Digital twinning-based global traffic tourism guide sign setting method

Abstract

The invention relates to the technical field of intelligent traffic and digital twinning, in particular to a global traffic tourism guide sign setting method based on digital twinning. The method comprises the following steps: constructing a digital twinborn model of a global traffic scene; calculating a node passing rate based on historical trajectory data taking the target scenic spot as an end point, and preliminarily setting a guide identifier coordinate; setting an expected external probability, arranging a plurality of simulation starting points in the model to carry out verification simulation, driving the feature points to move and recording a track; calculating a coincidence length ratio, an average distance and average time consumption according to the simulated trajectory, and determining a guiding characterization value by combining a preset initial value; according to the method, simulation verification and multi-dimensional evaluation are carried out by constructing the digital twinborn model, dynamic optimization and accurate layout of guide identification setting are realized, and the guide efficiency and reliability of tourism traffic are effectively improved.

Description

Technical Field

[0001] This invention relates to the fields of intelligent transportation and digital twin technology, and in particular to a method for setting up all-area traffic and tourism guidance signs based on digital twins. Background Technology

[0002] In the tourism transportation sector, the proper placement of wayfinding signs is crucial for guiding tourists to their destinations efficiently and safely. Traditional wayfinding signage deployment relies heavily on human experience and static traffic data, lacking systematic analysis and simulation of dynamic traffic flow, tourist behavior preferences, and complex road conditions. This can lead to problems such as low guidance efficiency, incomplete coverage, and difficulty adapting to real-time traffic changes, impacting tourist travel experience and regional traffic flow.

[0003] With the development of digital twin technology, it has become possible to simulate, analyze, and optimize real-world traffic systems by constructing virtual-real traffic scene models. However, existing technologies have not fully integrated historical trajectory mining, multi-starting-point simulation, and multi-dimensional guidance effectiveness evaluation, making it difficult to accurately evaluate and dynamically optimize wayfinding signage placement schemes. Therefore, there is an urgent need for a method that can scientifically set up and continuously optimize all-area tourism traffic wayfinding signage based on a digital twin environment, through a combination of data-driven and simulation verification.

[0004] Chinese Patent Application No. CN112036621B discloses a method and device for setting up directional signs in a transportation hub. The method includes: Step 1, using a comprehensive correlation clustering model to divide the functional areas within the transportation hub; Step 2, organizing passenger flow lines using the shortest passenger movement trajectory lines to form passenger movement flow lines; the passenger movement flow lines include: normal movement flow lines and emergency evacuation flow lines; Step 3, using space syntax theory to optimize the functional area division and the passenger movement flow lines; Step 4, setting up directional signs within the transportation hub with the optimized functional area layout based on the optimized passenger movement flow lines. This invention helps improve the transfer and evacuation efficiency of passengers within the hub, and enhances the level of passenger flow organization and management and the ability to ensure order management.

[0005] However, existing technologies still have the following problems: Given the relatively closed and structured internal environment of transportation hubs, it is difficult to directly apply to open and complex all-area tourism transportation scenarios, and it lacks the ability to optimize the layout of wayfinding signs based on actual historical trajectory data and dynamic simulation verification using digital twins. Summary of the Invention

[0006] To address this, the present invention provides a method for setting up all-area transportation and tourism guidance signs based on digital twins, in order to overcome the problems of existing technologies that are geared towards the relatively closed and structured internal environment of transportation hubs, making them difficult to directly apply to open and complex all-area tourism transportation scenarios, and lacking the ability to optimize the layout of guidance signs based on actual historical trajectory data and dynamic simulation verification of digital twins.

[0007] To achieve the above objectives, the present invention provides a method for setting up comprehensive transportation and tourism wayfinding signs based on digital twins, comprising: Step S1: Construct a digital twin model of the entire traffic scenario. The digital twin model includes an environment model and several node models distributed within it, and establish a mapping relationship between each node model and the corresponding physical node in the actual traffic scenario. Step S2: Obtain historical trajectory data of vehicles with the target tourist attraction as the destination. For each node in the digital twin model, calculate the ratio of the number of historical trajectories passing through that node to the total number of historical trajectories. Based on the ratio, determine and record the setting coordinates of the guide sign in the digital twin model. Step S3: Set the expected external probability value. In the digital twin model, set multiple simulation starting points based on a preset discrete value distribution. Place the feature points at each of the simulation starting points, perform verification simulation based on the expected external probability value, drive each feature point to move towards the endpoint, and record the simulation trajectory of each feature point. Step S4: Based on each of the simulated trajectories, calculate the overlap length ratio, average distance and average time, and combine the expected external probability value, the preset initial distance and initial time to determine the guidance characterization value. Step S5: Determine whether the current wayfinding sign setting meets the guidance expectation based on the guidance characterization value. If the current wayfinding sign setting does not meet the guidance expectation, obtain each overlapping route and determine the reason why the current wayfinding sign setting does not meet the guidance expectation based on the average value of each overlapping route.

[0008] Further, in step S4, the formula for calculating the guiding characterization value is: ; Where R is the guiding characteristic value, B is the overlap length ratio, L is the average distance, L0 is the initial distance, t is the average time, t0 is the initial time, and η is the expected external probability.

[0009] Further, in step S5, determining whether the current wayfinding sign setting meets the wayfinding expectation based on the wayfinding representation value includes: If the guidance representation value is greater than the preset guidance representation value, then it is determined that the current guidance identifier setting meets the guidance expectation; If the guidance representation value is less than or equal to the preset guidance representation value, it is determined that the current guidance identifier setting does not meet the guidance expectation.

[0010] Further, in step S5, each overlapping route is acquired, and the reason why the current wayfinding signage setting does not meet the guidance expectation is determined based on the average value of each overlapping route, including: If the average value is greater than the preset average value, the reason why the current directional sign settings do not meet the directional expectations is determined based on the average pace. If the average value is less than or equal to the preset average value, it is determined that the reason why the current wayfinding sign settings do not meet the guidance expectations is that the distribution of wayfinding signs does not meet the standard, and the overlap length ratio is adjusted.

[0011] Furthermore, based on average pace, the reasons why the current directional signage settings do not meet directional expectations are determined, including: The speed of each feature point is calculated based on the ratio of the simulated trajectory length of each feature point to the corresponding time consumption. Calculate the average pace of all characteristic points to obtain the average pace; If the average pace is greater than the preset average pace, it is determined that the reason why the current directional sign setting does not meet the directional expectations is that the directional sign distribution does not meet the standard, and the overlap length ratio is adjusted. If the average pace is less than or equal to the preset average pace, the cause is determined to be road condition factors, and the initial time consumption is adjusted.

[0012] Further, adjusting the initial time consumption includes: Calculate the variance of the pace at each feature point, and adjust the initial time based on the variance; The smaller the variance, the greater the reduction in initial time consumption.

[0013] Further, after adjusting the initial time based on the variance, the process includes: The initial reduction in the initial time consumption is obtained based on the variance of the speed distribution at each feature point. Obtain the average distance of the simulated trajectory; The initial reduction rate is corrected based on the average distance to obtain the corrected reduction rate. The initial time consumption is adjusted according to the corrected reduction amount; The smaller the average distance, the greater the reduction in the corrected reduction relative to the initial reduction.

[0014] Further, adjusting the overlap length ratio includes: Calculate the difference between the current guidance representation value and the preset guidance representation value; The reduction in the percentage of overlapping lengths is determined based on the difference, wherein the larger the difference, the greater the reduction. The percentage of overlapping lengths is adjusted according to the reduction rate.

[0015] Furthermore, after adjusting the overlap length ratio, it also includes: Based on the adjusted overlap length ratio, a new guidance characterization value is recalculated; If the new guidance representation value is still less than or equal to the preset guidance representation value, it is determined that the reason why the current guidance identifier setting does not meet the guidance expectation is that the verification simulation does not meet the standard. In response to the determination that the verification simulation does not meet the standard, the number of simulation starting points is adjusted based on the average distance. The greater the average distance, the greater the increase in the number of simulated starting points.

[0016] Furthermore, in step S5, after adjusting the number of simulated starting points based on the average distance, the method further includes: In response to adjusting the number of simulation starting points, the preset discrete value distribution is adjusted based on the adjusted number; The larger the adjusted quantity, the greater the adjustment range of the preset discrete value distribution.

[0017] Compared with existing technologies, the beneficial effects of this invention lie in the fact that it directly uses successful case data from "the local area or similar scenarios" for calibration. The determined preset guidance characterization values ​​integrate localized factors such as road characteristics, driving habits, and tourist behavior in specific areas, making the evaluation criteria for guidance effectiveness more aligned with actual application scenarios. This improves the targeting of the optimization and the final implementation effect. Using the average level of historical successful cases as the expected target for the new solution means that the new solution only needs to reach a known and verified effective level of effectiveness to be considered qualified. This sets a realistic and feasible optimization target, avoiding both under-optimization due to overly low targets and unnecessary cost investment or technical difficulties due to overly high targets, achieving a good balance between effectiveness and cost.

[0018] Furthermore, this invention transforms the determination of the preset average pace from relying on subjective experience or a single data source to a quantitative calculation based on objective road design standards (design speed / legal speed limit) and an empirical discount coefficient reflecting actual traffic conditions. By utilizing existing road segment attribute information (grade, speed limit) in the digital twin model and the mileage ratio of the simulated trajectory for weighting, and adjusting it in conjunction with statistically reasonable empirical coefficients, Vpreset, as a key diagnostic threshold, has a solid physical basis and practical significance, thereby significantly improving the objectivity, scientificity, and credibility of subsequent cause diagnosis (way signage issues vs. road condition issues). By comparing the average pace Vavg of the simulation results with the reasonable benchmark Vpreset calculated based on the inherent attributes of the road, a precise preliminary diagnosis of the cause of "way signage failure" is achieved: if Vavg > Vpreset, it indicates that the vehicle can reach or even exceed a reasonable speed level under the existing road network conditions, thus locking the root cause of the problem into the efficiency of path selection (i.e., unreasonable distribution of way signs leading to detours), and precisely triggering the optimization adjustment of "overlapping length ratio B". If Vavg≤Vpreset, it indicates that the vehicle speed has not reached the level that the road network itself can support, thus pointing the root cause of the problem to the road conditions themselves (such as congestion, complex road conditions), and accurately triggering the optimization adjustment of "initial time t0"; avoiding blind optimization, making subsequent adjustment measures more targeted, and significantly improving the efficiency and effectiveness of the overall optimization process.

[0019] Furthermore, this invention introduces pace variance (S 2Variance is used as a core indicator to measure the "universality" and "severity" of road condition impact. Its underlying logic is: the smaller the variance, the lower the speed of all simulated vehicles, indicating a generally low speed across the region (such as widespread congestion or severe weather). Therefore, a significant downward correction is needed to the initial optimistic time estimate (t0). Conversely, a large variance indicates high speed dispersion, which may be a special case for individual vehicles or road sections, and the overall road condition expectation does not require significant adjustment. This dynamic adjustment based on data statistical characteristics (variance) upgrades the parameter optimization process from experience-driven to data and model-driven, significantly improving the scientific rigor and accuracy of the adjustment actions. After obtaining the initial adjustment range based on variance, this method further introduces average distance (Lavg) as a correction factor, reflecting a profound consideration of the physical reality of traffic travel. Its core insight is that for short-distance travel (smaller Lavg), the total time base is smaller, making it more sensitive to road condition fluctuations, and travelers have different psychological tolerance levels. Applying large adjustments derived from long-distance travel to such trips could lead to overcalibration of the model, deviating from reality. Therefore, by designing the f(Lavg) function, the smaller the average distance, the greater the reduction in adjustment relative to the initial magnitude, ensuring that the adjustment magnitude matches the reasonableness of the travel distance. This dual adjustment mechanism (variance first, then distance) greatly enhances the adaptability and robustness of the entire optimization algorithm, avoiding miscalibration or overcalibration in complex scenarios.

[0020] Furthermore, when the initial adjustment of B based on ΔR still yields unsatisfactory results (Rnew≤Rset), the system can automatically determine that the problem has progressed from the "label placement" level to the "reliability of the simulation model." This embodies an intelligent optimization system with deep diagnostic capabilities: it first attempts to solve the most likely problem (label distribution), and if ineffective, automatically traces back to a more fundamental aspect (the representativeness of the simulation samples). This multi-layered, top-down closed-loop optimization logic significantly enhances the method's ability to cope with complex real-world scenarios, avoiding the predicament of traditional methods falling into local optima or ineffective adjustments. Adjusting the number of starting points based on the average distance (Lavg): the underlying logic is that the longer the average distance, the higher the potential spatial diversity of travel starting points, requiring more samples to cover this diversity to ensure statistical significance. This makes the simulation sample size no longer a static value, but dynamically adjusted according to the travel characteristics revealed by the evaluation, fundamentally improving the scientific rigor and adaptability of simulation sampling. Based on the new quantity linkage adjustment of dispersion (σ): while increasing the number of samples, the spatial distribution dispersion (σnew) is appropriately expanded proportionally, ensuring that the newly added sample points can effectively expand the spatial coverage of the simulation, rather than just being densely concentrated in the original area. This coordinated adjustment of "quantity" and "distribution" avoids the information redundancy problem that may be caused by simply increasing the number of samples, and optimizes the simulation sample set in both "quantity" and "quality", thereby ensuring that subsequent evaluation is based on a more representative traffic flow origin assumption. Attached Figure Description

[0021] Figure 1 This is a flowchart illustrating the workflow of the digital twin-based method for setting up all-area traffic and tourism wayfinding signs according to the present invention. Figure 2 This is a flowchart illustrating the process of determining whether the current wayfinding sign setting meets the expected guidance in the digital twin-based method for setting up all-area traffic and tourism wayfinding signs. Detailed Implementation

[0022] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0023] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0024] Please see Figures 1-2 As shown, Figure 1 This is a flowchart illustrating the workflow of the digital twin-based method for setting up all-area traffic and tourism wayfinding signs according to the present invention. Figure 2 This is a flowchart illustrating the process of determining whether the current wayfinding sign setting meets the expected guidance in the digital twin-based method for setting up all-area traffic and tourism wayfinding signs.

[0025] This invention relates to a method for setting up comprehensive transportation and tourism wayfinding signs based on digital twins, comprising: Step S1: Construct a digital twin model of the entire traffic scenario. The digital twin model includes an environment model and several node models distributed within it, and establish a mapping relationship between each node model and the corresponding physical node in the actual traffic scenario. Step S2: Obtain historical trajectory data of vehicles with the target tourist attraction as the destination. For each node in the digital twin model, calculate the ratio of the number of historical trajectories passing through that node to the total number of historical trajectories. Based on the ratio, determine and record the setting coordinates of the guide sign in the digital twin model. Step S3: Set the expected external probability value. In the digital twin model, set multiple simulation starting points based on a preset discrete value distribution. Place the feature points at each of the simulation starting points, perform verification simulation based on the expected external probability value, drive each feature point to move towards the endpoint, and record the simulation trajectory of each feature point. Step S4: Based on each of the simulated trajectories, calculate the overlap length ratio, average distance and average time, and combine the expected external probability value, the preset initial distance and initial time to determine the guidance characterization value. Specifically, in this embodiment, the overlap length ratio is the proportion of the overlap length between the simulated trajectory and the theoretical shortest path to the total distance.

[0026] Step S5: Determine whether the current wayfinding sign setting meets the guidance expectation based on the guidance characterization value. If the current wayfinding sign setting does not meet the guidance expectation, obtain each overlapping route and determine the reason why the current wayfinding sign setting does not meet the guidance expectation based on the average value of each overlapping route.

[0027] In this embodiment of the invention, firstly, high-precision Geographic Information System (GIS) data, road network data, traffic facility data, and surrounding environmental data of the target area are collected. Using 3D modeling software (such as CityEngine, 3ds Max) or professional simulation platforms (such as AnyLogic, SUMO combined with 3D rendering), a refined environmental model is constructed, including elements such as terrain, roads, intersections, bridges, parking lots, and service areas. In this environmental model, each road intersection, ramp, and important path decision point is abstracted as a node model. Each node model includes its 3D coordinates in digital space, type attributes (such as crossroads, T-junctions), and topological relationships with other connected nodes (i.e., road segments). Subsequently, a unique identifier (ID) is assigned to each node model, and a mapping relationship is established with its corresponding physical location in the actual traffic scenario (associated through latitude and longitude coordinates). Historical trajectory data of all vehicles with the parking lot of the scenic area as the endpoint within the past year are obtained from traffic management departments or vehicle GPS service platforms. Each trajectory data contains an ordered sequence of latitude and longitude coordinates. In the digital twin model, for each node model (e.g., node Ni), the following calculations are performed: Counting the number of passes: Traversing all historical trajectories, using a map matching algorithm to match the coordinate sequence of each trajectory to the road network, and determining whether the trajectory "passes" the node. If the path sequence of a trajectory contains node Ni, it is counted as one pass; Calculating the passing ratio: Let the total number of historical trajectories be Mtotal, and the number of trajectories passing node Ni be Mi. Then the passing ratio Pi of node Ni is calculated using the formula: Pi = Mi / Mtotal; Determining the identifier and setting coordinates: A ratio threshold Pthreshold (e.g., 0.7) is preset. This ratio threshold can be determined based on statistical analysis of historical trajectory data, for example, by calculating the distribution of the passing ratio Pi of all nodes and setting its upper quartile (i.e., the 75th percentile) as Pthreshold. This ensures that signs are placed on core path nodes with concentrated traffic, balancing coverage and setup cost. For any node Ni, if its traversal rate Pi ≥ Pthreshold, then the node is determined to be a key guidance node, and its coordinates in its digital twin model (e.g., (xi,yi,zi)) are recorded as guidance sign placement coordinates, indicating that a guidance sign pointing to the scenic area is recommended to be placed near the corresponding physical location. All coordinates that meet the conditions constitute the initial guidance sign layout scheme. An expected probability value is set: an expected probability value η is set, for example, η = 0.05. This expected probability value can be set based on historical trajectory analysis. For example, analyzing the abnormal turning rate at key nodes where the shortest path direction was not chosen in historical trajectories, and taking its statistical average as the reference benchmark for η. Typically, this value can be set empirically between 0.01 and 0.1 to simulate different degrees of path selection uncertainty.This parameter represents the probability that, during simulated driving, the driver will not follow directional signs or the shortest path logic and will randomly choose a direction (i.e., the probability of "taking the wrong road" or "exploring a new road"). The simulation starting point is configured as follows: within the area covered by the digital twin model, a preset discrete value distribution is defined to determine the spatial distribution of the simulation starting points. For example, a two-dimensional normal distribution is used, with its mean located at the geometric center of the area, and the standard deviation set according to the size of the area. This allows for the setting of multiple simulation starting points at the model's edges and center (e.g., generating 100 starting point coordinates, which can be empirically determined based on the area covered by the digital twin model and the complexity of the path network). Each starting point represents a possible tourist departure location. A verification simulation is then performed: a feature point (representing a simulated vehicle) is initialized at each simulation starting point, and all feature points are driven to move towards the target destination (scenic area). The movement rules are based on classic shortest path algorithms (such as Dijkstra's algorithm), but also introduce an unexpected probability value η for intervention: at each path decision node, the feature point has a probability of (1-η) to choose the current shortest path direction to the destination, and a probability of η to randomly choose a direction from all feasible directions; the complete path node sequence from the starting point to the destination of each feature point and the simulation time of each road segment are fully recorded to form a simulated trajectory dataset for subsequent analysis.

[0028] Specifically, in step S4, the formula for calculating the guiding characterization value is: ; Where R is the guiding characteristic value, B is the overlap length ratio, L is the average distance, L0 is the initial distance, t is the average time, t0 is the initial time, and η is the expected external probability.

[0029] In this embodiment of the invention, the overlap length ratio is the ratio of the length of the overlapping route to the average distance. The length of the overlapping route refers to the fact that for a route with overlapping parts, if the ratio of the number of overlapping simulated trajectories to the total number of simulated trajectories in the route is greater than a preset value, then the route is recorded as an overlapping route. The average distance refers to the average distance traveled by each feature point from its own starting point to its end point. The average time taken is the average time taken by each feature point to complete its own simulated trajectory.

[0030] Specifically, in step S5, determining whether the current guidance identifier setting meets guidance expectations based on the guidance representation value includes: If the guidance representation value is greater than the preset guidance representation value, then it is determined that the current guidance identifier setting meets the guidance expectation; If the guidance representation value is less than or equal to the preset guidance representation value, it is determined that the current guidance identifier setting does not meet the guidance expectation.

[0031] In this embodiment of the invention, the preset guidance characterization value can be determined based on the following method: collecting relevant data from existing wayfinding signage systems in the local area or similar scenarios that have been proven to have good guidance effects (such as high tourist satisfaction and few congestion complaints). Substituting the key parameters of these historical success cases and general preset values ​​into the wayfinding characterization value calculation formula, a series of wayfinding characterization values ​​are calculated, and the average value is taken as the preset wayfinding characterization value.

[0032] This invention uses successful case data from local or similar scenarios for calibration. The determined preset guidance characterization values ​​incorporate localized factors such as road characteristics, driving habits, and tourist behavior in a specific area. This makes the evaluation criteria for guidance effectiveness more aligned with actual application scenarios, improving the targeting of the optimization and the final implementation effect. Using the average level of historical successful cases as the expected target for the new solution means that the new solution only needs to reach a known and verified level of effectiveness to be considered qualified. This sets a realistic and feasible optimization target, avoiding both under-optimization due to overly low targets and unnecessary cost or technical difficulty due to overly high targets, achieving a good balance between effectiveness and cost.

[0033] Specifically, in step S5, each overlapping route is acquired, and the reason why the current wayfinding signage setting does not meet the guidance expectations is determined based on the average value of each overlapping route, including: If the average value is greater than the preset average value, the reason why the current directional sign settings do not meet the directional expectations is determined based on the average pace. If the average value is less than or equal to the preset average value, it is determined that the reason why the current wayfinding sign settings do not meet the guidance expectations is that the distribution of wayfinding signs does not meet the standard, and the overlap length ratio is adjusted.

[0034] In this embodiment of the invention, the preset average value can be determined based on the ideal or standard path length in the digital twin model, combined with historical simulation data statistics. For example, to calculate the ideal path length: in the digital twin model, a path planning algorithm (such as Dijkstra's algorithm) is used to calculate the theoretical shortest path from all simulated starting points in step S3 to the endpoint, and the average of these theoretical path lengths is calculated, denoted as Lidealavg; to collect historical simulation data: run multiple (e.g., N) benchmark simulations based solely on the shortest path (without unexpected external probability interference, i.e., η=0), obtaining a set of simulated trajectories each time. Calculate the percentage B of the overlap length between all simulated trajectories and the corresponding theoretical shortest path in each simulation, and obtain the average value Bavgk (k=1,2,...,N) of the overlap length percentage of all trajectories in that simulation; to determine the preset average value: calculate the arithmetic mean Bavghistorical of Bavgk obtained from all N benchmark simulations, and set it as the preset average value. That is: preset average value = Bavghistorical. This value represents the average overlap between the simulated trajectory and the theoretical path under ideal guidance (completely following the shortest path) and without any signage setting errors. It can be used as a benchmark to evaluate whether the current actual signage setting scheme guides vehicles to travel effectively along the shortest (or most efficient) path.

[0035] Specifically, the reasons why the current directional signage settings do not meet directional expectations, determined based on average pace, include: The speed of each feature point is calculated based on the ratio of the simulated trajectory length of each feature point to the corresponding time consumption. Calculate the average pace of all characteristic points to obtain the average pace; If the average pace is greater than the preset average pace, it is determined that the reason why the current directional sign setting does not meet the directional expectations is that the directional sign distribution does not meet the standard, and the overlap length ratio is adjusted. If the average pace is less than or equal to the preset average pace, the cause is determined to be road condition factors, and the initial time consumption is adjusted.

[0036] In this embodiment of the invention, during the verification simulation in step S3, the total length of the simulated trajectory Ltotaln and the corresponding total simulation time Tn for each feature point (simulated vehicle) are recorded, and the pace Vn = Ltotaln / Tn for each feature point is calculated. For example, if feature point A travels 15 kilometers in 0.5 hours, its pace VA = 15km / 0.5h = 30km / h. The arithmetic mean of the pace Vn for all feature points (assuming a total of 100) is calculated to obtain the average pace Vavg for this simulation. The preset average pace is determined based on road design standards. In the digital twin model, each road segment has a corresponding road grade and design speed or legal speed limit (e.g., 60km / h for urban main roads and 30km / h for scenic area side roads). Take the weighted average of the speed limits of all simulated track segments (weighted according to the proportion of each segment in the total simulated mileage), and then multiply it by an empirical discount factor (such as 0.7-0.9) to simulate the average driving speed under actual non-ideal road conditions. The result can be used as the preset average pace Vpreset.

[0037] This invention transforms the determination of the preset average pace from relying on subjective experience or a single data source to a quantitative calculation based on objective road design standards (design speed / legal speed limit) and empirical discount coefficients reflecting actual traffic conditions. By utilizing existing road segment attribute information (grade, speed limit) in the digital twin model and the mileage ratio of the simulated trajectory for weighting, and adjusting it with statistically reasonable empirical coefficients, the Vpreset, as a key diagnostic threshold, has a solid physical basis and practical significance, thereby significantly improving the objectivity, scientificity, and credibility of subsequent cause diagnosis (way signage issues vs. road condition issues). By comparing the average pace Vavg of the simulation results with the reasonable benchmark Vpreset calculated based on the inherent road attributes, a precise preliminary diagnosis of the cause of "way signage failure" is achieved: if Vavg > Vpreset, it means that the vehicle can reach or even exceed a reasonable speed level under the existing road network conditions, thus locking the root cause of the problem into the efficiency of route selection (i.e., unreasonable distribution of way signs leading to detours), and precisely triggering the optimization adjustment of "overlapping length ratio B". If Vavg≤Vpreset, it indicates that the vehicle speed has not reached the level that the road network itself can support, thus pointing the root cause of the problem to the road conditions themselves (such as congestion, complex road conditions), and accurately triggering the optimization adjustment of "initial time t0"; avoiding blind optimization, making subsequent adjustment measures more targeted, and significantly improving the efficiency and effectiveness of the overall optimization process.

[0038] Specifically, adjusting the initial time consumption includes: Calculate the variance of the pace at each feature point, and adjust the initial time based on the variance; The smaller the variance, the greater the reduction in initial time consumption.

[0039] Specifically, after adjusting the initial time based on the variance, the process includes: The initial reduction in the initial time consumption is obtained based on the variance of the speed distribution at each feature point. Obtain the average distance of the simulated trajectory; The initial reduction rate is corrected based on the average distance to obtain the corrected reduction rate. The initial time consumption is adjusted according to the corrected reduction amount; The smaller the average distance, the greater the reduction in the corrected reduction relative to the initial reduction.

[0040] In this embodiment of the invention, based on the aforementioned judgment, the simulated average pace Vavg = 28 km / h is less than the preset average pace Vpreset = 32 km / h, which is determined to be due to road condition factors causing a discrepancy with the guidance expectation, requiring adjustment of the initial travel time t0; the pace Vn arrays for all feature points (e.g., 100) and the simulated average distance Lavg (i.e., the average value of all feature points Ltotaln) are already available; the variance S of the pace Vn for all feature points is calculated. 2 This is used to quantify the dispersion of the speeds of each simulated vehicle. Variance S 2 The smaller the value, the more it indicates that the speeds of all vehicles are generally low and to a similar degree, suggesting that poor road conditions are a common phenomenon; the reduction in initial time t0 (i.e., the amount that needs to be reduced) and variance S 2 Negative correlation. The smaller the variance, the greater the reduction, because the generally lower speed requires a greater correction to the initial optimistic time estimate; Preset value: variance threshold S 2 Threshold determination method: It can be based on the pace variance calculated from multiple historical simulations (or historical trajectory data). For example, calculate the upper quartile (i.e., the 75th quartile) of the historical variance dataset and set it as S. 2 This threshold is used to distinguish between "high" and "low" velocity dispersion; it utilizes the variance S. 2 And a preset reduction base value Δbase (e.g., Δbase = 0.05 × t0, indicating that 5% of t0 is used as the base adjustment unit), the initial reduction magnitude Δpreliminary is calculated through a mapping function. A simple linear relationship example: Δpreliminary = Δbase × (1 - S 2 / S 2 threshold), and ensure that when S 2 >=S 2 At the threshold, Δpreliminary is 0 or a minimum value; Example: Assume t0 = 1.0 hours, Δbase = 0.05 hours, S2 threshold = 25 (km / h) 2 , the variance S of the current simulation 2 = 4 (km / h) 2 . Then Δpreliminary = 0.05×(1 - 4 / 25) = 0.05×0.84 = 0.042 hours; the smaller the average distance Lavg, the shorter the total vehicle travel distance, and the relatively higher the sensitivity of the travel time to road condition fluctuations. Excessive drastic adjustments to the travel time may not conform to the actual situation of short-distance travel. Therefore, it is necessary to reduce the reduction amplitude; preset value: method for determining the average distance threshold Lthreshold: The average or median of the theoretical shortest path lengths from all simulation starting points to the end points can be taken as Lthreshold, representing the typical travel distance in this area; calculate the corrected reduction amplitude: Δfinal = Δpreliminary×f(Lavg). Among them, f(Lavg) is a correction coefficient function that satisfies: when Lavg >= Lthreshold, f ≈ 1 (basically no correction); when Lavg < Lthreshold, f(Lavg) decreases as Lavg decreases (for example, f = Lavg / Lthreshold). Example continuation: Assume Lthreshold = 20 km and the current simulated Lavg = 18 km. Then the correction coefficient f = 18 / 20 = 0.9. The corrected reduction amplitude Δfinal = 0.042 hours×0.9 = 0.0378 hours; finally, update the initial travel time t0 to: t0new = t0old - Δfinal. Example final adjustment: t0new = 1.0 - 0.0378 = 0.9622 hours.

[0041] The present invention introduces the pacing variance (S 2Variance is used as a core indicator to measure the "universality" and "severity" of road condition impact. Its underlying logic is: the smaller the variance, the lower the speed of all simulated vehicles, indicating a generally low speed across the region (such as widespread congestion or severe weather). Therefore, a significant downward correction is needed to the initial optimistic time estimate (t0). Conversely, a large variance indicates high speed dispersion, which may be a special case for individual vehicles or road sections, and the overall road condition expectation does not require significant adjustment. This dynamic adjustment based on data statistical characteristics (variance) upgrades the parameter optimization process from experience-driven to data and model-driven, significantly improving the scientific rigor and accuracy of the adjustment actions. After obtaining the initial adjustment range based on variance, this method further introduces average distance (Lavg) as a correction factor, reflecting a profound consideration of the physical reality of traffic travel. Its core insight is that for short-distance travel (smaller Lavg), the total time base is smaller, making it more sensitive to road condition fluctuations, and travelers have different psychological tolerance levels. Applying large adjustments derived from long-distance travel to such trips could lead to overcalibration of the model, deviating from reality. Therefore, by designing the f(Lavg) function, the smaller the average distance, the greater the reduction in adjustment relative to the initial magnitude, ensuring that the adjustment magnitude matches the reasonableness of the travel distance. This dual adjustment mechanism (variance first, then distance) greatly enhances the adaptability and robustness of the entire optimization algorithm, avoiding miscalibration or overcalibration in complex scenarios.

[0042] Specifically, adjusting the overlap length ratio includes: Calculate the difference between the current guidance representation value and the preset guidance representation value; The reduction in the percentage of overlapping lengths is determined based on the difference, wherein the larger the difference, the greater the reduction. The percentage of overlapping lengths is adjusted according to the reduction rate.

[0043] In this embodiment of the invention, after calculation in step S4, the current wayfinding sign setting scheme's wayfinding characteristic value Rcurrent = 0.72 is obtained, and the preset wayfinding characteristic value Rset = 0.85 (determined as described above, for example, based on the average calculation of historical successful case data). The difference between the current wayfinding characteristic value and the preset value is calculated: ΔR = Rset - Rcurrent = 0.85 - 0.72 = 0.13. The larger ΔR is, the greater the gap between the current setting effect and the expected goal. The method for determining the adjustment ratio coefficient k: This coefficient can be calibrated through historical optimization experience or simulation testing. A simple method is to select a baseline scenario in the digital twin model, manually adjust the overlap length ratio B within a small range, observe its impact on the guidance representation value R, calculate the approximate ratio ΔR / ΔB, and take its reciprocal or a conservative fraction (e.g., 0.5 to 0.8 times) as the initial k value. The k value is usually in the range of (0.1, 1.0). Calculate the reduction magnitude: the reduction magnitude ΔB of the overlap length ratio is proportional to the difference ΔR. The calculation formula can be set as: ΔB = k × ΔR. Example: Assuming the preset adjustment ratio coefficient k = 0.6, then the reduction magnitude ΔB of this adjustment is 0.6 × 0.13 = 0.078. The currently calculated overlap length ratio is Bcurrent = 0.70. After adjustment based on the reduction magnitude ΔB, a new overlap length ratio value is obtained: Bnew = Bcurrent - ΔB = 0.70 - 0.078 = 0.622. This new value Bnew will be used to update the evaluation benchmark of the wayfinding signage layout scheme, or directly guide the adjustment of the signage distribution density (for example, in subsequent steps, according to the new B value requirements, add or move the setting coordinates of wayfinding signs in the digital twin model).

[0044] Specifically, after adjusting the overlap length ratio, the method further includes: Based on the adjusted overlap length ratio, a new guidance characterization value is recalculated; If the new guidance representation value is still less than or equal to the preset guidance representation value, it is determined that the reason why the current guidance identifier setting does not meet the guidance expectation is that the verification simulation does not meet the standard. In response to the determination that the verification simulation does not meet the standard, the number of simulation starting points is adjusted based on the average distance. The greater the average distance, the greater the increase in the number of simulated starting points.

[0045] Specifically, in step S5, after adjusting the number of simulated starting points based on the average distance, the method further includes: In response to adjusting the number of simulation starting points, the preset discrete value distribution is adjusted based on the adjusted number; The larger the adjusted quantity, the greater the adjustment range of the preset discrete value distribution.

[0046] In this embodiment of the invention, after adjusting the overlap length ratio B according to the aforementioned embodiment, the guiding characterization value is recalculated based on the new B value, resulting in Rnew=0.80, while the preset guiding characterization value remains Rset=0.85. Since Rnew(0.80)≤Rset(0.85), the reason why the current guidance sign setting does not meet the guidance expectation is that the verification simulation does not meet the standard. That is, the existing simulated starting point distribution may not be sufficient to fully reflect the diversity of real travel starting points, resulting in evaluation distortion. The number of simulated starting points is adjusted based on the average distance: Preset value: the number of baseline starting points Nbase and the adjustment coefficient α, Nbase: the initial number of simulated starting points (e.g., 100), α: the adjustment coefficient, used to control the adjustment intensity as the average distance changes. It can be set based on experience. For example, α=0.05 means that for every 1 kilometer the average distance exceeds the threshold, the number of starting points increases by 5% of the baseline number. Preset value: the average distance threshold Ltrigger determination method: the median of the theoretical shortest path length from all simulated starting points to the destination or the average of the historical actual travel distance can be taken as Ltrigger to determine whether the trip is "long-distance". Calculate the number of new starting points: the number of new starting points Nnew=Nbase×(1+α×max(0,(Lavg-Ltrigger) / Ltrigger)), the formula reflects the logic of "the larger the average distance, the greater the increase". Example: Let Nbase=100, α=0.05, Ltrigger=15km, and current Lavg=18km, then Nnew=100×(1+0.05×((18-15) / 15))=100×(1+0.01)=101; if Lavg=30km, then Nnew=100×(1+0.05×((30-15) / 15))=100×(1+0.05)=105. Adjust the discrete value distribution based on the new quantity: Adjustment object: The preset discrete values ​​used in step S3 to determine the spatial distribution of the simulation starting point. Distribution parameters, such as the standard deviation σ of a two-dimensional normal distribution, can be increased to make the starting point distribution more dispersed. The preset value is the dispersion adjustment coefficient β. The method for determining β is as follows: β is a small constant, for example, β = 0.005, representing the percentage increase in standard deviation σ for every 1% increase in the number of starting points. The new dispersion is calculated as: new standard deviation σnew = σold × (1 + β × ((Nnew / Nbase) - 1)). This formula reflects the logic that "the larger the adjusted quantity, the larger the adjustment magnitude." Example: Continuing the previous example, Nnew = 105, σold = 2.0 km, β = 0.005. Then σnew = 2.0 × (1 + 0.005 × ((105 / 100) - 1)) = 2.0 × (1 + 0.005 × 0.05) ≈ 2.0005 km. Although the adjustment magnitude is small, the direction is clear, aiming to spread more starting points to the outer periphery and improve the comprehensiveness of the simulation coverage.Re-execute the simulation: Using the updated number of simulation start points Nnew and discrete value distribution parameters (such as σnew), re-execute steps S3 (verify the simulation) and S4 (calculate the guided characterization value) to obtain more reliable and representative evaluation results, thereby continuing subsequent optimization or verification.

[0047] When the initial adjustment of B based on ΔR still yields unsatisfactory results (Rnew≤Rset), the system can automatically determine that the problem has deepened from the "marker placement" level to the "simulation model reliability" level. This embodies an intelligent optimization system with deep diagnostic capabilities: it first attempts to solve the most likely problem (marker distribution), and if ineffective, automatically traces back to a more fundamental aspect (representativeness of simulation samples). This multi-layered, top-down closed-loop optimization logic significantly enhances the method's ability to cope with complex real-world scenarios, avoiding the predicament of traditional methods falling into local optima or ineffective adjustments. Adjusting the number of starting points based on average distance (Lavg): The underlying logic is that the longer the average distance, the higher the potential spatial diversity of travel starting points, requiring more samples to cover this diversity to ensure statistical significance. This makes the simulation sample size no longer a static value, but dynamically adjusted according to the travel characteristics revealed by the evaluation, fundamentally improving the scientific rigor and adaptability of simulation sampling. Adjusting the dispersion (σ) based on the new number of samples: While increasing the number of samples, the dispersion of their spatial distribution (σnew) is appropriately expanded proportionally, ensuring that the newly added sample points effectively expand the spatial coverage of the simulation, rather than simply being densely concentrated in the original area. This coordinated adjustment of "quantity" and "distribution" avoids the information redundancy problem that may result from simply increasing the number of samples, and optimizes the simulation sample set in both "quantity" and "quality" at the same time, thereby ensuring that subsequent evaluations are based on a more representative traffic flow origin assumption.

[0048] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

[0049] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention can have various modifications and variations. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for setting up comprehensive transportation and tourism guidance signs based on digital twins, characterized in that, include: Step S1: Construct a digital twin model of the entire traffic scenario. The digital twin model includes an environment model and several node models distributed within it, and establish a mapping relationship between each node model and the corresponding physical node in the actual traffic scenario. Step S2: Obtain historical trajectory data of vehicles with the target tourist attraction as the destination. For each node in the digital twin model, calculate the ratio of the number of historical trajectories passing through that node to the total number of historical trajectories. Based on the ratio, determine and record the setting coordinates of the guide sign in the digital twin model. Step S3: Set the expected external probability value. In the digital twin model, set multiple simulation starting points based on a preset discrete value distribution. Place the feature points at each of the simulation starting points, perform verification simulation based on the expected external probability value, drive each feature point to move towards the endpoint, and record the simulation trajectory of each feature point. Step S4: Based on each of the simulated trajectories, calculate the overlap length ratio, average distance and average time, and combine the expected external probability value, the preset initial distance and initial time to determine the guidance characterization value. Step S5: Determine whether the current wayfinding sign setting meets the guidance expectation based on the guidance characterization value. If the current wayfinding sign setting does not meet the guidance expectation, obtain each overlapping route and determine the reason why the current wayfinding sign setting does not meet the guidance expectation based on the average value of each overlapping route.

2. The method for setting up all-area traffic and tourism guidance signs based on digital twins according to claim 1, characterized in that, In step S4, the formula for calculating the guiding characterization value is: ； Where R is the guiding characteristic value, B is the overlap length ratio, L is the average distance, L0 is the initial distance, t is the average time, t0 is the initial time, and η is the expected external probability.

3. The method for setting up all-area traffic and tourism guidance signs based on digital twins according to claim 1, characterized in that, In step S5, determining whether the current wayfinding sign setting meets the expected wayfinding based on the wayfinding representation value includes: If the guidance representation value is greater than the preset guidance representation value, then it is determined that the current guidance identifier setting meets the guidance expectation; If the guidance representation value is less than or equal to the preset guidance representation value, it is determined that the current guidance identifier setting does not meet the guidance expectation.

4. The method for setting up all-area traffic and tourism guidance signs based on digital twins according to claim 1, characterized in that, In step S5, each overlapping route is acquired, and the reason why the current wayfinding signage setting does not meet the guidance expectations is determined based on the average value of each overlapping route, including: If the average value is greater than the preset average value, the reason why the current directional sign settings do not meet the directional expectations is determined based on the average pace. If the average value is less than or equal to the preset average value, it is determined that the reason why the current wayfinding sign settings do not meet the guidance expectations is that the distribution of wayfinding signs does not meet the standard, and the overlap length ratio is adjusted.

5. The method for setting up all-area traffic and tourism guidance signs based on digital twins according to claim 4, characterized in that, The reasons why the current directional signage settings do not meet directional expectations are determined based on the average pace, including: The speed of each feature point is calculated based on the ratio of the simulated trajectory length of each feature point to the corresponding time consumption. Calculate the average pace of all characteristic points to obtain the average pace; If the average pace is greater than the preset average pace, it is determined that the reason why the current directional sign setting does not meet the directional expectations is that the directional sign distribution does not meet the standard, and the overlap length ratio is adjusted. If the average pace is less than or equal to the preset average pace, the cause is determined to be road condition factors, and the initial time consumption is adjusted.

6. The method for setting up all-area transportation and tourism guidance signs based on digital twins according to claim 5, characterized in that, Adjusting the initial time consumption includes: Calculate the variance of the pace at each feature point, and adjust the initial time based on the variance; The smaller the variance, the greater the reduction in initial time consumption.

7. The method for setting up all-area transportation and tourism guidance signs based on digital twins according to claim 6, characterized in that, After adjusting the initial time based on the variance, the following is included: The initial reduction in the initial time consumption is obtained based on the variance of the speed distribution at each feature point. Obtain the average distance of the simulated trajectory; The initial reduction rate is corrected based on the average distance to obtain the corrected reduction rate. The initial time consumption is adjusted according to the corrected reduction amount; The smaller the average distance, the greater the reduction in the corrected reduction relative to the initial reduction.

8. The method for setting up all-area traffic and tourism guidance signs based on digital twins according to claim 5, characterized in that, Adjusting the overlap length ratio includes: Calculate the difference between the current guidance representation value and the preset guidance representation value; The reduction in the percentage of overlapping lengths is determined based on the difference, wherein the larger the difference, the greater the reduction. The percentage of overlapping lengths is adjusted according to the reduction rate.

9. The method for setting up all-area traffic and tourism guidance signs based on digital twins according to claim 3, characterized in that, After adjusting the overlap length ratio, the method further includes: Based on the adjusted overlap length ratio, a new guidance characterization value is recalculated; If the new guidance representation value is still less than or equal to the preset guidance representation value, it is determined that the reason why the current guidance identifier setting does not meet the guidance expectation is that the verification simulation does not meet the standard. In response to the determination that the verification simulation does not meet the standard, the number of simulation starting points is adjusted based on the average distance. The greater the average distance, the greater the increase in the number of simulated starting points.

10. The method for setting up all-area traffic and tourism guidance signs based on digital twins according to claim 1, characterized in that, In step S5, after adjusting the number of simulated starting points based on the average distance, the method further includes: In response to adjusting the number of simulation starting points, the preset discrete value distribution is adjusted based on the adjusted number; The larger the adjusted quantity, the greater the adjustment range of the preset discrete value distribution.

Images