Temperature control method, device and system for low-temperature equipment

Abstract

The invention relates to a temperature control method, device and system for low-temperature equipment. The method comprises the following steps: a temperature detection step: detecting expected temperature information and actual temperature information of the low-temperature equipment; when a difference value between the expected temperature information and the actual temperature information is not in a preset temperature range threshold value, processing the expected temperature information and the actual temperature information by using a pre-trained neural network to obtain a temperature adjustment influence parameter of the temperature controller; controlling a temperature controller to output a temperature control signal according to the temperature adjustment influence parameter; and after the temperature of the low-temperature equipment is adjusted according to the temperature control signal, returning to the temperature detection step until the difference value between the actual temperature information and the expected temperature information is within the preset temperature range threshold value. By adopting the method, the temperature of the low-temperature equipment can be accurately controlled.

Description

Technical Field

[0001] The present application relates to the field of quantum voltage technology, and in particular to a temperature control method, device and system for cryogenic equipment. Background Art

[0002] Temperature control of power electronics is crucial for ensuring safe operation and improving performance and reliability. For example, in cryostat temperature control for liquid helium-free quantum voltage devices, the cold-end temperature is affected by a variety of physical factors, including material thermal conductivity, specific heat capacity, and mass. These factors are not independent but rather coupled, resulting in significant nonlinear characteristics in the entire temperature control system, increasing the difficulty of precise temperature control.

[0003] Currently, PID (proportional-integral-derivative) controller technology is commonly used for temperature control. Among them, the PID controller is the most widely used classic control algorithm in the field of industrial control. Its core outputs temperature regulation influencing parameters through the coordinated action of the proportional, integral, and differential links. The temperature regulation influencing parameters include proportional information, integral information, and differential information. Based on the temperature regulation influencing parameters, precise regulation of the controlled object is achieved.

[0004] However, the current method of using a PID controller to achieve temperature control of cryogenic equipment still cannot accurately control the temperature of cryogenic equipment. Summary of the Invention

[0005] Based on this, it is necessary to provide an accurate temperature control method, device, system, computer equipment, computer-readable storage medium and computer program product for the above-mentioned technical problems.

[0006] In a first aspect, the present application provides a temperature control method for a cryogenic device, comprising:

[0007] Temperature detection step: detecting the expected temperature information and actual temperature information of the cryogenic equipment;

[0008] When the difference between the expected temperature information and the actual temperature information is not within the preset temperature range threshold, the expected temperature information and the actual temperature information are processed using a pre-trained neural network to obtain the temperature adjustment influencing parameters of the temperature controller;

[0009] According to the temperature adjustment influencing parameters, the temperature controller is controlled to output a temperature control signal;

[0010] After the temperature of the cryogenic device is adjusted according to the temperature control signal, the process returns to the temperature detection step until the difference between the actual temperature information and the expected temperature information is within a preset temperature range threshold.

[0011] In a second aspect, the present application further provides a temperature control device for cryogenic equipment, the device comprising:

[0012] The cryogenic equipment temperature detection module is used for the temperature detection step: detecting the expected temperature information and actual temperature information of the cryogenic equipment;

[0013] A neural network processing module is used to process the expected temperature information and the actual temperature information using a pre-trained neural network when the difference between the expected temperature information and the actual temperature information is not within a preset temperature range threshold, so as to obtain a temperature regulation influencing parameter of the temperature controller;

[0014] The temperature control signal output module is used to control the temperature controller to output the temperature control signal according to the temperature adjustment influencing parameters;

[0015] The temperature adjustment module is used to return to the temperature detection step after adjusting the temperature of the low-temperature equipment according to the temperature control signal until the difference between the actual temperature information and the expected temperature information is within a preset temperature range threshold.

[0016] In a third aspect, the present application further provides a temperature control system for a cryogenic device, the system comprising:

[0017] A low-temperature device, comprising a sample stage and a temperature-averaging module;

[0018] a first temperature sensor, the first temperature sensor being disposed on the sample stage and configured to detect actual temperature information of the cryogenic device and send the actual temperature information to the control component;

[0019] A heating module, comprising a heating rod and a heating block, wherein the heating rod is arranged on the sample stage, and the heating block is arranged on the temperature equalizing module, and the heating rod and the heating block are used to adjust the temperature of the low-temperature device;

[0020] Control components for:

[0021] Temperature detection step: detecting expected temperature information and actual temperature information of the cryogenic equipment;

[0022] When the difference between the expected temperature information and the actual temperature information is not within a preset temperature range threshold, the expected temperature information and the actual temperature information are processed using a pre-trained neural network to obtain a temperature regulation influencing parameter of a temperature controller;

[0023] controlling the temperature controller to output a temperature control signal to the heating rod and the heating block according to the temperature adjustment influencing parameter;

[0024] After the heating rod and the heating block adjust the temperature of the low-temperature device according to the temperature control signal, the process returns to the temperature detection step until the difference between the actual temperature information and the expected temperature information is within a preset temperature range threshold.

[0025] In a fourth aspect, the present application further provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0026] Temperature detection step: detecting the expected temperature information and actual temperature information of the cryogenic equipment;

[0027] When the difference between the expected temperature information and the actual temperature information is not within the preset temperature range threshold, the expected temperature information and the actual temperature information are processed using a pre-trained neural network to obtain the temperature adjustment influencing parameters of the temperature controller;

[0028] According to the temperature adjustment influencing parameters, the temperature controller is controlled to output a temperature control signal;

[0029] After the temperature of the cryogenic device is adjusted according to the temperature control signal, the process returns to the temperature detection step until the difference between the actual temperature information and the expected temperature information is within a preset temperature range threshold.

[0030] In a fifth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the following steps:

[0031] Temperature detection step: detecting the expected temperature information and actual temperature information of the cryogenic equipment;

[0032] When the difference between the expected temperature information and the actual temperature information is not within the preset temperature range threshold, the expected temperature information and the actual temperature information are processed using a pre-trained neural network to obtain the temperature adjustment influencing parameters of the temperature controller;

[0033] According to the temperature adjustment influencing parameters, the temperature controller is controlled to output a temperature control signal;

[0034] After the temperature of the cryogenic device is adjusted according to the temperature control signal, the process returns to the temperature detection step until the difference between the actual temperature information and the expected temperature information is within a preset temperature range threshold.

[0035] In a sixth aspect, the present application further provides a computer program product, comprising a computer program, which, when executed by a processor, implements the following steps:

[0036] Temperature detection step: detecting the expected temperature information and actual temperature information of the cryogenic equipment;

[0037] When the difference between the expected temperature information and the actual temperature information is not within the preset temperature range threshold, the expected temperature information and the actual temperature information are processed using a pre-trained neural network to obtain the temperature adjustment influencing parameters of the temperature controller;

[0038] According to the temperature adjustment influencing parameters, the temperature controller is controlled to output a temperature control signal;

[0039] After the temperature of the cryogenic device is adjusted according to the temperature control signal, the process returns to the temperature detection step until the difference between the actual temperature information and the expected temperature information is within a preset temperature range threshold.

[0040] The temperature control method, device, system, computer device, computer-readable storage medium, and computer program product for the cryogenic equipment described above pre-train a neural network so that the neural network supports accurate detection of the temperature control influencing parameters of the temperature controller. When the difference between the expected temperature information and the actual temperature information is less than a preset temperature threshold, the expected temperature information and the actual temperature information are processed using the pre-trained neural network to obtain accurate temperature control influencing parameters of the temperature controller, thereby controlling the temperature controller to output a temperature control signal and cyclically regulating the temperature of the cryogenic equipment until the difference between the actual temperature information and the expected temperature information of the cryogenic equipment is less than the preset temperature threshold. Throughout this process, compared to directly using a PID controller to output a temperature control signal, the present application achieves higher accuracy in temperature regulation of the cryogenic equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following briefly introduces the drawings required for use in the embodiments of the present application or related technical descriptions. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying any creative work.

[0042] Figure 1 This is a diagram of an application environment of a temperature control method for cryogenic equipment in one embodiment;

[0043] Figure 2 Schematic diagram of a flow chart of a temperature control method for a cryogenic device according to an embodiment;

[0044] Figure 3 A schematic flow chart of a temperature control method for cryogenic equipment in another embodiment;

[0045] Figure 4 Schematic diagram of the structure of a BP (Back Propagation Neural Network) neural network in one embodiment;

[0046] Figure 5 A schematic diagram of the basic structure of a quantum voltage device without liquid helium refrigeration in one embodiment;

[0047] Figure 6 A flowchart of an ISSA (Improved Sparrow Search Algorithm)-BP-PID temperature control algorithm in a detailed embodiment;

[0048] Figure 7 is a structural block diagram of a temperature control device for a cryogenic device in one embodiment;

[0049] Figure 8 A schematic diagram of a temperature control system of a cryostat in a liquid helium-free quantum voltage device in a specific application embodiment;

[0050] Figure 9 Schematic diagram of the spatial distribution of temperature sensors, heating blocks, and heating rods in a specific application embodiment;

[0051] Figure 10 FIG. 1 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0052] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are used to explain this application and are not intended to limit this application.

[0053] The temperature control method of the cryogenic equipment provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown, the terminal 102 communicates with the temperature control system 104 of the cryogenic device through a network, and the temperature control system 104 of the cryogenic device is used to control the temperature of the cryogenic device 106 .

[0054] The user triggers the temperature control control of the cryogenic device 106 on the temperature control interface of the terminal 102, and the terminal 102 responds to the trigger request of the temperature control control, generates a temperature control request for the cryogenic device 106, and sends the temperature control request of the cryogenic device 106 to the temperature control system 104 of the cryogenic device. The temperature control system 104 of the cryogenic device obtains the temperature control request of the cryogenic device 106 to detect the expected temperature information and the actual temperature information of the cryogenic device 106; when the difference between the expected temperature information and the actual temperature information is not within the preset temperature range threshold, the expected temperature information and the actual temperature information are processed using a pre-trained neural network to obtain the temperature adjustment influencing parameters of the cryogenic device 106; according to the temperature adjustment influencing parameters, the temperature of the cryogenic device 106 is adjusted, and the step of detecting the expected temperature information and the actual temperature information of the cryogenic device 106 is returned until the difference between the actual temperature information and the expected temperature information is within the preset temperature range threshold.

[0055] The terminal 102 may include, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices may include smart speakers, smart TVs, smart air conditioners, smart car devices, projectors, and the like. Portable wearable devices may include smart watches, smart bracelets, head-mounted devices, and the like. Head-mounted devices may include virtual reality (VR) devices, augmented reality (AR) devices, smart glasses, and the like. The temperature control system 104 for the cryogenic equipment may be a standalone physical server, a server cluster or distributed system consisting of multiple physical servers, or a cloud server providing cloud computing services.

[0056] In an exemplary embodiment, Figure 2 As shown, a temperature control method for low temperature equipment is provided, and the method is applied to Figure 1 The temperature control system 104 of the low-temperature equipment in the embodiment is used as an example for explanation.

[0057] S100, temperature detection step: detecting expected temperature information and actual temperature information of the cryogenic equipment.

[0058] Cryogenic equipment, such as cryostat or other equipment, is used in low-temperature environments. The desired temperature information for a cryogenic equipment refers to the temperature it needs to reach, representing the optimal operating temperature. Actual temperature information refers to the actual temperature detected by the cryogenic equipment, which can be measured using devices such as temperature sensors.

[0059] Specifically, the expected temperature information of the cryogenic equipment is obtained, which may be temperature information customized by the staff or temperature information automatically adjusted according to historical temperature information; and the actual temperature information of the cryogenic equipment is detected by a temperature sensor.

[0060] Furthermore, it is detected whether the difference between the expected temperature information and the actual temperature information is within the preset temperature range threshold. If the difference between the expected temperature information and the actual temperature information is not within the preset temperature range threshold, it means that the actual temperature of the cryogenic equipment is abnormal, and the temperature of the cryogenic equipment needs to be adjusted so that the actual temperature information of the cryogenic equipment is close to the expected temperature information; if the difference between the expected temperature information and the actual temperature information is within the preset temperature range threshold, it means that the actual temperature of the cryogenic equipment is normal, and there is no need to adjust the temperature of the cryogenic equipment.

[0061] More specifically, when the difference between the expected temperature information and the actual temperature information is not within the preset temperature range threshold, if the difference between the expected temperature information and the actual temperature information is positive, that is, the expected temperature information is greater than the actual temperature information, then the low-temperature equipment needs to be controlled to increase the temperature; if the difference between the expected temperature information and the actual temperature information is negative, that is, the expected temperature information is less than the actual temperature information, then the low-temperature equipment needs to be controlled to decrease the temperature. In actual applications, the low-temperature equipment is generally controlled to increase the temperature.

[0062] S200 , when the difference between the expected temperature information and the actual temperature information is not within the preset temperature range threshold, the expected temperature information and the actual temperature information are processed using a pre-trained neural network to obtain a temperature adjustment influencing parameter of the temperature controller.

[0063] Among them, the neural network is used to detect the temperature regulation influencing parameters of cryogenic equipment.

[0064] Specifically, when the difference between the expected temperature information and the actual temperature information is greater than or equal to the preset temperature threshold, since the network parameters of the neural network are already pre-trained network parameters at this time, the neural network at this time will support more accurate temperature regulation influencing parameter detection. The expected temperature information and the actual temperature information can be input into the pre-trained neural network, and the pre-trained neural network will output the temperature regulation influencing parameters of the temperature controller.

[0065] Among them, the temperature regulation influencing parameters are device parameter information related to the temperature controller for temperature regulation of the low-temperature equipment. For example, when the temperature controller for temperature regulation of the low-temperature equipment is a PID controller, the temperature regulation influencing parameters can be the integral influence information, differential influence information and proportional influence information of the PID controller, etc. In addition, when the temperature controller for temperature regulation of the low-temperature equipment is other controllers, the temperature regulation influencing parameters can be a control quantity of other controllers.

[0066] In one embodiment, pre-training the neural network is essentially initializing the network parameters of the neural network and then iteratively adjusting the initialized initial network parameters.

[0067] Common network parameters in neural networks include at least weights, biases, learning rates, and activation functions. Weights are parameters that connect neurons and control the strength of input signals. Biases are additional parameters for each neuron and are used to adjust the output of activation functions. Learning rates are hyperparameters that control the step size of parameter updates in optimization algorithms (such as gradient descent). Activation functions introduce nonlinear characteristics, enabling neural networks to learn complex patterns.

[0068] Specifically, in order to more accurately utilize neural networks to detect temperature regulation influencing parameters of cryogenic equipment, the present application can update the network parameters of the neural network, that is, provide the neural network with optimal network parameters and enhance the neural network's ability to fit complex nonlinear relationships. Before updating the network parameters of the neural network, the network parameters of the neural network must first be initialized to obtain the initial network parameters of the neural network. At the same time, not only the neural network needs to be initialized, but also its optimization algorithm for the network parameters.

[0069] Then, various network parameter optimization algorithms are used to iteratively update the initial network parameters to obtain the optimal network parameters. In practical applications, network parameter optimization algorithms can be SSA (Sparrow Search Algorithm), ISSA algorithm, particle optimization algorithm, etc., which are not limited here.

[0070] Then, according to the network parameters in the optimal state, the network parameters of the neural network are adjusted, that is, the network parameters in the optimal state are determined as the network parameters of the neural network, and the neural network after adjusting the network parameters is obtained. At this time, the pre-training of the neural network is completed.

[0071] In practical applications, if the neural network is a BP neural network, the network parameters can be the weights and thresholds of the BP neural network, including the weights connecting the input layer to the hidden layer in the BP neural network, the threshold of the hidden layer neurons, the weights connecting the hidden layer to the output layer, and the output layer threshold.

[0072] S300 , controlling the temperature controller to output a temperature control signal according to the temperature adjustment influencing parameter.

[0073] Specifically, the temperature controller may be adjusted according to the temperature adjustment influencing parameter so that the temperature controller outputs a temperature control signal, and the temperature of the cryogenic device is adjusted according to the temperature control signal.

[0074] Furthermore, the temperature control signal can be a heating signal or a cooling signal. When the temperature control signal is a heating signal, the low-temperature equipment can be heated using a heating block, such as a heating block and a heating rod; when the temperature control signal is a cooling signal, the low-temperature equipment can be cooled using a cooling device, such as a refrigerator.

[0075] S400, after adjusting the temperature of the cryogenic device according to the temperature control signal, return to the temperature detection step until the difference between the actual temperature information and the expected temperature information is within a preset temperature range threshold.

[0076] Specifically, if after the temperature adjustment is completed, when the expected temperature information and the actual temperature information of the low-temperature equipment are re-detected, the difference between the expected temperature information and the actual temperature information is still not within the preset temperature range threshold, the above steps of using the pre-trained neural network to process the expected temperature information and the actual temperature information can be repeated until the difference between the actual temperature information and the expected temperature information of the low-temperature equipment is within the preset temperature range threshold.

[0077] After the temperature adjustment is completed, if the difference between the expected temperature information and the actual temperature information of the low-temperature device is within the preset temperature range threshold when the expected temperature information and the actual temperature information are re-detected, there is no need to adjust the temperature of the low-temperature device again.

[0078] In the temperature control method for cryogenic equipment described above, a neural network is pre-trained so that the neural network supports accurate detection of the temperature control parameters affecting the temperature controller. When the difference between the desired temperature information and the actual temperature information is less than a preset temperature threshold, the desired temperature information and the actual temperature information are processed using the pre-trained neural network to obtain accurate temperature control parameters affecting the temperature controller. This controls the temperature controller to output a temperature control signal and cyclically adjusts the temperature of the cryogenic equipment until the difference between the actual temperature information and the desired temperature information of the cryogenic equipment is less than the preset temperature threshold. Throughout this process, compared to directly using a PID controller to output a temperature control signal, the present application achieves higher accuracy in temperature regulation of cryogenic equipment.

[0079] In an exemplary embodiment, Figure 3 As shown, S200 includes:

[0080] S220 , when the difference between the expected temperature information and the actual temperature information is not within the preset temperature range threshold, detecting temperature error information of the cryogenic device according to the expected temperature information and the actual temperature information.

[0081] S240 , processing the expected temperature information, the actual temperature information, and the temperature error information according to the pre-trained neural network to obtain temperature regulation influencing parameters of the temperature controller.

[0082] Specifically, the temperature control influencing parameters are affected not only by the desired temperature information and the actual temperature information, but also by the difference between the desired and actual temperature information. Therefore, the difference between the desired and actual temperature information can be detected and used as the temperature error information of the cryogenic equipment. The desired temperature information, actual temperature information, and temperature error information are input into a pre-trained neural network for processing to obtain the temperature control influencing parameters of the temperature controller. The temperature control influencing parameters are then used to control the temperature controller to output a temperature control signal to adjust the temperature of the cryogenic equipment.

[0083] Taking the BP neural network as an example and the PID controller as the temperature controller, the BP neural network adopts a three-layer structure, in which the input layer contains 3 neurons, the hidden layer contains 5 neurons, and the output layer contains 3 neurons. The input of the BP neural network is the actual temperature information y(k) of the cryostat, the temperature error information e(k), and the set desired temperature information. The output of the BP neural network is the influencing parameter K of the dynamically adjusted PID controller. P , K I , K D In the temperature control process, during forward propagation, the input signal is nonlinearly mapped by the hidden layer and then outputs the PID parameters, driving the PID controller to generate the temperature control signal u(k); during reverse propagation, the square of the error is used as the performance indicator, and the weights of the BP neural network are updated by gradient descent combined with the inertia term. The structural diagram of the BP neural network used in this application is as follows Figure 4 As shown, h i is the state representation of the hidden layer.

[0084] In the above embodiment, not only the expected temperature information and the actual temperature information can be input into the pre-trained neural network, but also the temperature error information can be input into the pre-trained neural network, so that the pre-trained neural network can process the expected temperature information, the actual temperature information and the temperature error information to obtain accurate temperature regulation influencing parameters of the temperature controller.

[0085] In an exemplary embodiment, the training process corresponding to the pre-trained neural network includes:

[0086] Detection step: determine the current number of iterations, and based on the current number of iterations, detect the search weight information corresponding to the current network parameters of the neural network; based on the search weight information, update the current network parameters to obtain new network parameters; if the neural network meets the preset iteration termination conditions, the neural network with the new network parameters is determined as the pre-trained neural network; if the neural network does not meet the preset iteration termination conditions, the new network parameters are used as the current network parameters again, and return to execute the detection step until the neural network meets the preset iteration termination conditions.

[0087] Specifically, as mentioned in the above embodiments, the network parameter optimization algorithm may be an SSA algorithm (Sparrow Search Algorithm), an ISSA algorithm (Improved Sparrow Search Algorithm), a particle optimization algorithm, or the like.

[0088] The Sparrow Search Algorithm (SSA) primarily optimizes by mimicking the foraging strategies of sparrow colonies. It achieves high efficiency while ensuring a good global search. The population is divided into discoverers, joiners, and watchers based on fitness. However, when dealing with complex, high-dimensional optimization problems, the SSA algorithm is prone to becoming trapped in local optimal solutions, resulting in reduced search efficiency. Therefore, to enhance the global optimization capabilities of temperature control algorithms for cryogenic equipment, this application proposes an improved Sparrow Search Algorithm.

[0089] The algorithm adjusts the weight of the global search, which is defined as an exponential decay calculation method and can be determined by the current number of iterations of the neural network. That is, according to the current number of iterations, the search weight information of the current network parameters of the neural network in the current iteration process can be detected.

[0090] Specifically, the calculation expression of the search weight information w is as follows:

[0091]

[0092] Where t represents the current iteration number of the neural network.

[0093] As can be seen, the search weight information w is a dynamically adjusted strategy that exponentially decays with the number of iterations. As iterations progress, w is initially larger, enhancing the random search capabilities of individual sparrows (searchers), helping the algorithm escape local optima and explore the solution space more broadly. As iterations proceed, w gradually decreases, shortening the individual sparrows' movement steps. The algorithm then focuses more on a detailed search within the neighborhood of the current optimal solution. In other words, this weight update method gradually reduces the searcher's search pace during the development phase. This approach ensures extensive exploration in the early stages of the algorithm while focusing on local optimization in the later stages.

[0094] Therefore, based on the search weight information corresponding to the current network parameters of the neural network, the current network parameters in the current iteration process can be updated to obtain new network parameters.

[0095] Finally, when the neural network meets the preset iteration termination condition, the iteration of the neural network is stopped, and the neural network with the latest network parameters is determined as the pre-trained neural network; when the neural network does not meet the preset iteration termination condition, the new network parameters are used as the current network parameters, and the detection step is returned to execute the next iteration process until the neural network meets the preset iteration termination condition.

[0096] In practical applications, the preset iteration termination condition may be at least one of the following: network parameter convergence, the accuracy of the neural network reaches a preset accuracy threshold, and the number of iterations of the neural network reaches a preset maximum number of iterations.

[0097] In the above embodiment, the network parameters of the neural network are iteratively adjusted so that the network parameters of the neural network are adjusted more accurately. Therefore, the neural network supports accurate detection of the temperature adjustment influencing parameters of the temperature controller. Furthermore, through the current number of iterations, the search weight information corresponding to the current network parameters of the neural network can be obtained, so that based on the search weight information, it can be ensured that the update of the current network parameters can be widely searched in the early stage and focused on local optimization in the later stage, thereby obtaining accurate new network parameters.

[0098] In an exemplary embodiment, the current network parameters are updated according to the search weight information to obtain new network parameters, including:

[0099] Obtain a current network parameter population of a current iteration process, and detect a current optimal network parameter from the current network parameter population; for each current network parameter in the current network parameter population, obtain current network parameter distance information based on the current optimal network parameter and the current network parameter; detect update control information of the current network parameter based on the current number of iterations and the maximum number of iterations; update the current network parameter distance information based on the update control information, and update the current network parameter based on the search weight information; and obtain new network parameters based on the updated current network parameter and the updated current network parameter distance information.

[0100] Specifically, the improved sparrow search algorithm optimizes parameters by mimicking the foraging strategy of a sparrow colony. Therefore, during each iteration of the neural network, the initial network parameters mimic the foraging behavior of a sparrow colony, resulting in multiple randomly updated network parameters, known as a network parameter population. During this iteration, the current network parameter population can be obtained and the optimal network parameters can be detected from this population.

[0101] For each current network parameter in the current network parameter population, the current network parameter distance information of each population individual in the current iteration process is obtained by calculating the difference between the current optimal network parameter and the current network parameter.

[0102] Furthermore, when obtaining the current network parameter distance information, a first random value can be generated within the interval [0, 2π] to reflect the impact of the current optimal network parameters on the discoverer, i.e., the current network parameters, during the current iteration. Therefore, the current network parameter distance information can be determined by adjusting the current optimal network parameters with the first random value and then subtracting the difference from the current network parameters.

[0103] Furthermore, in order to enhance the balance between exploration and development, the update control information r1 of the current network parameters is detected according to the current number of iterations and the maximum number of iterations. The specific expression is:

[0104]

[0105] Where: t means the tth iteration is currently being performed, The above formula allows the search strategy to be continuously adjusted during iteration, reducing the risk of the algorithm falling into a local optimum and achieving a balance between accuracy and search range.

[0106] Then, based on the search weight information, the current network parameters in the current iteration are updated, retaining some of the current network parameters' own information. This avoids completely discarding historical search experience during the update process, making the search process smoother. Furthermore, based on the update control information, the distance information of the current network parameters is updated. Finally, based on the updated current network parameters and the updated distance information of the current network parameters, the new network parameters corresponding to each individual in the population are obtained.

[0107] In the above embodiment, the current network parameters in the current iteration process are updated according to the search weight information, and the update control information of the current network parameters in the current iteration process is detected according to the current number of iterations and the maximum number of iterations, so as to update the current network parameter distance information based on the current optimal network parameters and the current network parameters, so that new network parameters are finally obtained based on the updated current network parameters and the updated current network parameter distance information, and the new network parameters retain some of the current network parameters and are close to the current optimal network parameters.

[0108] In an exemplary embodiment, updating the current network parameter distance information according to the update control information includes:

[0109] Obtain the environmental random value associated with the current network parameter distance information when it is updated; when the environmental random value is less than the preset environmental alert value, update the current network parameter distance information using the global search function according to the update control information; when the environmental random value is greater than or equal to the preset environmental alert value, update the current network parameter distance information using the local search function according to the update control information.

[0110] Specifically, the network parameter distance information can be updated by combining a global search function and a local search function to dynamically adjust the search direction and step size. More specifically, the environment random value associated with the current network parameter distance information update is obtained, and the current network parameter distance information is updated in sections based on the environment random value. That is, when the environment random value is less than a preset environment alert value, the current network parameter distance information is updated using the global search function according to the update control information; when the environment random value is greater than or equal to the preset environment alert value, the current network parameter distance information is updated using the local search function according to the update control information.

[0111] The environmental random value is a random number uniformly distributed in the interval [0, 1], representing the uncertainty of the environment (such as the probability of natural enemies appearing, the randomness of food distribution, etc.). The preset environmental alert value is usually set to , such as 0.5, 0.6, 0.8, etc., are the sensitivity thresholds of network parameters to risks, which are preset or dynamically adjusted by the algorithm.

[0112] In practice, the global search function is a sine function, and the local search function is a cosine function. This means that when the environmental random value is less than the preset environmental alert value, the sine function is used to update the current network parameter distance information, and the algorithm conducts a more extensive exploration. When the environmental random value is greater than or equal to the preset environmental alert value, the cosine function is used to update the current network parameter distance information, and the algorithm is in a local optimization state. This method balances comprehensive search of the understanding space with fine-tuning of local areas, ensuring a balance between diversity and convergence.

[0113] In an exemplary embodiment, the current network parameters are updated according to the search weight information to obtain new network parameters, including:

[0114] The present invention provides a method for obtaining a current network parameter population of a current iteration process, detecting a current optimal network parameter from the current network parameter population; obtaining current network parameter distance information based on the current optimal network parameter and the current network parameter for each current network parameter in the current network parameter population; detecting update control information of the current network parameter based on the current number of iterations and the maximum number of iterations; obtaining an environmental random value associated with the update of the current network parameter distance information; when the environmental random value is less than a preset environmental alert value, updating the current network parameter distance information using a global search function according to the update control information; when the environmental random value is greater than or equal to the preset environmental alert value, updating the current network parameter distance information using a local search function according to the update control information; and updating the current network parameter according to the search weight information; obtaining a new network parameter based on the updated current network parameter and the updated current network parameter distance information.

[0115] For example, the new network parameters The update expression is:

[0116]

[0117] Where: w is the search weight information, is the position information of the mth data in the nth dimension at the tth iteration, that is, the current network parameters in the current iteration process, r1 is the update control information, and r3 is the first random value in the interval [0, 2π] to reflect the current optimal network parameters The impact on the position of the discoverer, r2 represents a second random value generated in the interval [0, 2π], which is used to adjust the pace of the discoverer. When r2 represents a second random value generated in the interval [0, 2π], its value range is , alternating between positive and negative. When r2 is in the range [0,π], sin(r2) is positive, guiding the individual toward the optimal position; when r2 is in the range [π,2π], sin(r2) is negative, guiding the individual toward the direction away from the optimal position, making the search direction have a periodic oscillation characteristic and preventing the algorithm from falling into a local optimum. R2 is an environmental random value between 0 and 1, ST is a preset environmental alert value, sin() is the global search function, and cos() is the local search function.

[0118] In the above embodiments, the ISSA algorithm proposed in this application maintains low computational complexity and fast convergence while effectively reducing the risk of falling into local optimum. At the same time, it can also better determine the network parameters of the neural network.

[0119] In an exemplary embodiment, controlling the temperature controller to output a temperature control signal according to the temperature adjustment influencing parameter includes:

[0120] According to the temperature regulation influencing parameters, the current temperature control information of the temperature controller during the current temperature detection cycle is detected, and the previous temperature control information during the previous temperature detection cycle is obtained; according to the temperature control change between the current temperature control information and the previous temperature control information, the temperature controller is controlled to output a temperature control signal.

[0121] Specifically, the temperature control parameters output by the neural network serve as influencing parameters when the temperature controller outputs temperature control information. Furthermore, the temperature controller detects the current temperature control information of the cryogenic device during the current temperature detection cycle based on the temperature control parameters and outputs a temperature control signal.

[0122] This application generally uses incremental adjustment when outputting a temperature control signal using the current temperature control information. That is, the incremental adjustment method used does not require the accumulation of temperature control information from all temperature detection cycles, but is only related to the error amount of the last three temperature detection cycles. In this case, it is necessary to obtain the previous temperature control information from the previous temperature detection cycle, and obtain the temperature control change based on the difference between the current temperature control information and the previous temperature control information. In this way, the temperature controller is controlled to output the temperature control signal based on the temperature control change.

[0123] The expression of its control law is: .

[0124] Among them, u(k) is the current temperature control information during the k-th temperature detection cycle, u(k-1) is the previous temperature control information during the k-1-th temperature detection cycle, is the difference between the current temperature control information during the kth temperature detection cycle and the previous temperature control information during the k-1th temperature detection cycle, that is, the temperature control information in the temperature control signal.

[0125] In the above embodiment, by controlling the temperature control change between the current temperature control information in the current temperature detection cycle and the previous temperature control information in the previous temperature detection cycle, the temperature controller can be controlled to output an accurate temperature control signal to accurately adjust the temperature of the low-temperature equipment.

[0126] In an exemplary embodiment, the temperature adjustment influence parameters include integral influence information, differential influence information, and proportional influence information; detecting current temperature control information of the temperature controller during the current temperature detection cycle according to the temperature adjustment influence parameters includes:

[0127] During the current temperature detection cycle, the temperature error information between the actual temperature information and the expected temperature information is detected, and the previous temperature error information between the previous actual temperature information and the expected temperature information in the previous temperature detection cycle, as well as the historical temperature cumulative error information from the first temperature detection cycle to the current temperature detection cycle are obtained; the actual temperature information is processed according to the proportional influence information to obtain the first temperature control information; the historical temperature cumulative error information is processed according to the integral influence information to obtain the second temperature control information; the difference between the temperature error information and the previous temperature error information is processed according to the differential influence information to obtain the third temperature control information; according to the first temperature control information, the second temperature control information and the third temperature control information, the current temperature control information of the temperature controller during the current temperature detection cycle is detected.

[0128] Specifically, in the current temperature detection cycle, the temperature error information between the actual temperature information and the expected temperature information is detected, and the previous temperature error information between the previous actual temperature information and the expected temperature information in the previous temperature detection cycle is obtained, as well as the historical temperature cumulative error information from the first temperature detection cycle to the current temperature detection cycle;

[0129] Furthermore, based on the temperature adjustment influencing parameters, the temperature error information, the previous temperature error information, and the historical temperature accumulated error information are processed to obtain the current temperature control information of the temperature controller during the current temperature detection cycle.

[0130] In the case where the temperature controller is a PID controller, the temperature regulation influencing parameters may include integral influence information, differential influence information, and proportional influence information of the temperature controller.

[0131] Therefore, the actual temperature information can be processed according to the proportional influence information to obtain the first temperature control information; the historical temperature accumulated error information can be processed according to the integral influence information to obtain the second temperature control information; and the difference between the temperature error information and the previous temperature error information can be processed according to the differential influence information to obtain the third temperature control information.

[0132] Furthermore, current temperature control information of the temperature controller during the current temperature detection cycle is determined according to the first temperature control information, the second temperature control information, and the third temperature control information.

[0133] The specific expression for determining the current temperature control information can be:

[0134]

[0135] Among them, e(k) is the temperature error information between the actual temperature information and the expected temperature information during the current k-th temperature detection cycle; e(k-1) is the temperature error information between the previous actual temperature information and the expected temperature information during the previous temperature detection cycle, that is, the k-1-th temperature detection cycle; e(j) is the historical temperature cumulative error information from the first temperature detection cycle to the current k-th temperature detection cycle, that is, all error sequences from system startup to the current moment, and the accumulation from j=0 to k reflects the core idea of ​​control, that is, eliminating steady-state deviations through the accumulation of historical errors; K P is the proportional coefficient, which determines the system's response to the current error change. is the first temperature control information, K I is the integral coefficient, used to eliminate steady-state errors, is the second temperature control information, T I is the integration time constant, K D is the differential coefficient, reflecting the error change trend, is the third temperature control information, T D is the differential time constant, and T is the sampling period.

[0136] In the above embodiment, the output of the neural network is used as the parameter optimization value of the PID controller to adjust the temperature control information output by the PID controller, so that the temperature control information output by the PID controller is more accurate.

[0137] The following describes the temperature control method of the cryostat in a quantum voltage device without liquid helium refrigeration based on the Programmable Josephson Voltage Standard (PVJS) as an example using a specific application example.

[0138] The basic structure of the quantum voltage device without liquid helium refrigeration is shown in the figure below. Figure 5 As shown, it includes a low-temperature thermostat, DC voltage precision measurement, bias current source, microwave radio frequency module, temperature control circuit, control system and software, vacuum molecular pump and He helium compressor, etc., wherein the low-temperature thermostat includes at least a junction array and a refrigerant cold head. More specifically, the junction array is a chip junction array, and the refrigerant cold head includes a first-level cold head and a second-level cold head.

[0139] Quantum voltage devices based on liquid helium-free refrigeration face two challenges due to inherent temperature fluctuations in their primary and secondary cold heads. Furthermore, the heat transfer path uses solid-state materials, which are affected by the thermal inertia of the materials. First, the periodic temperature fluctuations of the cold heads are transferred layer by layer along the heat conduction path to the chip, causing chip temperature fluctuations. Second, the heat generated by the chip during operation (such as the thermal effects of microwave power and bias signals) must pass through several layers of solid-state materials before reaching the cold heads. This slow heat diffusion leads to localized accumulation, resulting in a temperature gradient between the chip and the cold heads. This combined effect of temperature unevenness and fluctuations significantly impacts the phase stability of the quantum voltage signal. Therefore, minimizing temperature fluctuations in the cryostat of a liquid helium-free refrigeration system is crucial for achieving stable quantum voltage signal output. Therefore, temperature control methods for the cryostat in liquid helium-free quantum voltage devices are crucial.

[0140] PID controller technology is currently commonly used for temperature control. However, in traditional PID control methods, once the control parameters are fixed, it is difficult for the system to adapt to the dynamic changes of the controlled object in real time. As a result, the accuracy of temperature regulation control is often not ideal and the response speed is slow.

[0141] In order to solve the above problems and improve the speed, stability and accuracy of temperature control of the cryostat in the liquid helium-free quantum voltage device, this application proposes a temperature control method for the cryostat in the liquid helium-free quantum voltage device based on the ISSA-BP-PID algorithm. The sparrow search algorithm is optimized using an improved strategy to enhance its global search capability and form an improved sparrow search algorithm. The ISSA is used to provide the optimal initial weights and thresholds for the BP neural network, thereby enhancing the BP neural network's ability to fit complex nonlinear relationships. The ISSA-BP algorithm is combined to optimize and adjust the core parameter K of the incremental PID controller. P, K I , K D , thereby improving the control accuracy and stability of the liquid helium-free quantum voltage cryogenic system. Simulation and experimental results show that the algorithm can control temperature fluctuations within ±1mK within the temperature range of 3.8K to 20K.

[0142] The flow chart of the ISSA-BP-PID temperature control algorithm is as follows Figure 6 shown.

[0143] Specifically, the parameters of the ISSA optimization algorithm (population size, maximum number of iterations, etc.) and the network parameters of the BP neural network (learning rate, number of iterations, etc.) are first initialized, and then the ISSA optimization algorithm is used to find the optimal network parameters of the BP neural network, such as the optimal weights and thresholds.

[0144] The expected temperature information and actual temperature information of the cryostat in the liquid helium-free quantum voltage device are then sampled, and the temperature error information formed between the expected temperature information and the actual temperature information is calculated. When the difference between the expected temperature information and the actual temperature information does not meet the requirements, the expected temperature information, the actual temperature information, and the temperature error information are input into the BP neural network for processing. The BP neural network outputs the temperature adjustment influencing parameter K of the PID controller. P , K I , K D .

[0145] The PID controller is an incremental control, that is, the influencing parameter K is adjusted according to the temperature P , K I , K D After detecting the current temperature control information of the temperature controller during the current temperature detection cycle, it is also necessary to obtain the previous temperature control information during the previous temperature detection cycle; according to the temperature control change between the current temperature control information and the previous temperature control information, the temperature controller is controlled to output a temperature control signal, so as to control the low-temperature thermostat in the liquid helium-free quantum voltage device based on the temperature control signal to continuously adjust the temperature of the refrigerator or the temperature of the heating rod and the heating block, so as to control the temperature of the low-temperature thermostat in the liquid helium-free quantum voltage device.

[0146] It is determined whether the difference between the expected temperature information and the actual temperature information after this temperature control meets the requirements. If so, the temperature control is terminated. If not, the process returns to the step of sampling the expected temperature information and the actual temperature information of the cryostat in the liquid helium-free quantum voltage device, and a new round of temperature control is performed again until the difference between the expected temperature information and the actual temperature information meets the requirements.

[0147] Furthermore, for the ISSA optimization algorithm, the optimal initial values ​​of the BP neural network are determined by the improved ISSA algorithm in this paper. This paper proposes an improved sparrow search algorithm (ISSA) to solve the initial weights and thresholds of the BP neural network. Its core is to simulate the foraging and vigilance behavior of a sparrow colony, assigning the network parameters corresponding to the final sparrow position that converges to the optimal fitness as the initial weights and thresholds to the BP neural network (the weights connecting the input layer to the hidden layer, the thresholds of the hidden layer neurons, the weights connecting the hidden layer to the output layer, and the output layer threshold). The specific process can be as follows:

[0148] First, the parameters of the sparrow search algorithm are initialized. The fitness of individual sparrows is then calculated to update their positions. Based on the current iteration number, the algorithm then checks the search weights for the initial network parameters during the current iteration. Based on this search weight information, the algorithm dynamically adjusts the search direction and step size using sine and cosine functions to determine the optimal network parameters for the current iteration. This process continues until the maximum number of iterations is reached. The algorithm then outputs the optimal network parameters from the last iteration and updates the global optimal fitness value.

[0149] By means of the temperature control method of the cryostat in the liquid helium-free quantum voltage device, the rapidity, stability and accuracy of the temperature control of the cryostat in the liquid helium-free quantum voltage device are improved.

[0150] It should be understood that, although the steps in the flowcharts of the above embodiments are shown in sequence as indicated by the arrows, these steps are not necessarily performed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps can be performed in other orders. Moreover, at least a portion of the steps in the flowcharts of the above embodiments may include multiple steps or multiple stages, and these steps or stages are not necessarily performed at the same time, but can be performed at different times. The execution order of these steps or stages is not necessarily to be performed in sequence, but can be performed in turn or alternately with other steps or at least a portion of steps or stages in other steps.

[0151] Based on the same inventive concept, embodiments of the present application also provide a cryogenic device temperature control apparatus for implementing the aforementioned cryogenic device temperature control method. The solution provided by this apparatus is similar to the solution described in the aforementioned method. Therefore, the specific limitations of one or more cryogenic device temperature control apparatus embodiments provided below can be found in the aforementioned limitations of the cryogenic device temperature control method, and will not be further elaborated here.

[0152] In an exemplary embodiment, Figure 7As shown, a temperature control device for cryogenic equipment is provided, comprising: a cryogenic equipment temperature detection module 100, a neural network processing module 200, a temperature control signal output module 300 and a temperature adjustment module 400, wherein:

[0153] The cryogenic equipment temperature detection module 100 is used for the temperature detection step: detecting the expected temperature information and the actual temperature information of the cryogenic equipment;

[0154] A neural network processing module 200 is configured to process the desired temperature information and the actual temperature information using a pre-trained neural network to obtain a temperature control influencing parameter of the temperature controller when the difference between the desired temperature information and the actual temperature information is not within a preset temperature range threshold;

[0155] The temperature control signal output module 300 is used to control the temperature controller to output a temperature control signal according to the temperature adjustment influencing parameter;

[0156] The temperature adjustment module 400 is configured to, after adjusting the temperature of the cryogenic device according to the temperature control signal, return to the temperature detection step until the difference between the actual temperature information and the expected temperature information is within a preset temperature range threshold.

[0157] In one embodiment, the neural network processing module 200 is also used to detect the temperature error information of the low-temperature equipment based on the expected temperature information and the actual temperature information; according to the pre-trained neural network, the expected temperature information, the actual temperature information and the temperature error information are processed to obtain the temperature adjustment influencing parameters of the temperature controller.

[0158] In one embodiment, the temperature control device of the cryogenic equipment further includes a neural network parameter adjustment module, which is used for the detection step: determining the current number of iterations, and detecting the search weight information corresponding to the current network parameters of the neural network based on the current number of iterations; updating the current network parameters based on the search weight information to obtain new network parameters; if the neural network meets the preset iteration termination conditions, the neural network with the new network parameters is determined as a pre-trained neural network; if the neural network does not meet the preset iteration termination conditions, the new network parameters are re-used as the current network parameters, and the detection step is returned to be executed until the neural network meets the preset iteration termination conditions.

[0159] In one embodiment, the neural network parameter adjustment module is further used to obtain the current network parameter population of the current iterative process, and detect the current optimal network parameter from the current network parameter population; for each current network parameter in the current network parameter population, obtain the current network parameter distance information based on the current optimal network parameter and the current network parameter; detect the update control information of the current network parameter based on the current number of iterations and the maximum number of iterations; update the current network parameter distance information based on the update control information, and update the current network parameter based on the search weight information; and obtain the new network parameter based on the updated current network parameter and the updated current network parameter distance information.

[0160] In one embodiment, the neural network parameter adjustment module is also used to obtain the environmental random value associated with the current network parameter distance information when it is updated; when the environmental random value is less than the preset environmental alert value, the current network parameter distance information is updated using the global search function according to the update control information; when the environmental random value is greater than or equal to the preset environmental alert value, the current network parameter distance information is updated using the local search function according to the update control information.

[0161] In one embodiment, the temperature control signal output module 300 is also used to detect the current temperature control information of the temperature controller during the current temperature detection cycle according to the temperature adjustment influencing parameters, and obtain the previous temperature control information during the previous temperature detection cycle; according to the temperature control change between the current temperature control information and the previous temperature control information, control the temperature controller to output a temperature control signal.

[0162] In one embodiment, the temperature regulation influencing parameters include integral influence information, differential influence information and proportional influence information; the temperature control signal output module 300 is also used to detect the temperature error information between the actual temperature information and the expected temperature information during the current temperature detection cycle, and obtain the previous temperature error information between the previous actual temperature information and the expected temperature information during the previous temperature detection cycle, as well as the historical temperature cumulative error information from the first temperature detection cycle to the current temperature detection cycle; according to the proportional influence information, the actual temperature information is processed to obtain the first temperature control information; according to the integral influence information, the historical temperature cumulative error information is processed to obtain the second temperature control information; according to the differential influence information, the difference between the temperature error information and the previous temperature error information is processed to obtain the third temperature control information; according to the first temperature control information, the second temperature control information and the third temperature control information, the current temperature control information of the temperature controller during the current temperature detection cycle is detected.

[0163] Each module in the temperature control device of the cryogenic equipment described above may be implemented in whole or in part through software, hardware, or a combination thereof. Each module may be embedded in or independent of a processor in a computer device in the form of hardware, or may be stored in a memory in the computer device in the form of software, so that the processor can call and execute the corresponding operations of each module.

[0164] In an exemplary embodiment, a temperature control system for a cryogenic device is further provided, the system comprising:

[0165] A low-temperature device, comprising a sample stage and a temperature-averaging module;

[0166] a first temperature sensor, the first temperature sensor being disposed on the sample stage and configured to detect actual temperature information of the cryogenic device and send the actual temperature information to the control component;

[0167] A heating module, comprising a heating rod and a heating block, wherein the heating rod is arranged on the sample stage, and the heating block is arranged on the temperature equalizing module, and the heating rod and the heating block are used to adjust the temperature of the low-temperature device;

[0168] Control components for:

[0169] Temperature detection step: detecting expected temperature information and actual temperature information of the cryogenic equipment;

[0170] When the difference between the expected temperature information and the actual temperature information is not within a preset temperature range threshold, the expected temperature information and the actual temperature information are processed using a pre-trained neural network to obtain a temperature regulation influencing parameter of a temperature controller;

[0171] controlling the temperature controller to output a temperature control signal to the heating rod and the heating block according to the temperature adjustment influencing parameter;

[0172] After the heating rod and the heating block adjust the temperature of the low-temperature device according to the temperature control signal, the process returns to the temperature detection step until the difference between the actual temperature information and the expected temperature information is within a preset temperature range threshold.

[0173] Specifically, the low-temperature device includes a sample table and a temperature equalizing module. The sample table is equipped with a first temperature sensor and a heating rod. The heating rod is used to adjust the temperature of the low-temperature device under the action of the temperature control signal output by the temperature controller. The temperature equalizing module is also equipped with a heating block. The heating block is used to adjust the temperature of the low-temperature device under the action of the temperature control signal output by the temperature controller. The temperature adjustment here generally refers to heating the low-temperature device. The first temperature sensor is used to detect the actual temperature information of the low-temperature device and send the detected actual temperature information of the low-temperature device to the control component.

[0174] The control component obtains the actual temperature information sent by the first temperature sensor, and obtains the expected temperature information for detecting the low-temperature device. When the difference between the expected temperature information and the actual temperature information is not within the preset temperature range threshold, the expected temperature information and the actual temperature information are processed using a pre-trained neural network to obtain the temperature adjustment influencing parameters of the temperature controller; according to the temperature adjustment influencing parameters, the temperature controller is controlled to output a temperature control signal to the heating rod and the heating block; after the heating rod and the heating block adjust the temperature of the low-temperature device according to the temperature control signal, the temperature detection step is returned until the difference between the actual temperature information and the expected temperature information is within the preset temperature range threshold. The specific steps are described in the above method embodiments and will not be repeated here.

[0175] In the above embodiment, by providing the first temperature sensor, the actual temperature information of the sample stage can be accurately detected, and by providing the heating module, the low-temperature device can be effectively heated under the control of the temperature regulator.

[0176] In an exemplary embodiment, the system further includes a second temperature sensor installed in the temperature equalization module, the second temperature sensor being used to detect comparative temperature information of the cryogenic device and send the comparative temperature information to the control component;

[0177] The control component is further used to:

[0178] The temperature regulation effect of the cryogenic device is detected according to the actual temperature information and the comparison temperature information.

[0179] Specifically, the temperature control system of the cryogenic device is provided with two temperature sensors, the first temperature sensor is provided on the sample stage of the cryogenic device, and the second temperature sensor is provided on the temperature equalization module of the cryogenic device.

[0180] The first temperature sensor detects the actual temperature information of the sample table, and the second temperature sensor detects the comparative temperature information of the temperature equalization module. There is a gradient relationship between the actual temperature information and the comparative temperature information. The temperature regulation effect of the low-temperature equipment is detected through the gradient relationship between the actual temperature information and the comparative temperature information. When the gradient is small, the temperature regulation effect of the high-temperature equipment is better.

[0181] In the above embodiment, by providing a second temperature sensor installed in the temperature equalization module, the temperature regulation effect of the low-temperature equipment can be accurately detected.

[0182] like Figure 8 As shown, taking the temperature control system of a cryostat in a liquid helium-free quantum voltage device as an example, the temperature of the cryostat in the liquid helium-free quantum voltage device is controlled by the temperature control system.

[0183] There are two temperature sensors in the figure. Temperature sensor 1 is placed on the sample stage, and temperature sensor 2 is placed on the temperature equalization module for measuring temperature. The chip array is placed on the sample stage and connected to temperature sensor 1. The sample stage and the temperature equalization module are both placed in a low-temperature thermostat. Among them, the temperature equalization module is made of high-purity indium sheet, low-temperature grease and aluminum-coated PET film. The function of the temperature equalization module is to reduce the temperature gradient between the secondary cold head and the sample stage; a heating block is placed on the temperature equalization module, and the other heating rod is placed in the center of the sample stage, as an actuator to increase the temperature; the driving circuit serves as the driving mechanism for temperature sensor 1, temperature sensor 2, heating block, and heating rod; the main control system performs calculation analysis and control of the ISSA-BP-PID algorithm. The spatial distribution of temperature sensors, heating blocks, and heating rods is shown in the figure. Figure 9 As shown in (a) and (b).

[0184] When the temperature controller outputs a temperature control signal, if the temperature control signal is a cooling signal, the refrigerator is controlled to cool down. If the temperature control signal is a heating signal, the heating block and heating rod are used to precisely increase the temperature. It can be understood that the refrigerator controls the temperature to decrease (coarse adjustment), and the heating rod + heating block controls the temperature to increase (fine adjustment).

[0185] In an exemplary embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. Figure 10As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O) and a communication interface. The processor, memory and input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The database of the computer device is used to store data such as pre-trained neural networks. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a temperature control method for a cryogenic device is implemented.

[0186] Those skilled in the art will understand that Figure 10 The structure shown in the figure is a block diagram of a partial structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0187] In one embodiment, a computer device is further provided, including a memory and a processor. The memory stores a computer program, and the processor implements the steps in the above method embodiments when executing the computer program.

[0188] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0189] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0190] Those skilled in the art will understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, database or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The databases involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the various embodiments provided herein may be, but are not limited to, general-purpose processors, central processing units (CPUs), graphics processing units (GPUs), digital signal processors (DSPs), programmable logic devices (PLDs), quantum computing-based data processing logic devices, artificial intelligence (AI) processors, and the like.

[0191] The technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0192] The above embodiments merely illustrate several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art may make various modifications and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

Claims

1. A temperature control method for cryogenic equipment, characterized in that: The method comprises: Temperature detection step: detecting the expected temperature information and actual temperature information of the cryogenic equipment; When the difference between the expected temperature information and the actual temperature information is not within a preset temperature range threshold, the expected temperature information and the actual temperature information are processed using a pre-trained neural network to obtain a temperature regulation influencing parameter of a temperature controller; controlling the temperature controller to output a temperature control signal according to the temperature adjustment influencing parameter; After the temperature of the cryogenic device is adjusted according to the temperature control signal, the process returns to the temperature detection step until the difference between the actual temperature information and the expected temperature information is within a preset temperature range threshold.

2. The method according to claim 1, characterized in that The method of processing the expected temperature information and the actual temperature information using a pre-trained neural network to obtain the temperature adjustment influencing parameters of the temperature controller includes: detecting temperature error information of the cryogenic device according to the expected temperature information and the actual temperature information; The expected temperature information, the actual temperature information and the temperature error information are processed according to a pre-trained neural network to obtain the temperature regulation influencing parameters of the temperature controller.

3. The method according to claim 1, characterized in that The training process corresponding to the pre-trained neural network includes: Detection step: determining a current number of iterations, and detecting search weight information corresponding to current network parameters of the neural network based on the current number of iterations; updating the current network parameters according to the search weight information to obtain new network parameters; If the neural network meets the preset iteration termination condition, the neural network with the new network parameters is determined as the pre-trained neural network; If the neural network does not meet the preset iteration termination condition, the new network parameters are used as the current network parameters again, and the detection step is returned to be executed until the neural network meets the preset iteration termination condition.

4. The method according to claim 3, characterized in that The updating of the current network parameters according to the search weight information to obtain new network parameters includes: Obtain the current network parameter population of the current iteration process, and detect the current optimal network parameters from the current network parameter population; For each current network parameter in the current network parameter population, obtaining current network parameter distance information according to the current optimal network parameter and the current network parameter; Detecting update control information of the current network parameters according to the current number of iterations and the maximum number of iterations; updating the current network parameter distance information according to the update control information, and updating the current network parameter according to the search weight information; New network parameters are obtained according to the updated current network parameters and the updated current network parameter distance information.

5. The method according to claim 4, characterized in that The updating of the current network parameter distance information according to the update control information includes: Obtaining an environmental random value associated with the current network parameter distance information update; When the environmental random value is less than a preset environmental alert value, updating the current network parameter distance information using a global search function according to the update control information; When the environmental random value is greater than or equal to a preset environmental alert value, the current network parameter distance information is updated using a local search function according to the update control information.

6. The method according to claim 1, characterized in that The step of controlling the temperature controller to output a temperature control signal according to the temperature adjustment influencing parameter comprises: detecting current temperature control information of the temperature controller during a current temperature detection cycle according to the temperature adjustment influencing parameter, and obtaining previous temperature control information during a previous temperature detection cycle; The temperature controller is controlled to output a temperature control signal according to a temperature control variation between the current temperature control information and the previous temperature control information.

7. The method according to claim 6, characterized in that The temperature adjustment influencing parameters include integral influence information, differential influence information, and proportional influence information; detecting the current temperature control information of the temperature controller during the current temperature detection cycle according to the temperature adjustment influencing parameters includes: During the current temperature detection cycle, detecting temperature error information between the actual temperature information and the expected temperature information, and obtaining previous temperature error information between the previous actual temperature information and the expected temperature information during the previous temperature detection cycle, as well as historical temperature cumulative error information from the first temperature detection cycle to the current temperature detection cycle; processing the actual temperature information according to the proportional influence information to obtain first temperature control information; Processing the historical temperature accumulated error information according to the integral influence information to obtain second temperature control information; processing a difference between the temperature error information and the previous temperature error information according to the differential influence information to obtain third temperature control information; The current temperature control information of the temperature controller in the current temperature detection cycle is detected according to the first temperature control information, the second temperature control information and the third temperature control information.

8. A temperature control device for cryogenic equipment, characterized in that: The device comprises: A cryogenic equipment temperature detection module is used in the temperature detection step to detect the expected temperature information and actual temperature information of the cryogenic equipment; a neural network processing module, configured to, when a difference between the expected temperature information and the actual temperature information is not within a preset temperature range threshold, process the expected temperature information and the actual temperature information using a pre-trained neural network to obtain a temperature regulation influencing parameter of a temperature controller; a temperature control signal output module, configured to control the temperature controller to output a temperature control signal according to the temperature adjustment influencing parameter; The temperature adjustment module is configured to adjust the temperature of the cryogenic device according to the temperature control signal and then return to the temperature detection step until the difference between the actual temperature information and the expected temperature information is within a preset temperature range threshold.

9. A temperature control system for cryogenic equipment, characterized in that: The system comprises: A low-temperature device, comprising a sample stage and a temperature-averaging module; a first temperature sensor, the first temperature sensor being disposed on the sample stage and configured to detect actual temperature information of the cryogenic device and send the actual temperature information to the control component; A heating module, comprising a heating rod and a heating block, wherein the heating rod is arranged on the sample stage, and the heating block is arranged on the temperature equalizing module, and the heating rod and the heating block are used to adjust the temperature of the low-temperature device; Control components for: Temperature detection step: detecting expected temperature information and actual temperature information of the cryogenic equipment; When the difference between the expected temperature information and the actual temperature information is not within a preset temperature range threshold, the expected temperature information and the actual temperature information are processed using a pre-trained neural network to obtain a temperature regulation influencing parameter of a temperature controller; controlling the temperature controller to output a temperature control signal to the heating rod and the heating block according to the temperature adjustment influencing parameter; After the heating rod and the heating block adjust the temperature of the low-temperature device according to the temperature control signal, the process returns to the temperature detection step until the difference between the actual temperature information and the expected temperature information is within a preset temperature range threshold.

10. The system according to claim 9, characterized in that The system further includes a second temperature sensor installed in the temperature equalization module, the second temperature sensor is used to detect the comparative temperature information of the cryogenic device and send the comparative temperature information to the control component; The control component is further used to: The temperature regulation effect of the cryogenic device is detected according to the actual temperature information and the comparison temperature information.

Images