Method, device and medium for optimizing generation parameters of waveforms driving quantum gates

Abstract

The application discloses an optimization method and device of a generation parameter of a waveform of a quantum gate and a medium, and relates to the technical field of quantum computing. The optimization method comprises the following steps: determining a waveform function of a target quantum gate, wherein the waveform function comprises a low-order power base and a high-order orthogonal correction base function; and updating a generation parameter. The step of updating the generation parameter comprises the following steps: executing the target quantum gate according to the waveform function; performing calibration testing on the executed target quantum gate to obtain a calibration testing result of the target quantum gate; and sequentially updating a first parameter of the low-order power base and a second parameter corresponding to the high-order orthogonal correction base function according to the calibration testing result to generate an updated waveform function. The optimization method can satisfy the expression ability of a low-dimensional physical parameter method, and can also reduce the cost of high-dimensional waveform optimization measurement.

Description

Technical Field

[0001] This application relates to the field of quantum computing technology, and more specifically, to a method, apparatus, and medium for optimizing the generation parameters of waveforms driving quantum gates. Background Technology

[0002] Quantum gates are the fundamental execution units for quantum algorithms and quantum error correction. As quantum processors increase in scale and circuit depth, gate fidelity directly determines the effective circuit depth, algorithm success rate, and whether error correction codes can enter a stable operating range. Especially in critical gate operations such as two-qubit entanglement gates and controlled phase gates, factors such as the amplitude, frequency / detuning, phase, envelope shape, and timing alignment of the driving waveform all contribute to gate error. Furthermore, given the crosstalk, frequency congestion, non-ideal response, and device bandwidth limitations in quantum control systems, gate errors often exhibit multi-parameter coupling characteristics, making gate calibration a long-term bottleneck in experimental operations.

[0003] In existing quantum control systems, a common approach to quantum gate calibration is to describe the quantum gate driving waveform using a small number of physically interpretable parameters, and then calibrate key parameters through experimental sequence extraction or parameter-by-parameter scanning. These parameters include, for example, the Rabi frequency corresponding to the amplitude, frequency / detuning, phase bias, pulse width, rise / fall time, and compensation terms (such as first / second-order AC Starkshift compensation). This method is characterized by clear physical meaning of the parameters and simplicity of implementation, making it suitable for initial coarse calibration and small-scale system maintenance. However, when quantum control systems exhibit complex errors such as bandwidth limitations, nonlinear responses, crosstalk, frequency congestion, and higher-order distortions caused by drift, relying solely on a small number of low-order parameters is often insufficient to characterize the residual error structure. This leads to calibration remaining at a "local optimum" or requiring frequent manual intervention. Furthermore, multi-parameter coupling reduces the efficiency of dimension-by-dimensional scanning.

[0004] Another common approach to quantum gate calibration involves introducing orthogonal basis functions or dimensionality reduction techniques. This involves projecting the control waveform onto mutually orthogonal or approximately orthogonal bases (e.g., orthogonal polynomials, Fourier orthogonal bases, Walsh basis functions), or constructing principal subspaces (e.g., principal component directions, effective subspaces) based on historical data or sensitivity information. This reduces the effective search dimensionality and improves parameter correlation. However, this approach typically emphasizes "mathematical decoupling" but lacks a unified organization and hierarchical strategy for physically interpretable parameters (amplitude, detuning, phase, AC Stark compensation, etc.). If orthogonal basis parameterization is used exclusively, it is often difficult to guarantee the rapid locking of key physical quantities, and in engineering practice, it is not easy to diagnose "which type of error is dominant." Furthermore, if the selection of orthogonal basis is not closely integrated with hardware constraints (boundary continuity, bandwidth, amplitude-phase coupling, quantization error, etc.), realizability and convergence problems can easily arise.

[0005] Therefore, current quantum gate calibration methods suffer from technical problems such as insufficient expressive power of low-dimensional physical parameters, high cost of high-dimensional waveform optimization measurement, and unstable convergence.

[0006] The content of the background section is merely technology known to the public and does not necessarily represent existing technology in the field. Summary of the Invention

[0007] This application aims to provide a method, apparatus, and medium for optimizing the generation parameters of the waveform driving a quantum gate, in order to solve the technical problems of insufficient expressive power of low-dimensional physical parameter methods, high cost of high-dimensional waveform optimization measurement, and unstable convergence in the current gate calibration methods.

[0008] According to one aspect of this application, a method for optimizing the generation parameters of a waveform driving a quantum gate is provided. The optimization method includes: determining a waveform function driving a target quantum gate, wherein the waveform function includes a low-order power basis and a high-order orthogonal correction basis function; and updating the generation parameters. The updating generation parameter step includes: executing the target quantum gate according to the waveform function; performing a calibration test on the executed target quantum gate to obtain the calibration test result of the target quantum gate; and sequentially updating the first parameter of the low-order power basis and the second parameter corresponding to the high-order orthogonal correction basis function according to the calibration test result to generate an updated waveform function.

[0009] According to some embodiments of this application, executing a target quantum gate based on a waveform function includes: determining a waveform modulation signal based on the waveform function; and executing the target quantum gate under the waveform modulation signal.

[0010] According to some embodiments of this application, determining a waveform modulation signal based on a waveform function includes: generating a digital waveform based on the waveform function; and determining a waveform modulation signal based on the digital waveform.

[0011] According to some embodiments of this application, updating the first parameter of the lower-order power basis and the second parameter corresponding to the higher-order orthogonal modified basis function in sequence according to the calibration test results includes: updating the first parameter according to the calibration test results when the second parameter remains unchanged or the constraint conditions change, so that the first parameter satisfies the first near-optimal region condition; and updating the second parameter according to the calibration test results when the first parameter satisfies the first near-optimal region condition, so that the second parameter satisfies the second near-optimal region condition.

[0012] According to some embodiments of this application, after updating the first parameter of the lower-order power basis and the second parameter corresponding to the higher-order orthogonal correction basis function in sequence according to the calibration test results to generate the updated waveform function, the optimization method further includes: determining whether the calibration test results meet the preset stopping condition; if not, taking the updated waveform function as the current waveform function, and executing the parameter update generation step again until the calibration test results meet the preset stopping condition.

[0013] According to some embodiments of this application, after sequentially updating the first parameter of the low-order power basis and the second parameter corresponding to the high-order orthogonal correction basis function based on the calibration test results to generate the updated waveform function, the optimization method further includes: determining whether the calibration test results meet the preset stopping condition; if not, using the updated waveform function as the current waveform function, executing the parameter update generation step again until the calibration test results meet the preset stopping condition; wherein, sequentially updating the first parameter of the low-order power basis and the second parameter corresponding to the high-order orthogonal correction basis function based on the calibration test results includes: updating the first parameter and the second parameter based on the updated calibration test results of the target quantum gate using a preset adaptive algorithm.

[0014] According to some embodiments of this application, updating the first and second parameters based on the calibration test results of the updated target quantum gate, based on a preset adaptive algorithm, includes: updating the first and second parameters based on the calibration test results of the updated target quantum gate using a preset adaptive algorithm and a preset higher-order mode activation algorithm. According to one aspect of this application, this application provides an optimization device for the generation parameters of the waveform driving a quantum gate. The optimization device includes a waveform parameter unit and a quantum gate execution unit. The waveform parameter unit determines the waveform function driving the target quantum gate, wherein the waveform function includes a low-order power basis and a higher-order orthogonal correction basis function. The waveform parameter unit and the quantum gate execution unit perform a parameter update step, including: the quantum gate execution unit executes the target quantum gate according to the waveform function; the quantum gate execution unit performs a calibration test on the executed target quantum gate to obtain the calibration test results of the target quantum gate; the waveform parameter unit sequentially updates the low-order power basis and the higher-order orthogonal correction basis function according to the calibration test results to generate an updated waveform function.

[0015] According to some embodiments of this application, when the second parameter remains unchanged or the limiting conditions change, the waveform parameter unit updates the first parameter based on the calibration test results so that the first parameter satisfies the first near-optimal region condition; when the first parameter satisfies the first near-optimal region condition, the waveform parameter unit updates the second parameter based on the calibration test results so that the second parameter satisfies the second near-optimal region condition.

[0016] The technical solution of this application can determine the waveform function that drives the target quantum gate. This technical solution can execute the target quantum gate using the waveform function and perform calibration tests on the target quantum gate to obtain calibration test results. This technical solution can then use the calibration test results to sequentially update the first parameter of the lower-order power basis and the second parameter corresponding to the higher-order orthogonal correction basis function to generate the updated waveform function.

[0017] The technical solution of this application can be formed by combining a low-order power basis (physically interpretable principal parameters) with a high-order orthogonal correction basis (residual error subspace) to form a waveform function, which has a unified parameterized calibration framework.

[0018] The technical solution of this application can first update the first parameter during the waveform function optimization and calibration process to quickly lock the dominant physical quantities such as amplitude, detuning, and phase, and then update the second parameter to efficiently absorb the high-order residuals introduced by hardware link distortion, crosstalk, and drift in the orthogonal subspace. This can not only satisfy the expression capability of the low-dimensional physical parameter method, but also reduce the cost of high-dimensional waveform optimization measurement and improve convergence stability, thereby improving the fidelity and long-term stability of the target quantum gate. Attached Figure Description

[0019] 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.

[0020] Figure 1 A flowchart illustrating an optimization method 1000 according to an embodiment of this application is shown;

[0021] Figure 2 A flowchart illustrating step S120 according to an embodiment of this application is shown;

[0022] Figure 3 A flowchart illustrating step S121 according to an embodiment of this application is shown;

[0023] Figure 4 A flowchart illustrating step S1211 according to an embodiment of this application is shown;

[0024] Figure 5 A flowchart illustrating step S123 according to an embodiment of this application is shown;

[0025] Figure 6 A flowchart illustrating the optimization method 2000 according to an embodiment of this application is shown;

[0026] Figure 7A flowchart illustrating the optimization method 3000 according to an embodiment of this application is shown;

[0027] Figure 8 A flowchart illustrating step S320 according to an embodiment of this application is shown;

[0028] Figure 9 A schematic diagram of the structure of an optimization device for generating waveforms of a driving quantum gate according to an embodiment of this application is shown.

[0029] Figure 10 A schematic diagram of the structure of a quantum gate execution unit according to an embodiment of this application is shown.

[0030] Explanation of reference numerals in the attached figures:

[0031] 40. Optimization device; 41. Waveform parameter unit; 42. Quantum gate execution unit.

[0032] 421. Waveform generation module; 422. Digital-to-analog front-end module; 423. Quantum processor; 424. Measurement and data acquisition module. Detailed Implementation

[0033] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the embodiments set forth herein; rather, they are provided so that this application will be thorough and complete, and will fully convey the concept of the exemplary embodiments to those skilled in the art. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted.

[0034] The described features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. Numerous specific details are provided in the following description to give a full understanding of embodiments of this disclosure. However, those skilled in the art will recognize that the technical solutions of this disclosure can be practiced without one or more of these specific details, or other methods, components, materials, devices, etc. In these cases, well-known structures, methods, devices, implementations, materials, or operations will not be shown or described in detail.

[0035] Furthermore, the terms “comprising” and “having”, and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the steps or units listed, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus.

[0036] The terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish different objects, not to describe a specific order.

[0037] The technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0038] The English terms used in this application, their full English names, and their corresponding Chinese definitions are as follows:

[0039] DDS, Direct Digital Synthesis;

[0040] NCO stands for Numerically Controlled Oscillator.

[0041] LUT, Look-Up Table;

[0042] DAC, Digital-to-Analog Converter;

[0043] AC Stark shift.

[0044] Figure 9 This diagram illustrates the structure of an optimization device for generating waveforms of a driving quantum gate according to an embodiment of this application; see also... Figure 9 The optimization device 40 includes a waveform parameter unit 41 and a quantum gate execution unit 42.

[0045] See Figure 10 The quantum gate execution unit 42 may include a waveform generation module 421, a digital-to-analog front-end module 422, a quantum processor 423, and a measurement and data acquisition module 424.

[0046] The following is combined with Figure 9 and Figure 10 This application describes a method 1000 for optimizing the generation parameters of waveforms driving quantum gates. See also... Figure 1 The optimization method 1000 may include steps S110-S120.

[0047] In step S110, the waveform function driving the target quantum gate is determined.

[0048] According to the example embodiment, the target quantum gate can be the quantum gate to be executed. The target quantum gate can be a single-qubit gate or a multi-qubit gate. The waveform function can be the mathematical expression of the time-varying control signal that implements the operation of the target quantum gate. The waveform function includes low-order power bases and high-order orthogonal correction basis functions. The low-order power base can be a function of parameters (e.g., amplitude, phase, frequency, detuning, bandwidth, etc.) with corresponding physical meanings in the waveform function. The low-order power base can describe the overall trend and slowly varying part of the waveform function.

[0049] Higher-order orthogonal correction basis functions can be functions of the residual error subspace parameters in the waveform function. They are a set of mutually orthogonal functions within a specific interval (such as the target quantum gate operating time) and typically exhibit oscillatory characteristics. Higher-order orthogonal correction basis functions can be used as correction terms to achieve finer and more complex control.

[0050] Configuration parameters can be parameters of waveform functions. For example, configuration parameters may include the target quantum gate type, initial parameters, scanning strategy, and preset stopping conditions. The target quantum gate type can be a single-qubit quantum gate or a two-qubit quantum gate, etc. Initial parameters can be initial parameters for amplitude, phase, frequency, bandwidth, boundary conditions, and their respective constraints. The scanning strategy can be the scanning range and scanning step size of the parameters to be optimized. Preset stopping conditions can be preset conditions for the generation parameters of the optimized quantum gate waveform. For example, preset stopping conditions may include calibration test results falling below a threshold, reaching the required number of optimization iterations, etc.

[0051] For example, in step S110, the waveform parameter unit 41 can receive user instructions through a user interface to determine the waveform function driving the target quantum gate. The user instructions can be the form of the waveform function and the initial values ​​of the configuration parameters determined by the user based on the theoretical design, simulation calculation, prior knowledge, or experimental parameters of the target quantum gate. The waveform parameter unit 41 can determine the waveform function based on the form of the waveform function (including the form of lower-order power basis functions and higher-order orthogonal correction basis functions) and the initial values ​​of the configuration parameters (including the configuration parameters of lower-order power basis functions and higher-order orthogonal correction basis functions) given by the user instructions.

[0052] Step S120 is the step to update the generated parameters. See also... Figure 2 Step S120 may include steps S121-S123.

[0053] In step S121, the target quantum gate is executed according to the waveform function.

[0054] For example, in step S121, the quantum gate execution unit 42 can determine the waveform modulation signal based on the waveform function. The quantum gate execution unit 42 can then execute the target quantum gate under the waveform modulation signal to obtain observation data.

[0055] In step S122, the target quantum gate after execution is calibrated to obtain the calibration test results of the target quantum gate.

[0056] According to an example embodiment, calibration testing can be a test to adjust the execution quality of a target quantum gate. The calibration test result can be an evaluation of the execution of the target quantum gate after execution, and can be used as a mathematical scalar function to quantify the difference or closeness between the actually implemented target quantum gate operation and the ideal target quantum gate. The calibration test result may include gate fidelity and its associated parameters.

[0057] For example, calibration tests can include quantum process tomography, random benchmarking, and channel spectrum benchmarking.

[0058] In step S122, the quantum gate execution unit 42 can perform a calibration test on the executed target quantum gate to obtain the calibration test results of the target quantum gate (e.g., gate fidelity).

[0059] In step S123, the first parameter of the lower-order power basis and the second parameter of the higher-order orthogonal correction basis function are updated sequentially according to the calibration test results to generate the updated waveform function.

[0060] According to the example embodiment, the first parameter can be a vector parameter formed by the various physical configuration parameters (e.g., amplitude, phase, frequency, detuning, bandwidth, etc.) in the lower-order power basis. The second parameter can be the configuration coefficients corresponding to the various parameters of the higher-order orthogonal correction basis function.

[0061] For example, in step S123, the waveform parameter unit 41 can update the first parameter based on the calibration test results, provided that the second parameter remains unchanged or the limiting conditions change, so that the first parameter satisfies the first near-optimal region condition. The waveform parameter unit 41 can also update the second parameter based on the calibration test results, provided that the first parameter satisfies the first near-optimal region condition, so that the second parameter satisfies the second near-optimal region condition, thereby sequentially updating the first and second parameters to generate the updated waveform function.

[0062] Through the above embodiments, the technical solution of this application can determine the waveform function driving the target quantum gate. The technical solution of this application can execute the target quantum gate through the waveform function and perform calibration tests on the target quantum gate to obtain calibration test results. The technical solution of this application can sequentially update the first parameter of the lower-order power basis and the second parameter corresponding to the higher-order orthogonal correction basis function based on the calibration test results to generate an updated waveform function.

[0063] The technical solution of this application can be formed by combining a low-order power basis (physically interpretable principal parameters) with a high-order orthogonal correction basis (residual error subspace) to form a waveform function, which has a unified parameterized calibration framework.

[0064] The technical solution of this application can first update the first parameter during the waveform function optimization and calibration process to quickly lock the dominant physical quantities such as amplitude, detuning, and phase, and then update the second parameter to efficiently absorb the high-order residuals introduced by hardware link distortion, crosstalk, and drift in the orthogonal subspace. This can not only satisfy the expression capability of the low-dimensional physical parameter method, but also reduce the cost of high-dimensional waveform optimization measurement and improve convergence stability, thereby improving the fidelity and long-term stability of the target quantum gate.

[0065] Alternatively, the waveform function can be represented as follows:

[0066]

[0067] in, For waveform functions; For the low-order parameter vector The determined lower-order power basis, lower-order parameter vector That is, the first parameter; N is the order of the lower-order power basis.

[0068] These are higher-order orthogonal corrected basis functions; The coefficients corresponding to the higher-order orthogonal correction basis functions are the second parameter. M is the total number of available higher-order orthogonal modes.

[0069] t is time, T represents the duration of the target quantum gate.

[0070] The waveform functions described above can be used for both the amplitude channel and the phase / frequency channel. For example, for carrier drive, the amplitude envelope and phase can be represented using the waveform functions described above, and the frequency offset can also be represented using the waveform functions described above and the phase obtained by integration.

[0071] Alternatively, a low-order power basis can be represented as follows:

[0072]

[0073] in, It is a low-order power; To normalize time, , .

[0074] For the sub-parameters of the k-th lower-order parameter vector, Mapping to parameters with clear physical meaning, such as amplitude calibration coefficients, detuning compensation, phase offset, pulse width, simple rising / falling edge parameters, and first / second order AC Stark shift compensation terms, enables the first layer scan to quickly lock onto the dominant physical quantities.

[0075] Optionally, higher-order orthogonal correction basis functions can be generated using orthogonal basis generation, for example, in We select orthogonal polynomials (e.g., Legendre polynomials, Chebyshev polynomials) or families of orthogonal functions, and ensure that the inner products of the weight functions are orthogonal. Higher-order orthogonal modified basis functions can be expressed as follows:

[0076]

[0077] in, These are higher-order orthogonal corrected basis functions; The weight function; For Kroneck's symbol.

[0078] The waveform parameter unit 41 can be obtained by starting from the candidate function set (piecewise polynomials, splines, window functions, Fourier / Walsh candidates, etc.) and performing Gram-Schmidt orthogonalization on the candidate functions according to the discrete-time grid. .

[0079] The waveform parameter unit 41 can construct a sensitivity matrix based on the local sensitivity information of the waveform parameters according to the calibration test results, and orthogonalize its principal direction to obtain a result that more closely approximates the "error reduction direction". This is to improve the effectiveness of each scan.

[0080] Optionally, see Figure 3 Step S121 may include steps S1211-S1212.

[0081] In step S1211, the waveform modulation signal is determined based on the waveform function.

[0082] According to an example embodiment, the waveform modulation signal can be a signal that controls the waveform of the target quantum gate. For example, the waveform modulation signal can be a radio frequency modulation signal, a microwave modulation signal, and an optical modulation signal.

[0083] Optionally, see Figure 4 Step S1211 may include steps S12111 and S12112.

[0084] In step S12111, a digital waveform is generated based on the waveform function.

[0085] According to the example embodiment, the digital waveform can be a discrete control signal that executes the target quantum gate. For example, in step S12111, the waveform generation module 421 generates a digital waveform according to a waveform function. The waveform generation module 421 can be a DDS module, an NCO module, an amplitude envelope generator, a phase accumulator, a LUT module, or a piecewise polynomial calculation module, etc.

[0086] In step S12112, the waveform modulation signal is determined based on the digital waveform.

[0087] For example, in step S12112, the analog-to-digital front-end module 422 can determine the waveform modulation signal based on the digital waveform. The analog-to-digital front-end module 422 can generate the waveform modulation signal after performing digital-to-analog conversion, amplification, mixing (or modulation), and filtering on the digital waveform.

[0088] According to an example embodiment, the digital-to-analog and analog front-end module 422 may include circuits such as a DAC, amplifier, mixer / modulation link, and filter.

[0089] In step S1212, the target quantum gate is executed under the waveform modulation signal.

[0090] According to the example embodiment, the observation data can be the observable measurement results output after performing the target quantum gate operation. The observation data may include data such as state occupancy, benchmark counts, and error synthesis indices.

[0091] For example, in step S1212, the quantum processor 423 can execute the target quantum gate under the waveform signal to manipulate the state of the qubit. After executing the target quantum gate, the quantum processor 423 outputs the observation data.

[0092] The measurement and data acquisition module 424 can perform calibration tests on the target quantum gate after execution based on the observation data to obtain the calibration test results of the target quantum gate (e.g., gate fidelity).

[0093] Through the above embodiments, the technical solution of this application can determine the waveform modulation signal through a waveform function. The technical solution of this application can execute a target quantum gate under a waveform modulation signal.

[0094] The technical solution of this application generates waveforms in a parameterized manner, reducing the overhead of sending complete waveform data. This helps to improve the closed-loop iteration rate and reduce the impact of delay on the calibration effect, thereby supporting higher frequency online maintenance and rapid calibration.

[0095] Optionally, see Figure 5 Step S123 may include steps S1231 and S1232.

[0096] In step S1231, if the second parameter remains unchanged or the limiting conditions change, the first parameter is updated according to the calibration test results so that the first parameter satisfies the first near-optimal region conditions.

[0097] According to the example embodiment, the limiting condition can be a preset constraint on the second parameter, for example, the limiting condition can be the numerical range of each coefficient in the second parameter. The first near-optimal region condition can be the range of the first parameter when the waveform function is close to the ideal waveform.

[0098] For example, in step S1231, the waveform parameter unit 41 can adjust the first parameter based on the calibration test results, provided that the second parameter remains unchanged or the limiting conditions change. Perform a scan or optimization update to obtain... This ensures that the first parameter satisfies the conditions of the first near-optimal region.

[0099] Scanning methods can include parameter-by-parameter scanning, coordinate descent, small-scale grid, and local quadratic fitting.

[0100] Optionally, the waveform parameter unit 41 may employ a local search of a small number of sampling points for each sub-parameter in the first parameter to reduce the number of measurements.

[0101] In step S1232, if the first parameter satisfies the first near-optimal region condition, the second parameter is updated according to the calibration test results so that the second parameter satisfies the second near-optimal region condition.

[0102] According to the example embodiment, the second near-optimal region condition can be the range of the second parameter when the waveform function is close to the ideal waveform. For example, in step S1231, the waveform parameter unit 41 can adjust the second parameter based on the calibration test results when the first parameter satisfies the first near-optimal region condition. Perform a scan or optimization update to obtain... This ensures that the second parameter satisfies the second near-optimal region condition.

[0103] Optionally, the waveform parameter unit 41 may update the second parameter using coordinates along a single transverse direction, or update the second parameter using coordinates along a single transverse direction in conjunction with small-scale multimodal analysis.

[0104] Through the above embodiments, the technical solution of this application can update the first parameter based on the calibration test results while keeping the second parameter unchanged or changing the limiting conditions, so that the first parameter satisfies the first near-optimal region condition. The technical solution of this application can also update the second parameter based on the calibration test results while ensuring the first parameter satisfies the first near-optimal region condition, so that the second parameter satisfies the second near-optimal region condition.

[0105] The technical solution of this application can reduce the number of measurements required for calibration and shorten the calibration time by first updating the first parameter (which has dominant physical parameters such as amplitude, detuning, and phase) and then making a small correction in the orthogonal subspace (i.e., updating the second parameter). This reduces the invalid scans and repeated measurements caused by parameter coupling during the update process, thereby reducing the number of measurements required for calibration and shortening the calibration time, thus improving efficiency.

[0106] The technical solution of this application can absorb residual errors such as waveform distortion and crosstalk introduced by the hardware link by introducing high-order orthogonal correction basis functions. On the basis of the alignment of low-order parameters, it further reduces gate error, making the final gate performance more stable and batch consistency better.

[0107] The following is combined with Figure 9 and Figure 10 This application describes a method for optimizing the generation parameters of waveforms driving quantum gates (2000). See also... Figure 6 The optimization method 2000 may include steps S210-S240.

[0108] Referring to the figure, steps S210-S220 are the same as steps S110-S120, so they will not be described again.

[0109] In step S230, it is determined whether the calibration test results meet the preset stop conditions.

[0110] According to the example embodiment, the preset stopping condition can be a preset condition for the generation parameters of the optimized quantum gate waveform. For example, the preset stopping condition may include the calibration test result being lower than a threshold, reaching the optimization iteration number, or the calibration test result being lower than the threshold for a certain number of consecutive times.

[0111] If the waveform parameter unit 41 determines that the calibration test result meets the preset stop condition, then the update and optimization of the first and second parameters can be stopped.

[0112] If the waveform parameter unit 41 determines that the calibration test result does not meet the preset stop condition, then step S240 is executed.

[0113] In step S240, the updated waveform function is used as the current waveform function, and the parameter update generation step is executed again until the calibration test result meets the preset stop condition.

[0114] For example, in step S240, the waveform parameter unit 41 uses the updated waveform function as the current waveform function and executes step S220 again until the calibration test result meets the preset stop condition.

[0115] Through the above embodiments, the technical solution of this application can determine whether the calibration test result meets the preset stop condition. If it is determined that the calibration test result does not meet the preset stop condition, the technical solution of this application can use the updated waveform function as the current waveform function and execute the parameter update generation step again until the calibration test result meets the preset stop condition.

[0116] The following is combined with Figure 9 and Figure 10 This application describes a method 3000 for optimizing the generation parameters of waveforms driving quantum gates. See also... Figure 7 The optimization method 3000 may include steps S310-S340.

[0117] Referring to the figure, steps S310-S330 are the same as steps S210-S230, so they will not be described again.

[0118] If the waveform parameter unit 41 determines that the calibration test result meets the preset stop condition, then the update and optimization of the first and second parameters can be stopped.

[0119] If the waveform parameter unit 41 determines that the calibration test result does not meet the preset stop condition, then step S340 is executed.

[0120] In step S340, the updated waveform function is used as the current waveform function, and the parameter update generation step is executed again until the calibration test result meets the preset stop condition.

[0121] For example, in step S340, the waveform parameter unit 41 uses the updated waveform function as the current waveform function and executes step S320 again until the calibration test result meets the preset stop condition.

[0122] See Figure 8 Step S320 may include steps S321-S323. In the case of the first execution of step S320, steps S321-S323 may be the same as steps S121-S123. In subsequent executions of step S320, step S323 may specifically involve: updating the first and second parameters based on a preset adaptive algorithm and the updated calibration test results of the target quantum gate.

[0123] According to the example embodiment, the preset adaptive algorithm can be a preset algorithm that updates the first parameter and the second parameter based on the calibration test results. The preset adaptive algorithm may include the order in which the first parameter and the second parameter are updated in the current iteration, the scan range and scan step size of the first parameter, and the scan range and scan step size of the second parameter, etc.

[0124] For example, in step S323, the waveform parameter unit 41 can determine the scanning range and scanning step size of the first parameter, the scanning range and scanning step size of the second parameter, and the order of updating the first parameter and the second parameter in the current iteration based on the calibration test results of the updated target quantum gate through a preset adaptive algorithm.

[0125] For example, the waveform parameter unit 41 can determine the scanning range and scanning step size of the first parameter, the scanning range and scanning step size of the second parameter, and the order of updating the first and second parameters in the current iteration based on the improvement of the calibration test results of the most recent rounds, through a preset adaptive algorithm.

[0126] Through the above embodiments, the technical solution of this application can adjust the order of updating the first parameter and the second parameter, the scanning range and scanning step size of the first parameter, and the scanning range and scanning step size of the second parameter in the current iteration by using a preset adaptive algorithm, so as to quickly locate and compensate for the main change direction when the parameters drift over time.

[0127] Optionally, step S323 may further be specifically: based on a preset adaptive algorithm and a preset higher-order mode activation algorithm, update the first parameter and the second parameter according to the calibration test results of the updated target quantum gate.

[0128] According to the example embodiment, the preset higher-order mode activation algorithm can be a preset coefficient algorithm that selects a portion of the second parameters corresponding to the higher-order orthogonal correction basis function. For example, if the higher-order orthogonal correction basis function includes M modes, the waveform parameter unit 41 can use the preset higher-order mode activation algorithm to select a portion of the second parameters corresponding to the first m modes to form an activation set for updating. The activation set can be represented as: .

[0129] The activation criteria for selecting the first m modes can include one or more of the following: sensitivity criterion, contribution criterion, and threshold criterion.

[0130] The sensitivity criterion can be: by making small perturbations to each mode. Estimated calibration test results disturbance Select The largest of the top m modes.

[0131] The contribution criterion can be: by calculating the decrease in calibration test results caused by each mode in the most recent iterations, select the top m modes with the largest contribution to the bad behavior.

[0132] The threshold criterion can be: calculate the perturbation of the calibration test results corresponding to each mode. Less than the corresponding preset threshold, or the modal coefficient (i.e., the second parameter). If the value is less than the corresponding preset threshold for several rounds, then this mode is selected.

[0133] For example, in step S323, the waveform parameter unit 41 can select the mode coefficients of the first m modes according to the calibration test results of the updated target quantum gate through a preset high-order mode activation algorithm, and determine the scanning range and scanning step size of the first parameter, the scanning range and scanning step size of the second parameter, and the order of updating the first and second parameters in the current iteration through a preset adaptive algorithm.

[0134] Through the above embodiments, the technical solution of this application can reduce the need for frequent full recalibration and improve long-term operational stability by iteratively updating the modal coefficients (i.e., the second parameter) of the first m modes through a preset high-order modal activation algorithm. The technical solution of this application can be applied to pulse / waveform control links commonly found in various quantum computing physics platforms (e.g., neutral atoms, trapped ions, superconductors, etc.), and is suitable for rapid tuning and drift compensation of gate parameters in microwave / RF driving, optical driving, and their combined driving scenarios.

[0135] The following describes the optimization method of this application in detail, taking a single-qubit or two-qubit quantum gate as an example, including the following steps.

[0136] S1: The waveform parameter unit determines whether the target quantum gate is a single-bit gate or a two-bit gate, and determines the target quantum gate as T.

[0137] S2: The waveform parameter unit can determine the first parameter p of the low-order power basis, which includes amplitude calibration coefficients, detuning compensation terms, and phase bias terms (or equivalent low-order power basis coefficients). The waveform parameter unit can construct M high-order orthogonal modified basis functions on [0, T] (which can be orthogonal polynomials or orthogonal functions obtained through Gram-Schmidt orthogonalization), initially... This generates a complete waveform function.

[0138] S3: The quantum gate execution unit can determine the waveform modulation signal based on the waveform function. The quantum gate execution unit can execute the target quantum gate under the waveform modulation signal, and perform calibration tests on the executed target quantum gate to obtain the calibration test results of the target quantum gate.

[0139] S4: The waveform parameter unit can fix q, and perform a small number of point scans on p based on the calibration test results to find the p that minimizes the calibration test results, thus obtaining the updated p. The waveform parameter unit can be updated by making small perturbations to each mode. Estimated calibration test results disturbance Select The top m largest modes form the activation set A. With p fixed, only the... Perform coordinate updates or a small-range scan to obtain the updated q.

[0140] S5: The waveform parameter unit determines whether the calibration test results meet the preset stop conditions.

[0141] S6: If the waveform parameter unit determines that the calibration test results meet the preset stop conditions, then the waveform parameter unit will output the updated p and q.

[0142] S7: If the waveform parameter unit determines that the calibration test result does not meet the preset stop condition, then repeat S3-S5 to continue the iteration.

[0143] S8: During the operation of the optimization device, the above S3-S7 updates can be repeated according to a set period or triggering conditions (such as an increase in error index or an increase in calibration test results) to track drift.

[0144] Through the above embodiments, the technical solution of this application can quickly align the dominant physical quantities through low-order power basis, absorb residual errors in orthogonal coordinates with fewer measurement times through high-order orthogonal correction basis functions, and avoid high-dimensional full search through preset high-order modal activation algorithm.

[0145] According to another aspect of this application, this application also provides a non-volatile computer-readable storage medium storing a computer program thereon, which, when executed by a processor, is capable of implementing the method for optimizing the generation parameters of the waveform driving the quantum gate as described above.

[0146] According to another aspect of this application, this application also provides an electronic device, including: one or more processors; and a storage device for storing one or more programs, which, when executed by the one or more processors, enable the one or more processors to implement the optimization method for generating waveforms of driving quantum gates as described above.

[0147] According to another aspect of this application, this application also provides a computer program product, comprising: a computer program stored on a computer-readable storage medium; the computer program includes program instructions that, when executed by a computer, cause the computer to perform the optimization method for generating waveforms driving quantum gates as described above.

[0148] Finally, it should be noted that the above description is merely a preferred embodiment of this application and is not intended to limit this application. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions of the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for optimizing the generation parameters of the waveform driving a quantum gate, characterized in that, The optimization method includes: Determine the waveform function driving the target quantum gate, wherein the waveform function includes a low-order power basis and a high-order orthogonal correction basis function; The steps to update the generated parameters include: Execute the target quantum gate according to the waveform function; The target quantum gate is calibrated after execution to obtain the calibration test results of the target quantum gate; Based on the calibration test results, the first parameter of the lower-order power basis and the second parameter corresponding to the higher-order orthogonal correction basis function are updated sequentially to generate the updated waveform function.

2. The optimization method according to claim 1, characterized in that, The step of executing the target quantum gate according to the waveform function includes: The waveform modulation signal is determined based on the waveform function; The target quantum gate is executed under the waveform modulation signal.

3. The optimization method according to claim 2, characterized in that, Determining the waveform modulation signal based on the waveform function includes: Generate a digital waveform based on the waveform function; The waveform modulation signal is determined based on the digital waveform.

4. The optimization method according to claim 1, characterized in that, The step of sequentially updating the first parameter of the low-order power basis and the second parameter corresponding to the high-order orthogonal correction basis function based on the calibration test results includes: If the second parameter remains unchanged or the constraints change, the first parameter is updated according to the calibration test results so that the first parameter satisfies the first near-optimal region condition. If the first parameter satisfies the first near-optimal region condition, the second parameter is updated according to the calibration test results so that the second parameter satisfies the second near-optimal region condition.

5. The optimization method according to claim 1, characterized in that, After updating the first parameter of the lower-order power basis and the second parameter corresponding to the higher-order orthogonal correction basis function sequentially according to the calibration test results to generate the updated waveform function, the optimization method further includes: Determine whether the calibration test results meet the preset stop conditions; If not, using the updated waveform function as the current waveform function, execute the update parameter generation step again until the calibration test result meets the preset stop condition.

6. The optimization method according to claim 1, characterized in that, After updating the first parameter of the lower-order power basis and the second parameter corresponding to the higher-order orthogonal correction basis function sequentially according to the calibration test results to generate the updated waveform function, the optimization method further includes: Determine whether the calibration test results meet the preset stop conditions; If not, using the updated waveform function as the current waveform function, execute the update parameter generation step again until the calibration test result meets the preset stop condition; The step of sequentially updating the first parameter of the lower-order power basis and the second parameter corresponding to the higher-order orthogonal correction basis function based on the calibration test results includes: Based on a preset adaptive algorithm, the first parameter and the second parameter are updated according to the calibration test results of the updated target quantum gate.

7. The optimization method according to claim 6, characterized in that, The step of updating the first parameter and the second parameter based on the calibration test results of the updated target quantum gate using a preset adaptive algorithm includes: Based on the preset adaptive algorithm and the preset higher-order mode activation algorithm, the first parameter and the second parameter are updated according to the calibration test results of the updated target quantum gate.

8. A device for optimizing the generation parameters of the waveform driving a quantum gate, characterized in that, The optimization device includes a waveform parameter unit and a quantum gate execution unit; The waveform parameter unit determines the waveform function driving the target quantum gate, wherein the waveform function includes a low-order power basis and a high-order orthogonal correction basis function; The waveform parameter unit and the quantum gate execution unit perform the parameter update generation step, including: The quantum gate execution unit executes the target quantum gate according to the waveform function, and the quantum gate execution unit performs a calibration test on the executed target quantum gate to obtain the calibration test result of the target quantum gate; The waveform parameter unit updates the first parameter of the low-order power basis and the second parameter corresponding to the high-order orthogonal correction basis function in sequence according to the calibration test results to generate the updated waveform function.

9. The optimization device according to claim 8, characterized in that, When the second parameter remains unchanged or the limiting conditions change, the waveform parameter unit updates the first parameter according to the calibration test results so that the first parameter satisfies the first near-optimal region condition. When the first parameter satisfies the first near-optimal region condition, the waveform parameter unit updates the second parameter according to the calibration test result so that the second parameter satisfies the second near-optimal region condition.

10. A non-volatile computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the method for optimizing the generation parameters of the waveform driving the quantum gate as described in any one of claims 1-7.

Images