Multimodal transportation-based route planning methods, systems, devices, and storage media

Abstract

This invention relates to the field of navigation technology, specifically providing a method, system, device, and storage medium for path planning based on multimodal transportation. The method includes: collecting real-time data on multimodal transportation; updating node attribute values ​​and edge weight values ​​related to traffic state in a pre-constructed dynamic directed weighted graph based on the real-time data; using a path search algorithm, searching for K shortest paths between the starting point and the destination as candidate paths based on the updated dynamic directed weighted graph; acquiring user preference data, calculating index scores for the candidate paths, and selecting the optimal path from the candidate paths based on the preference data and index scores. This invention solves three major pain points in traditional multimodal path planning: lagging dynamic response, limited solution options, and insufficient personalization, achieving accurate and efficient personalized navigation services.

Description

Technical Field

[0001] This invention belongs to the field of navigation technology, specifically relating to a path planning method, system, device, and storage medium based on multimodal transportation. Background Technology

[0002] With the acceleration of urbanization, residents' travel needs are becoming more diversified and dynamic. Traditional navigation systems mainly rely on static route planning for a single mode of transportation (such as driving or public transportation), which makes it difficult to meet the challenges posed by real-time variables in modern travel scenarios, such as subway delays, imbalances in shared bicycle scheduling, and dynamic pricing for ride-hailing services.

[0003] While some existing navigation software supports multi-mode transportation planning, they do not fully integrate real-time operational status (such as bus arrival time deviations) and transfer costs (such as the distance to shared bicycle parking spots). These limitations result in insufficient reliability of recommended routes in scenarios such as sudden congestion and capacity interruptions. There is an urgent need to establish a route planning system that integrates real-time multi-source data, dynamic weight adjustments, and personalized decision-making. Summary of the Invention

[0004] In view of the above-mentioned shortcomings of the prior art, the present invention provides a route planning method, system, device and storage medium based on multimodal transportation to solve the above-mentioned technical problems.

[0005] In a first aspect, the present invention provides a route planning method based on multimodal transportation, comprising:

[0006] Collect real-time data of multi-modal vehicles, including location and operating status;

[0007] Based on the real-time data, update the node attribute values ​​and edge weight values ​​related to traffic status in the pre-constructed dynamic directed weighted graph; wherein, the edge weight values ​​include at least travel time or cost; and the node attribute values ​​include at least the waiting time at transfer points.

[0008] Receive the start and end points input by the user;

[0009] Using a path search algorithm, based on the updated dynamic directed weighted graph, search for K shortest paths between the starting point and the ending point as candidate paths;

[0010] Obtain user preference data, calculate the indicator score of candidate paths, and select the optimal path from the candidate paths based on the preference data and indicator score.

[0011] In one optional implementation, real-time data on multimodal transportation is collected, including:

[0012] The system collects real-time subway train schedule data and real-time location data through vehicle-to-ground wireless communication and smart card swiping recording system, and uses a sliding window to determine the departure interval based on the real-time location.

[0013] The system uses a positioning chip to obtain the real-time location of the bus, and uses onboard sensors to obtain the real-time speed and direction angle of the bus. It also uses an infrared counter on an electronic bus stop sign to obtain the number of people queuing at the station.

[0014] Vehicle status data of shared bicycles are obtained through a low-power wide-area network, including location and battery level.

[0015] By relying on a perception network composed of floating car data and roadside radar, the congestion index of each road segment is obtained.

[0016] In an optional implementation, the method further includes:

[0017] A timestamp synchronization mechanism is used to align the collected real-time data in time and space.

[0018] In one alternative implementation, the method for constructing a dynamically directed weighted graph includes:

[0019] Define the basic parameters for modeling, including:

[0020] The node attributes include a basic type identifier and spatiotemporal feature labels. The spatiotemporal feature labels include: the three-dimensional coordinates of the transportation station, the entrance and exit distribution data, and the transfer passage length and real-time availability data of escalators or elevators of the transfer node.

[0021] Edge classification and weight calculation include:

[0022] Same pattern edge: The weight is dynamically calculated by real-time traffic data, where the weight of the subway section edge = basic section time × (1 + 0.1 × line congestion coefficient), and the weight of the bus segment edge is calculated by segmented accumulation.

[0023] Transfer edge: The weight includes walking time and waiting time penalties, where the weight of the shared bicycle riding edge = distance / (base riding speed × weather correction factor × road condition factor);

[0024] A distributed graph database is used to store the network topology, a secondary index is established for frequently accessed nodes and edges, and a spatiotemporal partitioning storage strategy is implemented.

[0025] In an optional implementation, based on the real-time data, the node attribute values ​​and edge weight values ​​related to traffic state in the pre-constructed dynamic directed weighted graph are updated, including:

[0026] Node attributes are updated dynamically:

[0027] For subway station nodes, based on real-time train schedules and arrival prediction data, update the platform congestion label and the expected first train arrival time field, and trigger a red warning sign when there is a delay;

[0028] For bus stop nodes, based on vehicle positioning and electronic bus stop data, the estimated number of waiting passengers and the number of stops for the next bus to arrive are dynamically updated, and a countdown reminder is started when the number of stops is ≤2.

[0029] For shared bicycle stations, the ratio of available bicycles to total capacity is updated every 30 seconds. When the ratio is less than 0.2, a bicycle shortage is marked, and the number of available parking spaces is updated based on the electronic fence signal. If the ratio is 0, the coordinates of backup parking points are associated.

[0030] For transfer nodes, the availability status of vertical transfer facilities is updated in real time, and backup passages are activated in case of failure; the length of transfer passages is dynamically adjusted according to real-time passenger flow density.

[0031] Real-time calculation and updating of edge weights:

[0032] The weight of a subway section is calculated as follows: Basic section time × [1 + 0.1 × max(full load rate / 80%, 1)], plus the delay coefficient.

[0033] The weight of a bus route segment is equal to the cumulative value of each segment. In the event of a road emergency, the weight of the segment where the incident occurred is multiplied.

[0034] The weighting of pedestrian transfer sides incorporates weather correction and dynamic waiting penalty coefficients;

[0035] The weight of shared bicycle riding is calculated as distance / [5m / s × road condition coefficient × weather coefficient]. The weight of no-parking zones is set to infinity.

[0036] In an optional implementation, a path search algorithm is used to search for K shortest paths between the starting point and the ending point as candidate paths based on the updated dynamically directed weighted graph, including:

[0037] The heuristic function is dynamically selected based on the route optimization objective: when the optimization objective is the shortest time, a heuristic function based on geographical straight-line distance and theoretical fastest traffic speed is used; when the optimization objective is the fewest transfers, a heuristic function based on the estimated minimum number of transfers is used.

[0038] Simultaneously, the four path evaluation dimensions were optimized, including total path time, total number of transfer operations, total path carbon emissions, and total walking distance; the four path evaluation dimensions were combined into a comprehensive evaluation index according to user-configured weights.

[0039] During the search process, Pareto non-dominated solutions are retained in real time, path schemes whose comprehensive evaluation index is surpassed are eliminated, and finally a specified number of candidate paths are output.

[0040] In one optional implementation, acquiring user preference data, calculating the index score of candidate paths, and selecting the optimal path from the candidate paths based on the preference data and the index score includes:

[0041] The four core indicators of the route are normalized and transformed into scores. The four core indicators include total time, travel cost, transfer convenience and real-time reliability.

[0042] The initial weights of each indicator are determined through expert evaluation, and preference features are identified based on users' historical behavior. Personalized weights are generated based on the preference features, and dynamic weights are generated based on the personalized weights and the initial weights.

[0043] The standardized scores obtained from the normalized scores of each indicator are combined with the dynamic weights and summed to generate a comprehensive score.

[0044] Candidate paths are sorted in descending order of comprehensive score. When the score difference is less than the set tolerance, a second sorting based on reliability indicators is activated.

[0045] Secondly, the present invention provides a route planning system based on multimodal transportation, comprising:

[0046] The data acquisition module is used to collect real-time data of multi-mode vehicles, including location and operating status.

[0047] The graph update module is used to update the node attribute values ​​and edge weight values ​​related to traffic status in the pre-constructed dynamic directed weighted graph based on the real-time data; wherein, the edge weight values ​​include at least travel time or cost; and the node attribute values ​​include at least the waiting time at transfer points.

[0048] The request receiving module is used to receive the start and end points of user input.

[0049] The path search module is used to search for K shortest paths between the starting point and the ending point as candidate paths based on the updated dynamic directed weighted graph using a path search algorithm.

[0050] The path filtering module is used to acquire user preference data, calculate the index score of candidate paths, and filter the optimal path from the candidate paths based on the preference data and index score.

[0051] Thirdly, a device is provided, comprising:

[0052] Memory for storing route planning programs based on multimodal transportation;

[0053] A processor is configured to implement the steps of the multimodal traffic-based route planning method as provided in the first aspect when executing the multimodal traffic-based route planning program.

[0054] Fourthly, a computer-readable storage medium is provided, on which a multimodal traffic-based route planning program is stored, wherein when the multimodal traffic-based route planning program is executed by a processor, the steps of the multimodal traffic-based route planning method provided in the first aspect are implemented.

[0055] The beneficial effects of this invention are as follows: The multimodal transportation-based path planning method, system, device, and storage medium provided by this invention continuously update node attributes (such as transfer waiting time) and edge weights (such as travel time) based on real-time location / status data of transportation vehicles, ensuring that the road network representation is dynamically synchronized with the physical world; using an updated dynamic directed weighted graph as the search basis, combined with a K-shortest path generation strategy, it outputs diverse candidate solutions while ensuring computational efficiency; and by integrating user preference data and path indicator scores for joint decision-making, the final output optimal path combines objective quantitative advantages with personalized needs. This invention solves the three major pain points of traditional multimodal path planning: lagging dynamic response, limited solution options, and insufficient personalization, achieving accurate and efficient personalized navigation services.

[0056] Furthermore, the design principle of this invention is reliable, the structure is simple, and it has a very wide range of application prospects. Attached Figure Description

[0057] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0058] Figure 1 This is a schematic flowchart of a method according to an embodiment of the present invention.

[0059] Figure 2 This is a schematic block diagram of a system according to an embodiment of the present invention.

[0060] Figure 3 This is a schematic diagram of the structure of a device provided in an embodiment of the present invention. Detailed Implementation

[0061] To enable those skilled in the art to better understand the technical solutions of this invention, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this invention.

[0062] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

[0063] The multimodal traffic-based route planning method provided in this embodiment of the invention is executed by a computer device, and correspondingly, the multimodal traffic-based route planning system runs on the computer device.

[0064] Figure 1 This is a schematic flowchart illustrating a method according to an embodiment of the present invention. Wherein, Figure 1 The implementing entity can be a multimodal transportation-based route planning system. Depending on different requirements, the order of steps in this flowchart can be changed, and some steps can be omitted.

[0065] like Figure 1 As shown, the method includes:

[0066] S1. Collect real-time data of multi-mode vehicles, including location and operating status;

[0067] S2. Based on the real-time data, update the node attribute values ​​and edge weight values ​​related to traffic status in the pre-constructed dynamic directed weighted graph; wherein, the edge weight values ​​include at least travel time or cost; and the node attribute values ​​include at least the waiting time at transfer points;

[0068] S3. Receive the start and end points input by the user;

[0069] S4. Using a path search algorithm, based on the updated dynamic directed weighted graph, search for K shortest paths between the starting point and the ending point as candidate paths;

[0070] S5. Obtain user preference data, calculate the indicator score of the candidate path, and select the optimal path from the candidate paths based on the preference data and indicator score.

[0071] In one embodiment of the present invention, based on step S1, the following will provide a possible embodiment and describe its specific implementation in a non-limiting manner.

[0072] Multimodal transportation, such as subways, buses, and shared bicycles. Data collection for these modes of transportation includes, for example:

[0073] Subway and bus data: A dual-channel acquisition mechanism is constructed based on vehicle-to-ground wireless communication (LTE-M) and smart card swiping records (AFC system). Subway real-time train schedule data achieves sub-meter level positioning through trackside sensors and train transponders. The departure interval calculation adopts a sliding window algorithm (window size set to 5 trains) to balance real-time performance and stability. Bus vehicle positioning integrates GPS and Beidou dual-mode positioning (positioning frequency 1Hz), and collects real-time speed and direction angle through on-board OBU equipment. Combined with the infrared counter of the electronic bus stop sign, the number of people queuing at the station is obtained (error controlled within ±5 people).

[0074] Shared bike data: NB-IoT low-power wide area network is used to report vehicle status in real time (heartbeat packet every 30 seconds), including battery level (for assessing availability), GPS positioning and electronic fence trigger signals (to determine parking compliance); the station vehicle capacity utilization rate (current number of vehicles / maximum station capacity) is statistically analyzed in real time through edge computing nodes, and an early warning mechanism is triggered when the utilization rate exceeds 80%.

[0075] Urban traffic and weather data: Relying on a perception network composed of floating car data (FCD) and roadside radar, the congestion index (0-10 levels, based on the ratio of average vehicle speed to free-flow vehicle speed) is calculated by dividing the road into segments (500 meters / segment); weather data is obtained by connecting to the meteorological department's API, focusing on extracting parameters such as precipitation intensity (mm / h) and wind force level, and establishing a mapping model with the cycling resistance coefficient (e.g., when precipitation intensity > 5 mm / h, the cycling speed correction coefficient is set to 0.7).

[0076] Spatiotemporal alignment of the collected data:

[0077] A timestamp synchronization mechanism (based on BeiDou time synchronization with an error of ≤100ms) is adopted to normalize the spatiotemporal coordinates of data from different sources (unified to the WGS-84 coordinate system) and solve the time offset problem caused by the underground signal blind zone of the subway (by back-calculating the interval running time through the entry / exit timestamp).

[0078] Traffic condition estimation is achieved based on an improved Kalman filter algorithm. For example, it integrates bus GPS positioning data with road congestion index to correct vehicle arrival time prediction (mean absolute error controlled within 30 seconds). A weighted voting mechanism (weights based on device historical reliability) is used to resolve conflicting data (such as the availability status reported by shared bicycles versus actual failed scanning).

[0079] In one embodiment of the present invention, based on step S2, the following will provide a possible embodiment and describe its specific implementation in a non-limiting manner.

[0080] First, a multimodal travel network graph is constructed, and a Dynamic Directed Weighted Graph (DDWG) model is used to achieve a unified representation of the multimodal transportation network. Specific methods include:

[0081] Refined definition of nodes and edges:

[0082] Node attributes: In addition to basic types (subway / bus station, shared bicycle station, etc.), spatiotemporal feature labels are added. For example, subway station nodes include platform location (3D coordinates) and entrance / exit distribution; transfer nodes are labeled with transfer passage length and escalator / elevator availability (real-time status is collected by IoT sensors).

[0083] Edge classification and weight calculation:

[0084] For edges with the same pattern, such as subway section edges, the weight is the dynamic running time (based on the real-time train schedule and section travel time model, the formula is: running time = basic section time × (1 + 0.1 × line congestion coefficient)); the weight of bus segment edges integrates real-time vehicle location and road condition data, and adopts a segmented cumulative calculation method (the time for each 500-meter segment is calculated separately and then summed).

[0085] Transfer edge: such as the walking transfer edge from subway to bus, the weight includes walking time (calculated based on path length and average pedestrian speed of 1.2m / s) and waiting time penalty (if the bus arrival interval is >10 minutes, the penalty coefficient increases by 0.3); the weight of the shared bicycle riding edge takes into account real-time road conditions (riding resistance coefficient) and weather effects (rainy day correction coefficient of 1.5), the formula is: riding time = distance / (base riding speed × weather correction coefficient × road condition coefficient).

[0086] A distributed graph database (such as Neo4j Cluster) is used to store the network topology. Secondary indexes are created for frequently accessed transfer nodes and popular road segments. Combined with a spatiotemporal partitioning strategy (storage is divided into zones according to city administrative districts), the efficiency of node query during path search is improved (response time <100ms).

[0087] The core advantage of Dynamic Directed Weighted Graph (DDWG) lies in its ability to respond in real time to changes in multi-source traffic data collected in step S1, achieving precise evolution of network topology through dynamic mapping of node attributes and adaptive calculation of edge weights. Its update logic is built on a closed-loop process of "data triggering - model calculation - graph structure correction," and the specific technical implementation is as follows:

[0088] S201. Dynamic update mechanism for node attributes.

[0089] The updating of node attributes is driven by the collected real-time status data and is achieved through the following technical path:

[0090] For transportation hub nodes (subway / bus stations):

[0091] Subway Nodes: Integrate real-time subway schedules and arrival prediction data, and add a "Current Platform Congestion" label to the node attributes (calculated based on the ratio of passenger flow to platform capacity in the AFC system, divided into low / medium / high levels); when a train delay occurs, automatically update the "Estimated First Train Arrival Time" field (accuracy ±30 seconds), and enhance visual differentiation with a red warning sign.

[0092] Bus stop node: Based on bus vehicle positioning data and electronic bus stop feedback, add "estimated number of waiting passengers" (inferred by infrared counter and vehicle occupancy rate) and "number of stops from the next bus" to the node attributes. When the number of stops is ≤2, start the countdown reminder (the data refresh frequency is synchronized with the bus positioning frequency, i.e., 1Hz).

[0093] For shared bicycle station nodes: Integrate shared bicycle availability data and parking compliance information, and dynamically update the "number of available vehicles / total capacity" ratio in the node attributes (refreshed every 30 seconds). When the ratio is <0.2, it is marked as "vehicle shortage". At the same time, a signal is triggered by the electronic fence, and "current number of available parking spaces" is added to the attributes. If it is 0, the coordinates of the backup parking spaces within 50 meters are automatically associated.

[0094] For transfer nodes: Based on the escalator / elevator operation status collected by IoT sensors, the "Vertical Transfer Facility Availability" (normal / faulty / under maintenance) is updated in real time in the node attributes. In the faulty state, the path weight of the backup pedestrian passage is automatically activated. The length of the transfer passage is dynamically corrected according to the real-time passenger flow density. When the density is >0.8 people / ㎡, the physical length is multiplied by a congestion coefficient of 1.2 (based on pedestrian dynamics model calibration).

[0095] S202. Real-time calculation and updating model of edge weights.

[0096] The dynamic adjustment of edge weights is directly related to the multi-source data collected in step S1. The specific mapping rules are as follows:

[0097] Same-pattern edge weight update:

[0098] Subway section edge: The basic section travel time is taken from the historical average travel time (stored by weekday / weekend / holiday time period). The congestion coefficient is linked in real time to the carriage load factor data (obtained through image recognition from onboard cameras). When the load factor is >80%, the congestion coefficient is 1.0; otherwise, it is 0.5 (the formula is modified to: travel time = basic section travel time × (1 + 0.1 × max(load factor / 80%, 1))). When a subway train delay is detected in step 2, the weight of the affected section edge is automatically multiplied by the delay coefficient (coefficient of 1.2 for delays within 5 minutes, coefficient of 1.5 for delays of 5-10 minutes).

[0099] Along bus routes: The time calculation for every 500-meter segment integrates three data points: ① Base travel time (based on free-flow speed); ② Road condition impact time (based on the city's congestion level, with an additional 1.5 times the base time during congestion); ③ Stop time (based on the number of people waiting at the current stop, with an additional 2 seconds per person). The segment time is recalculated every 30 seconds. When a sudden road event (such as a traffic accident) is detected in step 2, the trigger weight for the affected segment is temporarily doubled (coefficient 2.0 within the accident's impact range).

[0100] Changeover edge weight update:

[0101] For pedestrian transfers: the average speed of 1.2 m / s is still used as the basis for calculating walking time, but weather data correction is introduced: the speed is corrected to 1.0 m / s on rainy days and reduced to 0.8 m / s during heavy rain; the waiting time penalty coefficient is dynamically linked to the real-time bus arrival interval. When the bus arrival interval collected in step 2 is greater than 10 minutes, the penalty coefficient is increased by 0.3 on the original basis (i.e., the total penalty coefficient = base value + 0.3), and gradually decreases as the interval shortens (increases by 0.15 when the interval is 5-10 minutes).

[0102] Shared bike riding edge: The base riding speed is set to 5m / s. The road condition coefficient is linked to the urban road condition data from step 2 (coefficient 0.8 for main roads when congested, 0.9 for secondary roads, and 1.0 for side roads). The weather correction coefficient is dynamically adjusted according to the precipitation intensity: 1.2 for light rain (≤2mm / h), 1.5 for moderate rain (2-10mm / h), and 2.0 for heavy rain (>10mm / h) (synchronizing with the weather data sampling frequency of step 2). When step 2 detects that shared bikes are prohibited from parking on a certain road segment, the weight of the riding edge for that segment is automatically set to infinity (i.e., unavailable).

[0103] Update trigger mechanism:

[0104] Establish a data change threshold triggering mechanism: When the amount of data change collected in step 2 exceeds the preset threshold (such as bus location deviation > 50 meters, traffic congestion level jump, weather warning upgrade), the weight recalculation of the relevant edges is immediately triggered; when there is no significant change, periodic updates are adopted (5 minutes / time for subway edge, 1 minute / time for bus edge, and 30 seconds / time for shared bicycle edge).

[0105] All edge weight updates are implemented through a distributed lock mechanism to ensure that the weight values ​​read by the path search algorithm remain consistent during the incremental graph update process (the dynamic update mechanism in step 3) (avoiding calculation errors caused by read-write conflicts).

[0106] S203. Technical support for data linkage.

[0107] To achieve efficient linkage between the data collected in step S1 and the DDWG model, the system adopts the following technical architecture:

[0108] Message queue middleware: Establishes a data subscription-publishing mechanism through Kafka, with real-time data (such as bus location and traffic updates) as message producers and the DDWG model as consumers, keeping message transmission latency within 100ms.

[0109] Caching acceleration layer: Frequently accessed node attributes (such as subway congestion coefficient, bus arrival interval) are stored in the Redis cluster, using a two-dimensional "time-space" sharding strategy (divided by region + data type) to ensure millisecond-level data retrieval during weight calculation.

[0110] Consistency check: Every hour, the node / edge data of the DDWG model is checked for consistency with the original data source in step S1. When the deviation rate is greater than 5%, full synchronization is initiated to ensure the consistency between the graph model and the actual traffic conditions.

[0111] Through the above mechanism, the DDWG model can achieve a sub-second response to the data collected in step S1, improve the accuracy of edge weight updates, and provide high-precision basic data support for subsequent path planning algorithms.

[0112] In one embodiment of the present invention, based on step S3, the following will provide a possible embodiment and describe its specific implementation in a non-limiting manner.

[0113] Receive requests sent by the client and parse the start and end point information (such as coordinates) from them.

[0114] In one embodiment of the present invention, based on step S4, the following will provide a possible embodiment and describe its specific implementation in a non-limiting manner.

[0115] The path search is based on the improved multi-objective weighted Dijkstra (IMWD) algorithm, and the core technical details are as follows:

[0116] Heuristic information introduction: A dynamic heuristic function h(n) is added to the traditional Dijkstra algorithm to optimize heuristic information for different path objectives. For example, h(n) for the shortest time path is set as the straight-line distance from node n to the destination divided by the speed of the fastest mode of transportation; h(n) for the path with the fewest transfers is set as the lower limit of the estimated number of transfers to reduce invalid search branches.

[0117] Multi-objective optimization mechanism: Construct an objective function vector F=[f1,f2,f3,f4], where f1 is the total time (including travel, waiting, and transfers), f2 is the number of transfers, f3 is the total carbon emissions (calculated based on the unit carbon emission factor of different modes of transportation: subway 62gCO2 / person·km, bus 82gCO2 / person·km, shared bicycle 5gCO2 / person·km), and f4 is the walking distance. The multi-objective optimization is transformed into a single-objective optimization using a weighted summation method, with the weight coefficients w1-w4 dynamically adjusted according to user preferences (e.g., in the low-carbon priority path, the weight of w3 is increased to 0.4).

[0118] Path pruning strategy: Set Pareto optimal solution set filter to retain non-dominated solutions generated during the algorithm search process and remove dominated solutions (such as a path that is inferior to another path in both time and number of transfers). The final solution set size is controlled within 20 paths to balance computational efficiency and solution diversity.

[0119] In one embodiment of the present invention, based on step S5, a possible embodiment will be given below, and its specific implementation will be described in a non-limiting manner.

[0120] To achieve objective and quantitative evaluation of path solutions, the system constructs a multi-level indicator quantification system. Through standardization and dynamic weight allocation, it generates a comprehensive score that can be directly used for ranking. This system adopts a three-tiered technical framework of "indicator normalization - dynamic weight calibration - comprehensive score synthesis," and the specific implementation details are as follows:

[0121] S501. Indicator standardization processing mechanism.

[0122] (1) Calculation of total time spent

[0123] Data preprocessing: Historical travel time data for similar routes (same origin-destination, same travel time) is cleaned to remove outliers (such as data exceeding 3 standard deviations due to traffic accidents). The 95th percentile value is used to determine the reference travel time (denoted as T). ref This ensures that the maximum time consumption is covered in extreme scenarios.

[0124] Negative linear mapping model:

[0125] Scoring formula:

[0126] Among them, T actual The total travel time (including travel, waiting, and transfer times) is estimated in real time for the route. When T actual >T ref When this happens, St is forcibly assigned the value 0 to avoid excessive interference from extreme values ​​in the scoring.

[0127] Time granularity adaptation: The calculation accuracy is dynamically adjusted according to the total time spent on the path. For example, short-distance paths (<30 minutes) retain two decimal places, while long-distance paths (>1 hour) use integer scores to balance the requirements of calculation efficiency and accuracy.

[0128] (2) Calculation of travel expense score

[0129] Multi-mode cost integration model:

[0130] Subway fare: Based on distance-based tiered pricing formula C sub =a+b×L (a is the starting price, b is the price per unit distance, and L is the actual distance traveled), and the billing parameters are obtained in real time through the subway ticketing system API.

[0131] Bus fares: A fixed price is used based on distance, and the preset fare C is read according to the route type (regular bus / Bus Rapid Transit). bus .

[0132] Shared bike fees: calculated based on "basic fee + time-based fee": C bike =c+d×t (c is the base fee within 30 minutes, d is the fee per minute after the time limit, and t is the estimated riding time).

[0133] Total cost: C total =C sub +C bus +C bike .

[0134] Cost standardization:

[0135] Maximum reference fee C ref The scoring formula is to find the maximum cost among all solutions of the same path. .

[0136] When C total =0 (e.g., a purely walking path), force assignment of S. c =1 to reflect the cost advantage.

[0137] (3) Fuzzy evaluation model for transfer convenience

[0138] Quantification of indicator dimensions:

[0139] Transfer walking distance: Set piecewise function mapping score S d :

[0140]

[0141] Transfer waiting time: Based on real-time bus / subway arrival data, score S w calculate:

[0142]

[0143] Completeness of transfer facilities: using 0-1 variable identification, S when there is an elevator / escalator. f =1, otherwise S f =0.5 (The penalty coefficient was determined through user research).

[0144] The score is synthesized using a weighted average operator:

[0145] S t =0.4×S d +0.4×S w +0.2×S f The weights are determined by traffic engineering experts using the Delphi method to ensure they align with users' perceptions of transfer convenience.

[0146] (4) Calculation of real-time reliability score

[0147] Data Sample Construction: Collect a sample set of path time {T1,T2,...,T} for the same time period (e.g., weekday morning rush hour 7:30-8:30) within the past 30 days. n When the sample size is insufficient (n<50), the average fluctuation coefficient of the regional road network is introduced to supplement it.

[0148] Coefficient of variation method:

[0149] Calculate the sample standard deviation with the mean Scoring formula:

[0150]

[0151] when At that time, forced S r =0 indicates that the path time fluctuates wildly and has extremely poor reliability.

[0152] Real-time correction mechanism: The base score is dynamically calibrated by incorporating factors that cause sudden changes in current road conditions (such as adding a 0.2 weight for heavy rain and a 0.3 weight for road construction).

[0153] ,in The influence coefficients of each mutation factor are given.

[0154] S502. Dynamic weight allocation strategy.

[0155] Basic weight determination: A judgment matrix was constructed using the Analytic Hierarchy Process (AHP). Ten transportation experts were invited to conduct pairwise comparisons of four indicators (time travel, cost, transfer convenience, and reliability). After passing the consistency test (CR < 0.1), the basic weight vector W0 = [w t0 ,w c0 ,wt0 ,w r0 In typical scenarios, the basic weights are: time consumption (0.4), cost (0.2), transfer convenience (0.2), and reliability (0.2).

[0156] User preference calibration: A preference feature vector is constructed based on the user's historical selection behavior. For example, if the frequency of a user choosing the "least walking" route exceeds 60% in the past 30 days, a weight adjustment mechanism is triggered.

[0157] Extract user preference tags (such as "time-sensitive" and "cost-sensitive") and use collaborative filtering algorithms to match the weight distribution of similar user groups.

[0158] Dynamic weight calculation formula: Its W u For user personalization weights, α is the fusion coefficient (α=0.8 for new users, α=0.8 for active users).

[0159] α=0.3).

[0160] Adaptive adjustment for specific scenarios: The weights are automatically adjusted for special scenarios. For example, during heavy rain, the reliability weight is increased to 0.3, while the weight of cycling-related transfer convenience is reduced to 0.1. During nighttime hours (22:00-6:00), the implicit safety factor (indirectly reflected through the reliability index) is increased. The weight adjustment range is controlled within ±0.1 to avoid instability of the scoring system.

[0161] S503. Comprehensive scoring synthesis and application.

[0162] Overall score , where w i For dynamic weights, S i Standardize the scores for each indicator and round the results to two decimal places (accurate to 0.01).

[0163] Rating verification mechanism: Set a rating threshold range [0.3, 0.95]. Paths with a rating below 0.3 are automatically marked as "not recommended", and paths with a rating above 0.95 are labeled as "preferred" to ensure the practicality of the recommendation results.

[0164] Sorting Application: The system sorts the path schemes in descending order of comprehensive score. When the score difference is less than 0.03, a secondary sorting rule is activated (prioritizing the path with higher reliability) to avoid sorting distortion caused by close scores.

[0165] In some embodiments, the multimodal transportation-based route planning system may include multiple functional modules composed of computer program segments. The computer programs for each segment of the multimodal transportation-based route planning system may be stored in the memory of a computer device and executed by at least one processor to perform (see details). Figure 1 (Description) Functionality based on multimodal transportation route planning.

[0166] In this embodiment, the multimodal transportation-based route planning system can be divided into multiple functional modules according to its functions, such as... Figure 2 As shown. The module referred to in this invention is a series of computer program segments that can be executed by at least one processor and perform a fixed function, and is stored in memory. In this embodiment, the functions of each module will be described in detail in subsequent embodiments.

[0167] The data acquisition module is used to collect real-time data of multi-mode vehicles, including location and operating status.

[0168] The graph update module is used to update the node attribute values ​​and edge weight values ​​related to traffic status in the pre-constructed dynamic directed weighted graph based on the real-time data; wherein, the edge weight values ​​include at least travel time or cost; and the node attribute values ​​include at least the waiting time at transfer points.

[0169] The request receiving module is used to receive the start and end points of user input.

[0170] The path search module is used to search for K shortest paths between the starting point and the ending point as candidate paths based on the updated dynamic directed weighted graph using a path search algorithm.

[0171] The path filtering module is used to acquire user preference data, calculate the index score of candidate paths, and filter the optimal path from the candidate paths based on the preference data and index score.

[0172] Figure 3 The multimodal traffic-based route planning method provided in the embodiments of this application can be applied to devices. Those skilled in the art will understand that the device structures involved in the embodiments of this invention do not constitute a limitation on the device. A device may include more or fewer components than illustrated, or combine certain components, or have different component arrangements. In the embodiments of this invention, the device includes, but is not limited to, laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the embodiments of this application described and / or claimed herein.

[0173] The device 300 may include a processor 310, a memory 320, and a communication unit 330. These components communicate via one or more buses. Those skilled in the art will understand that the server structure shown in the figure does not constitute a limitation of the present invention. It may be a bus topology or a star topology, and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0174] The memory 320 can be used to store execution instructions of the processor 310. The memory 320 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. When the execution instructions in the memory 320 are executed by the processor 310, the device 300 is able to perform some or all of the steps in the above method embodiments.

[0175] The processor 310 serves as the control center of the storage device, connecting various parts of the electronic device via various interfaces and lines. It executes software programs and / or modules stored in the memory 320, and calls data stored in the memory to perform various functions of the electronic device and / or process data. The processor can be composed of integrated circuits (ICs), such as a single packaged IC or multiple packaged ICs with the same or different functions connected together. For example, the processor 310 may consist only of a central processing unit (CPU). In this embodiment of the invention, the CPU may have a single processing core or include multiple processing cores.

[0176] The communication unit 330 is used to establish a communication channel, enabling the storage device to communicate with other devices. It can receive user data sent by other devices or send user data to other devices.

[0177] The present invention also provides a computer storage medium, wherein the computer storage medium may store a program, which, when executed, may include some or all of the steps provided in the embodiments of the present invention. The storage medium may be a magnetic disk, an optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0178] Those skilled in the art will clearly understand that the techniques in the embodiments of the present invention can be implemented using software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solutions in the embodiments of the present invention, or the parts that contribute to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, or any other medium capable of storing program code. It includes several instructions to cause a computer device (which may be a personal computer, a server, or a second device, network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention.

[0179] The same or similar parts between the various embodiments in this specification can be referred to mutually. In particular, the device embodiments are basically similar to the method embodiments, so the description is relatively simple, and the relevant parts can be referred to the description in the method embodiments.

[0180] In the embodiments provided by this invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative. For instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between systems or modules may be electrical, mechanical, or other forms.

[0181] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; that is, they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0182] In addition, the functional modules in the various embodiments of the present invention can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0183] Although the present invention has been described in detail with reference to the accompanying drawings and preferred embodiments, the present invention is not limited thereto. Various equivalent modifications or substitutions can be made to the embodiments of the present invention by those skilled in the art without departing from the spirit and essence of the invention, and such modifications or substitutions should all be within the scope of the present invention. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should also be covered within the protection scope of the present invention.

Claims

1. A path planning method based on multi-modal transportation, characterized in that, include: Collect real-time data of multi-modal vehicles, including location and operating status; Based on the real-time data, update the node attribute values ​​and edge weight values ​​related to traffic status in the pre-constructed dynamic directed weighted graph; wherein, the edge weight values ​​include at least travel time or cost; and the node attribute values ​​include at least the waiting time at transfer points. Receive the start and end points input by the user; Using a path search algorithm, based on the updated dynamic directed weighted graph, search for K shortest paths between the starting point and the ending point as candidate paths; Obtain user preference data, calculate the indicator scores of candidate paths, and select the optimal path from the candidate paths based on the preference data and indicator scores; Methods for constructing dynamic directed weighted graphs include: Define the basic parameters for modeling, including: The node attributes include a basic type identifier and spatiotemporal feature labels. The spatiotemporal feature labels include: the three-dimensional coordinates of the transportation station, the entrance and exit distribution data, and the transfer passage length and real-time availability data of escalators or elevators of the transfer node. Edge classification and weight calculation include: Same pattern edge: The weight is dynamically calculated by real-time traffic data, where the weight of the subway section edge = basic section time × (1 + 0.1 × line congestion coefficient), and the weight of the bus segment edge is calculated by segmented accumulation. Transfer edge: The weight includes walking time and waiting time penalties, where the weight of the shared bicycle riding edge = distance / (base riding speed × weather correction factor × road condition factor); A distributed graph database is used to store the network topology, a secondary index is established for frequently accessed nodes and edges, and a spatiotemporal partitioning storage strategy is implemented. Based on the real-time data, update the node attribute values ​​and edge weight values ​​related to traffic state in the pre-constructed dynamic directed weighted graph, including: Node attributes are updated dynamically: For subway station nodes, based on real-time train schedules and arrival prediction data, update the platform congestion label and the expected first train arrival time field, and trigger a red warning sign when there is a delay; For bus stop nodes, based on vehicle positioning and electronic bus stop data, the estimated number of waiting passengers and the number of stops for the next bus to arrive are dynamically updated, and a countdown reminder is started when the number of stops is ≤2. For shared bicycle stations, the ratio of available bicycles to total capacity is updated every 30 seconds. When the ratio is less than 0.2, a bicycle shortage is marked, and the number of available parking spaces is updated based on the electronic fence signal. If the ratio is 0, the coordinates of backup parking points are associated. For transfer nodes, the availability status of vertical transfer facilities is updated in real time, and backup passages are activated in case of failure; the length of transfer passages is dynamically adjusted according to real-time passenger flow density. Real-time calculation and updating of edge weights: The weight of a subway section is calculated as follows: Basic section time × [1 + 0.1 × max(full load rate / 80%, 1)], plus the delay coefficient. The weight of a bus route segment is equal to the cumulative value of each segment. In the event of a road emergency, the weight of the segment where the incident occurred is multiplied. The weighting of pedestrian transfer sides incorporates weather correction coefficients and dynamic waiting penalty coefficients; The weight of shared bicycle riding is calculated as distance / [5m / s × road condition coefficient × weather correction coefficient]. The weight of no-parking zones is set to infinity. Acquire user preference data, calculate indicator scores for candidate paths, and select the optimal path from the candidate paths based on the preference data and indicator scores, including: The four core indicators of the route are normalized and transformed into scores. The four core indicators include total time, travel cost, transfer convenience and real-time reliability. The initial weights of each indicator are determined through expert evaluation, and preference features are identified based on users' historical behavior. Personalized weights are generated based on the preference features, and dynamic weights are generated based on the personalized weights and the initial weights. The standardized scores obtained from the normalized scores of each indicator are combined with the dynamic weights and summed to generate a comprehensive score. Candidate paths are sorted in descending order of comprehensive score. When the score difference is less than the set tolerance, a second sorting based on reliability indicators is activated.

2. The method of claim 1, wherein, Collect real-time data on multiple modes of transportation, including: The system collects real-time subway train schedule data and real-time location data through vehicle-to-ground wireless communication and smart card swiping recording system, and uses a sliding window to determine the departure interval based on the real-time location. The system uses a positioning chip to obtain the real-time location of the bus, and uses onboard sensors to obtain the real-time speed and direction angle of the bus. It also uses an infrared counter on an electronic bus stop sign to obtain the number of people queuing at the station. Vehicle status data of shared bicycles are obtained through a low-power wide-area network, including location and battery level. By relying on a perception network composed of floating car data and roadside radar, the congestion index of each road segment is obtained.

3. The method of claim 2, wherein, The method further includes: A timestamp synchronization mechanism is used to align the collected real-time data in time and space.

4. The method of claim 1, wherein, Using a path search algorithm, based on the updated dynamically directed weighted graph, K shortest paths between the starting point and the ending point are searched as candidate paths, including: The heuristic function is dynamically selected based on the route optimization objective: when the optimization objective is the shortest time, a heuristic function based on geographical straight-line distance and theoretical fastest traffic speed is used; when the optimization objective is the fewest transfers, a heuristic function based on the estimated minimum number of transfers is used. Simultaneously, the four path evaluation dimensions were optimized, including total path time, total number of transfer operations, total path carbon emissions, and total walking distance; the four path evaluation dimensions were combined into a comprehensive evaluation index according to user-configured weights. During the search process, Pareto non-dominated solutions are retained in real time, path schemes whose comprehensive evaluation index is surpassed are eliminated, and finally a specified number of candidate paths are output.

5. A multi-modal transportation based path planning system for implementing the steps of the multi-modal transportation based path planning method as claimed in any one of claims 1 to 4, wherein, include: The data acquisition module is used to collect real-time data of multi-mode vehicles, including location and operating status. The graph update module is used to update the node attribute values ​​and edge weight values ​​related to traffic status in the pre-constructed dynamic directed weighted graph based on the real-time data; wherein, the edge weight values ​​include at least travel time or cost; and the node attribute values ​​include at least the waiting time at transfer points. The request receiving module is used to receive the start and end points of user input. The path search module is used to search for K shortest paths between the starting point and the ending point as candidate paths based on the updated dynamic directed weighted graph using a path search algorithm. The path filtering module is used to acquire user preference data, calculate the index score of candidate paths, and filter the optimal path from the candidate paths based on the preference data and index score.

6. A path planning device based on multi-modal transportation, characterized by, include: Memory for storing route planning programs based on multimodal transportation; A processor, configured to implement the steps of the multimodal traffic-based route planning method as described in any one of claims 1-4 when executing the multimodal traffic-based route planning program.

7. A computer readable storage medium storing a computer program, characterized in that, The readable storage medium stores a multimodal traffic-based route planning program, which, when executed by a processor, implements the steps of the multimodal traffic-based route planning method as described in any one of claims 1-4.

Images