-th neuron (i=1,2,3,4) of the input layer receives the N-th temperature, and the signal of each neuron is related to the weight of the h-th neuron (h=1,2,3,4,5) of the hidden layer. Multiply and sum, then add the bias of the hidden layer neurons. This forms the net input to the h-th neuron in the hidden layer. The hidden layer input is then transformed by the Sigmoid activation function before being output. Similarly, the output signal of each neuron in the hidden layer is related to the weights of the j-th neuron (j=1,2,3) in the output layer. Multiply and sum, then add the bias of the output layer neuron. This forms the net input to the j-th neuron in the output layer, and finally outputs the proportional coefficient, integral coefficient, and differential coefficient through the Softplus activation function. However, the embodiments of the present invention are not limited to this. In one embodiment of the present invention, the neural network can also output a power correction value ΔP for the heating power. m This allows the power correction value to be used to correct the proportional coefficient, integral coefficient, and derivative coefficient, thereby correcting the heating power and improving the accuracy of temperature control.

[0062] In the training process of the aforementioned neural network, the initial neural network can be used to tune the input control coefficients online. Subsequently, the neural network undergoes self-learning and parameter adjustment to slightly optimize the output control coefficients. This allows the network parameters to quickly reach their optimal values, thus completing the training process and improving the network's performance.

[0063] Specifically, during the weight and bias update process, the gradient of each layer can be calculated using a loss function, and the weights and biases can be updated using gradient descent. For example, the loss function can generate an error signal for the output layer based on the negative gradient of the net input of each neuron in the output layer. The output layer weights and biases are then updated based on the error signal, the output data of the corresponding hidden layer neurons, the learning rate, and the momentum term. Furthermore, the error signal of the output layer can be propagated back to the hidden layer along the output layer weights and multiplied by the derivative of the hidden layer activation function to obtain the error signal of each neuron in the hidden layer. The weights and biases from the input layer to the hidden layer are then updated based on the error signal, the corresponding input layer signal, the learning rate, and the momentum term. In this way, the training process of the aforementioned neural network can be completed, enabling the trained neural network to process the Nth temperature and obtain the (N+1)th control coefficient.

[0064] Based on this, the controller in this embodiment of the invention can fine-tune and optimize the control coefficients using a trained neural network based on N temperature error values, and then output an accurate pulse width modulation signal to the heating circuit through a timer pin. It should be noted that the controller in this embodiment of the invention can also be implemented based on a microcontroller. After training the neural network, the trained neural network can be deployed on the microcontroller. Thus, this embodiment of the invention can at least partially guarantee the computational speed of the neural network while consuming fewer resources. Furthermore, by using a microcontroller as the controller, this embodiment of the invention achieves a high degree of integration in circuit processing, offering significant advantages in terms of resource consumption and operational portability.

[0065] Furthermore, in another embodiment of the present invention, the power correction value can be calculated as follows: When the Nth temperature falls within a first range, the controller calculates the rate of change of the Nth temperature based on the temperature difference between the Nth and (N-1)th temperatures and the time interval; based on the rate of change of the Nth temperature and the reference temperature rate of change, a power correction value is calculated, and the power correction value is used to correct the (N+1)th control coefficient. The reference temperature rate of change is determined based on the rate of temperature change of the current environment per unit time when the current ambient temperature falls within a second range; the first range is lower than the second range. For example, the first temperature range can be 0~100 K. The second temperature range can be 100~300 K.

[0066] Specifically, in cryogenic environments, the specific heat capacity of materials decreases sharply as the temperature drops. To reduce the impact on the control system, an estimated power correction value is introduced to address, at least partially, the problems of insufficient model adaptability and disturbance errors in wide temperature ranges, thus enabling more precise control. While it is difficult to obtain the exact specific heat capacity of materials (e.g., copper blocks) within the range of 10–300 K, the temperature drop curve of the material can be measured while keeping the heating power of the heating circuit constant, thus obtaining the rate of temperature change at different temperatures. With the heating power of the heating circuit constant and the heating circuit operating for the same duration, the larger the absolute value of the rate of temperature change, the smaller the specific heat capacity. Based on this, in cryogenic environments (e.g., below 100 K), the power correction value can vary with the rate of temperature change to correct for the impact of the decreased specific heat capacity.

[0067] Figure 5 A temperature calibration curve for a second temperature range according to an embodiment of the present invention is shown. Figure 6 A temperature calibration curve for a first temperature range according to an embodiment of the present invention is shown. It should be understood that this is merely an example and is not intended to be elaborated upon.

[0068] like Figure 5 As shown, above 100 K, the temperature calibration curve is basically linear. Figure 5 The leftmost part of the curve corresponds to the state when cooling is off), and the rate of temperature change remains almost constant, therefore no power correction is needed. Based on this, [further adjustments can be made]. Figure 5 The temperature calibration curve shown selects one point as a reference point, and calculates the rate of change of the reference temperature at that reference point. For example... Figure 6 As shown, the rate of temperature change changes significantly below 100K. Therefore, the temperature calibration curve can be divided into segments of 1K each, and the rate of temperature change at each 1K interval can be calculated.

[0069] For any two temperatures, the rate of temperature change between them can be calculated using the following formula:

[0070] (5)

[0071] Where K represents the rate of temperature change. T m and T n Let t represent the temperature at time m and time n, respectively. m is a positive integer less than N, and m ≠ n. For example, n can range from 15 K to 100 K. m t represents time m. n Let ΔT represent time n. ΔT represents time T. m and T n The temperature change between t, Δt represents t m and t n The amount of change over time.

[0072] The power correction value can then be calculated using the following formula:

[0073] (6)

[0074] Where K0 represents the rate of change of the reference temperature. N ΔP represents the rate of temperature change of the Nth temperature. N This represents the power correction value corresponding to the Nth rate of temperature change.

[0075] Based on this, the pulse width modulation signal generated using the corrected control coefficients is shown below:

[0076] (7)

[0077] Where OUTPUT represents the pulse width modulation signal. X1 can represent the Nth temperature error value. X2 represents the sum of the above values, that is, the cumulative value of the temperature error values ​​from time 1 to time N. X3 represents the difference between the Nth temperature error value at time N and the (N-1)th temperature error value at time N-1. KP represents the proportional coefficient. KI represents the integral coefficient. KD represents the derivative coefficient. ΔP N This represents the power correction value corresponding to the Nth rate of temperature change.

[0078] Based on this, the heating circuit can be accurately controlled to generate heat energy using this pulse width modulation signal. Thus, this embodiment of the invention at least partially solves the problems of insufficient model adaptability and disturbance errors in wide temperature range environments, achieving high-precision temperature control in ultra-low temperature environments. Furthermore, as... Figure 5 and Figure 6 As shown, since this embodiment of the invention uses a silicon diode as a temperature sensor, the temperature measurement curve of this embodiment largely overlaps with that of some high-precision temperature measuring instruments. Therefore, it can be seen that the silicon diode in this embodiment of the invention can achieve high-precision temperature detection in ultra-low temperature environments.

[0079] Figure 7 A comparison graph is shown showing the temperature control curves of some PID controllers at 60K and the temperature control curves of the controller of this embodiment at 60K and 62K. Figure 8 A comparison graph of the temperature control curves of some PID controller schemes and the controller of the present invention embodiment at 295K is shown.

[0080] refer to Figure 7 It is known that for some PID controllers, the temperature control performance is relatively poor at 60K, with relatively large temperature fluctuations. In comparison, the controller of this embodiment of the invention exhibits relatively good temperature control performance at 60K, with relatively small temperature fluctuations. Furthermore, the controller of this embodiment of the invention not only has good temperature control performance at 60K, but also at temperatures different from 60K (e.g., 62K). Similarly, refer to... Figure 8 It can be seen that at a temperature of 295K, the controller of this embodiment still has a better temperature control effect than some PID controllers.

[0081] Furthermore, in this embodiment of the invention, the heating circuit includes a connected gate driving unit, a switching unit, a heating element, and a power supply unit for supplying power to the gate driving unit and the heating element. The gate driving unit, under the control of a pulse width modulation signal, controls the switching unit to output pulses to the heating element based on a corrected control coefficient, thereby causing the heating element to generate heat. The following is in conjunction with... Figure 9 Let's illustrate with examples.

[0082] Figure 9 A schematic diagram of a heating circuit according to an embodiment of the present invention is shown.

[0083] like Figure 9 As shown, the heating circuit in this embodiment may further include a signal isolation unit. The signal isolation unit can partially isolate the controller and the gate drive unit to at least partially prevent large signal backflow from damaging the controller in the event of abnormal operation of the gate drive unit.

[0084] The power supply unit may include an AC-to-DC unit and a DC-to-DC unit. The AC-to-DC unit may include two output signals. For example, the AC-to-DC unit can process 220 V AC voltage into 150 V DC voltage. Specifically, the AC-to-DC unit can provide 150 V DC voltage to the DC-to-DC unit. The DC-to-DC unit can also directly supply 150 V to the heating element through another branch to power the heating element. For example, the heating element can be a 100Ω ceramic heating element. Furthermore, the DC-to-DC unit transforms the 150 V DC voltage into 12 V DC voltage and provides 12 V DC voltage to the gate drive unit for normal operation. It should be noted that the branch providing the 150 V DC voltage and the branch providing the 12 V DC voltage are electrically isolated from each other.

[0085] The switching unit may include multiple transistors used as power transistors. The gate drive circuit can control the voltage at the gate of the transistor to control the transistor to turn on or off, thereby controlling the conduction state of the corresponding branch of the switching unit and controlling the 150V DC voltage supplied by the DC-to-DC unit to the heating element.

[0086] Furthermore, embodiments of the present invention may also include an overcurrent protection circuit. The fuse in the overcurrent protection circuit blows when the current exceeds a predetermined value to prevent excessive current from damaging the heating circuit.

[0087] Based on this, the maximum heating power of the heating circuit in this embodiment of the invention can be 225 W (watts).

[0088] Based on the above, the temperature control system of this invention can achieve wide-range temperature control in a high-vacuum environment, with a temperature control accuracy of ±0.05K and a steady-state standard deviation of 0.0059. Furthermore, this temperature control system can achieve precise temperature control in ultra-low temperature environments, providing a stable temperature control point over a wide temperature range from 15 K to 300 K.

[0089] Figure 10 A schematic diagram of a temperature control method according to an embodiment of the present invention is shown.

[0090] like Figure 10 As shown, the temperature control method of this embodiment includes operations S1010 to S1060.

[0091] In operation S1010, the Nth temperature measurement signal for the current environment is generated at time N.

[0092] In operation S1020, the Nth temperature measurement signal is analyzed to determine the Nth temperature of the current environment at time N.

[0093] In operation S1030, the trained neural network is used to process the N temperatures of the current environment at N times to obtain the (N+1)th control coefficient.

[0094] In operation S1040, the control coefficient of the N+1th temperature is corrected based on the rate of change of the Nth temperature from the (N-1)th temperature to the Nth temperature.

[0095] In operation S1050, a corresponding pulse width modulation signal is generated based on the corrected control coefficients.

[0096] In operation S1060, the temperature of the thermally conductive material in the current environment is controlled under the control of the pulse width modulation signal.

[0097] The following description is based on specific embodiments.

[0098] In this embodiment of the invention, a temperature sensor and a heating element are fixed in a vacuum chamber (without evacuation or refrigeration) using a thermally conductive copper block, while the remaining modules are placed at room temperature. The temperature sensor and sampling circuit are connected using a shielded signal cable. The heating element and heating circuit are connected using a power cable. Subsequently, a vacuum and ultra-low temperature environment can be constructed. Specifically, the vacuum chamber can be sealed, and an external vacuum pump can be used to evacuate the chamber until the vacuum level is ≤10⁻³ mbar. Afterward, the liquid helium refrigeration device is activated to construct the ultra-low temperature environment.

[0099] Subsequently, the temperature of the environment surrounding the heated object can be controlled using the aforementioned temperature control system. Furthermore, the controller of this temperature control system can display real-time information such as the temperature, the predetermined target temperature, the corrected control coefficient, and the heating power on the screen. If a temperature control curve is required, it can be transmitted to the host computer in real-time via a serial cable.

[0100] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0101] Those skilled in the art will understand that the features described in the various embodiments of the present invention can be combined and / or combined in various ways, even if such combinations or combinations are not explicitly described in the present invention. In particular, the features described in the various embodiments of the present invention can be combined and / or combined in various ways without departing from the spirit and teachings of the present invention. All such combinations and / or combinations fall within the scope of the present invention.

[0102] The embodiments of the present invention have been described above. However, these embodiments are merely illustrative and not intended to limit the scope of the invention. Although various embodiments have been described above, this does not mean that the measures in the various embodiments cannot be used advantageously in combination. Various substitutions and modifications can be made by those skilled in the art without departing from the scope of the invention, and all such substitutions and modifications should fall within the scope of the invention.

Claims

1. A temperature control system, characterized in that, The temperature control system includes connections to: A temperature measurement circuit is used to generate the Nth temperature measurement signal for the current environment at time N; N is an integer greater than 1. The signal processing circuit is used to analyze the Nth temperature measurement signal and determine the Nth temperature of the current environment at the Nth time. The controller is configured to use a trained neural network to process N temperatures at N times of the current environment to obtain the (N+1)th control coefficient, correct the (N+1)th control coefficient based on the rate of change of the Nth temperature from the (N-1)th temperature to the Nth temperature, and generate a corresponding pulse width modulation signal based on the corrected control coefficient. A heating circuit is used to generate corresponding heat energy under the control of the pulse width modulation signal; The controller is also used for: If the Nth temperature falls within the first range, the rate of change of the Nth temperature is calculated based on the temperature difference between the Nth temperature and the (N-1)th temperature and the unit time. Based on the Nth temperature change rate and the reference temperature change rate, a power correction value is calculated, and the N+1th control coefficient is corrected using the power correction value, wherein the first range is 0~100 K; The reference temperature change rate is determined based on the temperature change rate of the current environment per unit time, assuming the current ambient temperature falls within a second range; wherein the first range is lower than the second range.

2. The temperature control system according to claim 1, characterized in that, The controller is also used for: The trained neural network outputs the (N+1)th control coefficient based on the N temperature error values ​​between each of the N temperatures and the target temperature.

3. The temperature control system according to claim 2, characterized in that, The trained neural network includes an input layer; The input layer is used for: Determine the temperature error values ​​between each of the N temperatures and the target temperature to obtain the N temperature error values; Determine the sum of the N temperature error values; Determine the difference between the Nth temperature error value and the (N-1)th temperature error value; The controller is also configured to process the N temperature error values, the sum, the difference, and the Nth temperature using the trained neural network, and output the (N+1)th control coefficient.

4. The temperature control system according to claim 1, characterized in that, The temperature measuring circuit includes a connected constant current source circuit, a silicon diode, and a sampling circuit; The silicon diode is used to adjust the voltage drop between the first and second terminals of the silicon diode at the Nth time, based on the constant current of the constant current source circuit, under the influence of the temperature of the current environment, so that the sampling circuit generates the Nth temperature measurement signal based on the voltage drop.

5. The temperature control system according to claim 1, characterized in that, The signal processing circuit includes: The processor is used to determine the corresponding Nth temperature based on the received Nth temperature measurement signal using a pre-set temperature-voltage characteristic curve of a silicon diode.

6. The temperature control system according to claim 5, characterized in that, The signal processing circuit also includes a differential amplifier, a voltage follower, a filter circuit, and an analog-to-digital converter connected in sequence; The differential amplifier amplifies the Nth temperature measurement signal from the temperature measurement circuit and provides the amplified Nth temperature measurement signal to the filter circuit via the voltage follower, so that the filter circuit filters the amplified Nth temperature measurement signal, thereby causing the analog-to-digital converter to convert the type of the filtered Nth temperature measurement signal from an analog signal to a digital signal, and providing the Nth temperature measurement signal as a digital signal to the processor.

7. The temperature control system according to claim 5, characterized in that, The heating circuit includes a connected gate driving unit, a switching unit, a heating element, and a power supply unit for supplying power to the gate driving unit and the heating element. The gate driving unit is used to control the switching unit to output pulses to the heating element based on the corrected control coefficient under the control of the pulse width modulation signal, so that the heating element generates the heat energy.

8. A temperature control method, characterized in that, include: At time N, generate the Nth temperature measurement signal for the current environment; N is an integer greater than 1; Analyze the Nth temperature measurement signal to determine the Nth temperature of the current environment at the Nth time; The N+1th control coefficient is obtained by processing the N temperatures of the current environment at N times using a trained neural network. Based on the rate of change of the Nth temperature from the (N-1)th temperature to the Nth temperature, the N+1th control coefficient is corrected, and a corresponding pulse width modulation signal is generated based on the corrected control coefficient. Under the control of the pulse width modulation signal, the temperature of the thermally conductive material in the current environment is controlled; The temperature control method further includes: If the Nth temperature falls within the first range, the rate of change of the Nth temperature is calculated based on the temperature difference between the Nth temperature and the (N-1)th temperature and the unit time. Based on the Nth temperature change rate and the reference temperature change rate, a power correction value is calculated, and the N+1th control coefficient is corrected using the power correction value, wherein the first range is 0~100 K; The reference temperature change rate is determined based on the temperature change rate of the current environment per unit time, assuming the current ambient temperature falls within a second range; wherein the first range is lower than the second range.

Images