Integral sphere diffuse reflection laser cooling optimization method based on bayesian optimization

Abstract

This application provides a Bayesian optimization-based optimization method for integral sphere diffuse reflection laser cooling, which can find the optimal parameter combination in fewer iterations, improving the efficiency of experimental parameter optimization for integral sphere diffuse reflection laser cooling. The method includes: determining the optimization parameter range for the frequency and power of the cooling light, re-pump light, and probe light; within this range, selecting at least one set of initial parameters for the frequency and power of the cooling light, re-pump light, and probe light and experimenting in an integral sphere diffuse reflection laser cooling system to obtain the cost function value corresponding to each set of initial parameters; each set of initial parameters and its corresponding cost function value constitutes an initial dataset; based on the initial dataset, using a Bayesian optimization algorithm to fit the cost function and predict the optimal parameters; conducting experiments with the optimal parameters to obtain the cost function value corresponding to the optimal parameters; subsequently determining whether the cost function value corresponding to the optimal parameters meets preset conditions, thereby determining the final optimal parameters.

Description

Technical Field

[0001] This application relates to the field of cold atom frequency standard technology, specifically to a Bayesian-optimized method, system, device, and storage medium for integrating sphere diffuse reflection laser cooling optimization. Background Technology

[0002] The basic working principle of the integrating sphere cooling process is to lock the output frequency of the local oscillator to the transition lines of the cold atoms cooled by diffuse reflection laser, thereby enabling the local oscillator to output a highly stable and accurate frequency signal. Its frequency stability and uncertainty can reach extremely high levels, enabling more accurate time synchronization and measurement in fields with extremely high time accuracy requirements, such as scientific research and navigation positioning. Diffuse reflection laser cooling is a key technology for integrating sphere cold atom clocks, and optimizing the parameters in this process plays a crucial role in the cooling of the integrating sphere. The diffuse reflection light field within the integrating sphere includes: a probe beam, four cooling lasers, and a re-pump beam. To achieve efficient cooling of the atoms within the integrating sphere, it is necessary to optimize the frequencies, power, and incident directions of the probe, cooling, and re-pump beams. Experimental measurements of diffuse reflection laser cooling parameters reveal that the adjustment range of these parameters falls within a specific region based on physical theory. While phased optimization based on this theory is possible, adjustments to the frequencies and powers of the probe, cooling, and re-pump beams, as well as the incident directions of the various beams, are generally not independent. Experimental parameters obtained through empirical optimization are susceptible to external factors and subjective influences; the same parameters may yield different measurement results under different experimental conditions. Furthermore, optimizing parameters based on experimental results requires a significant investment of time and effort.

[0003] Currently, combining machine learning with physical experiments to optimize the parameters of diffuse reflection laser cooling can significantly save time and improve optimization efficiency. By using algorithms to optimize system parameters, combining the experimental system design with a cost function, and then using the optimization algorithm to find the maximum value of the cost function, the optimal experimental parameters can be determined. However, the main methods for online optimization of quantum system parameters using machine learning algorithms are artificial neural networks and differential evolution algorithms. These algorithms still have significant drawbacks. Artificial neural network algorithms typically require a large amount of training data to achieve good performance, and may suffer from overfitting when the amount of data is insufficient; differential evolution algorithms also depend on the quality and number of the initial population.

[0004] Therefore, in order to overcome the limitations of commonly used optimization algorithms, it is necessary to find better machine learning algorithms to combine with the experimental parameters of the integrating sphere diffuse reflection laser cooling, so as to improve the optimization efficiency of the experimental parameters of the integrating sphere diffuse reflection laser cooling. Summary of the Invention

[0005] This application provides a Bayesian optimization-based method, system, device, and storage medium for integrating sphere diffuse reflection laser cooling. By combining Bayesian optimization machine learning algorithms with physical experiments, the optimal parameter combination can be found efficiently in fewer iterations, thereby improving the optimization efficiency of integrating sphere diffuse reflection laser cooling experimental parameters.

[0006] To achieve the above objectives, this application adopts the following technical solution:

[0007] Firstly, this application provides a Bayesian optimization-based method for integrating sphere diffuse reflection laser cooling. The method includes: S1, determining the optimization parameter ranges for the frequencies and powers of the cooling light, re-pump light, and probe light; S2, within the optimization parameter range, selecting at least one set of initial parameters for the frequencies and powers of the cooling light, re-pump light, and probe light, and experimenting in an integrating sphere diffuse reflection laser cooling system to obtain the cost function value corresponding to each set of initial parameters. Each set of initial parameters and the corresponding cost function value constitute an initial dataset, wherein the cost function value corresponding to each set of initial parameters is the intensity value of the laser cooling absorption signal within the integrating sphere obtained from the experiment; S3, based on the initial dataset, using a Bayesian optimization algorithm to fit the cost function and predict the optimal... S4. Experiment with the optimal parameters in the integrating sphere diffuse reflection laser cooling system to obtain the cost function value corresponding to the optimal parameters; S5. Compare the cost function value corresponding to the optimal parameters with the cost function value in the initial dataset to see if the cost function value corresponding to the optimal parameters is the maximum value; S6. If yes, record the optimal parameters and further determine whether the preset exit condition is met. If the preset exit condition is met, output the optimal parameters. If the preset exit condition is not met, add the optimal parameters and the cost function value corresponding to the optimal parameters to the initial dataset and return to execute S3 until the preset exit condition is met; S7. If no, add the optimal parameters and the cost function value corresponding to the optimal parameters to the initial dataset and return to execute S3 until the preset exit condition is met.

[0008] One possible design approach, the first aspect of which also includes optimizing the frequency and power parameters of the cooling light to range from 115 MHz to 125 MHz and from 5 milliwatts to 85 milliwatts, respectively.

[0009] One possible design approach, the first aspect of which also includes optimizing the frequency and power parameters of the re-pumped light to range from 120 MHz to 130 MHz and from 3 milliwatts to 20 milliwatts, respectively.

[0010] One possible design approach, the first aspect of which also includes optimizing the frequency and power parameters of the probe light to range from 110 MHz to 120 MHz and from 0 mW to 10 mW, respectively.

[0011] One possible design scheme, the first aspect of the method, further includes: fitting a cost function using a Bayesian optimization algorithm based on an initial dataset and predicting the optimal parameters, including: using the initial dataset, constructing a surrogate model to approximate the objective function, wherein the objective function is the finally fitted cost function; selecting at least one set of parameters based on the acquisition function, calculating the cost function values ​​corresponding to each of the at least one set of parameters, optimizing the surrogate model to obtain the objective function; and predicting the optimal parameters based on the objective function.

[0012] One possible design scheme, the first aspect of the method, further includes exit conditions including: reaching a predetermined iteration number threshold, or, the cost function value corresponding to the optimal parameter in N consecutive predictions has not exceeded the previous maximum value, where N is a positive integer greater than or equal to zero.

[0013] Secondly, this application provides an integral sphere diffuse reflection laser cooling optimization system based on Bayesian optimization, the system comprising: a determination module, a processing module, and a judgment module;

[0014] The determination module is used to determine the optimal parameter range for the frequency and power of the cooling light, re-pump light, and probe light;

[0015] The processing module is used to select at least one set of initial parameters for the frequency and power of the cooling light, re-pump light, and probe light within the optimized parameter range, and conduct experiments in an integrating sphere diffuse reflection laser cooling system to obtain the cost function value corresponding to each set of initial parameters. Each set of initial parameters and the cost function value corresponding to each set of initial parameters constitute an initial dataset, wherein the cost function value corresponding to each set of initial parameters is the intensity value of the laser cooling absorption signal in the integrating sphere obtained from the experiment of each set of initial parameters; and is also used to fit the cost function using a Bayesian optimization algorithm based on the initial dataset and predict the optimal parameters; and to conduct experiments on the integrating sphere diffuse reflection laser cooling system with the optimal parameters to obtain the cost function value corresponding to the optimal parameters.

[0016] The judgment module compares the cost function value corresponding to the optimal parameter with the cost function value in the initial dataset to determine if the cost function value corresponding to the optimal parameter is the maximum value. If so, the optimal parameter is recorded, and the module further determines whether a preset exit condition is met. If the preset exit condition is met, the optimal parameter is output. If the preset exit condition is not met, the optimal parameter and its corresponding cost function value are added to the initial dataset, and the execution returns to step S3 until the preset exit condition is met. If not, the optimal parameter and its corresponding cost function value are added to the initial dataset, and the execution returns to step S3 until the preset exit condition is met.

[0017] Thirdly, a Bayesian-optimized integrated sphere diffuse reflection laser cooling optimization device is provided, which includes a module for performing the method described in the first aspect.

[0018] In one possible design, the Bayesian-optimized integrating sphere diffuse reflection laser cooling optimization device of the third aspect may further include a transceiver. This transceiver can be a transceiver circuit or an interface circuit. The transceiver can be used for communication between the Bayesian-optimized integrating sphere diffuse reflection laser cooling optimization device of the third aspect and other devices.

[0019] In one possible design, the Bayesian-optimized integral sphere diffuse reflection laser cooling optimization device of the third aspect may further include a memory. This memory may be integrated with the processor or disposed separately. The memory may be used to store instructions related to the method of the first aspect.

[0020] Fourthly, a Bayesian-optimized integrated sphere diffuse reflection laser cooling optimization device is provided. This Bayesian-optimized integrated sphere diffuse reflection laser cooling optimization device includes: a processor coupled to a memory, the processor executing instructions stored in the memory to cause the Bayesian-optimized integrated sphere diffuse reflection laser cooling optimization device to perform the method of the first aspect.

[0021] In one possible design, the Bayesian-optimized integrating sphere diffuse reflection laser cooling optimization device of the fourth aspect may further include a transceiver. This transceiver can be a transceiver circuit or an interface circuit. The transceiver can be used for communication between the Bayesian-optimized integrating sphere diffuse reflection laser cooling optimization device of the fourth aspect and other devices.

[0022] Fifthly, a Bayesian-optimized integral sphere diffuse reflection laser cooling optimization device is provided, comprising: a processor and a memory; the memory is used to store instructions, which, when executed by the processor, cause the Bayesian-optimized integral sphere diffuse reflection laser cooling optimization device to perform the method of the first aspect.

[0023] In one possible design, the Bayesian-optimized integrating sphere diffuse reflection laser cooling optimization device of the fifth aspect may further include a transceiver. This transceiver can be a transceiver circuit or an interface circuit. The transceiver can be used for communication between the Bayesian-optimized integrating sphere diffuse reflection laser cooling optimization device of the fifth aspect and other devices.

[0024] In a sixth aspect, a computer-readable storage medium is provided, the computer-readable storage medium including storage of a computer program or instructions that, when the computer program or instructions are run, cause the Bayesian-optimized integral sphere diffuse reflection laser cooling optimization method of the first aspect to be executed.

[0025] In this embodiment, a cost function is designed based on a method combining Bayesian optimization machine learning algorithms with physical experiments to form a multi-parameter autonomous optimization system, thereby deriving the optimal parameter combination for integrating sphere diffuse reflection laser cooling. This method not only ensures the reliability of the optimized parameter set, including the frequency and power of the cooling light, re-pump light, and probe light, in integrating sphere diffuse reflection laser cooling, but also efficiently finds the optimal parameter combination for integrating sphere diffuse reflection laser cooling within a fewer iterations, significantly improving optimization efficiency compared to traditional methods combining machine learning and physical experiments.

[0026] Other features and advantages of this application will be described in detail in the following detailed description section. Attached Figure Description

[0027] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0028] Figure 1 A schematic flowchart of the Bayesian optimization-based integral sphere diffuse reflection laser cooling optimization method provided in this application embodiment;

[0029] Figure 2 A schematic diagram of the Bayesian-optimized integrating sphere diffuse reflection laser cooling optimization device provided in this application embodiment. Figure 1 ;

[0030] Figure 3 A schematic diagram of the Bayesian-optimized integrating sphere diffuse reflection laser cooling optimization device provided in this application embodiment. Figure 2 . Detailed Implementation

[0031] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application. At the same time, in the description of the embodiments of this application, the terms "first," "second," etc., are only used to distinguish descriptions and should not be construed as indicating or implying relative importance. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0032] Figure 1 This is a schematic diagram of the process for optimizing diffuse laser cooling based on Bayesian optimization in an embodiment of this application.

[0033] The process for optimizing the diffuse laser cooling of the integrating sphere based on Bayesian optimization is as follows:

[0034] Step S101: Determine the optimal parameter range for the frequency and power of the cooling light, re-pump light, and probe light.

[0035] In other words, to ensure the high efficiency of Bayesian optimization-based integrated sphere diffuse reflection laser cooling optimization, the range of selectable optimization parameters must be determined before selecting the optimization parameters to ensure the effectiveness of Bayesian optimization sampling. Therefore, the cooling light frequency range is set to 115MHz-125MHz, the power range is set to 5-85mw, the pump light frequency range is set to 120MHz-130MHz, the power range is set to 3-20mw, and the probe light frequency range is set to 110MHz-120MHz, the power range is set to 0-10uw. Here, MHz represents megahertz and mw represents milliwatt.

[0036] Step S102: Within the optimized parameter range, select at least one set of initial parameters for the frequency and power of the cooling light, re-pump light, and probe light and conduct experiments in the integrating sphere diffuse reflection laser cooling system to obtain the cost function value corresponding to each set of initial parameters. Each set of initial parameters and the cost function value corresponding to each set of initial parameters constitute the initial dataset. The cost function value corresponding to each set of initial parameters is the intensity value of the laser cooling absorption signal in the integrating sphere obtained from the experiment of each set of initial parameters.

[0037] It is understood that, within the aforementioned optimization parameter range, at least one set of initial parameters for the frequency and power of the cooling light, re-pump light, and probe light is selected for experimentation in an integrating sphere diffuse reflection laser cooling system to obtain the cost function value corresponding to each set of initial parameters. In the example of this application, five sets of initial parameters x are selected. n Taking (n = 1, 2, 3, 4, 5) as an example, 5 sets of initial parameters x n Physical experiments were conducted on (n = 1, 2, 3, 4, 5) respectively, and the intensity y of the laser cooling absorption signal inside the integrating sphere was measured. n (n = 1, 2, 3, 4, 5), where y n This is the cost function value corresponding to the initial parameters. There are 5 sets of initial parameters x. n (n = 1, 2, 3, 4, 5) and the corresponding y n (n = 1, 2, 3, 4, 5) constitute the initial dataset.

[0038] Step S103: Based on the initial dataset, use the Bayesian optimization algorithm to fit the cost function and predict the optimal parameters.

[0039] This can be understood as follows: using the initial dataset, a surrogate model is constructed to approximate the objective function, where the objective function is the final fitted cost function; based on the acquisition function, at least one set of parameters is selected, and the cost function values ​​corresponding to each of the at least one set of parameters are calculated to optimize the surrogate model and obtain the objective function; based on the objective function, the optimal parameters are predicted.

[0040] Step S104: Experiment with the optimal parameters in the integrating sphere diffuse reflection laser cooling system to obtain the cost function value corresponding to the optimal parameters.

[0041] Step S105: Compare the cost function value corresponding to the optimal parameter with the cost function value in the initial dataset to determine whether the cost function value corresponding to the optimal parameter is the maximum value.

[0042] For example, if the cost function values ​​in the initial dataset are Y1, Y2, Y3, Y4, and Y5, and the cost function value corresponding to the optimal parameter is Y6, determine whether Y6 is greater than Y1, Y2, Y3, Y4, and Y5.

[0043] In step S106, if yes, record the optimal parameters and further determine whether the preset exit conditions are met. If the preset exit conditions are met, output the optimal parameters. If the preset exit conditions are not met, add the optimal parameters and the cost function value corresponding to the optimal parameters to the initial dataset, and return to execute S103 until the preset exit conditions are met.

[0044] In the example above, if Y6 is greater than Y1, Y2, Y3, Y4, and Y5, then record the optimal parameters corresponding to Y6, that is, the values ​​of the frequency and power of the cooling light, the re-pump light, and the probe light.

[0045] The exit conditions include: reaching a predetermined iteration threshold, or the cost function value corresponding to the optimal parameter in N consecutive predictions not exceeding the previous maximum value, where N is a positive integer greater than or equal to zero.

[0046] This can be understood as follows: the predetermined threshold for the number of iterations is 80. If the maximum value obtained in this iteration is acquired on the 60th iteration, the loop continues, adding the optimal parameter and its corresponding cost function value to the initial dataset, and then returning to execute S103 until the preset exit condition is met. If the maximum value obtained in this iteration is acquired on the 80th iteration, the preset exit condition is met, and the optimal parameter is output. The condition that the cost function value corresponding to the optimal parameter in N consecutive predictions does not exceed the previous maximum value can also be understood as follows: for example, if the preset condition is that the cost function value corresponding to the optimal parameter in 20 consecutive predictions does not exceed the previous maximum value (e.g., the optimal parameter Yl in the first prediction is less than the Yl in the subsequent 20 predictions), then the exit condition is met.

[0047] If not, in step S107, add the optimal parameters and the cost function value corresponding to the optimal parameters to the initial dataset, and return to execute S103 until the preset exit condition is met.

[0048] In summary, based on a method combining Bayesian optimization machine learning algorithms with physical experiments, a cost function is designed to form a multi-parameter autonomous optimization system, thereby deriving the optimal parameter combination for integrating sphere diffuse reflection laser cooling. This method not only ensures the reliability of the optimized parameter set, including the frequency and power of the cooling light, re-pump light, and probe light, in integrating sphere diffuse reflection laser cooling, but also efficiently finds the optimal parameter combination for integrating sphere diffuse reflection laser cooling within a fewer iterations. Compared to traditional methods combining machine learning and physical experiments, this significantly improves optimization efficiency.

[0049] The above combination Figure 1 This application provides a detailed description of the Bayesian optimization-based integral sphere diffuse reflection laser cooling optimization method provided in the embodiments of this application. The following details the Bayesian optimization-based integral sphere diffuse reflection laser cooling optimization system used to implement the embodiments of this application.

[0050] The system specifically includes: a determination module, a judgment module, and a decision module, as shown below.

[0051] The determination module is used to determine the optimal parameter range for the frequency and power of the cooling light, re-pump light, and probe light;

[0052] The determination module is used to select at least one set of initial parameters for the frequency and power of the cooling light, re-pump light, and probe light within the optimization parameter range, and conduct experiments in an integrating sphere diffuse reflection laser cooling system to obtain the cost function value corresponding to each set of initial parameters. Each set of initial parameters and the cost function value corresponding to each set of initial parameters constitute an initial dataset, wherein the cost function value corresponding to each set of initial parameters is the intensity value of the laser cooling absorption signal in the integrating sphere obtained from the experiment of each set of initial parameters; and is also used to fit the cost function using a Bayesian optimization algorithm based on the initial dataset and predict the optimal parameters; and is further used to conduct experiments on the integrating sphere diffuse reflection laser cooling system with the optimal parameters to obtain the cost function value corresponding to the optimal parameters;

[0053] The judgment module compares the cost function value corresponding to the optimal parameter with the cost function value in the initial dataset to determine if the cost function value corresponding to the optimal parameter is the maximum value. If so, the optimal parameter is recorded, and the module further determines whether a preset exit condition is met. If the preset exit condition is met, the optimal parameter is output. If the preset exit condition is not met, the optimal parameter and its corresponding cost function value are added to the initial dataset, and the execution returns to step S3 until the preset exit condition is met. If not, the optimal parameter and its corresponding cost function value are added to the initial dataset, and the execution returns to step S3 until the preset exit condition is met.

[0054] Furthermore, the specific implementation of the above system is basically similar to the method implementation, so the description is relatively simple. For relevant details, please refer to the description of the method implementation. Moreover, it should be noted that in the various modules of the system of this application, the components are logically divided according to the functions they are to perform. However, this application is not limited to this and can re-divide or combine the components as needed.

[0055] The above describes the Bayesian optimization-based integrating sphere diffuse reflection laser cooling optimization method and system provided in the embodiments of this application. The following, in conjunction with... Figures 2-3 This document provides a detailed description of the Bayesian-optimized integral sphere diffuse reflection laser cooling optimization device used to implement the embodiments of this application.

[0056] Figure 2 This is a schematic diagram of the structure of the Bayesian-optimized integrating sphere diffuse reflection laser cooling optimization device provided in the embodiments of this application. Figure 1 For example, such as Figure 2 As shown, the Bayesian-optimized integrating sphere diffuse reflection laser cooling optimization device 200 includes a transceiver module 201 and a processing module 202. For ease of explanation, Figure 2 Only the main components of this Bayesian-optimized integral sphere diffuse reflection laser cooling optimization device are shown.

[0057] The transceiver module 201 is used to perform the transceiver function of the above-mentioned Bayesian optimization-based integral sphere diffuse reflection laser cooling optimization method, and the processing module 202 is used to perform other functions of the above-mentioned Bayesian optimization-based integral sphere diffuse reflection laser cooling optimization method other than the transceiver function.

[0058] Optionally, the transceiver module 201 may include a sending module. Figure 2 (not shown in the image) and receiving module ( Figure 2 (Not shown in the image). The transmitting module is used to implement the transmitting function of the Bayesian-optimized integrating sphere diffuse reflection laser cooling optimization device 200, and the receiving module is used to implement the receiving function of the Bayesian-optimized integrating sphere diffuse reflection laser cooling optimization device 200.

[0059] Optionally, the Bayesian-optimized integrating sphere diffuse reflection laser cooling optimization device 200 may also include a storage module. Figure 2 (Not shown in the image), the storage module stores programs or instructions. When the processing module 202 executes the program or instructions, the Bayesian-optimized integrating sphere diffuse reflection laser cooling optimization device 200 can perform the Bayesian-optimized integrating sphere diffuse reflection laser cooling optimization method in the embodiments of this application.

[0060] The following is combined with Figure 3 Each component of the Bayesian-optimized integral sphere diffuse reflection laser cooling optimization device 300 is described in detail:

[0061] The processor 301 is the control center of the Bayesian-optimized integrating sphere diffuse reflection laser cooling optimization device 300. It can be a single processor or a collective term for multiple processing elements. For example, the processor 301 can be one or more central processing units (CPUs), application-specific integrated circuits (ASICs), or one or more integrated circuits configured to implement the embodiments of this application, such as one or more digital signal processors (DSPs), or one or more field-programmable gate arrays (FPGAs).

[0062] Optionally, the processor 301 can execute various functions of the Bayesian-optimized integrating sphere diffuse reflection laser cooling optimization device 300 by running or executing software programs stored in the memory 302 and calling data stored in the memory 302, such as executing the Bayesian-optimized integrating sphere diffuse reflection laser cooling optimization method in the embodiments of this application.

[0063] In a specific implementation, as one example, processor 301 may include one or more CPUs, for example... Figure 3 CPU0 and CPU1 are shown in the diagram.

[0064] In a specific implementation, as one example, the Bayesian-optimized integral sphere diffuse reflection laser cooling optimization device 300 may also include multiple processors, for example... Figure 3 The processors 301 and 304 are shown in the diagram. Each of these processors can be a single-core processor (single-CPU) or a multi-core processor (multi-CPU). Here, "processor" can refer to one or more devices, circuits, and / or processing cores for processing data (e.g., computer program instructions). The memory 302 is used to store the software program executing the scheme of this application, and its execution is controlled by the processor 301. Specific implementation methods can be found in the above method embodiments, and will not be repeated here.

[0065] Optionally, the memory 302 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto. The memory 302 may be integrated with the processor 301 or exist independently, and may be connected via the interface circuit of the Bayesian-optimized integrating sphere diffuse reflection laser cooling optimization device 300. Figure 3 (Not shown in the image) is coupled to processor 301, and this embodiment does not specifically limit this.

[0066] Transceiver 303 is used for communication with other communication devices. For example, in the Bayesian-optimized integral sphere diffuse reflection laser cooling optimization device 300, which is the first device, transceiver 303 can be used to communicate with a second device or a third device.

[0067] Optionally, transceiver 303 may include a receiver and a transmitter. Figure 3(Not shown separately). The receiver is used to implement the receiving function, and the transmitter is used to implement the sending function.

[0068] Optionally, the transceiver 303 can be integrated with the processor 301 or exist independently, and can be connected via the interface circuit of the Bayesian-optimized integrating sphere diffuse reflection laser cooling optimization device 300. Figure 3 (Not shown in the image) is coupled to processor 301, and this embodiment does not specifically limit this.

[0069] Understandable, Figure 3 The structure of the Bayesian-optimized integral sphere diffuse reflection laser cooling optimization device 300 shown does not constitute a limitation on the Bayesian-optimized integral sphere diffuse reflection laser cooling optimization device. The actual Bayesian-optimized integral sphere diffuse reflection laser cooling optimization device may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0070] Furthermore, the technical effects of the Bayesian-optimized integral sphere diffuse reflection laser cooling optimization device 300 can be referred to the technical effects of the method described in the above method embodiments, and will not be repeated here.

[0071] It should be understood that the processor in the embodiments of this application can be a central processing unit (CPU), but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0072] It should also be understood that the memory in the embodiments of this application can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDRSDRAM), enhanced synchronous DRAM (ESDRAM), synchronous linked DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0073] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

Claims

1. A Bayesian optimization-based method for optimizing diffuse laser cooling using an integrating sphere, characterized in that, The method includes: S1. Determine the optimal parameter range for the frequency and power of the cooling light, re-pump light, and probe light; S2. Within the range of the optimized parameters, at least one set of initial parameters for the frequency and power of the cooling light, re-pump light, and probe light are selected and tested in the integrating sphere diffuse reflection laser cooling system to obtain the cost function value corresponding to each set of initial parameters. The initial parameters and the cost function value corresponding to each set of initial parameters constitute the initial dataset. The cost function value corresponding to each set of initial parameters is the intensity value of the laser cooling absorption signal in the integrating sphere obtained by the experiment of each set of initial parameters. S3. Based on the initial dataset, use the Bayesian optimization algorithm to fit the cost function and predict the optimal parameters; S4. Experiment with the optimal parameters in the integrating sphere diffuse reflection laser cooling system to obtain the cost function value corresponding to the optimal parameters; S5. Compare the cost function value corresponding to the optimal parameter with the cost function value in the initial dataset to determine whether the cost function value corresponding to the optimal parameter is the maximum value. S6. If yes, record the optimal parameter and further determine whether the preset exit condition is met. If the preset exit condition is met, output the optimal parameter. If the preset exit condition is not met, add the optimal parameter and the cost function value corresponding to the optimal parameter to the initial dataset, return to execute S3, until the preset exit condition is met. S7. If not, add the optimal parameter and the cost function value corresponding to the optimal parameter to the initial dataset, return to execute S3, and continue until the preset exit condition is met.

2. The Bayesian optimization-based method for optimizing diffuse laser cooling of an integrating sphere as described in claim 1, characterized in that, The optimized parameters for the frequency and power of the cooling light are 115 MHz to 125 MHz and 5 mW to 85 mW, respectively.

3. The Bayesian optimization-based method for optimizing diffuse laser cooling of an integrating sphere as described in claim 1, characterized in that, The optimized parameters for the frequency and power of the re-pumped light are 120 MHz to 130 MHz and 3 mW to 20 mW, respectively.

4. The Bayesian optimization-based method for optimizing diffuse laser cooling of an integrating sphere as described in claim 1, characterized in that, The optimized parameters for the frequency and power of the probe light are 110 MHz to 120 MHz and 0 mW to 10 mW, respectively.

5. The Bayesian optimization-based method for optimizing diffuse laser cooling of an integrating sphere according to claim 1, characterized in that, The step of fitting a cost function using a Bayesian optimization algorithm based on the initial dataset and predicting the optimal parameters includes: Using the initial dataset, a surrogate model is constructed to approximate the objective function, where the objective function is the finally fitted cost function; Based on the acquisition function, at least one set of parameters is selected, the cost function value corresponding to each of the at least one set of parameters is calculated, the proxy model is optimized, and the objective function is obtained. Based on the objective function, the optimal parameters are predicted.

6. The Bayesian optimization-based method for optimizing diffuse laser cooling of an integrating sphere as described in claim 1, characterized in that, The exit conditions include: reaching a predetermined iteration number threshold, or the cost function value corresponding to the optimal parameter in N consecutive predictions not exceeding the previous maximum value, where N is a positive integer greater than or equal to zero.

7. A Bayesian-optimized integrated sphere diffuse reflection laser cooling optimization system, characterized in that, The system includes: The determination module is used to determine the optimal parameter range for the frequency and power of the cooling light, re-pump light, and probe light; The processing module is configured to: select at least one set of initial parameters for the frequency and power of the cooling light, re-pump light, and probe light within the optimized parameter range and experiment in an integrating sphere diffuse reflection laser cooling system to obtain the cost function value corresponding to each set of initial parameters. Each set of initial parameters and the corresponding cost function value constitute an initial dataset, wherein the cost function value corresponding to each set of initial parameters is the intensity value of the laser cooling absorption signal within the integrating sphere obtained from the experiment with each set of initial parameters. The module is also configured to: fit the cost function using a Bayesian optimization algorithm based on the initial dataset and predict the optimal parameters; and experiment with the optimal parameters in the integrating sphere diffuse reflection laser cooling system to obtain the cost function value corresponding to the optimal parameters. The judgment module is used to compare the cost function value corresponding to the optimal parameter with the cost function value in the initial dataset to determine whether the cost function value corresponding to the optimal parameter is the maximum value. If so, the optimal parameter is recorded, and it is further determined whether a preset exit condition is met. If the preset exit condition is met, the optimal parameter is output. If the preset exit condition is not met, the optimal parameter and the cost function value corresponding to the optimal parameter are added to the initial dataset, and execution returns to step S3 until the preset exit condition is met. If not, the optimal parameter and the cost function value corresponding to the optimal parameter are added to the initial dataset, and execution returns to step S3 until the preset exit condition is met.

8. A Bayesian-optimized integral sphere diffuse reflection laser cooling optimization device, characterized in that, The apparatus includes a module for performing the method as described in any one of claims 1-6.

9. A Bayesian-optimized integral sphere diffuse reflection laser cooling optimization device, characterized in that, The device includes a processor and a memory; the memory is used to store computer instructions, which, when executed by the processor, cause the Bayesian-optimized integral sphere diffuse reflection laser cooling optimization device to perform the method as described in any one of claims 1-6.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a computer program or instructions that, when executed, cause the method as described in any one of claims 1-6 to be performed.

Images