A traffic flow modeling method based on waveguide photon wandering

Abstract

The present invention discloses a traffic flow modeling method based on waveguide photon walks, comprising: dividing the traffic network into multiple sub-areas, extracting traffic flow data based on the sub-areas; generating driver dynamic probability patterns based on waveguide photon walks; screening the driver dynamic probability patterns for significant probability patterns to obtain significant probability patterns in each sub-area; realizing the mapping transformation between traffic flow data and driver significant probability patterns; mining the random features of the traffic network, defining the weighted average of the noise amplitude of the significant probability patterns as the driver randomness index; and defining the weighted average of the sum of the coupling strengths of the significant probability patterns as the traffic correlation index. The present invention innovatively applies the waveguide-based quantum walk system to the twin modeling of urban driver dynamics, effectively integrating driver randomness and traffic correlation in the traffic network, and providing a more accurate traffic volume modeling method.

Description

Technical Field

[0001] The present invention belongs to the fields of traffic geography and quantum theory, relates to the prediction and analysis of traffic flow, and specifically relates to a traffic flow modeling method based on waveguide light quantum wandering. Background Art

[0002] Traffic flow is a crucial indicator for analyzing and predicting urban traffic conditions. Traffic flow modeling, a key area of ​​intelligent transportation systems (ITS), provides theoretical support for applications such as traffic congestion analysis and route planning. However, as a complex system, the spatiotemporal evolution of traffic flow in urban transportation networks exhibits complex nonlinear characteristics, making it difficult to accurately model. This complexity stems from two random factors in transportation networks:

[0003] Driver randomness: Individual drivers have multiple travel routes to choose from. Due to the heterogeneity of individual behavioral preferences and the impossibility of capturing all individual micro-decisions, individual drivers' travel trajectories within the traffic network are random. Furthermore, we can assume that individual drivers appear at any traffic node with varying probabilities, thus forming a time-varying probability distribution within the traffic network. Furthermore, unexpected random events such as weather, traffic accidents, and congestion can lead to heterogeneous driver responses. This can cause drivers to deviate from their intended trajectories, resulting in random fluctuations in traffic flow evolution.

[0004] Traffic correlation: Traffic flow evolution patterns in different functional areas are correlated. Within a specific time window, traffic flows in connected areas exhibit a tendency to coevolve. For example, congestion at an intersection can affect upstream and downstream traffic conditions. Traffic correlation makes traffic flow changes in the transportation network holistic and correlated. Different functional areas exhibit significant spatial heterogeneity and temporal non-stationarity. Traffic correlation is difficult to characterize due to its spatial heterogeneity and temporal non-stationarity.

[0005] Taking into account driver randomness and traffic correlation in traffic flow modeling is crucial for improving traffic flow simulation accuracy and revealing the underlying patterns of traffic flow variations within traffic networks. Current traffic flow models can be categorized into macro and micro models based on their scale. Macro models, such as the Poisson model, regression model, and CTM model, account for driver randomness by replacing model parameters with random variables. However, macro models fail to consider the impact of different scenarios and infrastructure on individual driver behavior, resulting in limitations in simulating traffic flow at fine spatiotemporal resolution. Micro models, such as social force models and intelligent agent models, fully emphasize the influence of the external environment on individual driver behavior. However, these micro models, based on deterministic theories, assume that driver behavior is regular, describable, and predictable. However, driver trajectories in real traffic networks often do not meet these assumptions. In short, current traffic flow models fail to simultaneously account for driver randomness and traffic correlation, resulting in low traffic flow simulation accuracy. Summary of the Invention

[0006] Purpose of the Invention: To address the problem that existing traffic flow modeling methods cannot take into account both driver randomness and traffic correlation, a traffic flow modeling method based on optical waveguide quantum walks is proposed. The effect of noise and waveguide coupling on photon trajectories is analogized to the effect of individual randomness and traffic correlation on driver routes in urban traffic systems. This can achieve high-precision simulation of traffic flow and quantitative analysis of the random characteristics of driver randomness and traffic correlation in traffic networks.

[0007] Technical solution: To achieve the above objectives, the present invention provides a traffic flow modeling method based on waveguide light quantum wandering, comprising the following steps:

[0008] S1: Divide the traffic network into multiple sub-areas and extract traffic flow data based on the sub-areas;

[0009] S2: Generate driver dynamic probability pattern based on waveguide light quantum wandering;

[0010] S3: Screening the driver's dynamic probability pattern for significant probability patterns to obtain significant probability patterns in each sub-region;

[0011] S4: Achieve mapping transformation between traffic flow data and driver’s significant probability patterns;

[0012] S5: Conduct random feature mining of transportation networks;

[0013] The weighted average of the noise amplitude of the significant probability pattern is defined as the driver randomness index to measure the degree of random fluctuation of traffic volume caused by the randomness of drivers.

[0014] The weighted average of the sum of the coupling intensities of significant probability patterns is defined as the traffic correlation index to measure the importance of a region in the changes in traffic network flow.

[0015] Furthermore, traffic flow exhibits a continuous distribution in time and space, making direct study difficult. Therefore, to better explore the spatiotemporal evolution and regional differences in traffic flow, step S1 of the present invention employs a spatial partitioning approach, discretizing traffic flow at a specific longitude, latitude, and temporal resolution to form a grid-like traffic network. Within this network, the extracted traffic flow data represents a time-organized sequence of traffic flows distributed across each subregion of the network, with each data point representing the number of vehicles in that subregion within a specific time window. This constitutes the data foundation.

[0016] Furthermore, due to the randomness of the drivers, the positions of the drivers in the traffic network take the form of dynamic probability distribution. In theory, dynamic probability patterns can be generated by classifying the probability distribution of individual drivers, and then the traffic volume can be mapped as a superposition of different probability patterns to achieve traffic volume simulation. Under the constraints of the road network topology, the waveguide-based optical quantum wandering system can generate dynamic probability distributions. During the random wandering of photons, the introduction of noise causes the photons to deviate from the expected trajectory, while the waveguide coupling effect causes the probability of photons appearing on the coupled waveguides to be dynamically correlated. The influence of the above two physical parameters on the photon trajectory is similar to the influence of driver randomness and traffic correlation on the driver's trajectory in the traffic network;

[0017] In step S2, a waveguide array optical quantum walk consistent with the real traffic network topology is constructed, and the noise amplitude and coupling strength parameters are changed to generate driver dynamic probability patterns with different spatiotemporal structures in each sub-area.

[0018] Furthermore, the step S2 is specifically as follows:

[0019] The dynamic characteristics of quantum walk are controlled by Hamiltonian H. The Hamiltonian H of quantum walk of photons in a coupled array containing I waveguides is:

[0020]

[0021] Where Δβ is the noise amplitude, which is used to characterize the randomness intensity of different drivers; C ij is the coupling strength between waveguide i and waveguide j, which is used to characterize the traffic correlation strength between area i and area j in the corresponding traffic network; a i They are the annihilation operator and the generation operator respectively;

[0022] The state of a photon at time t is represented by a wave function:

[0023] ψ[(Δβ,C ij ),t]=e -iHt ψ(0) (2)

[0024] Among them, ψ[(Δβ,C ij ),t] is the wave function, which represents the state of the quantum system at time t; ψ(0) represents the state of the photon at the initial time; based on the wave function, the probability of a photon being in any waveguide at any time can be solved:

[0025] M i [(Δβ,C ij ),t]=ψ i [(Δβ,C ij ),t] 2 (3)

[0026] Among them, M i [(Δβ,C ij ),t] is the driver probability pattern, which represents the probability that the driver appears in sub-area i at time t; construct the light quantum walk parameter set By performing quantum walks under this parameter set, a series of probability patterns can be generated. These probability patterns represent the dynamic characteristics of drivers under different driver randomness and traffic-related conditions, thereby reflecting the complex structural characteristics of traffic flow.

[0027] Furthermore, in step S2, the light quantum wandering wave function is solved by using the light quantum platform FeynmanPAQ, specifically: first, the waveguide array structure is set based on the real traffic network topology; then, the noise amplitude Δβ and coupling strength C are input in the parameter setting panel. ij After submitting the parameters, the light quantum walk begins and the driver probability model M can be obtained. i [(Δβ,C ij ),t].

[0028] Furthermore, traffic node subregions contain multiple driver probability patterns, resulting in a highly complex structure. Furthermore, ride-hailing traffic exhibits significant spatial heterogeneity, meaning that the probability patterns in different subregions are not identical. Therefore, removing insignificant probability patterns within each subregion is crucial for traffic flow simulation and analysis of random characteristics of traffic networks. Therefore, under the constraints of real-world traffic flow data, we identify significant probability patterns within each subregion based on specific driver probability pattern screening rules. This facilitates subsequent exploration of the superimposed coupling relationship between traffic flow and multiple probability patterns.

[0029] The step S3 is specifically as follows: using the stepwise regression subset screening method as the screening method for the significant driver probability pattern, the actual observation of the traffic flow time series V in the sub-area i iUnder the constraint of (t), all possible driver probability patterns {M i [(Δβ,C ij )1,t],M i [(Δβ,C ij )2,t],...,M i [(Δβ,C ij ) N ,t]} perform stepwise regression subset screening.

[0030] Furthermore, the screening process in step S3 is expressed as:

[0031]

[0032] Based on the Akaike Information Criterion (AIC), all possible driver probability patterns are screened using the stepwise regression subset screening method in formula (4) to obtain a significant probability pattern set, which is denoted as: {M i [(Δβ,C ij ) n ,t]|n∈A i}, where A i is the significant probability pattern index of region i.

[0033] Furthermore, in step S4, based on the significant probability patterns in each area screened out in step S3, that is, the dynamically changing probability distribution obtained by solving the wave function, a multivariate linear regression model is used to construct a transformation mechanism between the dynamic evolution of traffic flow and the significant probability patterns, thereby realizing the simulation and optimization of traffic flow.

[0034] Furthermore, the step S4 is specifically as follows:

[0035] Based on the significant probability pattern {M i [(Δβ,C ij ) n ,t]|n∈A i}, a mapping transformation relationship between the sub-region traffic flow time series and the driver's significant probability pattern is constructed and expressed as:

[0036]

[0037] in, is the simulated traffic volume; A i is the significant probability pattern index obtained by screening in step S3; a i (n) and ξ i is the mapping parameter, a i (n) is the regression coefficient from the probability model to the traffic flow, ξ i is the random error term; (Δβ,C ij) n is the quantum walk parameter of the significant probability mode, which is an important parameter to characterize the randomness characteristics in the transportation network;

[0038] Applying steps S3 and S4 to all areas in the transportation network yields:

[0039]

[0040] Furthermore, to quantitatively describe the random characteristics of traffic networks, the present invention constructs measurement indices based on the noise amplitude and coupling strength of significant probability patterns: the driver randomness index and the traffic correlation index. The driver randomness index is used to measure the degree of random fluctuations in traffic volume caused by driver randomness (individual heterogeneity and environmental changes). The traffic correlation index is used to measure the importance of a sub-region in the traffic volume fluctuations of the traffic network.

[0041] The step S5 is specifically as follows:

[0042] The weighted average of the noise amplitude of the significant probability pattern is defined as the driver randomness index β to measure the degree of random fluctuation of traffic volume caused by the randomness of the driver. The formula is as follows:

[0043]

[0044] Among them, β i is sub-region N i Driver randomness index, Δβ n The noise amplitude representing the nth significant probability mode, a i (n) is the mapping parameter of the significant probability pattern, β i The larger it is, the more severe the random fluctuation of traffic volume caused by the randomness of drivers is;

[0045] The weighted average of the sum of the coupling intensities of significant probability patterns is defined as the traffic correlation index TC to measure the importance of a region in the change of traffic network flow. The formula is as follows:

[0046]

[0047] Among them, C k is the sum of the coupling strengths of the kth significant probability modes, TC i The larger it is, the more important sub-region i is in the traffic network flow change.

[0048] The present invention adopts a controllable quantum walk system - waveguide-based optical quantum walk, and uses waveguide-based optical quantum walk to perform twin modeling of the dynamics of drivers in the traffic network. In this quantum system, photons perform quantum walks through a waveguide array. Waveguides are precision optical devices specially used to guide and constrain the propagation path of photons. During the quantum walk, the trajectory of photons in the waveguide array is affected by noise and waveguide coupling, thereby generating a specific dynamic probability distribution. The impact of noise and waveguide coupling on the trajectory of photons can be compared to the impact of individual randomness and traffic correlation on the driver's route in an urban traffic system. In the present invention, the impact of noise and waveguide coupling on the trajectory of photons is compared to the impact of individual randomness and traffic correlation on the driver's route in an urban traffic system. Under the constraints of the road network topology, the present invention constructs optical quantum walks under the constraints of noise and waveguide coupling on the waveguide array, generates a dynamic probability pattern of drivers, and further constructs a mapping relationship between the probability pattern and the spatiotemporal evolution of traffic flow, thereby realizing the simulation of traffic volume and the analysis of random characteristics of the traffic network.

[0049] Beneficial effects: Compared with the existing technology, the present invention uses the wandering trajectory of photons in the waveguide array to fit the probability pattern of drivers in the real traffic network, and takes into account the randomness of drivers and traffic correlation through the physical parameter noise amplitude and coupling strength in the quantum walk system. Furthermore, a mapping relationship between the probability pattern and the spatiotemporal evolution of traffic flow is constructed, thereby realizing the simulation of traffic volume and the analysis of random characteristics of the traffic network. The method of the present invention achieves a high simulation accuracy and can effectively reveal complex structural features such as gradual / mutated changes, peaks / troughs in traffic flow changes. In addition, the method of the present invention reveals the random characteristics in the traffic network. The present invention innovatively applies the waveguide-based quantum walk system to the twin modeling of urban driver dynamics, effectively integrating the driver randomness and traffic correlation in the traffic network, and providing a more accurate traffic flow modeling method. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 Schematic diagram of the overall process of the method of the present invention;

[0051] Figure 2 This is a schematic diagram of the transportation network zoning;

[0052] Figure 3 Comparison chart between simulation results and measured data for each region;

[0053] Figure 4 Schematic diagram of traffic flow simulation accuracy in each area;

[0054] Figure 5 This is a schematic diagram of the randomness index of drivers in each area;

[0055] Figure 6Schematic diagram of traffic correlation index in each region. DETAILED DESCRIPTION

[0056] The present invention is further illustrated below with reference to the accompanying drawings and specific embodiments. It should be understood that these embodiments are only used to illustrate the present invention and are not used to limit the scope of the present invention. After reading the present invention, modifications of various equivalent forms of the present invention made by those skilled in the art all fall within the scope defined by the claims attached to this application.

[0057] The present invention provides a traffic flow modeling method based on waveguide light quantum wandering, such as Figure 1 As shown, it includes the following steps:

[0058] S1: Divide the traffic network into multiple sub-areas and extract traffic flow data based on the sub-areas;

[0059] Using the idea of ​​spatial partitioning, traffic flow is discretized at a certain longitude, latitude, and time resolution to form a grid-like traffic network. In this network, the extracted traffic flow data is a time-organized traffic flow sequence distributed over each sub-region of the traffic network. Each data point represents the number of cars in that sub-region within a specific time window, which forms the data foundation.

[0060] S2: Generate driver dynamic probability pattern based on waveguide light quantum wandering;

[0061] By constructing a waveguide array optical quantum walk consistent with the real traffic network topology, and changing the noise amplitude and coupling strength parameters, the driver dynamic probability pattern with different spatiotemporal structures is generated in each sub-area.

[0062] Specifically, the dynamic characteristics of quantum walk are controlled by Hamiltonian H. The Hamiltonian H of quantum walk of photons in a coupled array containing I waveguides is:

[0063]

[0064] Where Δβ is the noise amplitude, which is used to characterize the randomness intensity of different drivers; C ij is the coupling strength between waveguide i and waveguide j, which is used to characterize the traffic correlation strength between area i and area j in the corresponding traffic network; a i They are the annihilation operator and the generation operator respectively;

[0065] The state of a photon at time t is represented by a wave function:

[0066] ψ[(Δβ,C ij ),t]=e -iHt ψ(0) (2)

[0067] Among them, ψ[(Δβ,C ij ),t] is the wave function, which represents the state of the quantum system at time t; ψ(0) represents the state of the photon at the initial time; based on the wave function, the probability of a photon being in any waveguide at any time can be solved:

[0068] M i [(Δβ,C ij ),t]=ψ i [(Δβ,C ij ),t] 2 (3)

[0069] Among them, M i [(Δβ,C ij ),t] is the driver probability pattern, which represents the probability that the driver appears in sub-area i at time t; construct the light quantum walk parameter set By performing quantum walks under this parameter set, a series of probability patterns can be generated. These probability patterns represent the dynamic characteristics of drivers under different driver randomness and traffic-related conditions, thereby reflecting the complex structural characteristics of traffic flow.

[0070] In this embodiment, the FeynmanPAQ optical quantum platform is used to solve the quantum wandering wave function. Specifically, the waveguide array structure is first set based on the real traffic network topology; then, the noise amplitude Δβ and coupling strength C are input in the parameter setting panel. ij After submitting the parameters, the light quantum walk begins and the driver probability model M can be obtained. i [(Δβ,C ij ),t].

[0071] S3: Screening the driver's dynamic probability pattern for significant probability patterns to obtain significant probability patterns in each sub-region;

[0072] The stepwise regression subset screening method is used as the screening method for significant driver probability patterns. The actual traffic flow time series V in sub-area i is observed. i Under the constraint of (t), all possible driver probability patterns {M i [(Δβ,C ij )1,t],M i [(Δβ,C ij )2,t],...,M i [(Δβ,C ij ) N ,t]} perform stepwise regression subset screening;

[0073] The above screening process is expressed as:

[0074]

[0075] Based on the Akaike Information Criterion (AIC), all possible driver probability patterns are screened using the stepwise regression subset screening method in formula (4) to obtain a significant probability pattern set, which is denoted as: {M i [(Δβ,G ij ) n ,t]|n∈A i}, where A i is the significant probability pattern index of region i.

[0076] S4: Achieve mapping transformation between traffic flow data and driver’s significant probability patterns;

[0077] Based on the significant probability patterns in each area screened in step S3, that is, the dynamically changing probability distribution obtained by solving the wave function, a multivariate linear regression model is used to construct a transformation mechanism between the dynamic evolution of traffic flow and the significant probability patterns, thereby achieving simulation and optimization of traffic flow; specifically:

[0078] Based on the significant probability pattern {M i [(Δβ,C ij ) n ,t]|n∈A i}, a mapping transformation relationship between the sub-region traffic flow time series and the driver's significant probability pattern is constructed and expressed as:

[0079]

[0080] in, is the simulated traffic volume; A i is the significant probability pattern index obtained by screening in step S3; a i (n) and ξ i is the mapping parameter, a i (n) is the regression coefficient from the probability pattern to the traffic volume, which reflects the importance of the probability pattern in the traffic volume of the corresponding sub-region; i is the random error term; the two are obtained through forward stepwise regression. In the forward regression, the probability pattern is the feature vector and the traffic flow is the target vector. The forward stepwise regression can automatically obtain the regression coefficient and error; (Δβ, C ij ) n is the quantum walk parameter of the significant probability mode, which is an important parameter to characterize the randomness characteristics in the transportation network;

[0081] Applying steps S3 and S4 to all areas in the transportation network yields:

[0082]

[0083] S5: Conduct random feature mining of transportation networks;

[0084] To quantitatively describe the random characteristics of traffic networks, we construct measurement indices based on the noise amplitude and coupling strength of significant probability patterns: the driver randomness index and the traffic correlation index. The driver randomness index measures the degree of random fluctuations in traffic volume caused by driver randomness (individual heterogeneity and environmental changes); the traffic correlation index measures the importance of a sub-region in traffic volume fluctuations in the traffic network. Specifically,

[0085] The weighted average of the noise amplitude of the significant probability pattern is defined as the driver randomness index β to measure the degree of random fluctuation of traffic volume caused by the randomness of the driver. The formula is as follows:

[0086]

[0087] Among them, β i is sub-region N i Driver randomness index, Δβ n The noise amplitude representing the nth significant probability mode, a i (n) is the mapping parameter of the significant probability pattern, β i The larger it is, the more severe the random fluctuation of traffic volume caused by the randomness of drivers is;

[0088] The randomness of the driver in the present invention is reflected in the noise β i The introduction of noise causes the quantum system to deviate from the theoretical evolution path, which can be used to analogize the randomness of drivers in traffic systems;

[0089] The weighted average of the sum of the coupling intensities of significant probability patterns is defined as the traffic correlation index TC to measure the importance of a region in the change of traffic network flow. The formula is as follows:

[0090]

[0091] Among them, C k is the sum of the coupling strengths of the kth significant probability modes, TC i The larger it is, the more important sub-region i is in the traffic network flow change.

[0092] This invention innovatively applies a waveguide-based quantum walk system to the twin modeling of urban driver dynamics, effectively integrating driver randomness and traffic correlation in the traffic network, and providing a more accurate traffic flow modeling method.

[0093] In order to verify the effectiveness and practical effect of the method of the present invention, this embodiment was experimentally verified, as follows:

[0094] 1. Experimental Data

[0095] In this embodiment, the trajectory data of online ride-hailing vehicles in northeastern Chengdu, China (104.0421°E to 104.1221°E, 30.65294°N to 30.72294°N) published by Didi on November 1, 2016 is used as an example. Figure 2 As shown in the figure, the spatial resolution is set to 0.01°×0.01°, and the study area is divided into a transportation network consisting of 56 sub-areas, and the sub-areas are represented as N1, N2, ..., N 56 In terms of data processing, the time series of online ride-hailing traffic flow in each sub-region is summarized at a time granularity of 10 minutes.

[0096] 2. Experimental Content

[0097] The following experiments were conducted using the above data:

[0098] 1. Traffic flow simulation experiment: First, create an 8×7 waveguide array in FeynmanPAQS. Then, set 10 different coupling strengths C ij ; Third, for each C ij , set the noise amplitude Δβ to 0, 0.5, 1.5, 2, 2.5, 3, 3.5. This generates a total of 80 photon walk parameter sets (Δβ, C ij Fourth, 25 quantum walk experiments were conducted for each set of quantum walk parameters, resulting in 2,100 probabilistic patterns with different random characteristics. Finally, traffic flow simulation was performed based on stepwise regression.

[0099] 2. Simulation accuracy test: using the coefficient of determination (R 2 ) measures the simulation accuracy of the traffic flow modeling method based on waveguide light quantum wandering, and the formula is as follows:

[0100]

[0101] Among them, V i is the actual observed traffic flow time series; is the simulated traffic flow time series; The average value of observed traffic flow.

[0102] 3. Mining the randomness characteristics of the traffic network: Based on the formula in step 5, calculate the driver randomness index and traffic correlation index of each area.

[0103] 3. Experimental Results

[0104] The fluctuation of small-scale traffic flow is mainly caused by the randomness and heterogeneity of drivers. Whether these complex fluctuations can be well simulated is an important basis for judging the performance of the model. Figure 3 As shown, this embodiment demonstrates the model's ability to capture complex fluctuations in three typical areas (N18, N52, and N40) with the best, medium, and worst simulation accuracy. In areas with heavy traffic in N18 and N52, the model reflects the overall trend and structure of traffic volume changes throughout the entire period. Starting at 8:00, the traffic volume increases sharply; from 10:00 to 18:00, the traffic volume reaches a peak and fluctuates at a high level, which is the most active period for the urban population; after 19:00, the traffic volume gradually decreases. The model performs well in capturing peaks, troughs, and sudden changes in traffic volume on a small time scale (such as multiple irregular small peaks from 10:00 to 13:00). In the high-frequency oscillation area of ​​N40, the model also proves its ability to capture complex fluctuations.

[0105] The simulation accuracy of each region is as follows Figure 4 As shown, the modeling accuracy is high (0.84-0.99), with an average simulation accuracy of 0.986, and it can invert the complex fluctuation structure of traffic flow in different areas. The simulation accuracy of the model in N40, N48, and N32 is relatively low. This may be due to the following reasons: First, these areas are located on the edge of the study area, with a relatively sparse distribution of roads within them, consisting of only a few minor roads. Therefore, traffic volume in these areas is low and traffic volume fluctuations are irregular. Second, the number of significant patterns in these areas is relatively small, resulting in the omission of some significant probabilistic patterns.

[0106] The driver randomness index is used to measure the degree of random fluctuations in traffic flow caused by driver randomness (individual heterogeneity, environmental changes). Using the driver randomness index, we can explore the geographical connotations behind traffic flow changes. Figure 5 As shown, sub-regions with high driver randomness are often located at the edges of the study area. These areas are mainly residential, densely built, densely populated, and have an incomplete road system. In this complex traffic environment, driver behavior exhibits strong randomness, resulting in drastic fluctuations in regional traffic volume.

[0107] The traffic correlation index measures the importance of a region in the changes in traffic network flow. Figure 6 As shown, the southern and western "gateway areas" (N3, N4, N8, N27, and N25) have high traffic correlation indices. This is primarily due to the northwest-southeast orientation of the study area's main roads, which restrict vehicles to entering and exiting the area via the southern and western corridors. Clearly, traffic flows entering and exiting these "gateway areas" significantly impact the overall traffic flow within the study area, leading to high traffic correlation indices in these areas.

[0108] This embodiment also provides a traffic flow modeling system based on waveguide optical quantum wandering, which includes a network interface, a memory and a processor; wherein the network interface is used to realize signal reception and transmission in the process of sending and receiving information between other external network elements; the memory is used to store computer program instructions that can be run on the processor; and the processor is used to execute the steps of the above-mentioned consensus method when running the computer program instructions.

[0109] This embodiment also provides a computer storage medium that stores a computer program that can implement the method described above when a processor executes the computer program. The computer-readable medium can be considered to be tangible and non-transitory. Non-limiting examples of non-transitory tangible computer-readable media include non-volatile memory circuits (such as flash memory circuits, erasable programmable read-only memory circuits, or mask read-only memory circuits), volatile memory circuits (such as static random access memory circuits or dynamic random access memory circuits), magnetic storage media (such as analog or digital tapes or hard drives), and optical storage media (such as CDs, DVDs, or Blu-ray discs). The computer program includes processor-executable instructions stored on at least one non-transitory tangible computer-readable medium. The computer program may also include or rely on stored data. The computer program may include a basic input / output system (BIOS) that interacts with the hardware of a special-purpose computer, device drivers that interact with specific devices of the special-purpose computer, one or more operating systems, user applications, background services, background applications, etc.

[0110] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0111] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1 a process or multiple processes and / or boxes Figure 1A device that provides the functions specified in a block or multiple blocks.

Claims

1. A traffic flow modeling method based on waveguide light quantum wandering, characterized in that: The steps include: S1: Divide the traffic network into multiple sub-areas and extract traffic flow data based on the sub-areas; S2: Generate driver dynamic probability pattern based on waveguide light quantum wandering; S3: Screening the driver's dynamic probability pattern for significant probability patterns to obtain significant probability patterns in each sub-region; S4: Achieve mapping transformation between traffic flow data and driver’s significant probability patterns; S5: Conduct random feature mining of transportation networks; The weighted average of the noise amplitude of the significant probability pattern is defined as the driver randomness index to measure the degree of random fluctuation of traffic volume caused by the randomness of drivers. The weighted average of the sum of the coupling intensities of significant probability patterns is defined as the traffic correlation index to measure the importance of a region in the changes in traffic network flow.

2. The traffic flow modeling method based on waveguide light quantum wandering according to claim 1 is characterized in that: In step S1, traffic flow is discretized by longitude, latitude and time resolution to form a grid-like traffic network. In this network, the extracted traffic flow data is a time-organized traffic flow sequence distributed in each traffic network sub-area, where each data point represents the number of cars in the sub-area within a specific time window.

3. The traffic flow modeling method based on waveguide light quantum wandering according to claim 1 is characterized in that: In step S2, a waveguide array optical quantum walk consistent with the real traffic network topology is constructed, and the noise amplitude and coupling strength parameters are changed to generate driver dynamic probability patterns with different spatiotemporal structures in each sub-area.

4. The traffic flow modeling method based on waveguide light quantum wandering according to claim 3 is characterized in that: The step S2 is specifically as follows: The dynamic characteristics of quantum walk are controlled by Hamiltonian H. The Hamiltonian H of quantum walk of photons in a coupled array containing I waveguides is: Among them, β i is sub-region N i The driver randomness index; Δβ is the noise amplitude, which is used to characterize the randomness intensity of different drivers; C ij is the coupling strength between waveguide i and waveguide j, which is used to characterize the traffic correlation strength between area i and area j in the corresponding traffic network; a i They are the annihilation operator and the generation operator respectively; The state of a photon at time t is represented by a wave function: ψ[(Δβ,C ij ) , t]=e -iHt ψ(0) (2) Among them, ψ[(Δβ,C ij ),t] is the wave function, which represents the state of the quantum system at time t; ψ(0) represents the state of the photon at the initial time; based on the wave function, the probability of a photon being in any waveguide at any time can be solved: M i [(Δβ,C ij ) , t]=ψ i [(Δβ,C ij ) , [t] 2 (3) Among them, M i [(Δβ,C ij ),t] is the driver probability pattern, which represents the probability that the driver appears in sub-area i at time t; construct the light quantum walk parameter set By performing quantum walks under this parameter set, a series of probability patterns can be generated. These probability patterns represent the dynamic characteristics of drivers under different driver randomness and traffic-related conditions, thereby reflecting the complex structural characteristics of traffic flow.

5. The traffic flow modeling method based on waveguide light quantum wandering according to claim 4 is characterized in that: In step S2, the light quantum wandering wave function is solved by using the light quantum platform FeynmanPAQ, specifically: first, the waveguide array structure is set based on the real traffic network topology; then, the noise amplitude Δβ and coupling strength C are input in the parameter setting panel. ij After submitting the parameters, the light quantum walk begins and the driver probability model M can be obtained. i [(Δβ,C ij ) , t].

6. The traffic flow modeling method based on waveguide light quantum wandering according to claim 4 is characterized in that: The step S3 is specifically as follows: using the stepwise regression subset screening method as the screening method for the significant driver probability pattern, the actual observation of the traffic flow time series V in the sub-area i i Under the constraint of (t), all possible driver probability patterns {M i [(Δβ,C ij )1,t],M i [(Δβ,C ij )2,t],...,M i [(Δβ,C ij ) N ,t]} perform stepwise regression subset screening.

7. The traffic flow modeling method based on waveguide light quantum wandering according to claim 6 is characterized in that: The screening process in step S3 is expressed as: Based on the Akaike information criterion, all possible driver probability patterns are screened using the stepwise regression subset screening method in formula (4) to obtain a significant probability pattern set, which is denoted as: {Mi[(Δβ,C ij ) n ,t]|n∈A i }, where A i is the significant probability pattern index of region i.

8. The traffic flow modeling method based on waveguide light quantum wandering according to claim 1 is characterized in that: In step S4, based on the significant probability patterns in each area screened out in step S3, that is, the dynamically changing probability distribution obtained by solving the wave function, a multivariate linear regression model is used to construct a transformation mechanism between the dynamic evolution of traffic flow and the significant probability patterns, thereby realizing the simulation and optimization of traffic flow.

9. The traffic flow modeling method based on waveguide light quantum wandering according to claim 8 is characterized in that: The step S4 is specifically as follows: Based on the significant probability pattern {M i [(Δβ,C ij ) n ,t]|n∈A i }, a mapping transformation relationship between the sub-region traffic flow time series and the driver's significant probability pattern is constructed and expressed as: in, is the simulated traffic volume; A i is the significant probability pattern index obtained by screening in step S3; a i (n) and ξ i is the mapping parameter, a i (n) is the regression coefficient from the probability model to the traffic flow, ξ i is the random error term; (Δβ,C ij ) n is the quantum walk parameter of the significant probability mode, which is an important parameter to characterize the randomness characteristics in the transportation network; Applying steps S3 and S4 to all areas in the transportation network yields:

10. The traffic flow modeling method based on waveguide light quantum wandering according to claim 8 is characterized in that: The step S5 is specifically as follows: The weighted average of the noise amplitude of the significant probability pattern is defined as the driver randomness index β to measure the degree of random fluctuation of traffic volume caused by the randomness of the driver. The formula is as follows: Among them, β i is sub-region N i Driver randomness index, Δβ n The noise amplitude representing the nth significant probability mode, a i (n) is the mapping parameter of the significant probability pattern; The weighted average of the sum of the coupling intensities of significant probability patterns is defined as the traffic correlation index TC to measure the importance of a region in the change of traffic network flow. The formula is as follows: Among them, C n is the sum of the coupling strengths of the nth significant probability modes.

Images