A Global Collaborative Prediction Method for Traffic Signals Based on Quantum Entangled States

Abstract

This invention discloses a global collaborative prediction method for traffic signals based on quantum entanglement states. The method includes: establishing a quantumized model of the dynamic evolution path of the target traffic network within a future time window; constructing a spatiotemporal graph model; mapping the signal phase states of intersections to spatiotemporally integrated quantum states using a hierarchical quantum encoding scheme; constructing a Hamiltonian containing spatial coupling, temporal evolution, and cost terms based on a preset traffic optimization objective; solving for its ground state using a hybrid quantum-classical computation method to determine the optimal dynamic evolution path; encoding the current actual state of the traffic network as an initial state; decoding the collaborative signal control strategies for future time points using the optimal evolution operator and sequential measurements; and executing these strategies through a rolling time-domain control framework. This invention utilizes quantum parallelism for global optimization, enabling the acquisition of the optimal collaborative strategy across the entire spatiotemporal range in a single step, thereby improving the overall operational efficiency, predictability, and robustness of the traffic network.

Description

Technical Field

[0001] This invention relates to the field of intelligent transportation technology, and in particular to a method for global collaborative prediction of traffic signals based on quantum entangled states. Background Technology

[0002] Urban traffic signal control technology is the core of modern urban traffic management. Early traffic signal control primarily relied on fixed-time control, operating through preset fixed cycles and phase durations. While simple and reliable, this approach couldn't adapt to the dynamic fluctuations of traffic flow. To overcome this drawback, single-point sensor control emerged, using detectors to sense vehicle arrivals and dynamically adjust green light times, improving the local adaptability of intersections. However, with the increasing prominence of urban traffic networks, the limitations of single-point control became apparent, leading to the development of regional coordinated control technologies. Adaptive coordinated systems, such as SCOOT (Split CycleOffset Optimizing Technique) and SCATS (Sydney Coordinated Adaptive Traffic System), divide intersections within a region into sub-zones, achieving coordinated optimization of signal timing within each sub-zone, aiming to create "green waves" and reduce vehicle stops and delays. Furthermore, artificial intelligence and big data technologies have injected new vitality into traffic signal control. In particular, Multi-Agent Reinforcement Learning (MARL) methods can model each intersection as an agent, autonomously exploring the optimal control strategy through continuous interaction with the traffic environment and trial-and-error learning. These methods demonstrate great potential in handling complex, nonlinear, and high-dimensional traffic scenarios, enabling more refined and intelligent adaptive control, and have become a cutting-edge and hot research topic.

[0003] Despite significant advancements in existing traffic signal control technologies, their underlying paradigms still face insurmountable bottlenecks. These bottlenecks fundamentally stem from the inherent limitations of classical computing frameworks in handling large-scale, strongly coupled, and long-term optimization problems. First, whether it's traditional regional coordination systems or advanced multi-agent reinforcement learning, they are essentially "rolling" or "iterative" optimizations, typically making decisions within a relatively short future time window. This makes it difficult to truly capture and optimize the spatiotemporal global evolution of the entire traffic network over a longer period (e.g., 15-30 minutes). Furthermore, the decision-making process often involves "optimizing spatial coordination first, then moving to the next time point," or individual agents making decentralized decisions based on local observations. This results in a "global optimum" that is likely a greedy combination of local optima, easily falling into suboptimal solutions and failing to fundamentally predict and avoid potential future systemic congestion. Second, for the traffic network of a medium-to-large-scale city, the state space of signal phase combinations is astronomical. As the considered time window lengthens, the state space grows exponentially. Classical algorithms, including those based on heuristic search or reinforcement learning, either fail to provide solutions within practical timeframes due to computational resource exhaustion when faced with such massive discrete combinatorial optimization problems, or are forced to drastically simplify the model (e.g., reducing phases, ignoring secondary flow directions), sacrificing model fidelity and optimality of the solution. Finally, the end-to-end "black box" nature of the model makes its decision-making logic difficult to interpret, and debugging and optimization heavily rely on experience. While the model can learn effective policies, it struggles to embed complex traffic engineering constraints (such as minimum green light time, phase switching logic, and multi-objective tradeoffs) into its structure. This process often requires indirect guidance through complex reward functions, leading to unstable model performance and a lack of robustness. Summary of the Invention

[0004] The purpose of this section is to outline some aspects of embodiments of the present invention and to briefly describe some preferred embodiments. Simplifications or omissions may be made in this section, as well as in the abstract and title of this application, to avoid obscuring the purpose of these documents; however, such simplifications or omissions should not be construed as limiting the scope of the invention.

[0005] In view of the aforementioned existing problems, this invention is proposed. Therefore, this invention provides a global collaborative prediction method for traffic signals based on quantum entangled states to solve the problems mentioned in the background art.

[0006] To address the aforementioned technical problems, this invention provides the following technical solution: a global collaborative prediction method for traffic signals based on quantum entangled states, comprising:

[0007] Establish a quantum model of the dynamic evolution path of the target transportation network within a future time window;

[0008] Based on the preset traffic optimization objective, the quantized model is solved using quantum computing methods to determine the optimal dynamic evolution path for the time window;

[0009] From the optimal dynamic evolution path, the coordinated signal control strategy of the traffic network at each time point within the time window is decoded, thereby forming a global coordinated prediction covering the entire future time window.

[0010] As a preferred embodiment of the traffic signal global collaborative prediction method based on quantum entanglement states described in this invention, the method includes: establishing a quantized model, comprising:

[0011] Construct a spatiotemporal graph model that includes intersection nodes, road edges, and discretized time steps;

[0012] The signal phase state of each intersection in the spatiotemporal graph model at each discrete time step is mapped to a qubit, thereby constructing a spatiotemporal integrated quantum state space that represents all possible evolution paths.

[0013] As a preferred embodiment of the global collaborative prediction method for traffic signals based on quantum entanglement states described in this invention, wherein mapping the signal phase state to qubits includes:

[0014] A hierarchical quantum coding scheme is adopted, in which the first set of qubits is used to encode the phase of the ordinary signal, and at least one additional flag qubit is used to encode whether there is a special priority phase.

[0015] As a preferred embodiment of the traffic signal global collaborative prediction method based on quantum entanglement states described in this invention, wherein: the solution obtained through quantum computing methods includes:

[0016] Construct a Hamiltonian describing the total energy of the quantized model, the design of which is based on the preset traffic optimization objective;

[0017] Solve for the unitary evolution operator of the quantized model as it evolves from the initial state to the path with the lowest total energy.

[0018] As a preferred embodiment of the traffic signal global collaborative prediction method based on quantum entanglement states described in this invention, the Hamiltonian comprises at least one or more of the following combinations:

[0019] Spatial coupling terms represent the collaborative relationships between intersections within the same time step;

[0020] The time evolution term represents the state transition cost of the same intersection between adjacent time steps;

[0021] The cost term quantifies the traffic optimization objective as an energy penalty or reward.

[0022] As a preferred embodiment of the traffic signal global collaborative prediction method based on quantum entanglement states described in this invention, the method employs hybrid quantum-classical computation to solve the unitary evolution operator, including:

[0023] Design a parameterized quantum circuit with adjustable parameters based on Trotter-Suzuki decomposition to approximate the unitary evolution operator;

[0024] In a hybrid quantum-classical computing loop, the adjustable parameters are iteratively updated using the quantum natural gradient descent algorithm until convergence.

[0025] As a preferred embodiment of the traffic signal global cooperative prediction method based on quantum entanglement states described in this invention, the decoding cooperative signal control strategy includes:

[0026] The current actual state of the transportation network is encoded as the initial quantum state of the quantized model;

[0027] The quantum operation corresponding to the optimal dynamic evolution path is applied to the initial quantum state to obtain the final quantum state;

[0028] The final quantum state is sequentially measured, and the optimal signal control strategy for each intersection is analyzed step by step.

[0029] As a preferred embodiment of the global collaborative prediction method for traffic signals based on quantum entangled states described in this invention, the decoding step further includes:

[0030] Repeat the evolution and sequential measurement process to generate a candidate set of optimal evolution paths;

[0031] The paths in the candidate set are evaluated using a classic traffic simulation model, and the path with the best overall performance is selected as the final execution plan.

[0032] As a preferred embodiment of the global collaborative prediction method for traffic signals based on quantum entangled states described in this invention, the method is executed through a rolling time-domain control framework, that is, only the strategy for the first part of the time period in the collaborative signal control strategy is executed, and during the execution, the optimal solution for the next future time window is calculated using updated traffic state data.

[0033] Compared with existing technologies, the beneficial effects of the invention are as follows:

[0034] 1. This invention constructs a four-dimensional spatiotemporal quantum state space covering the entire target traffic network and a future time window, mapping all possible states of all intersections at all future time points to a unified quantum model at once. This breaks away from the limitations of traditional methods such as "rolling optimization" or "local coordination", and can predict and avoid long-term congestion caused by short-sighted decisions, thereby improving the operational efficiency of the entire traffic network.

[0035] 2. By designing the Hamiltonian and solving its ground state, this invention utilizes the parallelism of quantum evolution to efficiently explore the solution space, thereby enabling the discovery of high-quality collaborative control schemes for large-scale, high-dimensional urban traffic networks within an acceptable timeframe. This makes it possible to accurately optimize traffic systems over a wider range and longer periods, thus expanding the application boundaries of traffic signal control technology.

[0036] 3. The combinatorial construction method of Hamiltonian not only makes the optimization objective clear and interpretable, but also embeds complex real-world constraints into the physical laws of the model, ensuring the effectiveness and robustness of the solution. At the same time, when new optimization objectives (such as carbon emissions or public transport priority) need to be introduced, only the corresponding operator terms need to be added to Hamiltonian, thus making the model highly scalable. Attached Figure Description

[0037] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0038] Figure 1 This is a flowchart illustrating the overall process of a global collaborative prediction method for traffic signals based on quantum entangled states, as described in one embodiment of the present invention.

[0039] Figure 2 This is a flowchart of a hybrid quantum-classical computation loop for a traffic signal global collaborative prediction method based on quantum entangled states, as described in one embodiment of the present invention. Detailed Implementation

[0040] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0041] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0042] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0043] This invention is described in detail with reference to the schematic diagrams. When detailing the embodiments of this invention, for ease of explanation, the cross-sectional views illustrating the device structure may be partially enlarged, not adhering to the usual scale. Furthermore, the schematic diagrams are merely examples and should not be construed as limiting the scope of protection of this invention. In actual fabrication, the three-dimensional spatial dimensions of length, width, and depth should be included.

[0044] Furthermore, in the description of this invention, it should be noted that the terms "upper," "lower," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. These terms are used solely for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. In addition, the terms "first," "second," or "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0045] Unless otherwise explicitly specified and limited, the terms "installation," "connection," and "joining" in this invention should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; similarly, they can refer to mechanical connections, electrical connections, or direct connections, or indirect connections through an intermediate medium, or internal connections between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0046] Example 1

[0047] Reference Figure 1 and Figure 2 This is the first embodiment of the present invention, which provides a global collaborative prediction method for traffic signals based on quantum entangled states, including:

[0048] S1. Establish a quantum model of the dynamic evolution path of the target transportation network within a future time window;

[0049] It needs to be explained that by establishing a quantum model, the complex, physical problem of urban traffic networks can be transformed into a structured mathematical-physical model that can be modeled and solved on a quantum computer. This transformation process includes two sub-steps: constructing a spacetime graph model and quantum mapping.

[0050] Furthermore, a target urban traffic network area containing N intersections is selected. The total length of the prediction time window is set to T, and it is discretized into M time steps of equal length, each time step being denoted as Δt (e.g., Δt = 5 seconds). A four-dimensional spatiotemporal graph model G = (V, E, M) is constructed, where V is the set of nodes representing the state of all intersections at all discrete time steps, and a node v... i,t Let E represent the state of the i-th intersection (i∈[1,N]) at time step t (t∈[1,M]); E is the set of edges, which contains two types of edges: the first type is spatial edges, which connect two intersection nodes v that have a direct physical connection (adjacent) within the same time step t. i,t and v j,t The first type describes the spatial coordination between intersections (such as green waves and congestion diffusion), while the second type is the time edge, which connects nodes v of the same intersection i in adjacent time steps. i,t and v i,t+1 , used to describe the evolution and transition cost of the signal state over time;

[0051] Specifically, before constructing this four-dimensional spatiotemporal graph model, it is necessary to acquire and process real-time and predictive traffic data, including but not limited to: the topology of the traffic network (intersection locations, connectivity, road segment lengths), real-time traffic flow in each direction at each intersection, queue lengths, and arrival traffic flow within future time windows predicted based on historical data.

[0052] Furthermore, based on this spatiotemporal graph model, each node v i,t The classical state (i.e., signal phase) is mapped onto the quantum bit;

[0053] Specifically, for each intersection i, at any time step t, its signal phase set is P. i ={P1,P2,…,P K}, where K is the total number of phases at the intersection. In addition, in order to encode these K phases, at least ceil(log2K) qubits are needed, where ceil(·) represents the floor function.

[0054] It should be noted that the most common traditional method is flat encoding, which mixes all possible phases, whether regular or special, together and represents them using a binary sequence. For example, an intersection with 4 regular phases, 1 bus priority phase, and 1 emergency vehicle priority phase has a total of 6 states, requiring ceil(log 2 6) = 3 qubits are used for encoding; if the city manager decides to add a special "pedestrian-only" phase to the intersection, the total number of states will become 7, the original 3 qubits will not be enough for encoding, and the entire encoding table will have to be redesigned. All Hamiltonian terms that depend on the specific encoding must be scrapped and rebuilt, which cannot adapt to the ever-changing needs in the real world.

[0055] Preferably, the present invention employs a hierarchical quantum coding scheme, which is divided into conventional phase coding and special priority phase coding;

[0056] Specifically, conventional phase encoding: Assuming an intersection has 4 conventional signal phases (such as east-west straight, east-west left turn, north-south straight, and north-south left turn), then 2 qubits (e.g., |00>, |01>, |10>, |11>) are used to correspond to these 4 phases respectively;

[0057] Specifically, special priority phase encoding: In order to handle special cases, such as bus priority, emergency vehicle priority or pedestrian-only phase, an additional flag bit is introduced. When the flag bit is in the |0> state, it indicates that the current phase is the regular phase, and its specific phase is determined by the aforementioned regular phase bit. When the flag bit is activated to the |1> state, it indicates that the current phase is the special priority phase.

[0058] It should be noted that if multiple special phases exist, even more flag qubits can be used for encoding;

[0059] It should be noted that by employing a hierarchical quantum coding scheme, the state of an intersection i at time step t can be represented by a set of qubits |ψ> i,t To describe;

[0060] Specifically, the complete state of the entire transportation network within a future time window, i.e., the integrated spacetime quantum state space, is obtained by tensor product of the quantum states of all N intersections at all M time steps;

[0061] Specifically, the basis vectors of this space It is represented as a tensor product; the total number of quantum bits in this basis vector is equal to N×M× (the number of quantum bits required for each node). Each basis vector in this quantum state space corresponds to a complete evolution path of the entire traffic network within a future time window (i.e., the specific traffic light state of each intersection at each moment). Thus, the traffic scheduling problem in the physical world has been successfully transformed into a quantum many-body problem defined in Hilbert space.

[0062] S2. Based on the preset traffic optimization objectives, the quantized model is solved using quantum computing methods to determine the optimal dynamic evolution path within the time window;

[0063] It should be explained that the goal of this step is to combine the quantized model obtained in step S1 with the real-world traffic optimization objective, and to find the unique optimal solution from the model by leveraging the powerful capabilities of quantum computing.

[0064] Specifically, in quantum physics, the Hamiltonian operator represents the total energy of a system. An isolated quantum system will naturally evolve towards its lowest energy state (i.e., the ground state). This invention utilizes this physical principle to design a Hamiltonian so that the "quality" of a traffic evolution path is directly mapped to the "energy level" of its corresponding quantum state. If a traffic evolution path is more in line with traffic optimization goals (such as less congestion and higher traffic efficiency), then the energy of its corresponding quantum state will be lower. Therefore, solving for the ground state of the Hamiltonian is equivalent to finding the optimal traffic cooperative control strategy.

[0065] Furthermore, the total Hamiltonian is a weighted sum of multiple sub-terms with definite physical meaning;

[0066] Preferably, the formula for the total Hamiltonian can be expressed as:

[0067] H = ω c ×H cost +ω s ×H space +ω t ×H time

[0068] Where, ω c ω s ω t These represent the weighting coefficients of the cost term, spatial coupling term, and temporal evolution term, respectively, used to balance the importance of each term to the optimization objective. ω c +ω s +ω t =1; H cost Represented as a cost term, it is used to directly convert traffic optimization objectives into energy penalties; H space Represented as a spatial coupling term, used to encode the collaborative relationships between intersections; H timeIt is represented as a time evolution term, used to impose temporal logic constraints in traffic engineering to ensure the practical feasibility of the solution;

[0069] Furthermore, the cost term uses predictive traffic flow data based on a quantumized model. First, using classical traffic flow models (such as machine learning models based on historical data), it predicts the number of vehicles arriving at each approach lane of intersection i at time step t. For a specific phase p to be evaluated, it calculates the expected queue length Q(i,t,p) and delay D(i,t,p) if that phase is executed. For example, the queue length can be determined by the queue length of the previous time step, the number of arriving vehicles at the current time step, and the saturation flow rate (i.e., capacity) of phase p at the current time step. Then, the cost term is further designed to impose energy penalties on these indicators.

[0070] H cost =∑ i,t,p [Q(i,t,p)×Z i,t,p ] or H cost =∑ i,t,p [D(i,t,p)×Z i,t,p ]

[0071] Among them, Z i,t,p It is the Pauli-Z matrix acting on the qubits encoding phase p;

[0072] Furthermore, regarding spatial coupling terms, taking the implementation of a "green wave" as an example, for adjacent intersections i and j, if intersections i and j can form a green wave by taking phases p1 and p2 respectively at the same time step t (such as simultaneous straight-through traffic on the main road), then the energy in this situation should be reduced. This can be designed as follows:

[0073]

[0074] Among them, J s For the coupling strength, J s The larger the value, the more the system tends to sacrifice other objectives (such as delays at individual intersections) to meet spatial coordination requirements. This value is an adjustable hyperparameter that can be set according to specific traffic strategies (e.g., on urban arterial roads, where green wave coordination has a very high priority, a larger J value can be set). s Configure it;<i,j> This means traversing all adjacent intersection pairs (i, j). Here, "adjacent" usually refers to two intersections that are directly connected by a road in the traffic network topology. The traversal range defines which intersections need to be considered for spatial coordination. and Represented as the Pauli-Z matrix acting on the qubits encoding phases p1 and p2, in the form of the basis vectors {|0>, |1>}, this matrix is: Its function is as follows: when applied to a qubit, if the bit is in the |0> state, the result remains unchanged (multiplied by +1); if the bit is in the |1> state, the result becomes -|1> (multiplied by -1).

[0075] It needs to be explained that in physics, the lower the total energy of a system, the more stable it is. By introducing negative coupling strength, a term that is mathematically expected to have a large value (when the measurement results of two Z operators are the same) is transformed into an energy reward (negative energy contribution). Therefore, when the phases of adjacent intersections meet the coordination condition, this term will contribute a negative value, lowering the total energy.

[0076] Furthermore, for the time evolution term, the goal is to encode the temporal logic constraints in real-world traffic engineering into an energy term, ensuring that the decoded signal strategy is temporally coherent and executable. This term mainly includes the following two types of constraints:

[0077] Minimum Green Light Time Constraint: To prevent driver confusion and reduced traffic efficiency caused by frequent signal light switching, traffic rules require that once a phase is open, it must remain green for at least a minimum green light time (e.g., 10 seconds, corresponding to two Δt steps). A penalty is designed so that if a phase p is green at time t-1 but becomes non-green at time t, an energy penalty is applied. This penalty can be designed as follows:

[0078]

[0079] Where C is a very large positive constant, ensuring that paths violating this rule have extremely high energy and are naturally excluded when solving for the ground state. The meaning of this formula is that the energy level of a path is determined if and only if the phase p(Z) is at time t-1. i,t-1,p The expected value is -1) and time t is not the phase (Z). i,t,p Only when the expected value is +1 does this item contribute a large positive energy penalty;

[0080] Phase switching logic constraints: Some phases cannot be switched directly and must go through a clearing phase with all red or yellow lights. For example, east-west straight traffic cannot be switched directly to north-south straight traffic. In this case, energy penalties need to be applied to all disallowed phase transition pairs (p1→p2).

[0081]

[0082] Where D is the large penalty coefficient, P i,t,p This is a projection operator. Its value is 1 if and only if the system is in the state of intersection i, time t, and phase p, otherwise it is 0.

[0083] Furthermore, the time evolution term can be represented as:

[0084] H time =H time_min_green +H time_transition

[0085] Furthermore, to further understand this formula, the working principle and steps of this formula are explained as follows:

[0086] For each future time point t and each pair of adjacent intersections (i, j), iterate through them.

[0087] For each pair of intersections (i, j), define the phase combination (p1, p2) that can form a "green wave";

[0088] pass Check whether the system is simultaneously in the phase combination (p1,p2) at (i,t) and (j,t);

[0089] If the system is in this phase combination (p1, p2) (e.g., both corresponding to the encoding |1>), then The expected value of the measurement is (-1)×(-1)=+1, and after -J s This factor contributes -J to the total energy. s This is represented as an energy reward; if the system is in a non-cooperative state (e.g., one is |1>, the other is |0>), then... The expected value of the measurement is (-1)×(+1)=-1, and after -J s This term contributes +J to the total energy. s This represents an energy penalty;

[0090] Furthermore, hybrid quantum-classical computation is used to solve the unitary evolution operator;

[0091] It should be noted that since it is extremely difficult to directly calculate the unitary evolution operator that can evolve the system to the ground state, this invention adopts a variable quantum algorithm that is at the forefront of the current quantum computing field.

[0092] Specifically, a quantum circuit is constructed, the structure of which consists of fixed quantum gates and adjustable parameters (such as rotation angles);

[0093] Preferably, the circuit structure is designed based on Trotter-Suzuki decomposition, which provides a way to approximate the complex Hamiltonian into simple quantum gate operations. Theoretically, it guarantees that as long as the decomposition order is high enough, the quantum circuit can infinitely approximate the ideal evolution operator.

[0094] Furthermore, perform a hybrid quantum-classical computation loop, referencing Figure 2 :

[0095] S201. On a quantum computer, set an initial state (e.g., all qubits are in a superposition state), run a quantum circuit U(θ) with adjustable parameter θ, and measure the expected energy of the output state. Where U(θ)|ψ> represents the evolved quantum state, which is equivalent to a complete traffic signal coordination scheme to be evaluated, covering the entire traffic network and future time windows. U represents the unitary operator, and changing θ is equivalent to changing the scheme. H(U(θ)|ψ>) represents the total Hamiltonian applied to the evolved quantum state. This is represented as Hermitian conjugation, which is the inverse operation of quantum circuits;

[0096] S202. The measured expected energy value and adjustable parameters are transmitted to a classical computer.

[0097] S203. The classical computer executes an optimization algorithm to calculate how θ should be adjusted in the next step to achieve the desired energy value. <e>It dropped the fastest;

[0098] It should be noted that in the preferred embodiment of the present invention, the quantum natural gradient descent (QNGD) algorithm is adopted; because compared with the traditional gradient descent, the QNGD algorithm can better handle the complex geometric structure of the quantum parameter space, avoid getting trapped in local optima, and thus converge to the global energy minimum point at a faster speed and more stably.

[0099] S204. Return the updated θ from the optimization algorithm to the quantum computer, and repeat S201 to S203 until the expected energy value no longer decreases, reaching convergence.

[0100] S3. From the optimal dynamic evolution path, decode the coordinated signal control strategy of the traffic network at each time point within the time window, thereby forming a global coordinated prediction covering the entire future time window.

[0101] It should be noted that step S2 generates an optimal unitary evolution operator, i.e., U opt =U(θ) opt ), where θ opt Represented as the optimal adjustment parameters, the task of this step is to translate this abstract physical law into a real-world traffic controller that can understand and execute signal light switching instructions down to the second.

[0102] Furthermore, the current actual state of the transportation network is encoded as the initial quantum state of the quantized model;

[0103] Specifically, the traffic network status at the current moment is obtained through a real-time sensor network (such as inductive loops, video detectors, floating car data, etc.), including: the signal phase being executed at each intersection, and information such as queue length and traffic flow at major intersections; according to the hierarchical quantum coding scheme established in step S1, this information is converted into a definite quantum product state; for example, if intersection 1 is currently "north-south straight" (encoded as |01>) and intersection 2 is "east-west straight" (encoded as |00>), then the initial state is...

[0104] Furthermore, the quantum operation corresponding to the optimal dynamic evolution path (i.e., U) opt () Acting on the initial quantum state, the final quantum state is obtained;

[0105] Specifically, on a quantum computer, the optimal adjustment parameter θ has been fixed. opt Quantum circuits, and |ψ initial As input, the output is |ψ final >, where the |ψ final > is the final state, representing a complex superposition state that pervades the entire Hilbert space. However, unlike the initial uniform superposition state, the probability amplitude distribution of this final state is non-uniform, meaning that the ground state that conforms to the optimal evolution path (representing a specific and complete spatiotemporal evolution path) has the highest probability amplitude.

[0106] Furthermore, sequential measurements are performed on the final quantum state to resolve the optimal signal control strategy for each intersection step by step.

[0107] It should be noted that if we directly apply |ψ final If all the qubits are measured at once, then according to the collapse principle of quantum mechanics, the entire qubit state will randomly collapse to a ground state, obtaining a complete path but losing control and insight into the time evolution process. Therefore, sequential measurement is used to ensure that the decoded strategy has an inherent time causal logic by mimicking the unidirectional flow of time.

[0108] Specifically, only for |ψ final The measurement of all qubits representing the first future time step (t=1) causes this portion of the quantum state to collapse, giving a definite signal phase scheme, i.e. At this moment, the entire |ψ final The qubit has partially collapsed, and the probability distribution of the remaining unmeasured portion (t = 2, 3, ..., M) will be conditionally updated based on the measurement result at t = 1; subsequently, the qubit representing the second future time step (t = 2) will be measured to obtain the scheme. Repeat the above process, measuring step by step at t=3, t=4, ... and so on, until the last time step M, to obtain the cooperative control strategy sequence:

[0109] S=({p i (t=1)},{p i (t=2)},…,{p i (t=M)})

[0110] It should be noted that, in order to overcome the randomness of quantum measurement and to conduct a final "real-world test" of the quantum scheme using classical methods, the selected scheme is not only theoretically excellent but also more suitable for practical applications;

[0111] Furthermore, a candidate set of optimal evolutionary paths is generated;

[0112] Specifically, the optimal dynamic evolution path and sequential measurement process described above are repeated to generate a candidate set {S1, S2, ..., S...} containing K high-quality paths. K };

[0113] Furthermore, classic traffic simulation models (such as SUMO, VISSIM, etc.) are used to evaluate each path in the candidate set, and the path with the best overall performance is selected as the final execution plan.

[0114] Specifically, each path in the candidate set is taken as input and run in a pre-built micro-simulation environment that is completely consistent with the real road network. The simulator will output extremely detailed performance indicators, such as the average delay of the entire network, the total number of stops, the green wave width of the trunk line, the number of intersection overflows, carbon emissions, etc. According to the preset comprehensive evaluation function (for example, weighted summation of delay and number of stops), all paths are scored, and the path with the highest score is taken as the final globally optimal collaborative strategy to be adopted and deployed for execution.

[0115] Preferably, in practical applications, it is implemented through the Rolling Horizon Control (RHC) framework to build a closed-loop feedback control system that can continuously adapt to changes in traffic dynamics.

[0116] Specifically, only the first part of the global optimal coordination strategy is extracted, for example, the control instructions for the first 2 minutes (24 time steps), and sent to the urban traffic signal control system for execution. During the subsequent 2-minute execution period, the data acquisition module in the system background continuously collects the latest traffic status data of the entire network (real-time flow, queue length, etc.). When the 2-minute execution cycle is about to end, the latest traffic status at this moment is used as the new initial condition, and the complete calculation process from step S1 to step S3 is restarted to solve for a brand new optimal coordination strategy for the next 15 minutes from the current moment. The first 2 minutes of the optimal coordination strategy of this new strategy are executed, and the process is updated and calculated again.

[0117] It should be noted that, through the rolling cycle of "calculation-execution-recalculation", the present invention can continuously use the latest real-world data to revise its long-term predictions and plans, thereby effectively responding to random fluctuations in traffic flow and emergencies (such as traffic accidents), and ensuring that the control strategy always maintains a high degree of adaptability and robustness to the real world.

[0118] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of this application can be implemented using various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0119] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0120] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0121] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0122] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0123] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.< / e>

Claims

1. A global collaborative prediction method for traffic signals based on quantum entangled states, characterized in that, include: Establish a quantum model of the dynamic evolution path of the target transportation network within a future time window; Establishing the quantumization model includes: Construct a spatiotemporal graph model that includes intersection nodes, road edges, and discretized time steps; The signal phase state of each intersection in the spatiotemporal graph model at each discrete time step is mapped to a qubit, thus constructing a spatiotemporal integrated quantum state space that represents all possible evolution paths. Based on the preset traffic optimization objective, the quantized model is solved using quantum computing methods to determine the optimal dynamic evolution path for the time window; Solving using the quantum computing method includes: Construct a Hamiltonian describing the total energy of the quantized model, the design of which is based on the preset traffic optimization objective; Solve for the unitary evolution operator of the quantized model as it evolves from the initial state to the path with the lowest total energy; From the optimal dynamic evolution path, the coordinated signal control strategy of the target traffic network at each time point within the time window is decoded, thereby forming a global coordinated prediction covering the entire future time window; From the optimal dynamic evolution path, the coordinated signal control strategy of the target traffic network at each time point within the time window is decoded, including: The current actual state of the target traffic network is encoded as the initial quantum state of the quantized model; The quantum operation corresponding to the optimal dynamic evolution path is applied to the initial quantum state to obtain the final quantum state; The final quantum state is sequentially measured, and the optimal signal control strategy for each intersection is analyzed step by step.

2. The traffic signal global collaborative prediction method based on quantum entangled states as described in claim 1, characterized in that, Mapping the signal phase state to a quantum bit includes: A hierarchical quantum coding scheme is adopted, in which the first set of qubits is used to encode the phase of the ordinary signal, and at least one additional flag qubit is used to encode whether there is a special priority phase.

3. The traffic signal global collaborative prediction method based on quantum entangled states as described in claim 1, characterized in that, The Hamiltonian includes at least one or more of the following combinations: Spatial coupling terms represent the collaborative relationships between intersections within the same time step; The time evolution term represents the state transition cost of the same intersection between adjacent time steps; The cost term quantifies the traffic optimization objective as an energy penalty or reward.

4. The traffic signal global collaborative prediction method based on quantum entangled states as described in claim 1, characterized in that, Solving the unitary evolution operator using hybrid quantum-classical computation includes: Design a parameterized quantum circuit with adjustable parameters based on Trotter-Suzuki decomposition to approximate the unitary evolution operator; In a hybrid quantum-classical computing loop, the adjustable parameters are iteratively updated using the quantum natural gradient descent algorithm until convergence.

5. The traffic signal global collaborative prediction method based on quantum entangled states as described in claim 1, characterized in that, The decoding also includes: Repeat the process of determining the optimal dynamic evolution path and the sequential measurement to generate a candidate set of optimal evolution paths; The paths in the candidate set are evaluated using a classic traffic simulation model, and the path with the best overall performance is selected as the final execution plan.

6. The traffic signal global collaborative prediction method based on quantum entangled states as described in claim 1, characterized in that, The quantum entangled state-based global collaborative prediction method for traffic signals is executed through a rolling time-domain control framework. That is, it only executes the strategy for the first part of the time period in the collaborative signal control strategy, and during the execution, it uses updated traffic state data to continuously calculate the optimal solution for the next future time window.

Images