System noise risk feedback strategy optimization method and device, equipment and medium

Abstract

The invention provides a system noise risk feedback strategy optimization method and device, equipment and a medium, and the method comprises the following steps: obtaining noise data in a virtual-real experiment space; performing noise risk analysis on the noise data according to uncertainty risk factors of different types of quantum states to obtain a risk measurement value and a noise risk factor sequence; calling a corresponding risk control strategy according to the risk measurement value and the noise risk factor sequence, and performing optimization analysis on the risk control strategy; verifying the risk control strategy, if the risk control strategy meets a threshold requirement, directly outputting the risk control strategy, and if the risk control strategy does not meet the threshold requirement, carrying out feedback control and returning, and obtaining noise data again until the threshold requirement is met. The adaptability of the risk control strategy to the system noise risk is ensured through selection and optimization of the strategy, and the risk measurement process is optimized through dynamic iteration.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of quantum state preparation and manipulation, and particularly relates to a system noise risk feedback strategy optimization method, device, equipment and medium. BACKGROUND

[0002] Quantum computing takes quantum states as information carriers and relies on quantum superposition, entanglement and other quantum mechanics principles to realize information parallel computing. The core challenge of superconducting quantum computers lies in the influence of quantum bit noise and errors. Reducing such noise interference, building a standard quantum state hardware platform and implementing intelligent risk management throughout the process are key issues for quantum computing engineering. How to replace hardware functions with software-defined information systems to improve quantum computing fidelity is particularly important.

[0003] In related technologies, in the process of quantum state preparation and manipulation, there is a lack of top-level integrated design, and only functions are realized from a single link: some focus on quantum state preparation or manipulation, and some focus on measurement accuracy. The correspondence between physical experimental space and virtual simulation space is not established, and the strategy for system noise risk in the process of quantum state preparation and manipulation is not optimized and feedback controlled. SUMMARY

[0004] The purpose of the present application is to provide a system noise risk feedback strategy optimization method, device, equipment and medium to solve the above problems.

[0005] To this end, in a first aspect, the application provides a hybrid computing task classification method, comprising the following steps:

[0006] Obtain noise data in virtual and real experimental spaces;

[0007] Perform noise risk analysis on the noise data according to the uncertainty risk factors of different types of quantum states to obtain risk measure values and noise risk factor rankings;

[0008] Call corresponding risk control strategies according to the risk measure values and the noise risk factor rankings and perform optimization analysis on the risk control strategies;

[0009] Verify the risk control strategies. If the risk control strategies meet the risk threshold requirements, the risk control strategies are directly output. If the risk control strategies do not meet the threshold requirements, feedback control is performed, and noise data is reacquired until the threshold requirements are met.

[0010] Optionally, the type of quantum state at least includes one of a single-bit quantum state, a two-bit quantum state, and a multi-bit quantum system quantum state, and the uncertainty risk factor at least includes one of a microscopic noise interference uncertainty factor, a macroscopic noise interference uncertainty factor, and a quantum bit preparation manipulation means noise uncertainty factor.

[0011] Optionally, the noise risk analysis according to the uncertainty risk factors of different types of quantum states to obtain a risk measure value and a noise risk factor ranking includes:

[0012] identifying an uncertainty risk factor according to the type of quantum state;

[0013] determining a risk measure value of the different quantum states under the uncertainty risk factor based on an entropy algorithm according to the uncertainty risk factor;

[0014] obtaining a noise risk factor ranking based on a relative entropy ranking algorithm according to the risk measure value.

[0015] Optionally, the obtaining of the noise risk factor ranking based on the relative entropy ranking algorithm according to the uncertainty risk factor includes:

[0016] determining whether feedback control is needed for the current feedback control node, and if feedback control is needed, calling a standard threshold value from a risk threshold value library, calculating the relative entropy between the risk measure value and the standard threshold value, ranking the uncertainty risk factors according to the size of the relative entropy, and obtaining a noise risk factor ranking.

[0017] Optionally, the calling of the corresponding risk control strategy according to the risk measure value and the noise risk factor ranking includes:

[0018] calling a risk control strategy from a strategy library according to the risk measure value and the noise risk factor ranking.

[0019] Optionally, the calling of the risk control strategy from the strategy library according to the risk measure value and the noise risk factor ranking includes:

[0020] calling a corresponding strategy set from the strategy library according to the type of quantum state;

[0021] determining a strategy measure value of each risk control strategy in the strategy set according to quantum gate fidelity, coherence time, and background system noise;

[0022] selecting a risk control strategy with the highest strategy measure value in the strategy set.

[0023] Optionally, the feedback control and the return to reacquire noise data until the threshold value requirement is met if the risk control strategy does not meet the threshold value requirement include:

[0024] According to the noise risk factor ranking analysis factor correlation, a parameter early warning prompt of the virtual and real experimental space is given;

[0025] According to the parameter early warning prompt, noise data is reacquired in the virtual and real experimental space until the risk control strategy meets the threshold requirement.

[0026] In a second aspect, a system noise risk feedback strategy optimization device is provided, comprising:

[0027] A data acquisition module is configured to acquire noise data in a virtual and real experimental space;

[0028] A risk analysis module is configured to perform noise risk analysis on the noise data according to the uncertainty risk factors of different types of quantum states to obtain a risk measure value and a noise risk factor ranking;

[0029] A strategy calling module is configured to call a corresponding risk control strategy according to the risk measure value and the noise risk factor ranking and perform optimization analysis on the risk control strategy;

[0030] A strategy verification module is configured to verify the risk control strategy, and if the risk control strategy meets the threshold requirement, the risk control strategy is directly output, and if the risk control strategy does not meet the threshold requirement, feedback control is performed and noise data is reacquired until the threshold requirement is met.

[0031] In a third aspect, an electronic device is provided, comprising a memory and a processor;

[0032] The memory stores computer execution instructions;

[0033] The processor executes the computer execution instructions stored in the memory, so that the processor executes the method.

[0034] In a fourth aspect, a computer readable storage medium is provided, and the computer readable storage medium stores computer execution instructions. When the processor executes the computer execution instructions, the method is implemented.

[0035] Advantages:

[0036] (1) The present disclosure provides a system noise risk feedback strategy optimization method, device, equipment and medium. Through the selection and optimization of the strategy, the adaptability of the risk control strategy to the system noise risk is ensured, and the risk measurement process is optimized through dynamic iteration, the noise interference is continuously suppressed, and the stability and fidelity of the quantum state preparation and control process are effectively improved.

[0037] It is to be understood that the description of the background art is not an acknowledgement or consideration of related art prior to a priority date of the present application, and that the description is not admitted to be prior art by the present application. It is also to be understood that the description of the background art is not intended to limit or restrict the scope of the present application. BRIEF DESCRIPTION OF DRAWINGS

[0038] In order to more clearly illustrate the technical solutions of the embodiments of the present application or the prior art, the drawings required to be used in the embodiments or the prior art description will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor on the basis of these drawings.

[0039] Figure 1 Method flow chart of a system noise risk feedback strategy optimization method in the present disclosure;

[0040] Figure 2 Method flow chart of one embodiment of a system noise risk feedback strategy optimization method in the present disclosure;

[0041] Figure 3 Method flow chart of S203 of a system noise risk feedback strategy optimization method in the present disclosure;

[0042] Figure 4 Structural schematic diagram of a system noise risk feedback strategy optimization device in the present disclosure;

[0043] Figure 5 System structure diagram of an electronic device in the present disclosure.

[0044] In the figure, 101 is a data acquisition module, 102 is a risk analysis module, 103 is a strategy calling module, 104 is a strategy verification module, 200 is an electronic device, 201 is a processor, 202 is a memory, 203 is a communication component, and 204 is a bus. DETAILED DESCRIPTION

[0045] In order to make the purposes, technical solutions and advantages of the present application more clear, the technical solutions in the present application will be described clearly and completely below in combination with the drawings in the present application. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0046] The terms "first", "second", "third", "fourth", and the like in the description and in the claims of the present application and above-described drawings are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the use of the so-termed first, second, third, fourth, etc., names distinguishing between the objects in the appropriate cases might be interchanged under appropriate circumstances. For example, depending on the context, a first information might be termed a second information, and a second information might be termed a first information.

[0047] Depending on the context, the word "if" as used herein can be interpreted to mean "when" or "while" or "in response to determining".

[0048] Also, as used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context indicates otherwise.

[0049] It will be further understood that the terms "comprises" and "comprising", "includes" and "including", "has" and "having", when used herein, specify the presence of stated features, steps, operations, elements, components, items, categories, and / or groups but do not preclude the presence or addition of one or more other features, steps, operations, elements, components, items, categories, and / or groups thereof.

[0050] The terms "or" and "and / or" as used herein are to be interpreted as inclusive, i.e., as meaning one or any combination of the items. Thus, "A, B or C" or "A, B and / or C" means "any of the following: A; B; C; A and B; A and C; B and C; A, B and C". An exception to this definition will occur only when a combination of elements, functions, steps or acts are in some way inherently mutually exclusive.

[0051] The identification of sub-tasks in the related art is mostly based on keyword matching or function call graph analysis, which ignores the contextual semantic information when processing multi-paradigm or multi-language code, and cannot accurately distinguish the function types of custom functions, resulting in inaccurate task classification results.

[0052] To this end, in a first aspect, the present disclosure provides a system noise risk feedback strategy optimization method as shown in the following formula: Figure 1 The system noise risk feedback strategy optimization method comprises the following steps:

[0053] S101, acquiring noise data in a virtual-real experiment space;

[0054] The virtual and real experiment spaces are a one-to-one correspondence relationship based on digital twin technology to build an entity experiment space and a virtual experiment space, realize comprehensive collection and synchronization of noise data, and realize real-time collection and transmission of noise data by sensors in the quantum computer. The virtual experiment space forms a mapping relationship with the entity experiment space, builds a digital twin virtual model, synchronously receives and stores the noise data transmitted by the entity space, forms a virtual space noise data set corresponding to the entity space data in real time, provides a data basis for subsequent risk analysis, and solves the problems of incomplete data collection and strong lag in traditional single experiment space.

[0055] S102, performing noise risk analysis on the noise data according to the uncertainty risk factors of different types of quantum states to obtain risk measure values and noise risk factor rankings;

[0056] The type of quantum state at least includes one of a single-bit quantum state, a two-bit quantum state and a multi-bit system quantum state, and the uncertainty risk factor at least includes one of a microscopic noise interference uncertainty factor, a macroscopic noise interference uncertainty factor and a quantum bit preparation and control means noise uncertainty factor. There are differences in uncertainty risk factors for different types of quantum states, for example, a single-bit quantum state needs to pay attention to superposition state phase stability related noise, a two-bit quantum state needs to pay attention to inter-bit crosstalk noise, and a multi-bit system quantum state needs to pay attention to cumulative decoherence noise.

[0057] The risk measure value is used to measure the deviation degree of the noise data from the standard noise threshold, and provides a quantitative basis for evaluating the influence of noise in the preparation and control process of the quantum state. The noise risk factor ranking is based on the risk measure value, and each type of noise factor affecting the quantum state is sorted from small to large by a relative entropy sorting algorithm, which has the effect of clarifying the key risk source and providing a targeted basis for subsequent calling of risk control strategies, so as to ensure that the risk control measures can act on the most significant factors first, and improve the accuracy and efficiency of the strategies.

[0058] S103, calling a corresponding risk control strategy according to the risk measure value and the noise risk factor ranking and performing optimization analysis on the risk control strategy;

[0059] According to the risk ranking results of different quantum states, the AI calls a corresponding risk control strategy from a pre-set strategy library, for example, a single-bit superposition state focuses on a phase compensation strategy, a two-bit entangled state focuses on a crosstalk suppression strategy, and a multi-bit system coherent state focuses on a decoherence suppression strategy.

[0060] By optimizing and analyzing the risk control strategies, the optimal risk control strategy is selected to ensure that the strategy's control effectiveness against key risk factors is maximized. It should be noted that the relationship between the strategy library, strategy set, and risk control strategy is that the strategy library contains many strategy sets of different types, the strategy set contains multiple different risk control strategies, and the risk control strategy is a solution embedded in the risk control strategy library to deal with different risk factors.

[0061] Regarding risk control strategies, some examples are given below for illustration:

[0062] 1. Risk control strategies for addressing uncertainties related to microscopic noise interference include:

[0063] Control strategies for excitation noise risk factors: Excitation noise mainly affects the lifetime of qubits. The main noise is related to the processing technology. Therefore, the control strategies are: first, to select a better material Ta; second, to stabilize the production line with process parameters; and third, to recruit experienced processing personnel.

[0064] Strategies for controlling quasiparticle noise risk factors include: effective infrared radiation shielding, reducing the antenna modes of the qubit, and shielding different infrared radiation sources in the refrigerator. Other measures include reducing the transition energy of the superconducting bandgap, increasing the qubit lifetime, minimizing quasiparticle loss to the qubit lifetime T1, and reducing the quasiparticle concentration in the superconducting device.

[0065] Strategies for controlling charge fluctuation noise risk: Electric field noise is transverse to the quantization axis of the transmon qubit. The underlying reason is that the qubit frequency is highly sensitive to broadband charge fluctuations. Low-frequency charge noise will be longitudinally coupled to the bit and affect the decoherence time. Therefore, it is necessary to reduce charge fluctuations caused by temperature fluctuations and charge fluctuations caused by the coupling of mechanical vibrations with electrical signals.

[0066] Control strategies for magnetic flux noise risk factors: The direct impact of magnetic flux noise is an increase in decoherence time. The underlying reason is that for frequency-tunable qubits, using magnetic flux to adjust the bit frequency causes magnetic flux noise to be converted into bit frequency noise, leading to bit dephase. Therefore, to reduce noise intensity for frequency-tunable qubits, it is crucial to implement effective magnetic shielding measures, such as using encapsulation boxes, cryogenic shielding cylinders, and other noise-related components to mitigate the impact of cryogenic devices.

[0067] Control strategies for overheating noise risk factors: Focus on the uncertainty of thermal photon number and sample parameter changes. In the selection of thermal photon number and sample parameters, the main way to reduce the overheating effect is to improve the heat sink and keep the device temperature as low as possible.

[0068] Purcell effect control strategies for qubit-limiting noise risks: In the process of measuring superconducting qubits coupled to a readout cavity, to ensure readout speed, the qubit lifetime is also limited by the short lifetime of the cavity. The readout cavity needs a relatively long lifetime to allow light to leak out quickly. Focusing on chip design and the coupling strength between the qubit and the readout cavity, Purcell filters are used to suppress the Purcell effect.

[0069] 2. Risk control strategies for macroscopic noise interference uncertainty factors include:

[0070] Control strategies for vibration and noise risk factors: reduce vibration and noise in the experimental environment and minimize environmental vibrations that affect the preparation and manipulation of quantum states in the superconducting quantum computer system.

[0071] Control strategies for electromagnetic field noise risk factors: The way to eliminate its risk is to reduce various magnetic field-induced noises in the experimental environment and reduce the electromagnetic intensity generated by the superconducting quantum computer system for quantum state preparation and manipulation.

[0072] Control strategies for thermal noise risks: The quantum chip requires an extremely low temperature environment, with a minimum operating temperature ≤10mK; maximum cooling power ≥400μW@100mK; and a required ultra-low temperature cooling power ≥12μW@20mK. The method to eliminate these risks is to maintain environmental stability in the ultra-low temperature environment, i.e., the monitoring system for the ultra-low temperature environment must consistently provide the quantum chip with an ultra-low temperature environment, ensuring the state stability of the physical processing unit and maintaining its superconducting properties.

[0073] 3. Risk control strategies for uncertainties in quantum bit operation methods include:

[0074] Control strategies for microwave pulse stability risk factors: providing stable initial microwave pulses and stable microwave pulses throughout the quantum state preparation and manipulation process.

[0075] Control strategies for microwave pulse resolution risk factors: Provide stable initial microwave pulses for quantum state preparation and manipulation processes to improve microwave pulse resolution.

[0076] Control strategies for microwave pulse signal-to-noise ratio risk factors: Provide stable initial microwave pulses for quantum state preparation and manipulation processes, enhance signal strength, increase signal frequency, and improve microwave pulse signal-to-noise ratio.

[0077] Control strategies for microwave pulse delay risk factors: Control the total microwave pulse delay, that is, control the delay of each stage such as propagation, transmission, and queuing.

[0078] Control strategies for quantum nature risk factors in micro-electromagnetic fields: To reduce the impact of quantum nature noise in micro-electromagnetic fields on quantum state characteristics, the temperature of the refrigerator needs to be ≤10mK, the dynamic weak magnetic field noise needs to be low enough to improve quantum nature, the test points of 0 and 1 states in experimental measurements need to be more concentrated, and the accuracy of calculations needs to be higher.

[0079] Control strategies for quantum state coherence risk factors: In quantum states expressed using the two-level structure of qubits, a narrower energy level frequency width results in better qubit coherence, which is equivalent to a longer lifetime. Therefore, it is necessary to select qubits that meet the requirements for high capacitance and high survival rate.

[0080] Regarding the above example, it should be noted that the collection of strategy sets constitutes the strategy library. A strategy set is a set of countermeasures or solutions for different noise risk factors, and the risk control strategies in the strategy set are encapsulated in the strategy library.

[0081] S104. Verify the risk control strategy. If the risk control strategy meets the threshold requirement, output the risk control strategy directly. If the risk control strategy does not meet the threshold requirement, perform feedback control, return, and reacquire noise data until the threshold requirement is met.

[0082] If the strategy does not meet the threshold, the closed-loop feedback mechanism is activated, returning to S101 to re-collect noise data, and repeating the analysis and optimization process of S102-S103 until the strategy meets the verification indicators, and finally outputting the optimal risk control strategy.

[0083] The closed-loop verification mechanism ensures the effectiveness of the strategy, the stability and reliability of quantum state preparation and manipulation, and effectively improves the stability and fidelity of quantum state preparation and manipulation, thus solving the problem of strategy failure caused by the lack of closed-loop verification in traditional technologies.

[0084] In some publicly available implementations, a method for optimizing a system noise risk feedback strategy is also provided, including the following steps:

[0085] S201. Constructing a virtual and real experimental space;

[0086] The virtual and physical experimental spaces are established based on digital twin technology, creating a one-to-one correspondence between physical and virtual experimental spaces. The physical experimental space contains an entity database, such as a function library reflecting the physical experimental space's operating platform, a quantum state classification relation library for the physical experimental space's operating platform, and a sub-database of connection information regarding manipulated objects. The virtual experimental space, in response to the entity database of the physical experimental space, provides a database that corresponds one-to-one with the physical experimental space under virtual simulation conditions. This includes a mapping function library between the virtual and physical experimental spaces; a mapping relation library between the virtual and physical experimental spaces regarding the measurement parameters of the controlled objects; a quantum state classification relation library for the virtual experimental space's intelligent platform; and a mapping function library between the physical and virtual experimental spaces regarding the connection information sub-database of manipulated objects. By configuring embedded sensors in the quantum computer's low-temperature environment subsystem, measurement and control subsystem, chip subsystem, and testing laboratory environment, a data interface link is established between the sensor data interface of the physical experimental space system and the various data interface libraries of the virtual experimental space.

[0087] S202. Acquire noise data in a virtual and real experimental space;

[0088] Initialize the virtual and real experimental spaces, such as initializing the relevant environment, instrument preset parameters or static data of the physical experimental space, and acquiring noise data.

[0089] For the physical experimental space, the initial data for quantum state measurement and control corresponding to the physical experimental space are calibrated. Quantum systems are classified into single-qubit systems, two-qubit systems, and multi-qubit systems, and the state operation of quantum systems is calibrated and recorded in real time. For example, the process data of random noise measurement, real-time judgment, feedback control, cyclic repetition, and closed-loop verification of single-qubit quantum superposition state, two-qubit quantum entangled state, and multi-qubit system quantum coherent state are recorded. Relevant parameters characterizing quantum state properties, measurement and control data, and quantum state multidimensional attribute quantization information are calibrated and recorded. The above classification provides data support for subsequent parallel computing. The initial noise data received by embedded sensors in the low-temperature environment subsystem, measurement and control subsystem, chip subsystem, and test laboratory environment subsystem are calibrated.

[0090] For the virtual experimental space, a one-to-one correspondence is established between the virtual experimental space and the physical experimental space. The relevant environmental and instrument threshold parameters of the virtual experimental space are calibrated. The initialization processing of the static data for quantum state measurement and control corresponding to the virtual experimental space is calibrated, including calibrating and recording the initial data for quantum state measurement and control, relevant characteristic parameters, measurement and control data, and dynamic multidimensional attribute quantization information of the quantum state corresponding to the physical experimental space. At the same time, the initial noise data transmitted to the virtual experimental space by the embedded sensors of the low temperature environment subsystem, measurement and control subsystem, chip subsystem, and test laboratory environment subsystem are calibrated.

[0091] After initialization, noise data is obtained in the virtual and real experimental space.

[0092] S203. Based on the uncertainty risk factors of different types of quantum states, noise risk analysis is performed on the noise data to obtain risk measurement values ​​and noise risk factor rankings;

[0093] The noise risk analysis criteria for single-bit quantum states, two-bit quantum states, and multi-bit system quantum states are completed in parallel.

[0094] This includes the following steps:

[0095] S2031. Identify uncertainty risk factors based on the type of quantum state;

[0096] The types of quantum states include single-qubit quantum states, two-qubit quantum states, and multi-qubit system quantum states. The uncertainty risk factors include microscopic noise interference uncertainty factors, macroscopic noise interference uncertainty factors, and noise uncertainty factors of quantum bit preparation and manipulation methods.

[0097] S2032. Determine the risk measure value of the quantum state under different uncertain risk factors based on the entropy algorithm according to the uncertainty risk factors;

[0098] The entropy algorithm calculation formula is shown below.

[0099] R = -plnp

[0100] Where R is the risk measure value of the noise characteristic parameter, and p is the probability that the noise characteristic parameter brings risk.

[0101] A higher risk measure value means that the noise characteristic parameters are more chaotic and unstable, the noise uncertainty is greater, and the risk is higher; a lower risk measure value means that the noise characteristic parameters are more concentrated and stable, the noise uncertainty is smaller, and the risk is lower.

[0102] The risk measures of single-qubit quantum states, two-qubit quantum states, and multi-qubit system quantum states under the uncertainty factors of microscopic noise interference, macroscopic noise interference, and noise uncertainty factors of quantum bit preparation and manipulation methods are calculated according to the entropy algorithm.

[0103] S2033. Based on the risk metric values, the noise risk factors are ranked using a relative entropy ranking algorithm.

[0104] Determine whether the current feedback control node needs to perform feedback control. If feedback control is required, retrieve the standard threshold from the risk threshold library, calculate the relative entropy between the risk measure value and the standard threshold, and sort the uncertain risk factors according to the magnitude of the relative entropy to obtain the noise risk factor ranking.

[0105] Let the set of system noise risk factors be X = {x1, x2, ..., x...} n The calculation results of the quantum state preparation and manipulation process are intended to be analyzed. Regarding the feedback judgment of X, the feedback control process is divided into two states, and the state events are as follows:

[0106] E = {E1, E2}

[0107] State 1 (E1): a i If the results of quantum computing do not meet experimental requirements such as fidelity, error threshold, or error correction estimation, feedback judgment and control correction with respect to X are required, i.e.: a i It is a feedback control node, a i ∈E1;

[0108] State 2 (E2): a i Quantum computing results already meet experimental requirements such as fidelity, error threshold, or error correction estimation, and do not require feedback control, i.e.: a i It is not a feedback control node, a i ∈E2.

[0109] For arbitrary quantum computing results a in the process of quantum state preparation and manipulation i ∈E1 participates in feedback control, calls the risk threshold library to quantitatively identify and analyze risk factors, and provides a ranking of risk factors.

[0110] The risk factor set of system noise is denoted as X = {x1, x2, ..., x...} m The "standard threshold set" corresponding to the "risk threshold library" is denoted as Y = {y1, y2, ..., y}. m} T For any quantum computation result a i The risk factor measure set X is obtained by real-time measurement of quantum noise (∈E1, i=1,2,…,n). ai ={x i1 ,x i2 …,x im} T ,definition The relative entropy with respect to Y is

[0111]

[0112] Obviously, x ij ≥0,y j ≥0, i=1,2,…,n; j=1,2,…,m.

[0113] Where x ij / ∑x ij y j / ∑y jThese are probability measures, and they satisfy:

[0114] Entropy is a quantitative description of the uncertainty of an event; the lower the entropy, the lower the risk and the higher the accuracy. According to... Calculation results (h) i1 h i2 ,…,h im The "relative entropy sorting algorithm" is given:

[0115] By following this logic, any result 'a' can be obtained. i The system noise risk factors (i = 1, 2, ..., n) are sorted from smallest to largest:

[0116] {e i1 e i2 ,…,e im}

[0117] In some publicly available implementations, the steps of ranking risk factors are given according to the type of quantum state.

[0118] For the noise uncertainty risk analysis of the single-qubit quantum state preparation and manipulation process, in the process of single-qubit quantum superposition state preparation and manipulation, the corresponding threshold in the "risk threshold library" is called to quantitatively identify and analyze the risk factors, and the risk factors are ranked according to the relative entropy algorithm.

[0119] Risk measure set at time t in the verification process of manipulating a single-qubit quantum superposition state in a physical experiment Call the threshold of the single-qubit quantum superposition state stability parameter from the "Risk Threshold Library" result Compared to the standard threshold Y q ,calculate:

[0120]

[0121] According to the relative entropy sorting algorithm, the quantum noise risk factors of single-qubit quantum superposition states can be sorted from smallest to largest.

[0122] For the noise uncertainty risk analysis of the two-qubit quantum state preparation and manipulation process, the corresponding threshold in the risk threshold library is called to quantitatively identify and analyze the risk factors in the two-qubit quantum entangled state preparation and manipulation process, and the risk factors are ranked according to the relative entropy algorithm.

[0123] Risk measure set at time t in the physical experiment to verify the spatial manipulation of two-qubit quantum entangled states Call the "Threshold for Stability Parameter of Two-Bit Quantum Superposition State" from the "Risk Threshold Library". Comparison results Standard threshold Y 2q ,calculate

[0124]

[0125] Based on the relative entropy sorting algorithm, the quantum noise risk factors of two-qubit quantum entangled states are sorted from smallest to largest.

[0126] For the noise uncertainty risk analysis of the quantum state preparation and manipulation process of multi-qubit systems, the corresponding thresholds in the "risk threshold library" are called to quantitatively identify and analyze risk factors in the quantum coherent state preparation and manipulation process of multi-qubit systems, and the risk factors are ranked according to the relative entropy algorithm.

[0127] Suppose a risk measure set at time t during the verification process of the quantum coherent state of a multi-qubit system manipulated in a physical experiment. Call the "Threshold for Quantum Coherence State Stability Parameter of Multi-Bit Systems" from the "Risk Threshold Library". Comparison results Standard threshold Y kq ,calculate

[0128]

[0129] Based on the relative entropy sorting algorithm, the quantum noise risk factors of quantum coherent states in multi-qubit systems are sorted from smallest to largest.

[0130] S204. Based on the risk measurement value and the ranking of noise risk factors, call the corresponding risk control strategy and perform optimization analysis on the risk control strategy;

[0131] Risk control strategies are retrieved from the strategy library based on risk measurement values ​​and noise risk factors.

[0132] The corresponding strategy set is retrieved from the strategy library based on the type of quantum state. The strategy measure value of each risk control strategy in the measurement set is determined based on the quantum gate fidelity, coherence time, and background system noise. The risk control strategy with the highest strategy measure value in the strategy set is selected.

[0133] Arbitrary result of quantum computation for the preparation and manipulation of quantum states a i (i = 1, 2, ..., n) Strategy A, given constraint objective o ik The policy measure value for (k = 1, 2, ..., l) is {z} ik Therefore, as a i probability measure They are also independent discrete random variables.

[0134] In noisy environments, measurement fidelity becomes an important indicator for evaluating the accuracy of the entire measurement process. Let's assume the quantum gate fidelity is 0. i1 o i1 The higher the better; define o i2 For coherence time, o i2 The longer the better; {o ik-2 Let} be the target set corresponding to the ranking of noise risk factors. For any k noise target o ik-2 The smaller the risk measure value for (k = 1, 2, ..., l), the better.

[0135] For the constrained target o ik The policy measure value for (k = 1, 2, ..., l) is {z} ik}, under the constraint target o ik a under (k=1,2,…,l) i The integrated preference vector is denoted as Z. g The preference solution z is obtained by solving the following nonlinear programming problem. gi * .

[0136]

[0137] To solve the mathematical programming problem above, let:

[0138]

[0139] Among them, w k For b ik The corresponding weighting coefficients, ∑w k =1.

[0140] It is the globally optimal solution for P, that is, the feedback control node a. i The strategy A corresponding to (i = 1, 2, ..., n) is the optimal control strategy.

[0141] Among them, the determination of the quantum state fidelity attribute value:

[0142] Target o i1 The higher the quantum gate fidelity, the better. Measurement fidelity refers to the probability that a quantum state is correctly read out, that is, the probability that the ideal state of the system matches the actual measurement result in multiple measurements. Let Q be the ideal distribution, P be the statistically obtained distribution, and let the relative entropy z be... i1 (P / Q):

[0143]

[0144] Where: j is the number of experiments.

[0145] By z i1(P / Q) can quantify the difference between the calculated result and the theoretical expectation. This corresponds to feedback control node a. i ∈E1, the attribute value is denoted as

[0146] Determination of coherence property values ​​in quantum state manipulation processes:

[0147] Target o i2 A longer coherence time is better. Coherence time describes the ability of a qubit to maintain coherence in the presence of external noise. Choosing the phase coherence time T... x Characterizes the rate at which the phase information of a quantum bit is destroyed by the environment.

[0148] For a given coherence time T x For quantum bits, if the gate operation duration is long or too many gate operations are called sequentially, the cumulative coherence loss during evolution will be greater. The effective depth of quantum circuits is often limited by the coherence time.

[0149] Let T0 be the coherence time under ideal conditions, and T x Let the relative entropy z be the statistically obtained coherence time. i2 (T x / T0):

[0150]

[0151] Where: j is the number of experiments.

[0152] The attribute value is denoted as

[0153] The coherence time for a single-qubit gate is approximately 10-20 ns, while that for a two-qubit gate is approximately 100-200 ns. Each quantum gate is inherently affected by physical noise (such as decoherence, control errors, and crosstalk) during execution, causing the calculated results to deviate from the ideal unitary matrix. With increasing circuit depth, the cumulative effect of noise can be nonlinearly amplified, severely reducing the computational accuracy of the final state. The coherence time calculation for multi-qubit systems must consider the risk of accumulated errors.

[0154] Determination of background system noise attribute values:

[0155] It calls upon the risk management library to support the quantum state feedback control calculation at this time.

[0156] For any a i ∈E1, and k, h, let the target system be... The corresponding set of background quantum noise uncertainty risk factors ∈ {noise index system}, denoted as Its attribute value set is denoted as Obviously, The lower the noise, the better.

[0157] In some disclosed implementations, the following are included:

[0158] (1) Regarding the preparation and manipulation of single-qubit quantum superposition states

[0159] Arbitrary quantum computation results a for the preparation and manipulation of single-qubit quantum states i For risk factors ∈E1, the AI ​​calls upon the "single-qubit quantum superposition state strategy set" corresponding to the "quantum state measurement and control intervention strategy library system" to select the optimal intervention control strategy. Specific algorithm:

[0160] For any a i ∈E1 participates in feedback control, targeting When h = 1, it is denoted as the constraint target of the single-qubit quantum superposition state intervention and control strategy. The strategy measure for determining a single-qubit quantum superposition state is denoted as

[0161] Single-bit quantum gate fidelity measure Coherence time measure The set of quantum noise property values ​​is

[0162] make

[0163] in, for The corresponding weighting coefficients,

[0164] At this point z gi q* The corresponding strategy is to perform quantum computation on arbitrary results a in the process of preparing and manipulating a single-qubit quantum state. i The globally optimal strategy for ∈E1.

[0165] (2) Regarding the preparation and manipulation of two-qubit quantum entangled states

[0166] Arbitrary quantum computing result a for the preparation and manipulation of two-qubit quantum states i Based on the risk factor ranking results ∈E1, the AI ​​calls the corresponding "two-qubit quantum entangled state strategy set" from the "Quantum State Measurement and Control Intervention Strategy Library System" to select the optimal intervention and control strategy. Specific algorithm:

[0167] For any a i ∈E1 participates in feedback control. When h=2, it is denoted as the constraint target of the intervention control strategy given two-qubit quantum entangled states. The set of policy measure values ​​for determining two-qubit quantum entangled states is denoted as

[0168] make

[0169] Where k2 = 1, k2 = 2, for The corresponding weighting coefficients,

[0170] At this point z gi 2q* The corresponding strategy is to perform quantum computation on arbitrary results a in the process of preparing and manipulating two-qubit quantum states. i The globally optimal strategy for ∈E1.

[0171] (3) Regarding the preparation and manipulation of multi-qubit quantum entangled states

[0172] Arbitrary quantum computing results a for the preparation and manipulation of multi-qubit quantum states i Based on the risk factor ranking results ∈E1, the AI ​​calls the "Multi-bit System Quantum Coherent State Strategy Set" corresponding to the "Quantum State Measurement and Control Intervention Strategy Library System" to select the optimal intervention and control strategy. Specific algorithm:

[0173] For any a i ∈E1 participates in feedback control. When h=3, it is denoted as the constraint target of the quantum coherent state intervention control strategy for a given multi-qubit system. The strategy measure for determining the quantum coherent state of a multi-qubit system is denoted as

[0174] make

[0175] Where k3 = 1, k3 = 2, w h for The corresponding weighting coefficients,

[0176] At this point z gi kq* The corresponding strategy is to perform arbitrary quantum computations on the quantum state preparation and manipulation process of multi-qubit systems. i ∈E1 Global optimal strategy.

[0177] S205. Verify the risk control strategy;

[0178] Parallel judgment is performed on single-qubit quantum states, two-qubit quantum states, and multi-qubit system quantum states. If the risk control strategy meets the required quantum computing result at this time, proceed to step S207. When one or more results obtained by "parallel computing" of quantum state uncertainty risk do not meet the risk error threshold or the correction estimation requirement, feedback closed-loop control processing is performed, i.e., proceed to step S206.

[0179] S206. Perform feedback control;

[0180] Based on the ranking and analysis of noise risk factors, the correlation between factors is analyzed, and parameter warning prompts are given for the virtual and real experimental spaces. When one or more results in the risk control strategy for obtaining quantum state uncertainty risk do not meet the risk error threshold or the requirements of the correction estimation experiment, the risk factors are ranked, the larger risk factors are recorded, the correlation between factors is analyzed, and parameter warning prompts are given to prepare for the subsequent acquisition of initial data.

[0181] It should be noted that the analysis of the correlation between factors specifically includes ranking the noise risk factors, quantitatively giving the magnitude of the impact of different risk factors on noise, and whether they are positive or negatively correlated, and then determining the corresponding strategy in the risk control strategy library based on the ranking.

[0182] Based on the parameter warning prompts, readjust the parameters of the virtual and real experimental space, and return to step S202 to reacquire noise data.

[0183] S207, Output risk control strategies.

[0184] When the risk control strategy meets the risk threshold requirement of the required quantum computing result, the current risk control strategy is output.

[0185] Secondly, a risk feedback control device for the quantum state preparation and manipulation process is provided, comprising:

[0186] Data acquisition module 101 is used to acquire noise data in the virtual and real experimental space;

[0187] Risk analysis module 102 is used to perform noise risk analysis on the noise data based on the uncertainty risk factors of different types of quantum states to obtain risk measurement values ​​and noise risk factor ranking;

[0188] The strategy invocation module 103 is used to invoke the corresponding risk control strategy according to the risk measurement value and the noise risk factor sorting, and to perform optimization analysis on the risk control strategy;

[0189] The strategy verification module 104 is used to verify the risk control strategy. If the risk control strategy meets the threshold requirement, the risk control strategy is directly output. If the risk control strategy does not meet the threshold requirement, feedback control is performed, and noise data is reacquired until the threshold requirement is met.

[0190] Thirdly, such as Figure 5 As shown, an electronic device is provided, characterized in that it includes: a memory and a processor; The memory stores instructions that the computer executes; The processor executes computer execution instructions stored in memory, causing the processor to perform the above-described method.

[0193] In one embodiment, the electronic device 200 includes at least one processor 201 and a memory 202. Optionally, the electronic device 200 further includes a communication component 203. The processor 201, memory 202, and communication component 203 are connected via a bus 204.

[0194] In a specific implementation, at least one processor 201 executes computer execution instructions stored in memory 202, causing at least one processor 201 to perform the above-described method.

[0195] The specific implementation process of processor 201 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0196] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0197] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0198] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0199] Fourthly, this disclosure also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described method.

[0200] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0201] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.

[0202] The division of units is merely a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0203] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0204] In addition, the functional units in the various embodiments of this disclosure can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0205] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this disclosure, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause an electronic device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this disclosure. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0206] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0207] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A method for optimizing a system noise risk feedback strategy, characterized in that, Includes the following steps: Acquire noise data in a virtual and real experimental space; Based on the uncertainty risk factors of different types of quantum states, noise risk analysis is performed on the noise data to obtain risk measurement values ​​and noise risk factor rankings; Based on the risk measurement value and the noise risk factors, the corresponding risk control strategy is invoked and the risk control strategy is optimized and analyzed. The risk control strategy is verified. If the risk control strategy meets the threshold requirement, the risk control strategy is directly output. If the risk control strategy does not meet the threshold requirement, feedback control is performed, and the process is repeated to acquire noise data until the threshold requirement is met.

2. The method according to claim 1, characterized in that, The quantum state type includes at least one of the following: single-qubit quantum state, two-qubit quantum state, and multi-qubit system quantum state. The uncertainty risk factors include at least one of the following: microscopic noise interference uncertainty factors, macroscopic noise interference uncertainty factors, and noise uncertainty factors of quantum bit preparation and manipulation methods.

3. The method according to claim 1, characterized in that, The noise risk analysis based on the uncertainty risk factors of different types of quantum states to obtain risk measure values ​​and noise risk factor ranking includes: Identify uncertainty risk factors based on the type of quantum state; Based on the aforementioned uncertainty risk factors, the risk measure value of the quantum state under different uncertainty risk factors is determined using an entropy algorithm. The noise risk factors are ranked based on the risk metric values ​​using a relative entropy ranking algorithm.

4. The method according to claim 3, characterized in that, The step of ranking noise risk factors based on the relative entropy ranking algorithm according to the uncertainty risk factors includes: Determine whether the current feedback control node needs to perform feedback control. If feedback control is required, retrieve the standard threshold from the risk threshold library, calculate the relative entropy between the risk measure value and the standard threshold, and sort the uncertainty risk factors according to the magnitude of the relative entropy to obtain the noise risk factor ranking.

5. A method according to claim 1, characterized in that, The step of invoking the corresponding risk control strategy based on the risk measurement value and the noise risk factors includes: Risk control strategies are retrieved from the strategy library based on the risk metric value and the ranking of noise risk factors.

6. A method according to claim 5, characterized in that, The step of retrieving risk control strategies from the strategy library based on the risk measurement value and the noise risk factors includes: The corresponding policy set is retrieved from the policy library based on the type of quantum state; The strategy metric value for each risk control strategy in the measurement set is determined based on quantum gate fidelity, coherence time, and background system noise. Select the risk control strategy with the highest strategy measure value from the strategy set.

7. The method according to claim 1, characterized in that, The step of implementing feedback control if the risk control strategy does not meet the threshold requirement, and returning to reacquire noise data until the threshold requirement is met, includes: Based on the ranking and analysis of the noise risk factors, the correlation between the factors is analyzed, and parameter warning prompts for the virtual and real experimental spaces are given. Based on the parameter warning prompts, noise data is reacquired in the virtual and real experimental space until the risk control strategy meets the threshold requirements.

8. A system noise risk feedback strategy optimization device, characterized in that, include: The data acquisition module is used to acquire noise data in the virtual and real experimental space; The risk analysis module is used to perform noise risk analysis on the noise data based on the uncertainty risk factors of different types of quantum states to obtain risk measurement values ​​and noise risk factor rankings; The strategy invocation module is used to invoke the corresponding risk control strategies according to the risk measurement value and the noise risk factors, and to perform optimization analysis on the risk control strategies. The strategy verification module is used to verify the risk control strategy. If the risk control strategy meets the threshold requirement, the risk control strategy is directly output. If the risk control strategy does not meet the threshold requirement, feedback control is performed, and noise data is reacquired until the threshold requirement is met.

9. An electronic device, characterized in that, Including memory and processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform a method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, implement the method as described in any one of claims 1-7.

Images