A method for measuring the ion temperature in an ion trap through a neural network and image measurement

Abstract

The present invention discloses a method for measuring the ion temperature in an ion trap through a neural network and image measurement. The method includes: obtaining a photo of calcium ions within a specific temperature range; adjusting the temperature of the calcium ions within a set temperature range, and repeatedly obtaining multiple photos of the calcium ions at each specific temperature interval; using the multiple photos of the calcium ions and their corresponding temperatures to create a data set; inputting the created data set into a neural network to train the neural network until the accuracy rate is greater than a set value; and inputting a photo of calcium ions into the trained neural network to obtain the temperature of the calcium ions. The present invention uses a neural network to identify the imaging of ions, and can complete the measurement of the ion temperature when the ion imaging state is poor.

Description

Technical Field

[0001] The present invention relates to the technical field of quantum computing and quantum sensing based on trapped ions, and particularly relates to a method for measuring the temperature of ions in an ion trap through a neural network and image measurement, which is applicable to the rapid measurement of the temperature of ions in the ion trap. Background Art

[0002] In the application fields of quantum computing and quantum sensing, since the increase in ion temperature will lead to the enhancement of the decoherence effect during the information processing process and the reduction of the measurement accuracy of the quantum sensing system, people often need to adopt different methods to cool the ions and accurately measure the temperature of the ions. In addition, if the time consumed for measuring the ion temperature is too long, it will reduce the information processing speed during the quantum computing process or the sampling rate of the quantum sensor device. Therefore, it is particularly important to quickly and accurately measure the ion temperature in this field.

[0003] Currently, for ions at the millikelvin level, there are mainly two methods for measuring temperature: 1. The method of measuring the fluorescence spectral linewidth of ions, which uses the Doppler broadening effect of the spectral line to measure the ion temperature. However, this method not only has a very slow measurement speed (usually on the order of ten minutes), but also the ion temperature changes with the change of the laser frequency during the temperature measurement process; 2. The method of taking ion photos with a CCD, which obtains the ion temperature by analyzing the relationship between the ion spatial distribution region and the ion vibration amplitude in the photo. Although the measurement speed of this method for measuring the ion temperature can reach the minute level, if there is aberration in the fluorescence imaging system, this method is invalid. In addition, the measurement accuracy of this method highly depends on the resolution of the fluorescence imaging system. Summary of the Invention

[0004] Aiming at the deficiencies in the prior art, the present invention provides a method for measuring the temperature of ions in an ion trap through a neural network and image measurement, which uses a neural network to identify the ion imaging and can complete the measurement of the ion temperature when the ion imaging state is poor.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A method for measuring the ion temperature in an ion trap through a neural network and image measurement, which is applicable to an ion trap system. The ion trap system includes a calcium furnace, a vacuum chamber, an ion pump, and a micro-surface electrode ion trap. The calcium furnace is used to eject calcium atoms. The vacuum chamber has several observation windows. The micro-surface electrode ion trap is arranged in the vacuum chamber and is used to generate a three-dimensional confinement potential well. The calcium atoms ejected by the calcium furnace are ionized by photoionization to generate calcium ions and loaded into the three-dimensional confinement potential well. A part of the fluorescence emitted by the calcium ions enters the CCD through an imaging lens group, and the other part enters a photomultiplier tube. The signal output by the photomultiplier tube enters a photon counter for measurement.

[0007] The method includes:

[0008] Obtain a photo of calcium ions in a specific temperature range;

[0009] Adjust the temperature of calcium ions within a set temperature range, and repeat obtaining multiple photos of the calcium ions at each specific temperature range;

[0010] Use multiple photos of the calcium ions and their corresponding temperatures to make a data set;

[0011] Input the made data set into a neural network for training the neural network until the accuracy rate is greater than a set value;

[0012] Input a photo of calcium ions with the temperature to be measured into the trained neural network to obtain the temperature of the calcium ions.

[0013] Optionally, before obtaining a photo of calcium ions in a specific temperature range, it further includes the steps of:

[0014] Maintain the vacuum chamber in a set vacuum environment, apply a DC voltage and a RF high voltage on the micro-surface electrode ion trap, and use the micro-surface electrode ion trap to confine calcium ions;

[0015] Introduce a laser with a set wavelength through an observation window to cool the calcium ions below the set temperature;

[0016] At different laser powers, change the laser frequency and measure the fluorescence line shape of calcium ions with a photon counter, and repeat measuring the fluorescence line shape data multiple times at each laser power;

[0017] According to the measured fluorescence line shape data of calcium ions, calculate the saturation broadening and background at a specific laser power.

[0018] Optionally, obtain a photo of calcium ions at a specific laser power by running a first program, and the first program is executed by a line shape scanning module, which specifically includes the following steps:

[0019] Step 501: Generate a frequency list during the scanning line type process according to the detuning range and step size of the entire scanning process as input;

[0020] Step 502: Set the detuning of the probe light according to the list and measure the fluorescence signal of the ions at this time;

[0021] Step 503: Send a shooting command to the photo shooting module to make it record the photo of the ions at this time, and receive the photo number and the similarity of this photo returned by it;

[0022] Step 504: Change the detuning of the probe light to the next value in the frequency list, and repeat steps 502 and 503 of the operation process of this module until the corresponding ion fluorescence signals are measured for all the detunings in the frequency list;

[0023] Step 505: If the number of photos measured at this time is lower than the set number, continue to measure a certain number of photos, and save the photo numbers and similarities;

[0024] Step 506: Save the frequency list and the corresponding fluorescence signal values, save the similarities of the ion photos taken at each measurement point, and save the photo numbers and their corresponding similarities taken in step 505 of the operation process of this module.

[0025] Optionally, the first program is also executed by the photo shooting module, and the operation process is divided into a preparation stage and a measurement stage. The preparation stage specifically includes the following steps:

[0026] Step 511: Control the CCD to take several photos of the ions;

[0027] Step 512: Take the average of the brightness values of each pixel point on these ion photos to generate a new photo, and intercept a region of one pixel point around the ion on this new photo. This intercepted region is defined as target_photo;

[0028] The measurement stage specifically includes the following steps:

[0029] Step 521: Wait for the command from the line type scanning module;

[0030] Step 522: After receiving the shooting command, take a photo of the ions and number it;

[0031] Step 523: Locate the position of the ions and intercept a region of one pixel point around the ions. This intercepted region is defined as current_photo;

[0032] Step 524: Compare target_photo and current_photo and calculate the similarity;

[0033] Step 525: Save the captured photo and send the photo number and similarity to the line scan module;

[0034] Step 526: Return to running Step 521.

[0035] Optionally, a data set is made using multiple of the calcium ion photos by running a second program, and the second program specifically includes the following steps:

[0036] Step 601: Correlate all the data points obtained from scanning the line pattern, the ion photos captured when measuring this data point, and the similarity values calculated when taking the photos with each other;

[0037] Step 602: Check all the photos with a similarity lower than a certain value, and discard both the photos and their corresponding data points;

[0038] Step 603: Perform Lorentz fitting on each measured fluorescence line pattern, and calculate the temperature of the ions when scanning this fluorescence line pattern. Mark the temperature of the photos taken during each line pattern scanning process in the measurement as the temperature calculated at this time;

[0039] Step 604: Establish a data set using the ion photos and their corresponding temperatures;

[0040] Step 605: Randomly extract a small part of the data in the data set as the test group, and the remaining as the training group, and then save them separately.

[0041] Optionally, neural network training is performed by running a third program, and the third program specifically includes the following steps:

[0042] Step 801: Establish a neural network, where the pictures input into the neural network first pass through several convolutional layers, then through several Inception blocks in GoogLeNet, and finally the results are output through a fully connected layer;

[0043] Step 802: Set the loss function used during neural network training to cross entropy;

[0044] Step 803: Randomly extract several pictures with the same label in the training group data set, splice them into one picture as the input of the neural network, and then perform one training on the neural network;

[0045] Step 804: Repeat running Step 803. Every several training rounds, use the test group data set to test the accuracy until the accuracy is high enough;

[0046] Step 805: Save the neural network parameters at this time.

[0047] Optionally, the temperature of calcium ions is obtained by running a fourth program, and the fourth program specifically includes the following steps:

[0048] Step 911: Establish the same neural network as in the third program and load the neural network parameters obtained in the third program;

[0049] Step 912: Stitch a number of input pictures into one picture as the input of the neural network;

[0050] Step 913: Obtain the temperature of the ions from the output of the neural network.

[0051] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0052] (1) By using a neural network to identify the imaging of ions, the temperature of ions can be measured when the ion imaging state is poor.

[0053] (2) During the process of measuring the ion temperature, the laser power and frequency of the system remain unchanged, reducing the error caused by the laser.

[0054] (3) Without the condition of individual addressing of ions, the temperatures of multiple ions can be measured simultaneously. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0056] Figure 1 It is a schematic diagram of the overall concept of the method for measuring the temperature of ions in an ion trap by a neural network and an image in an embodiment of the present invention;

[0057] Figure 2 It is a schematic diagram of the composition of an ion trap system applicable to the method in an embodiment of the present invention; wherein, Figure 2 (a) is a schematic structural diagram of the ion trap system; Figure 2 (b) is Figure 2 (a) a schematic structural diagram of one perspective; Figure 2 (c) is Figure 2 (a) a schematic structural diagram of another perspective;

[0058] Figure 3 It is a schematic structural diagram of a micro surface electrode ion trap of an ion trap system applicable to the method in an embodiment of the present invention;

[0059] Figure 4Schematic diagram of an imaging system of an ion trap system applicable to the method of the embodiment of the present invention;

[0060] Figure 5 Schematic diagram of the operation logic of Program A;

[0061] Figure 6 Schematic diagram of the operation logic of Program B;

[0062] Figure 7 Schematic diagram of the operation logic of Program C;

[0063] Figure 8 Schematic diagram of the operation logic of Program D. Detailed implementation manners

[0064] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0065] Embodiment:

[0066] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily need to describe a specific order or sequence. It should be understood that such used data can be interchanged under appropriate circumstances so that the embodiments of the present invention described here can be implemented in an order other than those illustrated or described here. In addition, the terms "include" and "have" and any variations thereof in the embodiments of the present invention are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device including a series of steps or units does not necessarily have to be limited to those clearly listed steps or units, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0067] It should be understood that the terms "center", "longitudinal", "transverse", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise", "axial", "radial", "circumferential", etc. indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be construed as a limitation to the present invention.

[0068] In the description of the present invention, the meaning of "a plurality" is at least two, such as two, three, etc., unless otherwise specifically defined. In addition, unless otherwise clearly specified and defined, the terms "mounted", "connected", and "coupled" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific situations.

[0069] In the present invention, unless otherwise clearly specified and defined, the first feature being "on" or "under" the second feature can be that the first and second features are in direct contact, or the first and second features are in indirect contact through an intermediate medium. Moreover, the first feature being "above", "over", and "on top of" the second feature can be that the first feature is directly above or obliquely above the second feature, or merely indicates that the first feature has a higher horizontal height than the second feature. The first feature being "under", "beneath", and "underneath" the second feature can be that the first feature is directly below or obliquely below the second feature, or merely indicates that the first feature has a lower horizontal height than the second feature.

[0070] As Figure 1 shown, Figure 1 The figure shows a schematic diagram of a method for measuring the ion temperature in an ion trap through a neural network and image measurement. The overall idea of the method is specifically to first collect data from the ion trap system to create a data set, then use the data set to train the neural network, and finally use the trained neural network to measure the temperature of the ions.

[0071] A method for measuring the ion temperature in an ion trap through a neural network and image measurement, which is applicable to the ion trap system of the embodiment.

[0072] Among them, as Figure 2 shown, Figure 2 This is a schematic diagram of the composition of the vacuum part of the ion trap system applicable to this method provided by the embodiment of the present invention. The ion trap system includes a calcium furnace (or atomic furnace), a vacuum chamber 11, an ion pump 2, and a micro surface electrode ion trap. The calcium furnace is used to eject calcium atoms. The vacuum chamber 1 has several observation windows. The micro surface electrode ion trap is arranged inside the vacuum chamber 1; there are an ion trap electrode DC voltage signal connection port 5 and an ion trap electrode RF voltage signal connection port 6 outside the ion trap furnace.

[0073] Figure 3 This is a schematic diagram of the structure of the micro surface electrode ion trap of the ion trap system applicable to the method of the embodiment of the present invention. In the figure, RF is the radio frequency electrode, and A1, A7, B1, and B7 in the figure are cap electrodes, and the other electrodes are compensation electrodes.

[0074] Figure 4 Schematic diagram of an imaging system of an ion trap system adapted to the method of the embodiments of the present invention. As can be seen from Figure 4 it, the imaging system includes a lens group, a CCD camera, a photomultiplier tube, a photon counter, and a computer. Among them, the lens group is used to collect the fluorescence emitted by ions and form an image. The CCD camera is used to detect the fluorescence image of ions. The photomultiplier tube is used to receive and amplify the fluorescence signal of ions. The photon counter is used to measure the ion fluorescence signal output by the photomultiplier tube. The ions are located in a vacuum chamber. A part of the fluorescence emitted by the ions enters the CCD after passing through the imaging lens group, and the other part enters the photomultiplier tube. The signal output by the photomultiplier tube enters the photon counter for measurement. The computer is used to control the operation of devices such as the photon counter, CCD camera, signal source, and radio frequency switch.

[0075] An imaging device of an ion trap system, which includes: a vacuum chamber 1, an ion pump 2, a calcium furnace, a micro surface electrode ion trap, an imaging lens group, a CCD camera, a photomultiplier tube, and a photon counter; several observation windows are arranged on the outer surface of the vacuum chamber 1, and the several observation windows can be used for inputting laser light and detecting fluorescence signals; the ion pump 2 is connected to the vacuum chamber 1, and the function of the ion pump 2 is to maintain the vacuum degree in the vacuum chamber 1; the calcium furnace is located inside the vacuum chamber 1 and is connected to the vacuum chamber 1, and the function of the calcium furnace is to eject calcium atoms; the micro surface electrode ion trap is arranged in the vacuum chamber 1, and the function of the micro surface electrode ion trap is to generate a three-dimensional trapping potential well, and the calcium atoms ejected by the calcium furnace (or atomic furnace) are ionized by photoionization to generate 40 Ca + ions and loaded into the trapping potential well; the imaging lens group has a plurality of combined lenses for forming an optical path, and the combined lenses are used to collect the fluorescence emitted by ions and form an image; the CCD camera is used to photograph the fluorescence image of the ions passing through the imaging lens group, and the photomultiplier tube is used to receive and amplify the fluorescence signal of the ions; the photon counter can count the fluorescence signal of the ions output by the photomultiplier tube to obtain the fluorescence count of the ions.

[0076] In specific use, the ion trap system includes a calcium furnace, a vacuum chamber 1, an ion pump 2, and a microfabricated surface-electrode ion trap. The calcium furnace is used to eject calcium atoms. The vacuum chamber 1 has several observation windows. The microfabricated surface-electrode ion trap is disposed within the vacuum chamber 1. The vacuum chamber 1 is provided with an ion trap electrode DC voltage signal connection port 5 and an ion trap electrode RF voltage signal connection port 6. In this embodiment, an imaging lens group is used to collect and image the fluorescence emitted by the ions in the ion trap furnace. The CCD camera can detect the fluorescence image of the ions. The photomultiplier tube can receive and amplify the fluorescence signal of the ions. The photon counter is used to measure the ion fluorescence signal output by the photomultiplier tube. The ions are located in the ion trap within the vacuum chamber 1. Part of the fluorescence emitted by the ions enters the CCD after passing through the imaging lens group, and the other part enters the photomultiplier tube. The signal output by the photomultiplier tube enters the photon counter for measurement.

[0077] As an alternative embodiment, in some embodiments, the outer surface of the vacuum chamber 1 is provided with four CF16 quartz observation windows 3 and five CF63 quartz observation windows 4. Specifically, one CF63 quartz observation window 4 is used to detect the fluorescence signal, and the other CF63 quartz observation windows 4 and CF16 quartz observation windows 3 are used to input the 866 nm laser and the 397 nm laser required for laser cooling. Further, the ion trap furnace is provided with three pairs of magnetic field coils 8. As can be seen from Figure 3 the magnetic field coils 8 wound around the outer periphery of some of the said observation windows.

[0078] As an alternative embodiment, in some embodiments, it further includes two optical flats 9. The two optical flats 9 are horizontally arranged and have a stepped height. The bottom of the ion trap furnace is fixed by a bracket and forms a height difference with the ion pump 2.

[0079] As an alternative embodiment, in some embodiments, the microfabricated surface-electrode ion trap includes a central DC electrode, a ring RF electrode, and seven pairs of DC control electrodes. Referring to Figure 2 , the outer surface of the vacuum chamber 1 is provided with an ion trap electrode DC voltage signal connection port 5 and an ion trap electrode RF voltage signal connection port 6. Figure 5 It is a schematic diagram of the structure of the microfabricated surface-electrode ion trap of the ion trap system in the preferred embodiment. In the figure, RF is the RF electrode, A1, A7, B1, and B7 are the cap electrodes, and the other electrodes are compensation electrodes.

[0080] As an alternative embodiment, in some embodiments, the vacuum chamber 1 is provided with an electrode lead flange; the micro surface electrode ion trap is mounted on a socket, the electrodes of the micro surface electrode ion trap are connected to the electrodes of the socket, the pins of the socket are inserted into a PCB slot, the PCB slot is fixed on a fixing plate, a calcium furnace is fixed at the edge of the fixing plate, the fixing plate is fixed on the electrode lead flange, and the PCB slot is connected to the electrodes on the electrode lead flange. Further, the fixing plate is fixed on the electrode lead flange through a fixing rod. In the above embodiments, further, the socket is a ceramic socket, and the micro surface electrode ion trap is fixed on the ceramic socket through vacuum glue. In the above embodiments, further, the vacuum chamber 1 is provided with a voltage source connection flange, and the function of the voltage source connection flange is to connect with an external radio frequency voltage source.

[0081] As an alternative embodiment, in some embodiments, an ion pump 7 is further included, and the ion pump 7 is connected to the vacuum chamber 1. Specifically, the functions of the ion pump 2 and the ion pump 7 are to maintain the vacuum degree of the vacuum chamber 1 in the ion trap furnace. Further, the ion pump 2 and the ion pump 7 maintain the vacuum degree in the vacuum chamber 1 at 10 -9 Pa.

[0082] As an alternative embodiment, in some embodiments, a computer is further included, and the computer is connected to the CCD camera and the photon counter for controlling signals, and can further be connected to devices such as a signal source and a radio frequency switch for controlling signals.

[0083] In a complete embodiment, the method specifically may include the following steps:

[0084] Step 1: Maintain the vacuum chamber in a set vacuum environment, apply a DC voltage and a radio frequency high voltage to the micro surface electrode ion trap, and use the micro surface electrode ion trap to trap calcium ions.

[0085] In this step, the ion pump can be used to maintain the vacuum chamber in a 10 -9 Pa ultra-high vacuum environment, apply 15 independent and stable DC voltages and 1 radio frequency high voltage to 15 DC electrodes and radio frequency electrodes of the micro surface electrode ion trap respectively, so as to generate a three-dimensional trapping potential well on the electrode surface of the micro surface electrode ion trap, and ionize the calcium atoms ejected by the calcium furnace (or atomic furnace) through photoionization to generate 40 Ca + ions (or simply referred to as ions or calcium ions) and load them into the trapping potential well.

[0086] Step 2: Pass light and laser with a set wavelength through an observation window to cool the calcium ions to a temperature below the set temperature.

[0087] In this step, 397 nm light and 866 nm laser can be introduced through a CF35 quartz observation window to cool the ions to below 5 mK. Specifically, 397 nm light and 866 nm laser are introduced through the CF35 quartz observation window ( Figure 2-3 ) to perform Doppler cooling on the ions. The power of the 397 nm laser is set to 20 μW, the detuning is set to -30 MHz, and the power of the 866 nm laser is set to 200 μW. Then, by adjusting the voltage of the compensation electrode, the temperature of the ions is brought below 5 mK.

[0088] Step 3: At different laser powers, by changing the laser frequency and measuring the fluorescence line shape of calcium ions with a photon counter, and repeating the measurement data multiple times at each laser power.

[0089] In this step, at different 397 nm laser powers, the 397 nm laser frequency can be changed and the fluorescence line shape of the ions can be measured with a photon counter, and the fluorescence line shape is repeatedly measured multiple times at each 397 nm laser power.

[0090] This step can be divided into the following sub-steps:

[0091] Step 301: Use an optical switch to enable the power and frequency of the 397 nm laser to be switched between two sets of parameters, and the switching time is less than 10 μs. The power of one set of parameters is set to 20 μW and the detuning is set to -30 MHz. When the laser is under this parameter, it is defined as the cooling light. The power and detuning of the other set of parameters need to be adjusted at any time. When the laser is under this parameter, it is defined as the probe light;

[0092] Step 302: Let the detuning of the probe light vary from -50 MHz to +50 MHz. During this process, keep the power of the probe light at 1 μW. The change amount of the probe light detuning is set to +0.5 MHz. Each time the detuning changes, let the cooling light irradiate for 5 ms, then switch to the probe light to irradiate for 0.5 ms, and use a photon counter to record the number of photons emitted by the ions when the probe light irradiates. Take the detuning of the probe light as the x-axis and the measurement value of the photon counter as the y-axis to obtain the fluorescence line shape of the ions.

[0093] Step 4: According to the fluorescence line shape data of calcium ions obtained by measurement, calculate the saturation broadening and background at a specific laser power.

[0094] In this step, according to the data measured in Step 3, the saturation broadening and background can be calculated when the power of the 397 nm laser is 1 μW.

[0095] This step can be divided into the following sub-steps:

[0096] Step 401: Repeat Step 302 to measure 5 fluorescence line shapes respectively when the probe light power is 1 μW, 2 μW, 4 μW, 6 μW, 8 μW, and 10 μW. When there are no ions in the ion trap, repeat Step 4, and take the average of all the measured fluorescence line shapes at each power to obtain the background at each power.

[0097] Step 402: Calculate the saturation broadening and background when the power of the 397-nm laser scanning the ion fluorescence line shape is 1 μW based on the data measured in Step 302 and 401. Use the Lorentz function to fit the full width at half maximum (FWHM) of each fluorescence line shape. Let the horizontal axis be the cooling light power and the vertical axis be the average value of the measured FWHM, and then perform a linear fit to obtain the value of the FWHM when the laser power is zero. Subtract this value from the average value of the FWHM measured when the cooling light power is 1 μW to obtain the magnitude of the saturation broadening. 2 of the average value and then perform a linear fit to obtain the value of the FWHM when the laser power is zero. Subtract this value from the average value of the FWHM measured when the cooling light power is 1 μW to obtain the magnitude of the saturation broadening.

[0098] Step 5: Run the first program to obtain a calcium ion photo at a specific laser power.

[0099] In this step, Program A can be run. Set the power of the 397-nm light scanning the ion fluorescence line shape to 1 μW and measure at least 300 ion photos. Specifically, input the parameters measured in the above steps into Program A, set the power of the 397-nm light scanning the ion fluorescence line shape to 1 μW, run Program A, measure at least 300 ion photos, and set the output photo to a 20×20 pixel area centered on the ion.

[0100] Among them, Figure 5 is a schematic diagram of the operation logic of Program A. The main function of Program A is to scan the fluorescence line shape of ions while taking photos of the ions. Specifically, the entire program includes two main modules, the line shape scanning module and the photo taking module.

[0101] Specifically, the main function of the linear scanning module is to control the device to scan the fluorescence line shape of ions. The operation process of this module is as follows: 501. Generate a frequency list during the scanning line shape process according to the detuning range and step size of the entire scanning process input; 502. Set the detuning of the probe light according to the list and measure the fluorescence signal of the ions at this time; 503. Send a shooting command to the photo shooting module to make it record the photo of the ions at this time, and receive the photo number and the similarity of this photo returned by it; 504. Change the detuning of the probe light as the next value in the frequency list, and repeat steps 502 and 503 of the operation process of this module until the fluorescence signals of the corresponding ions are measured for all the detunings in the frequency list; 505. If the number of photos measured at this time is too small, continue to measure a certain number of photos, and save the photo number and similarity; 506. Save the frequency list and the corresponding fluorescence signal values, save the similarity of the ion photos taken at each measurement point, save the photo numbers taken in step 505 of the operation process of this module and their corresponding similarities.

[0102] Specifically, the main function of the photo shooting module is to control the CCD to take photos of ions, crop the part containing ions in the photo, and calculate the similarity. The operation process of this module is divided into a preparation stage and a measurement stage. The program will first run the preparation stage and then enter the measurement stage to wait for commands from the linear scanning module. The operation process of the preparation stage is as follows: 511. Control the CCD to take several photos of ions; 512. Take the average of the brightness values of each pixel point on these ion photos to generate a new photo, and intercept a small area of pixel points around the ion on this new photo. This intercepted area is called target_photo. The operation process of the measurement stage is as follows: 521. Wait for commands from the linear scanning module; 522. After receiving the shooting command, take a photo of the ions and number it; 523. Find the position of the ions and intercept a small area of pixel points around the ions. This area is called current_photo; 524. Compare target_photo and current_photo and calculate the similarity; 525. Save the taken photo and send the photo number and similarity to the linear scanning module; 526. Return to operation process 521.

[0103] Step 6: Adjust the temperature of the ions within the set temperature range, and measure the specific temperature of the ions and the corresponding calcium ion photos each time an adjustment is made.

[0104] In this step, the temperature of the ions can be changed within the range of 1 mK to 100 mK, and step 5 is repeated accordingly at each temperature. Specifically, adjust the voltage of a certain compensation electrode of the ion trap by 0.1 V, then repeat step 5 and measure at least 300 ion photos; repeat the above steps until at least 10,000 ion photos are measured in total.

[0105] Step 7: Run the second program to create a dataset;

[0106] In this step, run Program B to establish corresponding labels for the temperature corresponding to each ion photo. Those with a temperature between 0 - 10 mk are labeled 0, between 10 - 20 mk are labeled 1, between 20 - 30 mk are labeled 2, between 30 - 40 mk are labeled 3, and those greater than 40 mk are labeled 4.

[0107] Among them, Figure 6 is a schematic diagram of the operation logic of Program B. The main function of Program B is to create a dataset from the captured ion photos and the measured ion fluorescence line profiles.

[0108] Specifically, the operation process of the program is as follows: 601. Correlate the data points obtained from all scanned line profiles, the ion photos taken when measuring this data point, and the similarity values calculated when taking the photos; 602. Check all photos with a similarity lower than a certain value, and discard both the photos and their corresponding data points; 603. Perform Lorentzian fitting on each measured fluorescence line profile to calculate the temperature of the ions when scanning this fluorescence line profile. Mark the temperature of the photos taken after each scanned line profile during the measurement as the temperature calculated at this time; 604. Use the ion photos and their corresponding temperatures to create a dataset; 605. Randomly extract a small part of the data from the dataset as the test group, and the remaining as the training group, and then save them separately.

[0109] Step 8: Input the created dataset into the third program to train the neural network until the accuracy rate is greater than the set value;

[0110] In this step, input the created dataset into Program C to train the neural network until the accuracy rate is greater than 90%. The convolutional module of the established neural network contains four convolutional layers. The size of the convolutional kernels of the convolutional layers is 5×5, and the number of channels is 6, 16, 32, 64 respectively. The activation function is Relu. There are pooling layers and normalization layers between layers. The padding of the pooling layer is 2×1, and the stride is 2×1. The neural network contains an Inception module, and the parameters of the module are (128, (64, 128), (8, 32), 16). The linear module of the neural network contains 8 fully connected layers, and their number of channels is 4096, 2048, 1024, 258, 64, 32, 16, 5 respectively. The activation function is Relu, and there is a normalization layer between the fully connected layers. Each time of training, randomly select 300 ion photos with the same label and splice them vertically into one photo.

[0111] Among them, Figure 7It is a schematic diagram of the running logic of Program C. The main function of Program C is to train a neural network using a prepared dataset.

[0112] Specifically, the running process of the program is as follows: 801. Establish a neural network. The pictures input into the neural network first pass through several convolutional layers, then through several Inception blocks in GoogLeNet, and finally the results are output through a fully connected layer; 802. Set the loss function used during the training of the neural network as cross-entropy; 803. Randomly extract several pictures with the same label from the training set dataset, splice them into one picture as the input of the neural network, and then perform one training on the neural network; 804. Repeat the running process 803. Every time after training several rounds, use the test set dataset to test the accuracy until the accuracy is high enough; 805. Save the neural network parameters at this time.

[0113] Step 9. Input the ion photo of the temperature to be measured into the trained neural network and run Program D to obtain the calcium ion temperature.

[0114] After completing the training of the neural network, the measurement of the ion temperature can be completed only through the ion photo. The specific steps are as follows:

[0115] 901. Take an ion photo while keeping the ion trap confinement parameters and laser parameters unchanged. 902. Input the taken ion photo into Program D and run Program D to obtain the ion temperature.

[0116] Among them, Figure 8 It is a schematic diagram of the running logic of Program D. The main function of Program D is to use the trained neural network to judge the ion temperature according to the input ion photo.

[0117] Specifically, the running process of the program is as follows: 911. Establish the same neural network as in Program C and load the neural network parameters obtained in Program C; 912. Splice several input pictures into one picture as the input of the neural network; 913. Obtain the temperature of the ion from the output of the neural network.

[0118] In the description of this specification, the description referring to terms such as "one embodiment", "some embodiments", "example", "specific example", or "some examples", etc. means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0119] The above embodiments are only for illustrating the technical concept and features of the present invention, and the purpose is to enable those of ordinary skill in the art to understand the content of the present invention and implement it accordingly, and it cannot be used to limit the protection scope of the present invention. All equivalent changes or modifications made according to the essence of the content of the present invention should be covered within the protection scope of the present invention.

Claims

1. A method for measuring the ion temperature in an ion trap through a neural network and image measurement. The method is applicable to an ion trap system, which includes a calcium furnace, a vacuum chamber, an ion pump, and a micro surface electrode ion trap. The calcium furnace is used to eject calcium atoms. The vacuum chamber has several observation windows. The micro surface electrode ion trap is arranged in the vacuum chamber and is used to generate a three-dimensional confinement potential well. The calcium atoms ejected by the calcium furnace are ionized by photoionization to generate calcium ions and loaded into the three-dimensional confinement potential well. A part of the fluorescence emitted by the calcium ions enters a CCD through an imaging lens group, and the other part enters a photomultiplier tube. The signal output by the photomultiplier tube enters a photon counter for measurement. It is characterized in that, The method includes: Obtaining photos of calcium ions in a specific temperature range; Adjusting the temperature of calcium ions within a set temperature range, and repeatedly obtaining multiple photos of the calcium ions at each specific temperature range; Using the multiple photos of the calcium ions and their corresponding temperatures to make a data set; Inputting the made data set into a neural network for training the neural network until the accuracy rate is greater than a set value; Inputting a photo of calcium ions with the temperature to be measured into the trained neural network to obtain the temperature of the calcium ions.

2. The method for measuring the ion temperature in an ion trap through a neural network and image measurement according to claim 1, characterized in that, Before obtaining photos of calcium ions in a specific temperature range, it further includes the steps: Maintaining the vacuum chamber in a set vacuum environment, applying a DC voltage and a RF high voltage on the micro surface electrode ion trap, and using the micro surface electrode ion trap to confine calcium ions; Introducing a laser with a set wavelength through an observation window to cool the calcium ions to below the set temperature; At different laser powers, by changing the laser frequency and using a photon counter to measure the fluorescence line shape of calcium ions, and repeatedly measuring the fluorescence line shape data multiple times at each laser power; According to the measured fluorescence line shape data of calcium ions, calculating the saturation broadening and background at a specific laser power.

3. The method for measuring the ion temperature in an ion trap through a neural network and image measurement according to claim 1, characterized in that, Obtaining photos of calcium ions at a specific laser power by running a first program, which is executed by a line shape scanning module, specifically including the following steps: Step 501: Generating a frequency list during a scanning line shape process according to the detuning range and step size of the entire scanning process input; Step 502: Setting the detuning of the probe light according to the list and measuring the fluorescence signal of the ions at this time; Step 503: Sending a shooting command to a photo shooting module to make it record the photo of the ions at this time, and receiving the photo number and the similarity of this photo returned by it; Step 504: Changing the detuning of the probe light as the next value in the frequency list, and repeating steps 502 and 503 of the running process of this module until the fluorescence signals of the corresponding ions are measured for all the detunings in the frequency list of the probe light; Step 505: If the number of photos measured at this time is lower than the set number, continue to measure a certain number of photos and save the photo number and similarity; Step 506: Saving the frequency list and the values of the corresponding fluorescence signals, saving the similarity of the ion photos taken at each measurement point, and saving the photo numbers and their corresponding similarities of the photos taken in step 505 of the running process of this module.

4. The method for measuring the ion temperature in an ion trap by a neural network and image measurement according to claim 3, characterized in that, The first program is also executed by the photo shooting module. Its running process is divided into a preparation stage and a measurement stage. The preparation stage specifically includes the following steps: Step 511: Control the CCD to take several ion photos; Step 512: Take the average of the brightness values of each pixel point on these ion photos to generate a new photo, and intercept a one-pixel area around the ion on this new photo. The intercepted area is defined as target_photo; The measurement stage specifically includes the following steps: Step 521: Wait for the command from the line scan module; Step 522: After receiving the shooting command, take an ion photo and number it; Step 523: Locate the ion and intercept a one-pixel area around the ion. The intercepted area is defined as current_photo; Step 524: Compare target_photo and current_photo and calculate the similarity; Step 525: Save the taken photo and send the photo number and similarity to the line scan module; Step 526: Return to running step 521.

5. The method for measuring the ion temperature in an ion trap through a neural network and image measurement according to claim 1, characterized in that, By running the second program, a data set is made using multiple calcium ion photos. The second program specifically includes the following steps: Step 601: Correlate all the data points obtained from the scanned line patterns, the ion photos taken when measuring these data points, and the similarity values calculated when taking the photos; Step 602: Check all the photos with similarity lower than a certain value, and discard both the photos and their corresponding data points; Step 603: Perform Lorentz fitting on each measured fluorescence line pattern, and calculate the temperature of the ion when scanning this fluorescence line pattern; Mark the temperature of the photo taken after each line pattern scan during the measurement process as the temperature calculated at this time; Step 604: Use the ion photos and their corresponding temperatures to establish a data set; Step 605: Randomly extract some data from the data set as the test group, and the rest as the training group, and then save them respectively.

6. The method for measuring the ion temperature in an ion trap by a neural network and image according to claim 1, wherein By running the third program, the neural network is trained. The third program specifically includes the following steps: Step 801: Establish a neural network. Among them, the input pictures into the neural network first go through several convolutional layers, then go through several Inception blocks in GoogLeNet, and finally output the results through the fully connected layer; Step 802: Set the loss function used during the neural network training as cross entropy; Step 803: Randomly extract several pictures with the same label from the training group data set, splice them into one picture as the input of the neural network, and then perform one training on the neural network; Step 804: Repeat running step 803. Every time after training several rounds, use the test group data set to test the accuracy until the accuracy reaches the set standard; Step 805: Save the neural network parameters at this time.

7. The method for measuring the ion temperature in an ion trap through a neural network and image measurement according to claim 1, characterized in that, By running the fourth program, the temperature of the calcium ion is obtained. The fourth program specifically includes the following steps: Step 911: Establish the same neural network as in the third program and load the neural network parameters obtained in the third program; Step 912: Stitch a number of input pictures into one picture as the input of the neural network; Step 913: Obtain the temperature of the ions from the output of the neural network.

Images