Vehicle-road collaborative dynamic decision-making method and device based on quantum-classical hybrid architecture

Abstract

The invention relates to the technical field of intelligent traffic and quantum computing crossing, in particular to a vehicle-road collaborative dynamic decision-making method and device based on a quantum-classical hybrid architecture, and the method comprises the steps: carrying out the quantum principal component analysis of high-complexity multi-mode sensor data through a preset anti-vibration quantum processing unit, so as to obtain 64-dimensional data; a quantum annealing solution Ising model and a Pareto frontier screening method are used for carrying out combinatorial optimization problem solution on 64-dimensional data, so that an approximate optimal solution set is obtained; and performing KKT condition verification on the approximate optimal solution set based on a Q-Dyna hybrid decision algorithm, executing the approximate optimal solution set under the condition that the KKT condition verification is passed, and otherwise, performing reinforcement learning rollback on the approximate optimal solution set to output a suboptimal solution set. Therefore, the problems of computing power defect, quantum noise interference, communication protocol incompatibility and the like existing in an existing scheme for solving the real-time bottleneck of high-dimensional traffic flow optimization are solved.

Description

Technical Field

[0001] This invention relates to the field of intelligent transportation and quantum computing, and in particular to a vehicle-road cooperative dynamic decision-making method and device based on a quantum-classical hybrid architecture. Background Technology

[0002] Driven by both accelerated urbanization and a continuous increase in motor vehicle ownership, urban transportation systems are facing unprecedented complexity and dynamic challenges. Traditional traffic flow optimization methods are mostly based on low-dimensional spatial models, simplifying vehicle motion characteristics, road network topology, and interaction rules to construct macro- or meso-level control frameworks with flow, speed, and density as core parameters. However, with the increasing penetration of intelligent connected vehicles (ICVs) and the emergence of multi-modal transportation entities (such as autonomous vehicles, shared mobility tools, and drone delivery), transportation systems have evolved into multi-dimensional dynamic systems containing continuous spatiotemporal variables, heterogeneous behavioral decisions, and high-density coupling relationships. Existing optimization models are caught in a dilemma due to the curse of dimensionality: on the one hand, oversimplification leads to a disconnect between strategies and actual scenarios, making it difficult to cope with nonlinear events such as sudden congestion and accident disturbances; on the other hand, the introduction of high-precision simulation or reinforcement learning and other high-dimensional modeling techniques results in an exponential increase in computational complexity, failing to meet the millisecond-level response requirements of real-time decision-making. Especially during peak hours or under extreme weather conditions, delayed optimization may trigger a chain reaction of congestion, causing city-wide traffic paralysis. Therefore, overcoming the real-time bottleneck of high-dimensional traffic flow optimization has become a key technological challenge in building resilient urban transportation systems.

[0003] However, existing solutions to address the real-time bottleneck of high-dimensional traffic flow optimization have the following drawbacks: 1. Computing bottleneck: Traditional edge servers (such as NVIDIA DRIVE AGX Orin) have a latency of >100ms when processing 32-vehicle route optimization, which cannot meet the ASIL-D standard (≤10ms).

[0004] 2. Quantum noise interference: Vehicle vibration causes the decoherence time of the superconducting quantum chip to decrease to ≤5μs (compared to D-Wave laboratory data).

[0005] 3. Communication protocol incompatibility: 3GPP Release 16 does not define the quantum computing data field, and the quantum solution cannot be synchronized with classical control instructions.

[0006] 4. Bottlenecks in the adaptation of quantum computing to automotive applications (1) Material defects: Decoherence time decay model of traditional superconducting chips (such as D-Wave 5000Q) under vibration environment: τ=τ0 e γf (γ=0.023, f: vibration frequency) This results in a coherence time decay of >60% when vibrating at frequencies above 5Hz.

[0007] (2) Protocol limitations: Current V2X protocol field structure defects: # IEEE 1609.2 protocol field legacy_frame = { 'header': 24, # Fixed GPS timestamp 'data': 128,# Vehicle motion status 'crc': 16# Basic CRC checksum }# Quantum parameter capacity = 0” 5. Bottlenecks in pure classic architecture (1) Decision delay: Traditional GPU / CPU solutions (such as NVIDIA Drive AGX) take ≥120ms for path planning in a 100-node road network (SUMO simulation data), due to the complexity of Dijkstra's algorithm being O(n2). (2) High conflict rate: In multi-vehicle interaction scenarios (more than 15 obstacles per 100 meters), the path conflict rate is as high as 23% (CARLA platform test). (3) Energy consumption limitations: The 35W power consumption limits the battery life of in-vehicle terminals.

[0008] 6. Defects of pure quantum architecture (1) Insufficient number of bits: The D-Wave annealing machine has fewer than 1000 qubits, which cannot support large-scale road network decisions. (2) Interface missing: Quantum state encoding is incompatible with the automotive CAN bus protocol, resulting in a data conversion delay >50ms. (3) Poor reliability: The decoherence time of the superconducting quantum chip drops sharply by 80% in the temperature range of -40℃ to 125℃. Summary of the Invention

[0009] This invention provides a vehicle-road cooperative dynamic decision-making method and device based on a quantum-classical hybrid architecture to solve the problems of computing power deficiency, quantum noise interference, and communication protocol incompatibility in existing solutions to the real-time bottleneck of high-dimensional traffic flow optimization.

[0010] A first aspect of this invention provides a vehicle-road cooperative dynamic decision-making method based on a quantum-classical hybrid architecture, comprising the following steps: When the multimodal sensor data of the target vehicle is highly complex, a preset anti-vibration quantum processing unit is used to perform quantum principal component analysis on the multimodal sensor data to obtain 64-dimensional data. Quantum annealing is used to solve the Ising model and Pareto front screening method to solve the combinatorial optimization problem of 64-dimensional data, so as to obtain a set of approximate optimal solutions; Based on the Q-Dyna hybrid decision algorithm, KKT condition verification is performed on the near-optimal solution set. If the KKT condition verification passes, the near-optimal solution set is executed; otherwise, reinforcement learning is performed on the near-optimal solution set to back off and output a suboptimal solution set.

[0011] Optionally, the preset anti-vibration quantum processing unit includes a mechanical damping layer, a magnetically controlled stabilizing layer, and a quantum chip layer, wherein, The mechanical damping layer is used to generate a reverse mechanical wave by real-time acquisition of the vehicle body vibration spectrum to counteract the vibration energy in the multimodal sensor data. The magnetically controlled stabilizing layer is used to utilize the flux pinning effect of the high-temperature superconducting YBCO ring to suppress the phase drift of the quantum bits caused by geomagnetic disturbances in the multimodal sensor data. The quantum chip layer is used to improve the tunneling current stability of the multimodal sensor data by employing an asymmetric barrier design.

[0012] Optionally, it also includes: The multimodal sensor data is input into the preset vibration-resistant quantum processing unit using the optimized V2X frame structure for quantum principal component analysis. The optimized V2X frame structure has a protocol header of 24 bits (message type + GPS timestamp) + 4 bits QSI; The optimized V2X frame structure has a data field of 128 bits (vehicle position / speed) + 32 bits annealing parameter vector; The optimized V2X frame structure has a check field of 16-bit Cyclic Redundancy Check (CRC) + 24-bit Specific Control Code (STCC).

[0013] Optionally, the step of using quantum annealing to solve the Ising model and the Pareto front screening method to solve the combined optimization problem of 64-dimensional data to obtain a set of approximate optimal solutions includes: The 64-dimensional data is projected onto the feature space using quantum Fourier transform to obtain 64-dimensional data with reduced computational complexity. Quantum annealing is used to solve the Ising model to solve the combinatorial optimization problem of the 64-dimensional data with reduced computational complexity, so as to obtain a set of candidate solutions; The approximate optimal solution set is determined from the candidate solution set using the Pareto front screening method.

[0014] A second aspect of the present invention provides a vehicle-road cooperative dynamic decision-making device based on a quantum-classical hybrid architecture, comprising: The principal component analysis module is used to perform quantum principal component analysis on the multimodal sensor data of the target vehicle when the multimodal sensor data is highly complex, using a preset anti-vibration quantum processing unit to obtain 64-dimensional data. The problem-solving module is used to solve the Ising model using quantum annealing and the Pareto front screening method to solve the combined optimization problem of 64-dimensional data, so as to obtain a set of approximate optimal solutions; The validation model is used to perform KKT conditional validation on the near-optimal solution set based on the Q-Dyna hybrid decision algorithm. If the KKT conditional validation passes, the near-optimal solution set is executed; otherwise, reinforcement learning is performed on the near-optimal solution set to back up and output a suboptimal solution set.

[0015] Optionally, the preset anti-vibration quantum processing unit includes a mechanical damping layer, a magnetically controlled stabilizing layer, and a quantum chip layer, wherein, The mechanical damping layer is used to generate a reverse mechanical wave by real-time acquisition of the vehicle body vibration spectrum to counteract the vibration energy in the multimodal sensor data. The magnetically controlled stabilizing layer is used to utilize the flux pinning effect of the high-temperature superconducting YBCO ring to suppress the phase drift of the quantum bits caused by geomagnetic disturbances in the multimodal sensor data. The quantum chip layer is used to improve the tunneling current stability of the multimodal sensor data by employing an asymmetric barrier design.

[0016] Optionally, it also includes: The input module is used to input the multimodal sensor data into the preset vibration-resistant quantum processing unit using the optimized V2X frame structure for quantum principal component analysis. The optimized V2X frame structure has a protocol header of 24 bits (message type + GPS timestamp) + 4 bits QSI; The optimized V2X frame structure has a data field of 128 bits (vehicle position / speed) + 32 bits annealing parameter vector; The optimized V2X frame structure has a check field of 16-bit Cyclic Redundancy Check (CRC) + 24-bit Specific Control Code (STCC).

[0017] Optionally, the problem-solving module includes: The 64-dimensional data is projected onto the feature space using quantum Fourier transform to obtain 64-dimensional data with reduced computational complexity. Quantum annealing is used to solve the Ising model to solve the combinatorial optimization problem of the 64-dimensional data with reduced computational complexity, so as to obtain a set of candidate solutions; The approximate optimal solution set is determined from the candidate solution set using the Pareto front screening method.

[0018] A third aspect of the present invention provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the vehicle-road cooperative dynamic decision-making method based on a quantum-classical hybrid architecture as described in the above embodiments.

[0019] A fourth aspect of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described vehicle-road cooperative dynamic decision-making method based on a quantum-classical hybrid architecture.

[0020] The vehicle-road cooperative dynamic decision-making method and device based on a quantum-classical hybrid architecture proposed in this invention breaks through the decoherence time limit of quantum computing in vehicle scenarios (target ≥85μs); achieves a hybrid decision-making delay ≤10ms (32-vehicle scenario); and is compatible with the quantum data communication interface of existing V2X protocols.

[0021] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0022] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1 A flowchart illustrating a vehicle-road cooperative dynamic decision-making method based on a quantum-classical hybrid architecture according to an embodiment of the present invention; Figure 2 This is a schematic diagram illustrating the specific execution of a vehicle-road cooperative dynamic decision-making method based on a quantum-classical hybrid architecture according to an embodiment of the present invention. Figure 3 This is a schematic diagram of a V2X structure provided according to an embodiment of the present invention; Figure 4 This is a schematic diagram of the structure of a preset anti-vibration quantum processing unit according to an embodiment of the present invention; Figure 5 A two-domain flowchart of a Q-Dyna hybrid decision algorithm provided according to an embodiment of the present invention; Figure 6 This is a schematic diagram illustrating the execution of a Q-Dyna hybrid decision algorithm according to an embodiment of the present invention; Figure 7 This is a block diagram of a vehicle-road cooperative dynamic decision-making device based on a quantum-classical hybrid architecture according to an embodiment of the present invention; Figure 8 This is a schematic diagram of the structure of an electronic device according to an embodiment of the present invention.

[0023] Explanation of reference numerals in the attached figures: 70-Vehicle-Road Cooperative Dynamic Decision-Making Device Based on Quantum-Classical Hybrid Architecture, 701-Complexity Analysis Module, 702-Principal Component Analysis Module, 703-Problem Solving Module, 704-Verification Model, 801-Memory, 802-Processor, 803-Communication Interface. Detailed Implementation

[0024] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0025] The following description, with reference to the accompanying drawings, illustrates a vehicle-road cooperative dynamic decision-making method and apparatus based on a quantum-classical hybrid architecture. Addressing the issues mentioned in the background section regarding the lack of automotive-grade vibration-resistant quantum hardware design, the absence of a real-time verification mechanism for quantum-classical hybrid decision-making, and the lack of standardization for quantum data fields in V2X protocols, this invention provides a vehicle-road cooperative dynamic decision-making method based on a quantum-classical hybrid architecture. This method utilizes vibration-resistant quantum hardware (a three-stage vibration reduction structure) and a hybrid decision-making algorithm (Q-Dyna) to overcome the real-time bottleneck in high-dimensional traffic flow optimization, making it suitable for L3-L4 level autonomous driving scenarios.

[0026] It should be noted that the lack of automotive-grade vibration-resistant quantum hardware design specifically includes: (1) Detailed breakdown of current issues: Current automotive quantum hardware (such as D-Wave 5000Q) exhibits a decoherence time decay rate >60% under vibration environments (>5Hz), calculated using the following formula:

[0027] This results in a quantum bit failure rate of up to 12% under actual road conditions (common vibration frequencies of 20-80Hz), which fails to meet the ISO26262 ASIL-D reliability requirements.

[0028] (2) The essence of technological gaps: Material defects: Traditional superconducting chips (such as aluminum-based Josephson junctions) have a critical current fluctuation of >15% at -40℃, while automotive-grade requirements are <5%; Structural deficiencies: Lack of a three-stage piezoelectric ceramic anti-vibration array (attenuation rate ≥98%@80Hz) and a YBCO superconducting ring flux pinning design (pinning density >10). 4 An integrated solution (flux / cm²).

[0029] (3) Evidence of blank consequences: Actual test data shows that vibration increases the error rate of quantum path planning by 23% and the decision delay in emergency obstacle avoidance scenarios reaches 210ms (exceeding the safety threshold of 100ms).

[0030] The lack of a real-time verification mechanism for quantum-classical hybrid decision-making specifically includes: (1) Existing technological bottlenecks: Traditional serial verification processes (quantum optimization → classical verification) require multiple data transmissions, with latency accounting for over 40% of the total decision-making time. (2) Core gaps: Validation model defects: No real-time comparison mechanism was established between KKT conditions (Karush-Kuhn-Tucker) and quantum annealing results, resulting in a validation delay of >10ms; Dynamic switching is missing: There is no task rollback strategy when unexpected scenarios occur (such as the sudden appearance of obstacles), which increases the recalculation time of the classic algorithm to 60ms.

[0031] The lack of standardization for quantum data fields in V2X protocols specifically includes: (1) Protocol layer defect analysis: In the current IEEE 1609.2 protocol, the quantum parameter capacity is 0 in the 128-bit data field, resulting in: Key parameters such as annealing temperature ΔT (accuracy required to be 0.01K) and coupling strength Jij cannot be transmitted; The quantum-classical interface requires custom parsing, which adds a 20ms decoding latency.

[0032] (2) Consequences of standardization gaps: Custom fields used by various manufacturers (such as Huawei Q-V2X occupying 24 bits and ZTE QuanTE occupying 32 bits) have a cross-platform compatibility success rate of only 58%.

[0033] Specifically, Figure 1 This is a flowchart illustrating a vehicle-road cooperative dynamic decision-making method based on a quantum-classical hybrid architecture, provided in an embodiment of the present invention.

[0034] like Figure 1As shown, the vehicle-road cooperative dynamic decision-making method based on a quantum-classical hybrid architecture includes the following steps: In step S101, when the multimodal sensor data is highly complex, a preset anti-vibration quantum processing unit is used to perform quantum principal component analysis on the multimodal sensor data to obtain 64-dimensional data.

[0035] In some embodiments, the preset vibration-damping quantum processing unit includes a mechanical damping layer, a magnetically controlled stabilizing layer, and a quantum chip layer, wherein... The mechanical damping layer is used to generate reverse mechanical waves by acquiring the vibration spectrum of the vehicle body in real time, so as to counteract the vibration energy in the multimodal sensor data. A magnetically controlled stabilization layer is used to utilize the flux pinning effect of a high-temperature superconducting YBCO ring to suppress quantum bit phase drift caused by geomagnetic disturbances in multimodal sensor data. A quantum chip layer is used to improve the tunneling current stability of multimodal sensor data using an asymmetric barrier design.

[0036] In some embodiments, it also includes: The optimized V2X frame structure is used to input multimodal sensor data into a preset vibration-resistant quantum processing unit for quantum principal component analysis. The optimized V2X frame structure has a 24-bit header (message type + GPS timestamp) + 4-bit QSI; The optimized V2X frame structure has a data field of 128 bits (vehicle position / speed) + 32 bits annealing parameter vector; The optimized V2X frame structure has a 16-bit Cyclic Redundancy Check (CRC) + a 24-bit Specific Control Code (STCC) in its check field.

[0037] like Figure 2 As shown, multimodal sensor data of the target vehicle is collected, and scene complexity analysis is performed on the multimodal sensor data to determine whether the multimodal sensor data belongs to high complexity. If the multimodal sensor data does not belong to high complexity, the classical A* algorithm is used to process the multimodal sensor data. If the multimodal sensor data belongs to high complexity, the optimized V2X frame structure is used to input the multimodal sensor data into the preset anti-vibration quantum processing unit for quantum principal component analysis.

[0038] It should be noted that the original V2X frame structure suffers from defects in existing fields, missing quantum parameters, high redundancy, and spatiotemporal asynchrony. Existing field defects refer to three major problems exposed in traditional IEEE 1609.2 protocol frames under quantum hybrid scenarios: Missing quantum parameters refer to the inability to transmit key parameters such as annealing temperature ΔT and coupling strength Jij. High redundancy indicates that GPS timestamp duplication rate is 38% and data field utilization rate is less than 60%. Time and space asynchrony refers to the inability of classic CRC check to compensate for clock drift in the vehicle environment, with a synchronization error of ±10ns.

[0039] like Figure 3 As shown, this embodiment of the invention proposes an optimized V2X frame structure, the specific design and technical effects of which are as follows:

[0040] Example of dynamic parameter encoding: # Real-time generation code for quantum annealing parameters def encode_quantum_params(temperature=12.56, coupling=[0.8, 1.2, 0.5, 1.0]): temp_hex = int(temperature * 100) # 0.01K precision → 1256 → 0x1F3A coupling_hex = [int(c * 20) for c in coupling]# 0.05 precision → [0x0A,0x18, 0x0A, 0x14] return f"TEMP:{temp_hex:04X}, J:{''.join(f'{c:02X}' for c incoupling_hex)}" It should be noted that, in addition to using the optimized V2X frame structure, those skilled in the art can also use 5G NR-Uu.

[0041] Furthermore, such as Figure 4 As shown in the table below, the top-layer mechanical damping layer, which utilizes a pre-set anti-vibration quantum processing unit, employs phase reversal drive technology. It generates a reverse mechanical wave by real-time acquisition of the vehicle body vibration spectrum (5-200Hz), canceling over 80% of the vibration energy in the multi-modal sensor data. The intermediate magnetically controlled stabilizing layer utilizes the magnetic flux pinning effect (pinning density > 10) of a high-temperature superconducting YBCO ring. 4 The flux / cm²) suppresses the phase drift of qubits caused by geomagnetic disturbances in multimodal sensor data to Δφ < 0.01 rad; the Josephson junction of the bottom quantum chip layer adopts an asymmetric barrier design (AlO). x Thickness gradient 0.3nm→1.2nm) improves tunneling current stability (fluctuation <±0.05nA).

[0042]

[0043] It should be noted that those skilled in the art can use topological insulators (Bi2Se3) or molybdenum disulfide (MoS2) for the quantum chip layer.

[0044] Furthermore, the physical structure of the preset vibration-resistant quantum processing unit can adopt a 128-qubit superconducting chip (NbTiN superconducting circuit, critical current density ≥5kA / cm²), with a qubit array built using a Josephson structure, and an operating temperature of -196℃ (liquid nitrogen cooling); the vibration-resistant design of the preset vibration-resistant quantum processing unit can be a three-level piezoelectric ceramic array (top layer → middle layer YBCO superconducting ring → bottom layer quantum chip), with a vibration attenuation rate ≥98%@80Hz, ensuring a quantum coherence time ≥85μs; the coupling mechanism of the preset vibration-resistant quantum processing unit is to dynamically adjust the inter-qubit coupling strength J (range 0.1-10.0 GHz) through microwave pulses, supporting dynamic reconstruction of the Ising model.

[0045] In step S102, quantum annealing is used to solve the Ising model and Pareto front screening method to solve the combinatorial optimization problem of 64-dimensional data, so as to obtain a set of approximate optimal solutions.

[0046] In some embodiments, quantum annealing is used to solve the Ising model and the Pareto front screening method to solve a combinatorial optimization problem of 64-dimensional data, so as to obtain a set of approximate optimal solutions, including: The 64-dimensional data is projected onto the feature space through quantum Fourier transform to obtain 64-dimensional data with reduced computational complexity. Quantum annealing is used to solve the Ising model to solve the combinatorial optimization problem of 64-dimensional data with reduced computational complexity, so as to obtain a set of candidate solutions; The Pareto front screening method is used to determine the set of approximate optimal solutions from the candidate solution set.

[0047] like Figure 5 As shown, in actual execution, the 64-dimensional data is projected onto the feature space through quantum Fourier transform (QFT), which reduces the computational complexity from O(n2) to O(nlogn) and the data retention rate is ≥95%. Then, 4 qubits are used to compress the 16-dimensional data to obtain the 64-dimensional data with reduced computational complexity, i.e., 16-dimensional data, where the line depth is 8.

[0048] Furthermore, quantum annealing is used to solve the Ising model to perform a combinatorial optimization problem on the reduced 64-dimensional data, thereby obtaining a set of candidate solutions. The temperature decay rate during the quantum annealing stage is set to... To accelerate convergence, the classical simulated annealing stage is set to be triggered when the qubit has not converged, thus avoiding local optima.

[0049] Furthermore, multiple optimization objectives (such as cost, efficiency, quality, etc.) are defined. For each solution in the candidate solution set, its value on each objective is calculated. Next, a dominance relationship is determined. If solution A is not inferior to solution B on all objectives, and is strictly superior to solution B on at least one objective, then A is said to dominate B, and the dominated solution is eliminated. Then, the solutions not dominated by any other solution constitute the Pareto front, i.e., the set of approximately optimal solutions. .

[0050] In step S103, based on the Q-Dyna hybrid decision algorithm, KKT condition verification is performed on the near-optimal solution set. If the KKT condition verification passes, the near-optimal solution set is executed; otherwise, reinforcement learning backtracking is performed on the near-optimal solution set to output a suboptimal solution set. Figure 5 As shown, in actual implementation, this embodiment of the invention, under the hardware configuration of Xilinx UltraScale+ MPSoC (dual-core ARM Cortex-A76@2.4GHz + FPGA logic unit), uses the Q-Dyna hybrid decision algorithm to perform KKT condition verification on the approximate optimal solution set. The path feasibility verification based on KKT conditions (Karush-Kuhn-Tucker) uses the following constraint equations:

[0051] in, The objective function is (e.g., shortest path). For constraints (such as safe following distance).

[0052] The collision probability is also set according to the ASIL-D standard. Less than or equal to 10 -9 .

[0053] Furthermore, such as Figure 6 As shown, if the KKT condition verification passes, the decision of the near-optimal solution set is executed; otherwise, reinforcement learning is performed on the near-optimal solution set to backtrack, so that the suboptimal solution set in the candidate solution set is re-determined as the new near-optimal solution set, and the KKT condition verification is performed again on the new near-optimal solution set to finally execute the decision.

[0054] Furthermore, when making decisions about near-optimal solution sets, a dynamic task allocation model is used for coordination: Pq=τq β+τc (1 α)τc α Where Pq is the proportion of tasks allocated to quantum computing, τq is the quantum annealing time (linearly related to the problem size), Tc is the classical verification time (measured ≤ 5ms), α is the path optimization weight (0.6-0.9), and β is the real-time weight (0.1-0.4).

[0055] It should be noted that in a 32-vehicle high-speed scenario, the decision generated using the Q-Dyna hybrid decision-making algorithm has a latency of only 11.7ms. The following three specific embodiments will provide a detailed explanation of the vehicle-road cooperative dynamic decision-making method based on a quantum-classical hybrid architecture proposed in this invention.

[0056] Example 1: Highway merging scenario Quantum solution core code: sampler = EmbeddingSampler(solver='Advantage_system4.1') response = sampler.sample_ising(h, J, num_reads=100) # Annealing time=20μs Physical parameter record table:

[0057] Implementation 2: Tunnel Blind Spot Positioning Compensation (1) Input: Point cloud of LiDAR for 16 vehicles (bandwidth ≥ 1Gbps) (2) Migration: Generation time stamp synchronization error ±3ns (3) Recovery: 2cm (5σ) (no GPS scenario) Implementation 3: Coordinated lane changing for 5 vehicles on highways (1) Scenario parameters: 5-vehicle platoon, 3-vehicle continuous flow in the target lane, 5G URLLC communication latency of 50ms; (2) Dynamic task allocation: Initial number of quantum tasks: 16 (resource utilization rate: 72%) Increased traffic density detected → Q-learning model adjusted a=32 (utilization increased to 91%). Results: Lane change success rate was 98.7% (89.3% with traditional method), and the standard deviation of vehicle distance was reduced to 0.8m (42% improvement).

[0058] In summary, the vehicle-road cooperative dynamic decision-making method based on a quantum-classical hybrid architecture proposed in this embodiment of the invention solves the real-time bottleneck of high-dimensional traffic flow optimization by constructing a closed-loop decision-making mechanism of quantum annealing + KKT verification, a new V2X frame quantum field, and a dynamic task allocation model based on Q-learning. It also breaks through the decoherence time limit of quantum computing in vehicle scenarios (target ≥ 85μs), achieves a hybrid decision-making latency of ≤ 10ms (32-vehicle scenario), is compatible with the quantum data communication interface of existing V2X protocols, and is applicable to L3-L4 level autonomous driving scenarios.

[0059] Next, referring to the accompanying drawings, a vehicle-road cooperative dynamic decision-making device based on a quantum-classical hybrid architecture, according to an embodiment of the present invention, is described.

[0060] Figure 7 This is a block diagram of a vehicle-road cooperative dynamic decision-making device based on a quantum-classical hybrid architecture, provided in an embodiment of the present invention.

[0061] like Figure 7 As shown, the vehicle-road cooperative dynamic decision-making device 70 based on a quantum-classical hybrid architecture includes: a principal component analysis module 701, a problem-solving module 702, and a verification model 703.

[0062] The principal component analysis module 702 is used to collect multimodal sensor data of the target vehicle and perform scene complexity analysis on the multimodal sensor data to determine whether the multimodal sensor data belongs to high complexity. The problem-solving module 702 uses quantum annealing to solve the Ising model and Pareto front screening to solve the combined optimization problem of the 64-dimensional data, obtaining a set of approximate optimal solutions. The verification model 703 uses a Q-Dyna hybrid decision algorithm to perform KKT condition verification on the set of approximate optimal solutions. If the KKT condition verification passes, the approximate optimal solution set is executed; otherwise, reinforcement learning is used to backtrack the approximate optimal solution set to output a suboptimal solution set.

[0063] In some embodiments, the preset vibration-damping quantum processing unit includes a mechanical damping layer, a magnetically controlled stabilizing layer, and a quantum chip layer, wherein... The mechanical damping layer is used to generate reverse mechanical waves by acquiring the vibration spectrum of the vehicle body in real time, so as to counteract the vibration energy in the multimodal sensor data. A magnetically controlled stabilization layer is used to utilize the flux pinning effect of a high-temperature superconducting YBCO ring to suppress quantum bit phase drift caused by geomagnetic disturbances in multimodal sensor data. A quantum chip layer is used to improve the tunneling current stability of multimodal sensor data using an asymmetric barrier design.

[0064] In some embodiments, it also includes: The input module is used to input multimodal sensor data into a preset vibration-resistant quantum processing unit using an optimized V2X frame structure for quantum principal component analysis. The optimized V2X frame structure has a 24-bit header (message type + GPS timestamp) + 4-bit QSI; The optimized V2X frame structure has a data field of 128 bits (vehicle position / speed) + 32 bits annealing parameter vector; The optimized V2X frame structure has a 16-bit Cyclic Redundancy Check (CRC) + a 24-bit Specific Control Code (STCC) in its check field.

[0065] In some embodiments, the problem-solving module 703 includes: The 64-dimensional data is projected onto the feature space through quantum Fourier transform to obtain 64-dimensional data with reduced computational complexity. Quantum annealing is used to solve the Ising model to solve the combinatorial optimization problem of 64-dimensional data with reduced computational complexity, so as to obtain a set of candidate solutions; The Pareto front screening method is used to determine the set of approximate optimal solutions from the candidate solution set.

[0066] It should be noted that the foregoing explanation of the embodiment of the vehicle-road cooperative dynamic decision-making method based on quantum-classical hybrid architecture also applies to the vehicle-road cooperative dynamic decision-making device based on quantum-classical hybrid architecture in this embodiment, and will not be repeated here.

[0067] The vehicle-road cooperative dynamic decision-making device based on a quantum-classical hybrid architecture proposed in this embodiment of the invention solves the real-time bottleneck of high-dimensional traffic flow optimization by constructing a closed-loop decision-making mechanism of quantum annealing + KKT verification, a new V2X frame quantum field, and a dynamic task allocation model based on Q-learning. It breaks through the decoherence time limit of quantum computing in vehicle scenarios (target ≥85μs), achieves a hybrid decision-making latency of ≤10ms (32-vehicle scenario), is compatible with the quantum data communication interface of existing V2X protocols, and is applicable to L3-L4 level autonomous driving scenarios.

[0068] Figure 8 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0069] The electronic device may include: a memory 801, a processor 802, and a computer program stored on the memory 801 and capable of running on the processor 802.

[0070] When the processor 802 executes the program, it implements the vehicle-road cooperative dynamic decision-making method based on a quantum-classical hybrid architecture provided in the above embodiments.

[0071] Furthermore, electronic devices also include: Communication interface 803 is used for communication between memory 801 and processor 802.

[0072] The memory 801 is used to store computer programs that can run on the processor 802.

[0073] The memory 801 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0074] If the memory 801, processor 802, and communication interface 803 are implemented independently, then the communication interface 803, memory 801, and processor 802 can be interconnected via a bus to complete communication between them. 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 into address buses, data buses, control buses, etc. For ease of representation, Figure 8 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0075] Optionally, in a specific implementation, if the memory 801, processor 802, and communication interface 803 are integrated on a single chip, then the memory 801, processor 802, and communication interface 803 can communicate with each other through an internal interface.

[0076] The processor 802 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present invention.

[0077] This invention also provides a computer program product, which, when executed by a processor, implements the above-described vehicle-road cooperative dynamic decision-making method based on a quantum-classical hybrid architecture.

[0078] This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described vehicle-road cooperative dynamic decision-making method based on a quantum-classical hybrid architecture.

[0079] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0080] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "N" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0081] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or N executable instructions for implementing custom logic functions or processes, and the scope of preferred embodiments of the invention includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as will be understood by those skilled in the art to which embodiments of the invention pertain.

[0082] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0083] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. If implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0084] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0085] Furthermore, the functional units in the various embodiments of the present invention can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0086] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.

Claims

1. A vehicle-road cooperative dynamic decision-making method based on a quantum-classical hybrid architecture, characterized in that, Includes the following steps: When the multimodal sensor data of the target vehicle is highly complex, a preset anti-vibration quantum processing unit is used to perform quantum principal component analysis on the multimodal sensor data to obtain 64-dimensional data. Quantum annealing is used to solve the Ising model and Pareto front screening method to solve the combinatorial optimization problem of 64-dimensional data, so as to obtain a set of approximate optimal solutions; Based on the Q-Dyna hybrid decision algorithm, KKT condition verification is performed on the near-optimal solution set. If the KKT condition verification passes, the near-optimal solution set is executed; otherwise, reinforcement learning is performed on the near-optimal solution set to back off and output a suboptimal solution set.

2. The vehicle-road cooperative dynamic decision-making method based on a quantum-classical hybrid architecture according to claim 1, characterized in that, The preset anti-vibration quantum processing unit includes a mechanical damping layer, a magnetically controlled stabilizing layer, and a quantum chip layer, wherein... The mechanical damping layer is used to generate a reverse mechanical wave by real-time acquisition of the vehicle body vibration spectrum to counteract the vibration energy in the multimodal sensor data. The magnetically controlled stabilizing layer is used to utilize the flux pinning effect of the high-temperature superconducting YBCO ring to suppress the phase drift of the quantum bits caused by geomagnetic disturbances in the multimodal sensor data. The quantum chip layer is used to improve the tunneling current stability of the multimodal sensor data by employing an asymmetric barrier design.

3. The vehicle-road cooperative dynamic decision-making method based on a quantum-classical hybrid architecture according to claim 1, characterized in that, Also includes: The multimodal sensor data is input into the preset vibration-resistant quantum processing unit using the optimized V2X frame structure for quantum principal component analysis. The optimized V2X frame structure has a protocol header of 24 bits (message type + GPS timestamp) + 4 bits QSI; The optimized V2X frame structure has a data field of 128 bits (vehicle position / speed) + 32 bits annealing parameter vector; The optimized V2X frame structure has a check field of 16-bit Cyclic Redundancy Check (CRC) + 24-bit Specific Control Code (STCC).

4. The vehicle-road cooperative dynamic decision-making method based on a quantum-classical hybrid architecture according to claim 1, characterized in that, The method of using quantum annealing to solve the Ising model and Pareto front screening to solve the combinatorial optimization problem of 64-dimensional data, in order to obtain a set of approximate optimal solutions, includes: The 64-dimensional data is projected onto the feature space using quantum Fourier transform to obtain 64-dimensional data with reduced computational complexity. Quantum annealing is used to solve the Ising model to solve the combinatorial optimization problem of the 64-dimensional data with reduced computational complexity, so as to obtain a set of candidate solutions; The approximate optimal solution set is determined from the candidate solution set using the Pareto front screening method.

5. A vehicle-road cooperative dynamic decision-making device based on a quantum-classical hybrid architecture, characterized in that, include: The principal component analysis module is used to perform quantum principal component analysis on the multimodal sensor data of the target vehicle when the multimodal sensor data is highly complex, using a preset anti-vibration quantum processing unit to obtain 64-dimensional data. The problem-solving module is used to solve the Ising model using quantum annealing and the Pareto front screening method to solve the combined optimization problem of 64-dimensional data, so as to obtain a set of approximate optimal solutions; The validation model is used to perform KKT conditional validation on the near-optimal solution set based on the Q-Dyna hybrid decision algorithm. If the KKT conditional validation passes, the near-optimal solution set is executed; otherwise, reinforcement learning is performed on the near-optimal solution set to back up and output a suboptimal solution set.

6. The vehicle-road cooperative dynamic decision-making device based on a quantum-classical hybrid architecture according to claim 5, characterized in that, The preset anti-vibration quantum processing unit includes a mechanical damping layer, a magnetically controlled stabilizing layer, and a quantum chip layer, wherein... The mechanical damping layer is used to generate a reverse mechanical wave by real-time acquisition of the vehicle body vibration spectrum to counteract the vibration energy in the multimodal sensor data. The magnetically controlled stabilizing layer is used to utilize the flux pinning effect of the high-temperature superconducting YBCO ring to suppress the phase drift of the quantum bits caused by geomagnetic disturbances in the multimodal sensor data. The quantum chip layer is used to improve the tunneling current stability of the multimodal sensor data by employing an asymmetric barrier design.

7. The vehicle-road cooperative dynamic decision-making device based on a quantum-classical hybrid architecture according to claim 5, characterized in that, Also includes: The input module is used to input the multimodal sensor data into the preset vibration-resistant quantum processing unit using the optimized V2X frame structure for quantum principal component analysis. The optimized V2X frame structure has a protocol header of 24 bits (message type + GPS timestamp) + 4 bits QSI; The optimized V2X frame structure has a data field of 128 bits (vehicle position / speed) + 32 bits annealing parameter vector; The optimized V2X frame structure has a check field of 16-bit Cyclic Redundancy Check (CRC) + 24-bit Specific Control Code (STCC).

8. The vehicle-road cooperative dynamic decision-making device based on a quantum-classical hybrid architecture according to claim 1, characterized in that, The problem-solving module includes: The 64-dimensional data is projected onto the feature space using quantum Fourier transform to obtain 64-dimensional data with reduced computational complexity. Quantum annealing is used to solve the Ising model to solve the combinatorial optimization problem of the 64-dimensional data with reduced computational complexity, so as to obtain a set of candidate solutions; The approximate optimal solution set is determined from the candidate solution set using the Pareto front screening method.

9. A vehicle, characterized in that, include: The memory, the processor, and the computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the vehicle-road cooperative dynamic decision-making method based on a quantum-classical hybrid architecture as described in any one of claims 1-4.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, The program is executed by the processor to implement the vehicle-road cooperative dynamic decision-making method based on a quantum-classical hybrid architecture as described in any one of claims 1-4.

Images