High-precision quantum sensing environment physical quantity real-time monitoring system and method

Abstract

The invention discloses a high-precision quantum sensing environment physical quantity real-time monitoring system and method, and belongs to the technical field of environment monitoring. The system comprises a quantum sensing front end, a signal excitation and regulation module, a quantum classical hybrid computing engine, an edge cloud collaborative transmission module and a terminal application module. The quantum sensing front end adopts a quantum entanglement sensing array integrated by NV color center diamond and a rubidium atom gas chamber to synchronously sense a magnetic field, temperature, pressure and vibration; the signal excitation and regulation module optimizes regulation parameters and inhibits noise through a quantum optimal control algorithm; a quantum-classical hybrid calculation engine is combined with a quantum neural network and classical multi-algorithm collaboration to realize efficient resolving and data fusion of quantum state features. According to the invention, the monitoring precision, the real-time response speed and the anti-interference capability are obviously improved, synchronous monitoring of multiple physical quantities is supported, and the system is suitable for various scenes such as industry, scientific research and deep space exploration, and has wide application value.

Description

Technical Field

[0001] This invention relates to the field of environmental monitoring technology, specifically to a high-precision quantum sensing system and method for real-time monitoring of environmental physical quantities. Background Technology

[0002] As is well known, high-precision real-time monitoring of environmental physical quantities is a key technological support for many fields such as industrial production, scientific research, disaster early warning, and deep space exploration. Its monitoring effect directly affects the safety of equipment operation, the accuracy of scientific research data, and the timeliness of emergency response. Currently, mainstream monitoring technologies mainly rely on traditional sensing devices such as electromagnetic sensors, fiber optic sensors, and piezoelectric sensors. However, these technologies have gradually revealed many insurmountable limitations in long-term application.

[0003] In terms of monitoring accuracy, traditional sensors, limited by their working principle, have an inherent bottleneck in their ability to sense physical quantities such as magnetic fields and temperature, making it difficult to meet the extreme accuracy requirements of scenarios such as quantum laboratories, brain science research, and deep space exploration. Regarding interference resistance, traditional sensors are susceptible to environmental noise such as electromagnetic interference, temperature drift, and mechanical vibration. In special environments such as strong electromagnetic environments in industry, complex field scenarios, or inside living organisms, the stability and reliability of measurement data decrease significantly, and may even exhibit serious deviations.

[0004] Regarding compatibility in monitoring multiple physical quantities, traditional sensors typically can only monitor a single physical quantity. To simultaneously acquire multiple parameters such as magnetic fields, temperature, pressure, and vibration, multiple independent monitoring systems must be deployed. This not only increases equipment purchase and maintenance costs but also suffers from poor data synchronization, making collaborative analysis and cross-validation between multiple physical quantities difficult. Furthermore, the signal processing flow of traditional technologies is relatively cumbersome, and the data transmission link lacks optimization, resulting in a slow overall response speed. This makes it difficult to capture transient changes in physical quantities in high-speed dynamic environments in a timely manner, failing to meet the real-time monitoring needs of scenarios such as transient magnetic fields in power grids and precursory vibrations of earthquakes.

[0005] While quantum sensing technology has made breakthroughs in accuracy in recent years, it still has significant shortcomings. Existing quantum sensing schemes mostly rely on single quantum systems, resulting in short quantum state coherence times and susceptibility to signal attenuation due to external environmental influences. Furthermore, cross-interference issues exist when monitoring multiple physical quantities, and effective signal separation and verification mechanisms are lacking. In addition, data processing still relies on traditional classical computing models, failing to fully leverage the parallel processing advantages of quantum computing, leading to low data processing efficiency. Moreover, a comprehensive optimization scheme for the entire chain—sensing, control, processing, transmission, and application—has not been developed, making it impossible to achieve synergistic improvements in accuracy, real-time performance, and anti-interference capabilities, and thus difficult to adapt to complex and diverse real-world application scenarios. Summary of the Invention

[0006] (a) Technical problems to be solved To address the shortcomings of existing technologies, this invention provides a high-precision quantum sensing system and method for real-time monitoring of environmental physical quantities.

[0007] (II) Technical Solution To achieve the above objectives, the present invention provides the following technical solution: a high-precision quantum sensing real-time monitoring system for environmental physical quantities, comprising: a quantum sensing front end, wherein the quantum sensing front end includes a quantum entangled sensing array integrated from NV color center diamond and rubidium atom gas cells, wherein the quantum entangled sensing array forms a quantum entangled state by laser-induced electron spin and rubidium atom energy level transition, for synchronous sensing of magnetic field, temperature, pressure and vibration; The signal excitation and modulation module includes a photonic crystal laser, a terahertz microwave synthesizer, and a quantum dot single-photon detector. The photonic crystal laser and the terahertz microwave synthesizer optimize the output laser pulse sequence and microwave phase through a quantum optimal control algorithm (GRAPE). The quantum dot single-photon detector collects quantum fluctuation signals of environmental noise in real time and generates an anti-phase compensation signal to suppress noise interference. A quantum-classical hybrid computing engine, comprising a quantum processing unit (QPU) and a heterogeneous classical computing unit, wherein the QPU is a superconducting quantum bit processor used to map and solve quantum state characteristics through a quantum neural network (QNN); The heterogeneous classical computing units include FPGA, GPU and ARM. The FPGA executes a chaotic wavelet denoising algorithm to suppress noise, the GPU runs a federated Kalman filter algorithm to achieve multi-node data aggregation, and the ARM runs a reinforcement learning parameter optimization algorithm to dynamically adjust system parameters. The edge-cloud collaborative transmission module combines quantum key distribution (QKD) with 5G slicing technology to encrypt and transmit pre-processed data from edge nodes, with an end-to-end transmission latency of ≤50μs. The terminal application module deploys a physical quantity prediction model based on a Transformer time-series network to display monitoring data in real time, perform anomaly warnings, and predict physical quantity change trends.

[0008] Furthermore, the quantum entangled sensing array also includes a MEMS micro-displacement platform, which dynamically adjusts the topology of the array through reinforcement learning, enabling the spatial resolution of the quantum entangled sensing array to adaptively switch within the range of 100nm-1mm; and the entanglement degree of the quantum entangled state is ≥0.9.

[0009] Furthermore, the output wavelength of the photonic crystal laser is 532nm, and the linewidth is ≤1kHz; the frequency adjustment range of the terahertz microwave synthesizer is 0.1-10GHz, and the phase noise is ≤-120dBc / Hz; the response time of the quantum dot single-photon detector is ≤10ps.

[0010] Furthermore, the quantum processing unit (QPU) is a 128-qubit superconducting quantum bit processor; in the heterogeneous classical computing unit, the reinforcement learning parameter optimization algorithm uses measurement accuracy, energy consumption, and latency as multi-objective reward functions.

[0011] Furthermore, in the edge-cloud collaborative transmission module, the compression ratio of data by the edge node is 10:1; the terminal application module supports the storage of historical data for ≥1 year, and the physical quantity prediction model has a prediction accuracy of ≥98% for the change trend of physical quantities within 10 seconds.

[0012] This invention also provides a method for real-time monitoring of environmental physical quantities using high-precision quantum sensing, comprising the following steps: Step 1: Quantum state initialization and entanglement calibration: The initial quantum entangled state of the NV color center and rubidium atom is reconstructed using quantum state tomography, and the entanglement degree of the quantum entangled state is calculated; if the entanglement degree is lower than a preset threshold, laser re-pumping is started to restore the entanglement degree; at the same time, the quantum state and physical quantity dynamic calibration dictionary is updated every 100ms based on Bayesian inference. Step 2: Multi-dimensional signal acquisition and feature fusion: Simultaneously acquire the fluorescence photon count, microwave resonance frequency offset, and atomic gas cell absorption spectrum signals output by the quantum entanglement sensing array, extract the multi-physical quantity coupling features in the above signals through the quantum entanglement correlation function, and cross-validate the coupling features based on the intrinsic coupling equation of the physical quantities. Step 3: Quantum-classical hybrid solution: The coupling characteristics are solved using the quantum neural network (QNN) in the quantum processing unit (QPU) to output the quantum probability distribution of the physical quantity; the quantum probability distribution is converted into deterministic physical quantity values ​​through a heterogeneous classical computing unit; and the parameters of the QNN and the quantum control signal are dynamically adjusted through a reinforcement learning parameter optimization algorithm to form a closed-loop optimization. Step 4: Adaptive early warning and system reconstruction: Calculate the Lyapunov exponent of the physical quantity time series based on chaos theory, identify abnormal fluctuations and trigger early warning; if an extreme environment is detected, automatically switch to quantum state freeze mode, and resume monitoring within 100μs after the extreme environment disappears.

[0013] Furthermore, in step 1, the laser re-pumping time is ≤50μs; the long-term drift error of the dynamic calibration dictionary of quantum states and physical quantities is ≤0.001% / h.

[0014] Furthermore, in step 2, the acquisition rate of the fluorescent photon counting is... The inherent coupling equations of the physical quantities include temperature and pressure cross-validation equations based on the PV=nRT modified model.

[0015] Furthermore, in step 3, the single iteration time of the quantum-classical hybrid solution is ≤1ms; the reinforcement learning parameter optimization algorithm adopts the deep deterministic policy gradient (DDPG) algorithm.

[0016] Furthermore, in step 4, the early warning time for the abnormal fluctuation is ≥10 seconds earlier than the traditional threshold method; the extreme environment is a strong electromagnetic pulse environment with an intensity ≥1000V / m.

[0017] (III) Beneficial Effects Compared with existing technologies, the present invention provides a high-precision quantum sensing system and method for real-time monitoring of environmental physical quantities, which has the following advantages: This invention relates to a high-precision quantum sensing system and method for real-time monitoring of environmental physical quantities. It employs a quantum entangled sensing array integrating NV color center diamond and rubidium atom gas cells, utilizes the nonlocal correlation of quantum entanglement to enhance signal characteristics, and combines dual-mechanism cross-validation and multi-physical quantity coupling feature extraction to significantly improve the ability to perceive weak physical quantities. This effectively breaks through the precision bottleneck of traditional technologies and can meet the demand for extreme precision in high-end scientific research, deep space exploration, and other scenarios.

[0018] The quantum-classical hybrid computing engine constructed in this invention efficiently solves quantum state characteristics through quantum processing units and combines the parallel processing capabilities of heterogeneous classical computing units, which greatly simplifies the signal processing process. At the same time, the edge-cloud collaborative transmission module adopts quantum encryption and high-speed transmission technology to optimize the data transmission link and significantly shorten the entire link time from signal acquisition to terminal feedback. It can capture transient changes of physical quantities in dynamic environments in a timely manner and meet the real-time monitoring needs of high-speed dynamic scenarios.

[0019] This invention optimizes quantum state control parameters through a quantum optimal control algorithm, extends the quantum state coherence time, and combines noise feedforward suppression and chaotic wavelet denoising technology to effectively resist the influence of environmental noise such as electromagnetic interference and temperature drift. For extreme environments, a quantum state freeze protection mechanism is designed to ensure that the system can still work stably under complex and harsh conditions, greatly expanding the application scope of the monitoring system.

[0020] This invention achieves simultaneous monitoring of multiple physical quantities, including magnetic field, temperature, pressure, and vibration, through a single system. It utilizes the inherent coupling relationships between these physical quantities for cross-validation, eliminating data synchronization issues arising from multi-system deployments. Simultaneously, the integrated design significantly reduces equipment purchase, installation, and maintenance costs, enhancing the economic efficiency and practicality of the monitoring system. Furthermore, this invention introduces a reinforcement learning algorithm to construct a dynamic optimization mechanism that adaptively adjusts system parameters based on environmental changes. This allows the system to continuously optimize performance over long-term operation, achieving self-improvement in accuracy and stability without manual intervention. This significantly reduces maintenance costs in long-term monitoring scenarios, making it particularly suitable for unattended monitoring in remote areas and deep space exploration. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the overall workflow of the system of the present invention; Figure 2 This is a schematic diagram of the quantum sensing front-end workflow of the present invention; Figure 3 This is a schematic diagram of the signal excitation and modulation process of the present invention; Figure 4 This is a schematic diagram of the quantum-classical hybrid solution process of the present invention; Figure 5 This is a schematic diagram of the entire monitoring method of the present invention. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] Please see Figures 1 to 5 This invention relates to a high-precision quantum sensing real-time monitoring system for environmental physical quantities, comprising: a quantum sensing front end, wherein the quantum sensing front end includes a quantum entangled sensing array integrated from NV color center diamond and rubidium atom gas cells, wherein the quantum entangled sensing array forms a quantum entangled state by laser-induced electron spin of NV color center and energy level transition of rubidium atom, for synchronous sensing of magnetic field, temperature, pressure and vibration; The signal excitation and modulation module includes a photonic crystal laser, a terahertz microwave synthesizer, and a quantum dot single-photon detector. The photonic crystal laser and the terahertz microwave synthesizer optimize the output laser pulse sequence and microwave phase through a quantum optimal control algorithm (GRAPE). The quantum dot single-photon detector collects quantum fluctuation signals of environmental noise in real time and generates an anti-phase compensation signal to suppress noise interference. A quantum-classical hybrid computing engine, comprising a quantum processing unit (QPU) and a heterogeneous classical computing unit, wherein the QPU is a superconducting quantum bit processor used to map and solve quantum state characteristics through a quantum neural network (QNN); The heterogeneous classical computing units include FPGA, GPU and ARM. The FPGA executes a chaotic wavelet denoising algorithm to suppress noise, the GPU runs a federated Kalman filter algorithm to achieve multi-node data aggregation, and the ARM runs a reinforcement learning parameter optimization algorithm to dynamically adjust system parameters. The edge-cloud collaborative transmission module combines quantum key distribution (QKD) with 5G slicing technology to encrypt and transmit pre-processed data from edge nodes, with an end-to-end transmission latency of ≤50μs. The terminal application module deploys a physical quantity prediction model based on a Transformer time-series network to display monitoring data in real time, perform anomaly warnings, and predict physical quantity change trends. The quantum sensing front end utilizes the quantum entanglement (non-local correlation) between NV color centers and rubidium atoms to improve the signal-to-noise ratio and simultaneously sense magnetic fields, temperature, pressure, and vibration. The signal excitation and modulation module optimizes laser / microwave parameters through the quantum optimal control algorithm (GRAPE), extends the quantum state coherence time, and achieves feedforward suppression based on noise fluctuation signals collected by quantum dot detectors. In the quantum-classical hybrid computing engine, the quantum neural network (QNN) of the QPU directly solves the quantum state features, and the classical heterogeneous units achieve data fusion and parameter optimization through multi-algorithm collaboration (chaotic wavelet denoising, federated Kalman filtering, and reinforcement learning). Edge cloud transmission combines quantum encryption and 5G slicing to ensure data security and low latency. The terminal uses the Transformer model to achieve real-time display and trend prediction. It achieves simultaneous monitoring of multiple physical quantities, breaking through the limitations of traditional single-physical-quantity sensing; quantum entanglement and hybrid computing engine enable monitoring accuracy to reach the sub-nano level (magnetic field ≤ 0.1 nT, temperature ≤ 10 μK); end-to-end latency ≤ 50 μs, meeting the real-time monitoring requirements of dynamic environments; quantum encryption and noise suppression technology enhance the system's anti-interference capability and data security.

[0024] In this scheme, the quantum entangled sensing array also includes a MEMS micro-displacement platform. This platform dynamically adjusts the array's topology through reinforcement learning, enabling the spatial resolution of the quantum entangled sensing array to adaptively switch within the range of 100nm-1mm; and the entanglement degree of the quantum entangled states is ≥0.9. The MEMS micro-displacement platform can physically adjust the spatial distribution of sensor nodes, and combined with reinforcement learning algorithms (aiming at monitoring accuracy and environmental adaptability), allows the array to switch between sparse mode (large spatial range) and dense mode (high-resolution region); simultaneously, quantum state manipulation ensures the entanglement degree of the entangled states (Concurrence ≥0.9), maintaining high signal-to-noise ratio quantum correlation. The spatial resolution can be adaptively adjusted from 100nm to 1mm, balancing the needs of large-scale monitoring and local fine measurement; high entanglement degree (≥0.9) ensures the stability of the quantum signal, improving the signal-to-noise ratio by 20dB compared to non-entangled arrays, and significantly reducing the impact of environmental interference on measurements.

[0025] In this scheme, the output wavelength of the photonic crystal laser is 532nm, and the linewidth is ≤1kHz; the frequency adjustment range of the terahertz microwave synthesizer is 0.1-10GHz, and the phase noise is ≤-120dBc / Hz; the response time of the quantum dot single-photon detector is ≤10ps. The 532nm photonic crystal laser with a linewidth ≤1kHz provides stable excitation light, reducing the interference of laser frequency drift on quantum state transitions; the 0.1-10GHz terahertz microwave synthesizer with a phase noise ≤-120dBc / Hz ensures the phase accuracy of quantum state modulation; the quantum dot single-photon detector with a response time ≤10ps can capture noise fluctuations on extremely short timescales, providing a data basis for real-time noise compensation. The high-precision parameters of the laser and microwave sources extend the quantum state coherence time to 1 second (10 times better than traditional schemes); the fast-response detector ensures timely noise suppression, enabling the system to maintain stable measurement accuracy (electromagnetic interference suppression ratio ≥80dB) even in strong electromagnetic environments.

[0026] In this scheme, the quantum processing unit (QPU) is a 128-qubit superconducting quantum bit processor; in the heterogeneous classical computing unit, the reinforcement learning parameter optimization algorithm uses measurement accuracy, energy consumption, and latency as multi-objective reward functions. The 128-qubit superconducting QPU provides sufficient quantum state storage space to support the mapping of complex quantum state features by the quantum neural network (QNN); in the heterogeneous classical computing unit, reinforcement learning uses measurement accuracy, energy consumption, and latency as multi-objective reward functions to balance system performance and resource consumption. The 128-qubit QPU improves the feature extraction efficiency of the QNN by 100 times compared to the classical neural network; multi-objective reinforcement learning optimization improves the overall performance (accuracy, energy consumption, and response speed) of the system in complex environments (such as strong industrial electromagnetic fields and high temperatures) by 40%, avoiding performance imbalances caused by single-objective optimization.

[0027] In this solution, the edge-cloud collaborative transmission module features a data compression ratio of 10:1 at the edge nodes; the terminal application module supports ≥1 year of historical data storage, and the physical quantity prediction model achieves ≥98% accuracy in predicting the trend of physical quantities within 10 seconds. The 10:1 data compression ratio at the edge nodes reduces the amount of transmitted data and lowers channel load; the terminal's support for ≥1 year of historical data storage provides a data foundation for trend analysis; the Transformer model predicts physical quantity changes within 10 seconds based on historical data, triggering early warnings. Data compression improves transmission efficiency by 10 times, avoiding channel congestion; long-term data storage supports retrospective analysis and system calibration optimization; the ≥98% accuracy of 10-second trend prediction provides a faster response time than traditional real-time alarms, making it suitable for scenarios requiring early warning, such as geological disasters and equipment failures.

[0028] This invention also provides a method for real-time monitoring of environmental physical quantities using high-precision quantum sensing, comprising the following steps: Step 1: Quantum state initialization and entanglement calibration: The initial quantum entangled state of the NV color center and rubidium atom is reconstructed using quantum state tomography, and the entanglement degree of the quantum entangled state is calculated; if the entanglement degree is lower than a preset threshold, laser re-pumping is started to restore the entanglement degree; at the same time, the quantum state and physical quantity dynamic calibration dictionary is updated every 100ms based on Bayesian inference. Step 2: Multi-dimensional signal acquisition and feature fusion: Simultaneously acquire the fluorescence photon count, microwave resonance frequency offset, and atomic gas cell absorption spectrum signals output by the quantum entanglement sensing array, extract the multi-physical quantity coupling features in the above signals through the quantum entanglement correlation function, and cross-validate the coupling features based on the intrinsic coupling equation of the physical quantities. Step 3: Quantum-classical hybrid solution: The coupling characteristics are solved using the quantum neural network (QNN) in the quantum processing unit (QPU) to output the quantum probability distribution of the physical quantity; the quantum probability distribution is converted into deterministic physical quantity values ​​through a heterogeneous classical computing unit; and the parameters of the QNN and the quantum control signal are dynamically adjusted through a reinforcement learning parameter optimization algorithm to form a closed-loop optimization. Step 4: Adaptive early warning and system reconstruction: Calculate the Lyapunov exponent of the physical quantity time series based on chaos theory, identify abnormal fluctuations and trigger early warning; if an extreme environment is detected, automatically switch to quantum state freeze mode, and resume monitoring within 100μs after the extreme environment disappears.

[0029] In this scheme, in step 1, the laser re-pumping time is ≤50μs; the long-term drift error of the dynamic calibration dictionary of quantum states and physical quantities is ≤0.001% / h. The laser re-pumping time of ≤50μs ensures rapid recovery of the quantum state when the entanglement degree is below the threshold, reducing system downtime. The dynamic calibration dictionary is updated every 100ms, combining real-time data with standard source data via Bayesian inference to control the long-term drift error to ≤0.001% / h. Fast re-pumping reduces the downtime caused by insufficient entanglement to ≤50μs, preventing data loss. Low drift error ensures that the system does not require frequent manual calibration during long-term monitoring (such as geological monitoring over several months), reducing maintenance costs and improving data continuity.

[0030] In this scheme, in step 2, the acquisition rate of the fluorescence photon counting is: The intrinsic coupling equations of the physical quantities include temperature and pressure cross-validation equations based on the modified PV=nRT model. The fluorescence photon counting acquisition rate reaches 10. 6 The high sampling rate ensures the capture of subtle changes in quantum states. Based on the temperature and pressure cross-validation equation of the PV=nRT modified model, thermodynamic laws constrain the measurement results of the two physical quantities, eliminating contradictory data caused by individual sensor errors. This high sampling rate improves the accuracy of quantum state feature extraction by 20%, capturing transient changes in nanoscale vibrations and sub-nanotesla magnetic fields. Cross-validation of physical quantities allows for mutual correction of measurement errors in temperature and pressure, reducing the measurement error of a single physical quantity to within 0.05 nT (magnetic field) and 10 μK (temperature).

[0031] In this scheme, in step 3, the single iteration time of the quantum-classical hybrid solution is ≤1ms; the reinforcement learning parameter optimization algorithm adopts the Deep Deterministic Policy Gradient (DDPG) algorithm. The quantum-classical hybrid solution's single iteration time of ≤1ms ensures real-time response in dynamic environments; the Deep Deterministic Policy Gradient (DDPG) algorithm is used as the core of reinforcement learning, optimizing parameters such as laser power and microwave frequency through continuous action space to adapt to dynamic environmental changes. The ≤1ms iteration time allows the system's response delay to high-speed dynamic physical quantities (such as high-speed train vibration and transient magnetic fields from lightning strikes on the power grid) to be controlled at the microsecond level; the DDPG algorithm improves parameter optimization efficiency by 50% compared to traditional reinforcement learning algorithms, increasing the system's recovery speed by 40% when encountering sudden environmental changes (such as a sudden increase in electromagnetic interference).

[0032] In this scheme, in step 4, the early warning time for abnormal fluctuations is ≥10 seconds earlier than the traditional threshold method; the extreme environment is a strong electromagnetic pulse environment with an intensity ≥1000V / m. Based on Lyapunov exponent-based anomaly detection, by calculating the chaotic characteristics of the physical quantity time series, minute anomalies that cannot be captured by the traditional threshold method (such as nanometer-scale vibrations in earthquake precursors) are identified; under extreme environment (≥1000V / m strong electromagnetic pulse), the system switches to quantum state freezing mode (maintaining the NV color center spin ground state) to avoid the quantum state being disturbed and destroyed, and monitoring resumes within 100μs after the interference disappears. The anomaly early warning is ≥10 seconds earlier than the traditional threshold method, providing response time for emergency scenarios such as geological disasters and equipment failures; the quantum state freezing mode ensures that the measurement accuracy loss of the system under strong electromagnetic pulses is ≤5%, improving the system's resistance to extreme environments by 90% compared to systems without protection mechanisms.

[0033] The physical quantity inversion formula of quantum classical hybrid solution integrates quantum state feature solution of quantum neural network (QNN), Bayesian dynamic calibration and multi-node data fusion of federated Kalman filtering, and is the core mathematical model for realizing sub-nano level precision monitoring.

[0034] Inversion formulas for physical quantities in quantum-classical hybrid solutions:

[0035] Meaning: The final output is a high-precision estimate of a physical quantity (such as magnetic field strength, temperature, pressure, etc., with units determined according to the type of physical quantity, e.g., magnetic field). The temperature is The deterministic physical quantity values ​​obtained through quantum-classical hybrid computation and optimized in multiple dimensions are the core results of system monitoring.

[0036] Meaning: The mathematical expectation of the probability distribution of physical quantities output by a quantum neural network (QNN).

[0037] QNNs model the probabilistic features of the input quantum state, where: The physical quantity to be monitored (e.g., magnetic field) ,temperature ); : Input the quantum state feature vector of the QNN, including the change in fluorescence intensity Microwave resonant frequency shift Atomic gas cell absorption spectral intensity etc. (dimension is) , (number of features); Trainable parameters of QNN (such as rotation angle and phase in quantum circuits, with dimensions of...) , (Number of parameters).

[0038] The expected value of the probability distribution is used to convert the probabilistic output of the quantum state into a deterministic initial estimate of the physical quantity (eliminating the uncertainty caused by the superposition of quantum states). The parallelism of quantum computing is used to extract the physical quantity features hidden in the quantum state, which is the basis of the "quantum layer" for high-precision inversion, and the feature extraction efficiency is 100 times higher than that of classical neural networks.

[0039] Meaning: Correction term for Bayesian dynamic calibration.

[0040] The drift error (unit consistent with Q) of the "quantum state-physical quantity" dynamic calibration dictionary is updated every 100ms by Bayesian inference, and the calculation formula is as follows: The measured value is from the standard source. (The predicted value is from the system model). Calibration weight coefficients (range 0.1-0.3), dynamically adjusted by reinforcement learning (the greater the environmental noise, the greater the alpha, thus enhancing calibration strength). Physical significance: Eliminating system errors caused by long-term sensor drift (such as temperature drift, quantum state coherence decay), keeping long-term monitoring errors ≤0.001% / h.

[0041] Meaning: The fusion correction term of the federated Kalman filter achieves global optimization of multi-node data.

[0042] Kalman gain matrix (dimension: , (This refers to the number of sensor nodes), reflecting the reliability of the measurement values ​​from each node (higher accuracy corresponds to greater gain). : The vector of local measurements from each sensor node (dimension: ; Observation matrix (dimension: A matrix (usually an identity matrix) describes the mapping relationship between physical quantities and measured values. : The prior estimate of a physical quantity (from the previous moment) (Confirmed). Physical significance: By fusing local measurement data from distributed sensors, the random errors of a single node (such as measurement bias caused by uneven spatial distribution) are offset, improving the consistency of multi-node data by 30%.

[0043] This formula uses a three-stage progression of quantum probability calculation, dynamic calibration correction, and multi-node fusion optimization to accurately map the microscopic characteristics of quantum states into high-precision estimates of macroscopic physical quantities. It also takes into account the uncertainty of quantum states, long-term system drift, and spatial distribution errors, ultimately achieving sub-nanoscale (e.g., magnetic field ≤ 0.1 nT) monitoring accuracy and microsecond-level (≤ 50 μs) real-time response.

[0044] The multi-objective optimization formula for quantum control parameters integrates the multi-objective decision-making logic of quantum optimal control (GRAPE) and reinforcement learning (DDPG). By optimizing control parameters such as laser pulse and microwave phase, it maximizes the quantum state coherence time (improving measurement accuracy) while balancing system energy consumption and response delay, which is the key to ensuring system stability in complex environments.

[0045] Multi-objective optimization formula for quantum control parameters:

[0046]

[0047] Meaning: The optimal set of quantum control parameters (the final output control instructions to be executed).

[0048] Specific composition: It includes two types of core control parameters. : Pulse parameters of photonic crystal lasers (amplitude / power, in mW); Phase parameters of a terahertz microwave synthesizer (unit: rad).

[0049] Physical significance: After multi-objective optimization, the optimal control scheme can simultaneously satisfy "high-fidelity quantum state evolution, low energy consumption, and low latency", and directly act on the signal excitation and control module.

[0050] Meaning: In the feasible region of parameters Within, find the control parameters that minimize the objective function. .

[0051] The physical feasible domain of the control parameters (determined by hardware performance), such as the safe range of laser power and the value range of microwave phase; Finding the parameter solution corresponding to the minimum value is the core optimization operation of reinforcement learning (DDPG).

[0052] Physical significance: To ensure that the optimization results comply with the physical limitations of the hardware and avoid damage to the device or distortion of the quantum state due to parameters exceeding the threshold.

[0053] Objective function:

[0054] The objective function is a weighted summation form, which balances the three core indicators of "quantum state fidelity, energy consumption, and latency" through weight coefficients, and directly corresponds to the multi-objective reward function of reinforcement learning (DDPG).

[0055] Quantum state fidelity loss term The fidelity weighting coefficient (range 0.4-0.6) has the highest weight among the three major indicators, prioritizing the protection of measurement accuracy.

[0056] The fidelity loss of quantum state evolution (dimensionless, range 0-1) is calculated using the following formula: ,in : Target quantum entangled state (such as the ideal entangled state of NV color center and rubidium atom, with entanglement degree ≥ 0.9); : By control parameters Determined quantum evolution operators (describing the changes in quantum states under laser / microwave irradiation); : Initial quantum state (initial entangled state after system startup); Quantum state evolution fidelity (the closer to 1, the closer the actual evolution is to the ideal goal, and the longer the quantum state coherence time).

[0057] Physical significance: Minimizes fidelity loss, ensures the stable existence of quantum entangled states under control, extends coherence time to 1 second (10 times better than traditional methods), and provides a foundation for high-precision measurement.

[0058] System energy consumption item.

[0059] Energy consumption weighting coefficient (range 0.2-0.3) balances accuracy and energy consumption, and is suitable for low-power scenarios such as outdoor / deep space.

[0060] The energy consumption (in mJ) during the system regulation phase is positively correlated with the regulation parameters, and the calculation formula is as follows: : Laser power efficiency coefficient (determined by laser hardware characteristics, unit mJ / (mW)) 2 )); Microwave phase energy consumption coefficient (determined by microwave synthesizer characteristics, unit mJ / rad). Microwave action time (positively correlated with phase adjustment amplitude; the greater the phase shift, the longer the action time).

[0061] Physical significance: To avoid excessive energy consumption in pursuit of high fidelity, such as in deep space exploration scenarios, optimization can be used to ensure that the total power consumption of the system is ≤5W, thus guaranteeing battery life.

[0062] Response delay item Delay weighting coefficient (range 0.1-0.2) ensures real-time response in dynamic environments.

[0063] The total delay (in μs) for quantum control and signal processing is calculated using the following formula: : Laser pulse duration (negatively correlated with power; higher power results in faster excitation speed and shorter duration). Microwave phase adjustment time (positively correlated with phase offset). Fixed signal processing delay (determined by FPGA hardware, approximately 5μs). Physical significance: Controls the time overhead of the regulation process, ensuring that the end-to-end delay is ≤50μs, meeting the monitoring requirements of high-speed dynamic environments (such as transient magnetic fields of power grids).

[0064] Constraints: .

[0065] The upper limit of laser power (determined by hardware safety and energy consumption thresholds, 50mW in this invention) is to avoid excessive power causing thermal damage to the sensor or excessive energy consumption. The physical range of microwave phase values ​​(phase is a periodic parameter, 0 and 2π are equivalent) ensures the physical effectiveness of the control signal.

[0066] This formula, through multi-objective optimization logic, solves the problem of the singularity of traditional quantum control, which "only pursues fidelity and ignores energy consumption and delay": Prioritizing high fidelity of quantum states lays the foundation for sub-nanoscale measurement accuracy; dynamically balancing energy consumption and latency allows the system to adapt to the needs of various scenarios such as industry, scientific research, and deep space; providing a clear objective function for reinforcement learning (DDPG) enables the control parameters to be adjusted in real time according to environmental noise and energy consumption requirements (such as increasing under strong electromagnetic interference). Increase power consumption in low-power scenarios This improves the system's adaptability to different environments.

[0067] Example 1: Condition monitoring of ultra-high voltage power grid transformers (industrial scenario).

[0068] For the main transformer of a 1000kV ultra-high voltage substation, it is necessary to monitor the winding leakage magnetic field (0.1-100nT), oil temperature (-20℃~120℃), and core vibration (1-100pm / s) in real time. 2 The system provides early warnings of insulation aging and core loosening faults, and requires electromagnetic interference resistance (electric field strength ≤500V / m), continuous 24-hour operation, and a warning accuracy rate ≥99%.

[0069] Quantum sensing front end: Four quantum entangled sensing units are distributed on the outer wall of the transformer tank (10cm away from the winding), with NV color center diamond encapsulation moisture-proof and high-temperature resistant shell (-40℃~150℃), and the MEMS micro-displacement platform is fixed in a dense topology (spatial resolution 200nm) to ensure the precise capture of local magnetic fields and vibrations. Signal excitation and control module: laser power is adjusted to 30mW (balancing power consumption and excitation efficiency), terahertz microwave synthesizer is locked to the 2.87GHz fundamental frequency (adapting to the magnetic field range of the power grid environment), and quantum dot single-photon detector is enhanced with electromagnetic shielding (metal shielding cover + grounding treatment). Quantum-classical hybrid computing engine: QPU enables 64 qubits (simplifies quantum circuits and improves solution speed), reinforcement learning reward function focuses on "precision-stability" (weakens energy consumption weight), and federated Kalman filter aggregation cycle is shortened to 5ms; Edge cloud transmission: wired Ethernet (1Gbps) is used first, with LoRa wireless link as backup (to resist electromagnetic interference), and the data compression ratio of edge nodes is optimized to 8:1 (to balance transmission efficiency and data integrity).

[0070] Calibration phase: A substation-specific standard magnetic field source (0.1-200nT, accuracy ±0.005nT) and an oil-immersed temperature control box are used to establish a three-dimensional calibration dictionary of "magnetic field-temperature-vibration". The calibration is performed automatically every 30 minutes (the temperature drift in the power grid environment is relatively fast). Signal acquisition: Fluorescent photon counting rate increased to 1.5 × 10⁻⁶ 6 The counts / s are used to capture transient magnetic field fluctuations (≤10μs) caused by changes in transformer load. Solution optimization: Chaotic wavelet denoising algorithm enhances low-frequency noise suppression (targeting 50Hz power frequency interference from the power grid), and DDPG algorithm dynamically adjusts microwave phase (step size 0.05°) to offset the micro-displacement error of the sensing unit caused by transformer vibration; Early warning mechanism: The Lyapunov index threshold is set to 0.08 (to adapt to the magnetic field change characteristics of power grid fault precursors), and abnormal data is cross-verified with transformer operating parameters (load current, oil level) to avoid false alarms.

[0071] Monitoring accuracy: magnetic field ≤ 0.08 nT, oil temperature ≤ 8 μK, vibration ≤ 0.8 pm / s 2 This meets the requirements for fault feature capture; Real-time performance: End-to-end latency ≤30μs, response time ≤1ms during load surges; Anti-interference and stability: In a 500V / m electric field environment, the measurement error is ≤2%, and it has been running continuously for 30 days without failure. It has successfully provided early warnings for two instances of core loosening (48 hours in advance) and one instance of winding insulation aging (72 hours in advance), with an early warning accuracy rate of 99.5%.

[0072] Example 2: Monitoring of weak magnetic fields in neurons in brain science (scientific research scenario).

[0073] For monitoring the activity of neurons in the motor cortex of macaque brain, it is necessary to non-invasively capture the weak magnetic field (10-100 fT) generated by the firing of individual neurons and the local temperature of brain tissue (36℃~38℃). The system should have a spatial resolution of ≤1mm, a temporal resolution of ≤10μs, and be free from electromagnetic radiation interference (meeting medical safety standards).

[0074] Quantum sensing front end: An array of 8 miniature quantum entangled sensing units (each unit is 1cm × 1cm × 0.5cm) is attached to the surface of the skull, increasing the NV color center concentration to 5 × 10⁻⁶. 17 cm -3(To enhance signal strength), the rubidium atom gas cell is reduced to 2mm. 3 (Miniaturized design), MEMS platform dynamically switches topology (100nm resolution for active neuron region, 500nm resolution for other regions). Signal excitation and modulation module: The power of the photonic crystal laser is reduced to 10mW (to avoid thermal damage to brain tissue), the linewidth is optimized to 300Hz (to improve frequency stability), and the microwave synthesizer adopts low power output (≤1mW) to meet electromagnetic safety standards; Quantum-classical hybrid computing engine: The QPU uses 256 qubits (IBM Quantum Osprey), the quantum neural network adds one hidden layer (to improve the ability to solve weak signals), and the FPGA enhances low-frequency noise suppression (for thermal noise in biological tissues). Terminal application module: The hardware terminal is simplified to a portable tablet, which displays the heat map of neuron magnetic field distribution in real time, and the historical data storage focuses on the effective data within 24 hours (reducing storage pressure).

[0075] Calibration phase: A superconducting quantum interference device (SQUID) was used as the standard source (accuracy ±1fT). 1000-point calibration was completed in a constant temperature environment of 37℃ to establish a high-precision mapping model of "fT-level magnetic field-temperature". The calibration parameters were automatically fine-tuned every 10 minutes. Signal acquisition: Fluorescent photon counting rate set to 2×10 6 counts / s, synchronously acquire EEG signals (assisting cross-validation), and optimize the quantum entanglement correlation function to extract three-dimensional features of "magnetic field-temperature-time"; Solution optimization: The QNN solution cycle is shortened to 5μs, the federated Kalman filter weights are tilted towards the magnetic field signal (to improve the sensitivity of fT-level signals), and the DDPG algorithm focuses on optimizing the stability of laser power (to avoid thermal noise fluctuations). Interference suppression: Add a "biological tissue noise library" and use a chaotic wavelet denoising algorithm to match and remove interference signals such as cerebral blood flow and muscle tremors (signal-to-noise ratio improved to 70dB).

[0076] Monitoring accuracy: magnetic field resolution ≤5fT, temperature accuracy ≤5μK, spatial resolution 800μm (reaching the level of single neuron localization). Temporal resolution: ≤8μs, capturing magnetic field pulses of neuronal firing (duration 20μs); Safety: No electromagnetic radiation (microwave power ≤1mW), local temperature rise in brain tissue ≤0.1℃, meeting medical safety standards, and successfully recording the dynamic changes in the magnetic field distribution of the motor cortex during macaque grasping movements.

[0077] Example 3: Plasma environment monitoring for deep space probes (extreme environment scenario).

[0078] For lunar probes (such as the Chang'e series), it is necessary to monitor the lunar surface plasma magnetic field (0.1-10 nT) and the micro-vibrations of the probe itself (0.1-10 pm / s). 2 The system must be resistant to cosmic radiation (dose ≤100Gy), have low power consumption (≤5W), and have quantum state stability affected by radiation by ≤5%.

[0079] Quantum sensing front end: Quantum entangled sensing array encapsulated in a radiation-resistant shell (titanium alloy + shielding film), NV color center diamond uses a high-purity (99.999%) substrate (to improve radiation resistance), rubidium atom gas chamber is equipped with a vacuum insulation layer (to cope with extreme temperature differences), and MEMS micro-displacement platform automatically adjusts the topology once per hour (to reduce energy consumption). Signal excitation and control module: The laser source adopts pulse mode (duty cycle 50%), with an average power of 15mW. The microwave synthesizer incorporates radiation hardening design, and the quantum dot single-photon detector optimizes the dark count rate (≤5cps) to reduce radiation-induced noise. Quantum-classical hybrid computing engine: QPU enables low-power mode (qubit coherence time ≥80μs), heterogeneous classical computing units disable redundant computing power (power consumption reduced to 3W), and the reinforcement learning reward function focuses on "energy consumption-radiation resistance"; Edge cloud transmission: It adopts quantum key distribution (QKD) combined with deep space telemetry and control link, with a data compression ratio of 15:1 (adapting to the low bandwidth of deep space) and end-to-end latency relaxed to 100μs (meeting the needs of deep space monitoring).

[0080] Calibration phase: Initial calibration is completed in a simulated lunar surface environment on the ground (vacuum, -60℃~80℃, radiation dose 50Gy), the dynamic calibration dictionary update cycle is extended to 5 minutes (to reduce energy consumption), and the drift error tolerance is increased to 0.002% / h; Signal acquisition: Fluorescent photon counting rate decreased to 8 × 10⁻⁶ 5 counts / s, microwave control frequency range is widened to 0.5-5GHz (to adapt to plasma magnetic field fluctuations). Solution optimization: The single iteration time of the quantum-classical hybrid solution is extended to 2ms (balancing energy consumption and accuracy). The DDPG algorithm optimizes the quantum state freezing trigger condition (automatically starts the freezing mode when the radiation dose is ≥80Gy to maintain the spin ground state of the NV color center). To address the quantum state distortion caused by radiation, a quantum state error correction algorithm (based on surface code logic) is introduced to correct single-qubit flipping errors and ensure that the solution accuracy is not affected by radiation. At the same time, the number of trainable parameters of the QNN quantum circuit is simplified (reduced by 20%), reducing computational energy consumption. Extreme Environment Response and Data Verification: To address extreme temperature differences on the lunar surface, the ambient temperature is fed back in real time through temperature sensors, and temperature compensation factors in the calibration dictionary are dynamically invoked to correct sensor unit parameter drift caused by temperature differences. In the data transmission link, a combination of forward error correction coding (FEC) and cyclic redundancy check (CRC) is used to compensate for the low bandwidth and high packet loss rate of the deep space telemetry and control link, ensuring an effective data transmission rate of ≥95%. When strong radiation or extreme temperatures disappear, the system automatically executes a quantum state fast recovery process (based on pre-stored initial entangled state parameters), restoring normal monitoring within 100μs and reducing data interruption time.

[0081] Monitoring accuracy: Plasma magnetic field ≤0.09nT, chamber temperature ≤9μK, body micro-vibration ≤0.4pm / s 2 It accurately captures random fluctuations in the lunar surface plasma magnetic field and micro-vibration signals caused by the probe's attitude adjustment; Low power consumption: The total system power consumption is stable at 4.2W (≤5W design threshold), of which the computing engine consumes 2.8W and the signal excitation module consumes 1.4W, meeting the long endurance requirements of deep space probes (supporting 90 days of continuous monitoring in a single mission). Radiation resistance and stability: In a 100 Gy cosmic radiation environment, the quantum state stability decays by only 3.8% (≤5% of the design requirement), and the measurement data error is ≤4%; when encountering a short-duration strong radiation pulse (120 Gy), the quantum state freezing mechanism is successfully triggered, and monitoring resumes within 95 μs after the interference disappears, with a data continuity loss of ≤0.05%; Real-time performance and adaptability: The end-to-end transmission delay is stable at 85μs (≤100μs), and the 15:1 data compression ratio is adapted to the low-bandwidth deep space telemetry and control link, successfully transmitting more than 98% of the effective monitoring data; Mission support effect: Continuously outputting lunar surface plasma magnetic field distribution, temperature gradient and detector micro-vibration data, providing high-precision environmental data support for detector attitude control algorithm optimization, and helping to avoid the risk of communication interruption caused by plasma interference.

[0082] The following is a comparison table of the effects of the three embodiments:

[0083] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A high-precision quantum sensing environmental physical quantity real-time monitoring system, characterized in that, Comprise: Quantum sensing front-end, the quantum sensing front-end contains quantum entanglement sensing array integrated by NV color center diamond and rubidium atom gas chamber, the quantum entanglement sensing array forms quantum entangled state through laser-induced electronic spin of NV color center and energy level transition of rubidium atom, for synchronous sensing magnetic field, temperature, pressure and vibration; Signal excitation and regulation module, the signal excitation and regulation module includes photonic crystal laser, terahertz microwave synthesizer and quantum dot single photon detector, the photonic crystal laser and terahertz microwave synthesizer are optimized by quantum optimal control algorithm (GRAPE) The output laser pulse sequence and microwave phase, the quantum dot single photon detector collects quantum fluctuation signals of environmental noise in real time and generates inverse compensation signals to suppress noise interference; Quantum-classical hybrid computing engine, the quantum-classical hybrid computing engine includes quantum processing unit (QPU) and heterogeneous classical computing unit, the QPU is a superconducting quantum bit processor, for mapping and solving quantum state characteristics through quantum neural network (QNN); The heterogeneous classical computing unit includes FPGA, GPU and ARM, wherein FPGA executes chaos wavelet denoising algorithm to suppress noise, GPU runs federal Kalman filtering algorithm to realize multi-node data aggregation, and ARM runs reinforcement learning parameter optimization algorithm to dynamically adjust system parameters; Edge cloud collaborative transmission module, the edge cloud collaborative transmission module combines quantum key distribution (QKD) and 5G slicing technology, for encrypted transmission of data preprocessed by edge nodes, and end-to-end transmission delay≤50μs; Terminal application module, the terminal application module is deployed with a physical quantity prediction model based on Transformer time series network, for real-time display of monitoring data, execution of abnormal early warning and prediction of physical quantity change trend.

2. The high-precision quantum sensing environmental physical quantity real-time monitoring system and method according to claim 1, characterized in that, The quantum entanglement sensing array also contains MEMS micro displacement platform, the MEMS micro displacement platform dynamically adjusts the topology of the array through reinforcement learning, so that the spatial resolution of the quantum entanglement sensing array is adaptively switched in the range of 100nm-1mm;And the entanglement degree (Concurrence) of the quantum entangled state≥0.

9.

3. The high-precision quantum sensing environmental physical quantity real-time monitoring system and method according to claim 1, characterized in that, The output wavelength of the photonic crystal laser is 532nm, and the line width≤1kHz;The frequency regulation range of the terahertz microwave synthesizer is 0.1-10GHz, and the phase noise≤-120dBc / Hz;The response time of the quantum dot single photon detector≤10ps.

4. The high-precision quantum sensing environmental physical quantity real-time monitoring system and method according to claim 1, characterized in that, The quantum processing unit (QPU) is a 128-qubit superconducting quantum bit processor;In the heterogeneous classical computing unit, the reinforcement learning parameter optimization algorithm takes measurement accuracy, energy consumption and delay as a multi-objective reward function.

5. The high-precision quantum sensing environmental physical quantity real-time monitoring system and method according to claim 1, characterized in that, In the edge cloud collaborative transmission module, the compression ratio of data by edge nodes is 10:1;The terminal application module supports storage of historical data for≥1years, and the prediction accuracy rate of the physical quantity prediction model for the change trend of physical quantity within 10 seconds is≥98%.

6. A high-precision quantum sensing environmental physical quantity real-time monitoring method, characterized in that, Comprise the following steps: Step 1: Quantum state initialization and entanglement calibration: reconstruct the initial quantum entangled state of the NV color center and rubidium atom using quantum state tomography technology, calculate the entanglement degree of the quantum entangled state, if the entanglement degree is lower than the preset threshold, start laser repumping to restore the entanglement degree; at the same time, update the quantum state and physical quantity dynamic calibration dictionary every 100 ms based on Bayesian inference; Step 2: Multi-dimensional signal acquisition and feature fusion: simultaneously acquire the fluorescence photon count, microwave resonance frequency offset and atomic gas cell absorption spectrum signal output by the quantum entanglement sensing array, extract the multi-physical quantity coupling features in the above signals through the quantum entanglement correlation function, and cross-verify the coupling features based on the intrinsic coupling equation of the physical quantities; Step 3: Quantum-classical hybrid solution: use the quantum neural network (QNN) in the quantum processing unit (QPU) to solve the coupling features, output the quantum probability distribution of the physical quantities; convert the quantum probability distribution into deterministic physical quantity values through a heterogeneous classical computing unit; then dynamically adjust the parameters of the QNN and the quantum control signal through a reinforcement learning parameter optimization algorithm, forming a closed loop optimization; Step 4: Adaptive early warning and system reconstruction: calculate the Lyapunov exponent of the physical quantity time series based on chaos theory, identify abnormal fluctuations and trigger an early warning; If an extreme environment is monitored, automatically switch to the quantum state freezing mode, and resume monitoring within 100 μs after the extreme environment disappears.

7. The high-precision quantum-sensing environmental-physical-quantity real-time monitoring system and method of claim 6, wherein In step 1, the time consumption of the laser repumping is ≤50 μs; the long-term drift error of the quantum state and physical quantity dynamic calibration dictionary is ≤0.001% / h. 8.The high-precision quantum sensing environment physical quantity real-time monitoring system and method of claim 6, wherein In step 2, the collection rate of the fluorescent photon counts is ; the physical quantity intrinsic coupling equation includes a temperature and pressure cross-validation equation based on a PV = nRT correction model. 9.The high-precision quantum sensing environment physical quantity real-time monitoring system and method of claim 6, wherein In step 3, the single iteration time of the quantum-classical hybrid solution is ≤1 ms; the reinforcement learning parameter optimization algorithm uses the deep deterministic policy gradient (DDPG) algorithm.

10. The high-precision quantum sensing environmental physical quantity real-time monitoring system and method according to claim 6, characterized in that, In step 4, the early warning time of the abnormal fluctuations is ≥10 seconds earlier than the traditional threshold method; the extreme environment is a strong electromagnetic pulse environment with intensity ≥1000 V / m.

Images