Intelligent electronic station board guide system based on state interaction and dynamic information fusion

Abstract

The application relates to the technical field of public transport management, in particular to an intelligent electronic bus stop guide system based on state interaction and dynamic information fusion. The method comprises the following steps: obtaining passenger flow data of a target area; analyzing the passenger flow data to determine the people flow distribution characteristics of the target area; generating a dynamic traffic guide containing line planning and route smoothness according to the people flow distribution characteristics; obtaining a user-initiated travel guide; analyzing the travel guide to determine the travel demand characteristics; and generating a personalized guide route according to the travel demand characteristics and the dynamic traffic guide. The method avoids the mechanical processing caused by insufficient demand integration. By analyzing the travel guide and determining the travel demand characteristics, the demand characteristics are more in line with the actual scene, the fluency of personalized guidance is improved, and the route rigidity caused by mechanical mapping is avoided. The personalized guide route is generated, the personalization and real-time performance of the guidance are improved, the user experience is enhanced, and the response delay caused by the fragmentation of information fusion is solved.

Description

Technical Field

[0001] This application relates to the field of public transportation management technology, and in particular to a smart electronic bus stop signage system based on the fusion of dynamic interaction and dynamic information. Background Technology

[0002] With the advancement of smart city construction, smart electronic bus stop systems are playing an increasingly important role in the tourism guidance field. These systems integrate traffic information to provide real-time navigation services.

[0003] However, existing electronic bus stop signage often relies on simple data collection and mechanical calculations, resulting in limited information and a lack of real-time updates and accuracy. This makes personalized guidance impossible. These issues lead to inefficient guidance, failing to meet the diverse needs of modern passengers and negatively impacting the visitor experience. Summary of the Invention

[0004] This application provides a smart electronic bus stop signage system based on the fusion of dynamic interaction and dynamic information to solve the above problems.

[0005] In a first aspect, this application provides a smart electronic bus stop signage system based on the fusion of dynamic interaction and dynamic information, the system comprising: Acquire passenger flow data for the target area; analyze the passenger flow data to determine the passenger flow distribution characteristics of the target area; Based on the aforementioned pedestrian flow distribution characteristics, a dynamic traffic guide is generated, which includes route planning and route accessibility. Obtain travel guides actively imported by users; analyze the travel guides to determine the characteristics of travel needs; Personalized guided tour routes are generated based on the aforementioned tourism demand characteristics and the aforementioned dynamic transportation guide.

[0006] This solution acquires passenger flow data for the target area, avoiding prediction errors due to incomplete data and supporting the determination of high-precision passenger flow distribution characteristics. Analyzing passenger flow data determines the distribution characteristics of the target area, ensuring real-time reflection of gathering areas and evacuation routes, providing reliable input for dynamic traffic guide generation, and reducing the risk of recommending congested routes. Based on passenger flow distribution characteristics, a dynamic traffic guide including route planning and route accessibility is generated, improving the guide's reliability and providing an optimization basis for personalized tour routes. User-initiated travel guides are acquired, ensuring user needs are directly captured and avoiding mechanical processing due to insufficient demand integration. Travel guides are analyzed to determine travel demand characteristics, making these characteristics more aligned with actual scenarios, improving the fluency of personalized tours, and avoiding rigid routes caused by mechanical mapping. Based on travel demand characteristics and dynamic traffic guides, personalized tour routes are generated, improving the personalization and real-time nature of the tours, enhancing user experience, and resolving response delays caused by fragmented information integration.

[0007] Optionally, obtaining passenger flow data for the target area includes: Obtain basic passenger flow data within the target area through the interface of a third-party platform; Analyze the basic passenger flow data to determine the data type; The data type is compared with the preset data item type to determine whether there are any missing data items; If it exists, determine the associated data source based on the missing data item; Based on the associated data source, customer data is estimated to generate predicted data for the missing data items; The predicted data and the basic passenger flow data are used as the passenger flow data for the target area.

[0008] This solution acquires basic passenger flow data within the target area through a third-party platform interface, ensuring the reliability and timeliness of the data source. The basic passenger flow data is analyzed to determine the data type, clarify the data structure, eliminate data ambiguity, and ensure accurate comparison. The data type is compared with preset data item types to determine if any data items are missing, revealing data incompleteness. If so, the relevant data source is identified based on the missing data item, enabling targeted handling of the missing items, avoiding blind data queries, and ensuring efficient use of estimation resources. Passenger flow data is estimated based on the relevant data source, generating predicted data for the missing data items, ensuring the overall usability of passenger flow data and providing seamless input for final data integration. The predicted data and basic passenger flow data are used as the passenger flow data for the target area, ensuring overall consistency and direct usability of the passenger flow data.

[0009] Optionally, analyzing the passenger flow data to determine the pedestrian distribution characteristics of the target area includes: Based on the time series, the passenger flow data is analyzed to determine the passenger data and the number of passengers alighting; Obtain traffic information within the target area, and determine several available modes of transportation and the operating hours of each available mode of transportation within the target area; Analyze the operating period, determine the period characteristics, and based on the period characteristics, perform time decay weighted calculation on the passenger data and alighting data to obtain the calculation results; The target area is spatially gridded to obtain a unit grid, and a pedestrian density prediction model is established based on the unit grid. Based on the pedestrian density prediction model and the calculation results, the pedestrian distribution characteristics of the target area are determined.

[0010] This solution analyzes passenger flow data based on time series to determine passenger and alighting data, ensuring the preservation of the time-series characteristics of the passenger flow data. It acquires traffic information within the target area, identifies several available transportation modes and their operating hours, ensuring that weighted calculations reflect the actual dynamics of traffic operations and avoid detachment from reality. By analyzing operating hours and determining their characteristics, time-decay weighted calculations are applied to passenger and alighting data based on these characteristics, enhancing the timeliness of the data representation, reducing interference from historical data on current predictions, and providing more accurate input data for the passenger flow density prediction model. The target area is spatially gridded to obtain unit grids, and a passenger flow density prediction model is established based on these unit grids, achieving spatial discretization of passenger flow data and providing a quantifiable grid foundation for analyzing passenger flow distribution. Based on the passenger flow density prediction model and calculation results, the passenger flow distribution characteristics of the target area are determined, providing real-time passenger flow distribution data for dynamic traffic guidance.

[0011] Optionally, establishing a pedestrian density prediction model based on the cell grid includes: Acquire historical pedestrian flow data; analyze the historical pedestrian flow data to determine the pedestrian flow transmission patterns among several available modes of transportation; Acquire meteorological data, analyze the meteorological data and the available modes of transportation, and determine weather interference factors; Based on the aforementioned patterns of human flow transmission and the aforementioned weather interference factors, a human flow density prediction model is established.

[0012] This solution acquires historical pedestrian flow data, eliminating the static issues caused by missing third-party data. Analyzing historical pedestrian flow data identifies the pedestrian flow transmission patterns among several available transportation modes, providing logical support for dynamic prediction. Meteorological data is acquired and analyzed, along with several available transportation modes, to identify weather interference factors, eliminating the impact of ignoring weather on pedestrian movement speed and improving adaptability to dynamic interference. Based on pedestrian flow transmission patterns and weather interference factors, a pedestrian density prediction model is established, overcoming the inability to adaptively handle dynamically coupled data and achieving high-precision generation of pedestrian flow distribution characteristics.

[0013] Optionally, generating a dynamic traffic guide that includes route planning and route accessibility based on the pedestrian flow distribution characteristics includes: Get the user's current location and initial target coordinates; Based on the user's current location and the initial target coordinates, several candidate routes are generated according to a path planning algorithm; Based on the pedestrian flow distribution characteristics, calculate the travel time correction coefficient for each candidate route; Based on the travel time correction factor, the route and travel time of each candidate route are adjusted to obtain a dynamic traffic guide.

[0014] This solution obtains the user's current location and initial target coordinates, providing a starting point and target point benchmark for navigation calculations. Based on the user's current location and initial target coordinates, a route planning algorithm generates several candidate routes, providing a basic route set for assessing the impact of pedestrian flow. According to pedestrian flow distribution characteristics, a travel time correction coefficient is calculated for each candidate route, reflecting the proportion of correction to the basic travel time based on the degree of pedestrian congestion, providing data for route adjustments. Based on the travel time correction coefficient, the route and travel time of each candidate route are adjusted to obtain dynamic traffic guidance, avoiding highly congested areas and providing real-time and reliable navigation suggestions.

[0015] Optionally, generating personalized tour routes based on the tourism demand characteristics and the dynamic transportation guide includes: Based on the aforementioned tourism demand characteristics, determine the preferred duration of stay at each target tourist destination; The order of sightseeing spots is dynamically adjusted based on the population density and the preferred length of stay. A personalized tour route is generated based on the order of visits to attractions and the operating hours.

[0016] This solution determines the preferred dwell time at each target attraction based on tourism demand characteristics, quantifies users' personalized needs, and avoids mechanically using fixed duration values, thereby improving the personalization of routes and user satisfaction. Based on visitor density and dwell time preferences, the order of attractions is dynamically adjusted to avoid peak hours, reduce user waiting time, prevent congestion and delays, and maximize overall tour efficiency. Personalized guided tour routes are generated based on the attraction visit order and operating hours, ensuring the routes are feasible in terms of time, easy for users to use, and achieving personalized and efficient guided tour services.

[0017] Optionally, the generation of personalized guided tour routes based on the order of attraction visits and the operating hours includes: Based on the order of visits to the attractions and the preferred length of stay, predict the dining locations; Obtain restaurant data within the preset range of the dining attractions; Analyze the restaurant data to determine the characteristics of the restaurants; Based on the characteristics of the restaurants, a recommendation list is generated, and recommendation feedback is received; Based on the recommended feedback, determine the target dining location; Based on the pedestrian density, predict the expected queuing time at the target dining location; A personalized tour route is generated based on the order of visits to the attractions, the operating hours, and the expected queuing time.

[0018] This solution predicts dining options based on the order of sightseeing and preferred duration of stay at attractions, ensuring seamless integration of dining planning with the sightseeing sequence and preventing itinerary interruptions or time conflicts due to improper dining arrangements. It acquires restaurant data within a preset range of dining attractions, providing foundational data support for restaurant characteristic analysis and recommendations, ensuring the generated recommendation list is geographically relevant and practical. Analyzing restaurant data identifies restaurant characteristics, making recommendations more accurate and improving the applicability of dining options. Based on restaurant characteristics, a recommendation list is generated, and feedback is received, enabling dynamic interaction between user preferences and recommendations, ensuring personalized recommendations, and providing input for further identifying target dining locations. Based on recommendation feedback, target dining locations are identified, providing precise input for queue time prediction and route generation, avoiding recommendation redundancy. Based on crowd density, the expected queue time at target dining locations is predicted, quantifying the impact of crowd congestion on dining time, providing time parameters for route optimization, and reducing the risk of actual waiting for users. Personalized guided routes are generated based on the order of sightseeing, operating hours, and expected queue time, improving overall practicality and real-time performance.

[0019] Optionally, determining the associated data source based on the missing data item includes: Analyze the missing data items to determine the missing attribute monitoring time; Based on the missing attributes and the pattern of pedestrian flow, the associated transportation modes are determined; Based on the missing monitoring time, the operational data of the associated transportation mode is obtained as the associated data source.

[0020] This solution analyzes missing data items, determines the monitoring time of missing attributes, and identifies the type and timing of missing data, providing fundamental input for determining related data sources and ensuring a clear starting point for data completion. Based on missing attributes and passenger flow patterns, it identifies related transportation modes and uses dynamic passenger flow transfer patterns to narrow down the data source range, providing direction for obtaining alternative data. Based on the missing monitoring time, it obtains operational data for related transportation modes as related data sources, ensuring that alternative operational data is provided to support estimations when missing data items occur.

[0021] Optionally, the step of analyzing the travel guide to determine the characteristics of travel demand includes: Natural language processing was performed on the travel guide to extract keywords and time descriptions; Based on the keywords, match the POI data within the target area to determine candidate tourist spots; Analyze the POI data to determine historical visit patterns; Based on the historical travel patterns and the time descriptions, the expected duration of stay for each candidate tourist destination is determined. Obtain user historical behavior data, and determine user browsing preferences based on the user historical behavior data; Based on the user's travel preferences and expected length of stay, tourism demand characteristics are generated.

[0022] This solution uses natural language processing to extract keywords and time descriptions from travel guides, ensuring that user intent is accurately captured and represented. Based on keywords, it matches POI data within the target area to identify candidate tourist attractions, mapping user interests in the guide to actual geographical locations and providing a candidate set for attraction selection. It analyzes POI data to determine historical travel patterns, providing data support for predicting dwell time and ensuring that patterns are driven by historical data. Based on historical travel patterns and time descriptions, it determines the expected dwell time for each candidate tourist attraction, ensuring that the dwell time reflects historical patterns while meeting user time constraints, providing time parameters for demand characteristics. It acquires historical user behavior data and, based on this data, determines user travel preferences, personalizing demand characteristics and reflecting historical behavior patterns. Finally, it generates travel demand characteristics based on user travel preferences and expected dwell time, ensuring complete expression of demand information.

[0023] Optionally, predicting the expected queuing time at the target dining location based on the crowd density includes: Analyze the target dining locations to determine the characteristics of the food consumed; Based on the characteristics of the food and the population density, the expected queuing time at the target dining location is predicted.

[0024] This solution analyzes target dining locations and identifies the characteristics of their food offerings, ensuring accurate identification of these characteristics to provide a basis for predicting queue times. Based on the food characteristics and pedestrian density, it predicts the expected queue times for target dining locations, reflecting the impact of pedestrian density on actual queue conditions and providing time parameters for optimizing guided tour routes. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 This is a schematic diagram illustrating an application scenario provided in one embodiment of this application; Figure 2 A flowchart illustrating a smart electronic bus stop signage system based on state interaction and dynamic information fusion, provided as an embodiment of this application. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0028] Furthermore, the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article, unless otherwise specified, generally indicates that the preceding and following related objects have an "or" relationship.

[0029] The embodiments of this application will now be described in further detail with reference to the accompanying drawings.

[0030] Existing electronic bus stop signage often relies on simple data collection and mechanical calculations, resulting in limited information, a lack of real-time updates and accuracy, and an inability to provide personalized guidance. These issues lead to inefficient guidance, failing to meet the diverse needs of modern passengers and negatively impacting the visitor experience.

[0031] Based on this, this application provides a smart electronic bus stop sign navigation system based on dynamic interaction and dynamic information fusion. It acquires passenger flow data for the target area, avoiding prediction errors due to incomplete data, thus supporting the determination of high-precision passenger flow distribution characteristics. Analyzing passenger flow data determines the passenger flow distribution characteristics of the target area, ensuring real-time reflection of gathering areas and evacuation routes, providing reliable input for dynamic traffic guide generation, and reducing the risk of recommending congested routes. Based on passenger flow distribution characteristics, it generates dynamic traffic guides including route planning and route smoothness, improving the reliability of the guides and providing an optimization basis for personalized navigation routes. It acquires user-initiated travel guides, ensuring that user needs are directly captured, avoiding mechanical processing caused by insufficient demand integration. It analyzes travel guides to determine travel demand characteristics, making these characteristics more aligned with actual scenarios, improving the fluency of personalized navigation, and avoiding route rigidity caused by mechanical mapping. Based on travel demand characteristics and dynamic traffic guides, it generates personalized navigation routes, improving the personalization and real-time nature of the navigation, enhancing user experience, and solving response delays caused by fragmented information fusion.

[0032] Figure 1 This is a schematic diagram illustrating an application scenario provided by this application, showing the application of the method provided in this application when providing smart electronic bus stop guidance.

[0033] Specifically, the method provided in this application can be applied to any server, where the server interacts with a third-party platform to obtain real-time passenger flow data for the target area from the third-party platform. The passenger flow data is analyzed to determine the distribution characteristics of pedestrian traffic in the target area. Based on these characteristics, a dynamic traffic guide including route planning and route accessibility is generated. User-initiated travel guides are acquired to ensure that user needs are directly captured, avoiding mechanical processing due to insufficient demand integration. The travel guides are parsed to determine tourism demand characteristics, making these characteristics more aligned with actual scenarios, improving the fluency of personalized guides, and avoiding rigid routes caused by mechanical mapping. Based on the tourism demand characteristics and the dynamic traffic guide, personalized guide routes are generated, enhancing the personalization and real-time nature of the guides, improving user experience, and resolving response delays caused by fragmented information fusion. Specific implementation methods can be found in the following embodiments.

[0034] Figure 2 This is a flowchart illustrating a smart electronic bus stop sign navigation system based on state interaction and dynamic information fusion, provided as an embodiment of this application. The method of this embodiment can be applied to the server in the above scenario. Figure 2As shown, the system includes: S201. Obtain passenger flow data for the target area; analyze the passenger flow data to determine the passenger flow distribution characteristics of the target area; The target area can be a designated space area for city tourism and guided tours.

[0035] Passenger flow data can be raw information on the movement of people.

[0036] The characteristics of pedestrian flow distribution can be a quantitative representation of the spatial distribution of people within a target area.

[0037] Specifically, passenger flow data for the target area is obtained in real time by calling the API interface of a third-party platform. The passenger flow data is then input into a dynamic information fusion module for cleaning. Next, the cleaned passenger flow data is input into a pedestrian density prediction model based on an LSTM neural network to determine a spatial gridded heat map of the target area. Finally, features are extracted from the spatial gridded heat map to form pedestrian distribution characteristics.

[0038] S202. Based on the characteristics of pedestrian flow distribution, generate a dynamic traffic guide that includes route planning and route smoothness. Route planning can be a recommended sequence of tourist routes.

[0039] Route accessibility can be an indicator of how smoothly a route is passable.

[0040] Dynamic traffic guides can be navigation solutions that include route planning and route accessibility.

[0041] Specifically, based on the target area map data obtained from OpenStreetMap, parameters such as pedestrian density and clustering areas in the pedestrian distribution characteristics are combined; then, a path planning algorithm is used to calculate the shortest path, and the smoothness of traffic flow is evaluated in combination with the clustering areas; furthermore, real-time traffic data from third-party traffic API calls are combined to generate a dynamic traffic guide that includes route planning and route smoothness.

[0042] S203. Obtain travel guides actively imported by users; analyze the travel guides to determine the characteristics of travel needs; Travel guides can be unstructured descriptions of preferences and plans in text or document form.

[0043] Tourism demand characteristics can be structured demand data such as a list of attractions and length of stay.

[0044] Specifically, user-initiated travel guides are retrieved through the interactive interface of the smart electronic bus stop sign. Natural language processing technology is used to segment the travel guides into words and extract different entities; then, they are matched against a POI database to identify points of interest, and dynamically adjusted based on user historical behavior data to determine the characteristics of travel demand.

[0045] S204. Generate personalized tour routes based on tourism demand characteristics and dynamic transportation guides.

[0046] Personalized guided tours can be customized tour sequences generated by combining tourism demand characteristics and dynamic transportation guides.

[0047] Specifically, by integrating tourism demand characteristics with dynamic transportation guides, applying optimization algorithms to adjust route order and duration, and integrating derived needs based on POI data matching, personalized guided routes with text and map visualizations are generated.

[0048] This solution acquires passenger flow data for the target area, avoiding prediction errors due to incomplete data and supporting the determination of high-precision passenger flow distribution characteristics. Analyzing passenger flow data determines the distribution characteristics of the target area, ensuring real-time reflection of gathering areas and evacuation routes, providing reliable input for dynamic traffic guide generation, and reducing the risk of recommending congested routes. Based on passenger flow distribution characteristics, a dynamic traffic guide including route planning and route accessibility is generated, improving the guide's reliability and providing an optimization basis for personalized tour routes. User-initiated travel guides are acquired, ensuring user needs are directly captured and avoiding mechanical processing due to insufficient demand integration. Travel guides are analyzed to determine travel demand characteristics, making these characteristics more aligned with actual scenarios, improving the fluency of personalized tours, and avoiding rigid routes caused by mechanical mapping. Based on travel demand characteristics and dynamic traffic guides, personalized tour routes are generated, improving the personalization and real-time nature of the tours, enhancing user experience, and resolving response delays caused by fragmented information integration.

[0049] In some embodiments, basic passenger flow data within the target area is obtained through the interface of a third-party platform; the basic passenger flow data is parsed to determine the data type; the data type is compared with the preset data item type to determine whether there are any missing data items; if so, the associated data source is determined based on the missing data item; passenger flow data is estimated based on the associated data source to generate predicted data for the missing data item; the predicted data and basic passenger flow data are used as passenger flow data for the target area.

[0050] Third-party platforms can be external data sources that provide basic passenger flow data.

[0051] The interface can be a communication protocol or access channel used to obtain basic passenger flow data from a third-party platform.

[0052] Basic passenger flow data can be a collection of raw passenger flow information.

[0053] The data type can be the attribute category of each field in the basic passenger flow data.

[0054] The preset data item type can be a pre-defined set of necessary data items that should be included in complete passenger flow data. It is pre-stored in the server and retrieved when needed.

[0055] Missing data items can be preset data item types that are not included in the basic passenger flow data.

[0056] The associated data source can be a supplementary data provider corresponding to the missing data item.

[0057] Predictive data can be generated by estimating the value of missing data items by correlating with data sources.

[0058] Specifically, the process involves calling an API provided by a third-party platform, inputting the target area, and sending a data request. Upon receiving the API response, basic passenger flow data is received. The basic passenger flow data is then parsed to extract data fields. A data processing library is used to identify the data type of each field. Next, the statistical characteristics of historical passenger flow data and preset data item types based on industry-standard data are read. The data types are compared item by item with the preset data item types to determine if any data items are missing. If any are missing, a correlation mapping table built based on empirical association rules between different data items in domain knowledge and external data sources is queried to determine the relevant data source. The interface of the relevant data source is accessed, and a simple estimation model is used to process the raw data provided by the relevant data source to generate predicted data for the missing data items. Finally, the basic passenger flow data and the predicted data are merged to form a complete passenger flow dataset, which is stored in a structured format as the passenger flow data for the target area.

[0059] This solution acquires basic passenger flow data within the target area through a third-party platform interface, ensuring the reliability and timeliness of the data source. The basic passenger flow data is analyzed to determine the data type, clarify the data structure, eliminate data ambiguity, and ensure accurate comparison. The data type is compared with preset data item types to determine if any data items are missing, revealing data incompleteness. If so, the relevant data source is identified based on the missing data item, enabling targeted handling of the missing items, avoiding blind data queries, and ensuring efficient use of estimation resources. Passenger flow data is estimated based on the relevant data source, generating predicted data for the missing data items, ensuring the overall usability of passenger flow data and providing seamless input for final data integration. The predicted data and basic passenger flow data are used as the passenger flow data for the target area, ensuring overall consistency and direct usability of the passenger flow data.

[0060] In some embodiments, based on time series, passenger flow data is parsed to determine passenger data and alighting data; traffic information within the target area is acquired to determine several available modes of transportation and the operating hours of each available mode of transportation; the operating hours are analyzed to determine the time period characteristics, and based on the time period characteristics, time decay weighted calculations are performed on passenger data and alighting data to obtain calculation results; the target area is spatially gridded to obtain unit grids, and a passenger flow density prediction model is established based on the unit grids; based on the passenger flow density prediction model and calculation results, the passenger flow distribution characteristics of the target area are determined.

[0061] Time series can be the property of data being arranged in chronological order.

[0062] Passenger data can be structured information about the number of people boarding the vehicle.

[0063] Passenger disembarkation data can be structured information about the number of people getting off the bus.

[0064] Traffic information can include a list of modes of transportation and their operating hours within the target area.

[0065] Available modes of transportation can be a list of available transportation types within the target area.

[0066] Operating hours can be the start and end times and intervals for each available mode of transportation.

[0067] Time-period characteristics can be based on the time periods divided by operating hours and the average passenger flow density characteristics.

[0068] The calculation results can be weighted passenger data and disembarkation data datasets.

[0069] A unit grid can be a geographic grid unit of regular size that is divided into target areas through spatial gridding.

[0070] The pedestrian density prediction model can be a machine learning regression model used to predict the pedestrian density value for each cell grid.

[0071] Specifically, time-series information is extracted from passenger flow data. Then, a data processing library is used to parse the passenger flow data, separating the timestamp column and sorting it by timestamp. Next, the "number of passengers boarding" field is identified and extracted as passenger data, and the "number of passengers alighting" field is identified and extracted as alighting data. Subsequently, a query request is sent to the traffic information equipment in the target area, and traffic information is received. Then, the traffic information is parsed to determine the list of available transportation modes, and several available transportation modes and their operating hours for each mode are extracted from the target area. The operating hours are then analyzed, dividing the day into multiple time periods, and the average passenger flow density for each time period is calculated based on historical data as a time period feature. Then, a time decay weighting function is applied to weight each data point in the passenger and alighting data, calculating the difference between the timestamp of each data point and the current reference time, and substituting this difference into the decay function to obtain a weight value. Then, based on the weight values, weighted passenger and alighting data are generated. All weighted data are then summarized to determine the calculation result. Subsequently, based on the boundary coordinates of the target area, geospatial processing tools are used to spatially grid the target area, generating regular-sized cell grids. Then, a pedestrian density prediction model is constructed based on these cell grids. The calculation results are then input into the pedestrian density prediction model to perform predictions, generating predicted pedestrian density values ​​for each cell grid. Finally, based on the prediction results, the pedestrian distribution characteristics of the target area are determined.

[0072] This solution analyzes passenger flow data based on time series to determine passenger and alighting data, ensuring the preservation of the time-series characteristics of the passenger flow data. It acquires traffic information within the target area, identifies several available transportation modes and their operating hours, ensuring that weighted calculations reflect the actual dynamics of traffic operations and avoid detachment from reality. By analyzing operating hours and determining their characteristics, time-decay weighted calculations are applied to passenger and alighting data based on these characteristics, enhancing the timeliness of the data representation, reducing interference from historical data on current predictions, and providing more accurate input data for the passenger flow density prediction model. The target area is spatially gridded to obtain unit grids, and a passenger flow density prediction model is established based on these unit grids, achieving spatial discretization of passenger flow data and providing a quantifiable grid foundation for analyzing passenger flow distribution. Based on the passenger flow density prediction model and calculation results, the passenger flow distribution characteristics of the target area are determined, providing real-time passenger flow distribution data for dynamic traffic guidance.

[0073] In some embodiments, historical pedestrian flow data is acquired; the historical pedestrian flow data is analyzed to determine the pedestrian flow transmission pattern among several available modes of transportation; meteorological data is acquired, and the meteorological data and several available modes of transportation are analyzed to determine weather interference factors; based on the pedestrian flow transmission pattern and weather interference factors, a pedestrian density prediction model is established.

[0074] Historical pedestrian traffic data can be historical passenger flow records for the target area.

[0075] The pattern of passenger flow transmission can be described as the proportional relationship of passenger flow transfer among several available modes of transportation.

[0076] Meteorological data can include weather-related information such as temperature, precipitation, and wind speed.

[0077] Weather interference factor can be a quantitative parameter representing the impact of weather on passenger flow in transportation modes.

[0078] Specifically, historical pedestrian flow data for the target area is retrieved from historical data storage devices. Then, the transportation mode field is extracted from the historical pedestrian flow data; subsequently, the passenger flow change for each available transportation mode in adjacent time periods is statistically analyzed; then, based on the statistical results, the passenger flow transfer relationship between transportation modes is deduced, and the pedestrian flow transmission pattern among several available transportation modes is determined. Next, a query request is sent to an external meteorological service interface to receive meteorological data for the target area; then, the timestamps of the meteorological data are matched with the timestamps of the historical pedestrian flow data; then, the passenger flow deviation value for each available transportation mode under different weather conditions is calculated; then, based on the passenger flow deviation value, a parameter representing the intensity of weather impact is generated and determined as a weather interference factor. Subsequently, based on the spatial division of unit grids, the historical pedestrian flow data of each unit grid is associated; finally, based on the pedestrian flow transmission pattern and the weather interference factor, a regression algorithm is used to train the model, outputting the predicted pedestrian density value for each unit grid, thereby constructing a pedestrian density prediction model.

[0079] This solution acquires historical pedestrian flow data, eliminating the static issues caused by missing third-party data. Analyzing historical pedestrian flow data identifies the pedestrian flow transmission patterns among several available transportation modes, providing logical support for dynamic prediction. Meteorological data is acquired and analyzed, along with several available transportation modes, to identify weather interference factors, eliminating the impact of ignoring weather on pedestrian movement speed and improving adaptability to dynamic interference. Based on pedestrian flow transmission patterns and weather interference factors, a pedestrian density prediction model is established, overcoming the inability to adaptively handle dynamically coupled data and achieving high-precision generation of pedestrian flow distribution characteristics.

[0080] In some embodiments, the user's current location and initial target coordinates are obtained; based on the user's current location and initial target coordinates, several candidate routes are generated according to a path planning algorithm; based on the characteristics of pedestrian flow distribution, a travel time correction coefficient for each candidate route is calculated; based on the travel time correction coefficient, the route and travel time of each candidate route are adjusted to obtain a dynamic traffic guide.

[0081] The user's current location can be the user's geographic coordinates.

[0082] The initial target coordinates can be the geographic coordinates of the target destination.

[0083] A path planning algorithm can be an algorithm used to calculate a feasible path from the current location to the target location.

[0084] The candidate route can be multiple feasible path options.

[0085] The travel time correction factor can be a numerical factor used to correct the estimated base travel time of candidate routes.

[0086] A route can be a sequence of paths or a transportation route.

[0087] The travel time can be the estimated travel time.

[0088] Specifically, the system collects the user's current location in real time via the mobile device's GPS; simultaneously, it obtains initial target coordinates by parsing destination information from imported travel guides. The user's current location coordinates and initial target coordinates are input into the path planning algorithm; then, an initial path search is performed, considering shortest distance or time priority principles to generate several candidate routes. Next, the cell grid sequences traversed by these candidate routes are extracted; subsequently, based on the pedestrian flow distribution characteristics of each cell grid, a traffic speed attenuation factor is mapped; then, combining the route length and several available transportation modes, a travel time correction coefficient for each candidate route is calculated. Furthermore, based on the travel time correction coefficient, if the travel time correction coefficient indicates that a candidate route is highly congested, the route is replanned; simultaneously, the estimated base travel time for each candidate route is multiplied by the travel time correction coefficient to obtain the corrected travel time; finally, the adjusted candidate routes are integrated to generate a dynamic traffic guide.

[0089] This solution obtains the user's current location and initial target coordinates, providing a starting point and target point benchmark for navigation calculations. Based on the user's current location and initial target coordinates, a route planning algorithm generates several candidate routes, providing a basic route set for assessing the impact of pedestrian flow. According to pedestrian flow distribution characteristics, a travel time correction coefficient is calculated for each candidate route, reflecting the proportion of correction to the basic travel time based on the degree of pedestrian congestion, providing data for route adjustments. Based on the travel time correction coefficient, the route and travel time of each candidate route are adjusted to obtain dynamic traffic guidance, avoiding highly congested areas and providing real-time and reliable navigation suggestions.

[0090] In some embodiments, the preferred duration of stay at each target attraction is determined based on tourism demand characteristics; the order of attractions is dynamically adjusted based on visitor density and duration of stay preferences; and personalized guided tour routes are generated based on the order of attractions and operating hours.

[0091] The target tourist destination can be a specific attraction or point of interest extracted from the travel guide imported by the user.

[0092] Dwell duration preference can be a numerical preference value assigned to each target tourist destination based on tourism demand characteristics.

[0093] Pedestrian density can be a unit grid density value representing the degree of pedestrian congestion in different areas.

[0094] The order of sightseeing can be a sequence of target attractions.

[0095] Specifically, based on tourism demand characteristics, natural language processing (NLP) technology is used to analyze travel guide texts, identify keywords related to dwell time, and map these keywords to target attractions. Then, based on the mapping results, the preferred dwell time for each target attraction is determined. Furthermore, based on crowd density, the potential waiting time for each target attraction is calculated, and combined with dwell time preferences, a new sequence of attractions is generated. If a target attraction experiences high crowd density, the attraction sequence is dynamically adjusted. Finally, the planned arrival times in the attraction sequence are checked to ensure they fall within operating hours; if conflicts occur, the sequence is fine-tuned to generate a personalized guided tour route.

[0096] This solution determines the preferred dwell time at each target attraction based on tourism demand characteristics, quantifies users' personalized needs, and avoids mechanically using fixed duration values, thereby improving the personalization of routes and user satisfaction. Based on visitor density and dwell time preferences, the order of attractions is dynamically adjusted to avoid peak hours, reduce user waiting time, prevent congestion and delays, and maximize overall tour efficiency. Personalized guided tour routes are generated based on the attraction visit order and operating hours, ensuring the routes are feasible in terms of time, easy for users to use, and achieving personalized and efficient guided tour services.

[0097] In some embodiments, the system predicts dining locations based on the order of visits to attractions and preferred length of stay; obtains catering data within a preset range of dining locations; analyzes the catering data to determine catering characteristics; generates a recommendation list based on catering characteristics and receives recommendation feedback; determines target dining locations based on recommendation feedback; predicts the expected queuing time for target dining locations based on crowd density; and generates personalized guided tour routes based on the order of visits to attractions, operating hours, and expected queuing time.

[0098] Dining spots can be candidate dining locations.

[0099] The preset range can be a preset geographical distance range. It is stored in advance on the server and invoked when needed.

[0100] Catering data can be basic information about restaurants within a preset range.

[0101] Catering characteristics can be classified by analyzing catering data.

[0102] The recommended list can be a sorted list of restaurants.

[0103] Recommendation feedback can be input data provided by users to the recommendation list through an interactive interface.

[0104] The target dining location can be the final dining location determined based on recommendation feedback.

[0105] The expected queuing time can be a predicted numerical queuing wait time.

[0106] Specifically, based on dwell time preferences, the estimated end time of each target tourist spot is calculated, and time gaps are identified within the tourist attraction sequence to predict suitable candidate tourist spots for dining. Then, based on the geographical coordinates of the dining spots, restaurant data within a preset range is queuing. Furthermore, the menus, service descriptions, and other descriptive information in the restaurant data are analyzed to identify and determine the characteristics of the restaurants. Subsequently, the restaurants are categorized and sorted according to their characteristics, and a recommendation list is presented through a user interface; simultaneously, user feedback is received through the recommendation list. Then, user feedback is analyzed and mapped to corresponding items in the recommendation list to determine the target dining spots. Next, based on the pedestrian density corresponding to the location of the target dining spot, a density-duration mapping table is established, and a queuing model algorithm is applied to predict the expected queuing time for the target dining spot. Then, the arrival time of each target tourist spot in the tourist attraction sequence is verified and adjusted to confirm it is within the operating hours; finally, the expected queuing time is input into the dwell time of the target dining spot; thus, the final personalized guided tour route is output.

[0107] This solution predicts dining options based on the order of sightseeing and preferred duration of stay at attractions, ensuring seamless integration of dining planning with the sightseeing sequence and preventing itinerary interruptions or time conflicts due to improper dining arrangements. It acquires restaurant data within a preset range of dining attractions, providing foundational data support for restaurant characteristic analysis and recommendations, ensuring the generated recommendation list is geographically relevant and practical. Analyzing restaurant data identifies restaurant characteristics, making recommendations more accurate and improving the applicability of dining options. Based on restaurant characteristics, a recommendation list is generated, and feedback is received, enabling dynamic interaction between user preferences and recommendations, ensuring personalized recommendations, and providing input for further identifying target dining locations. Based on recommendation feedback, target dining locations are identified, providing precise input for queue time prediction and route generation, avoiding recommendation redundancy. Based on crowd density, the expected queue time at target dining locations is predicted, quantifying the impact of crowd congestion on dining time, providing time parameters for route optimization, and reducing the risk of actual waiting for users. Personalized guided routes are generated based on the order of sightseeing, operating hours, and expected queue time, improving overall practicality and real-time performance.

[0108] In some embodiments, missing data items are parsed to determine the missing monitoring time of the missing attributes; based on the missing attributes and the pattern of pedestrian flow, the associated transportation modes are determined; and based on the missing monitoring time, the operational data of the associated transportation modes are obtained as the associated data source.

[0109] Missing attributes can be specific attributes of missing data items.

[0110] Missing data monitoring time can be the point in time when the missing data item occurs.

[0111] The associated mode of transportation can be the mode of transportation related to the missing attribute.

[0112] Operational data can be operational data related to different modes of transportation.

[0113] Specifically, the missing data items are parsed using a log analyzer to extract the missing attributes and determine the occurrence time of the missing data items as the missing monitoring time. Then, based on the missing attributes, a pre-defined rule base is queried to determine pedestrian flow patterns. These patterns are then used to match and generate traffic modes related to the missing attributes as associated traffic modes. Finally, using the missing monitoring time as a query parameter, a data query request is sent to the interface of the associated traffic modes to retrieve their operational data, which is then used as the associated data source.

[0114] This solution analyzes missing data items, determines the monitoring time of missing attributes, and identifies the type and timing of missing data, providing fundamental input for determining related data sources and ensuring a clear starting point for data completion. Based on missing attributes and passenger flow patterns, it identifies related transportation modes and uses dynamic passenger flow transfer patterns to narrow down the data source range, providing direction for obtaining alternative data. Based on the missing monitoring time, it obtains operational data for related transportation modes as related data sources, ensuring that alternative operational data is provided to support estimations when missing data items occur.

[0115] In some embodiments, natural language processing is performed on travel guides to extract keywords and time descriptions; based on the keywords, POI data within the target area is matched to determine candidate tourist attractions; POI data is parsed to determine historical travel patterns; based on historical travel patterns and time descriptions, the expected duration of stay for each candidate tourist attraction is determined; user historical behavior data is obtained, and user travel preferences are determined based on the user historical behavior data; based on user travel preferences and expected duration of stay, travel demand characteristics are generated.

[0116] Keywords can be vocabulary units extracted from travel guides.

[0117] The time description can be a time expression string identified from travel guides.

[0118] POI data can be database entries for points of interest within the target area.

[0119] Candidate tourist spots can be a set of potential tourist spots.

[0120] Historical visit patterns can be statistical indicators calculated from historical visit records.

[0121] The expected duration of stay can be the time spent at each candidate tourist destination.

[0122] User history data can include a user's past browsing history and preference settings.

[0123] User browsing preferences can be derived from the preferences identified through analysis of users' historical behavior data.

[0124] Specifically, natural language processing is used to extract keywords and time descriptions from travel guides: First, word segmentation is performed to divide the text into independent lexical units; then, part-of-speech tagging technology is applied to identify nouns, verbs, and other parts of speech, and keywords are extracted based on named entity recognition; simultaneously, time descriptions are identified through a time expression parsing model. String matching algorithms are used to match keywords with POI data within the target area; then, successfully matched POI data is filtered to determine candidate tourist spots. Next, the POI data is accessed, and the historical dataset for each POI is parsed; furthermore, statistical indicators of the historical dataset are calculated to determine historical travel patterns. Then, based on each candidate tourist spot, combined with historical travel patterns and time descriptions, the expected duration of stay is determined: if the time description indicates a shorter stay, the duration is reduced; if the time description indicates a longer stay, the duration is increased, thus determining the expected duration of stay for each candidate tourist spot. Next, user historical behavior data is retrieved from the user database; then, the user historical behavior data is analyzed to identify preference features, and a rule engine is used to classify and output user travel preferences. Finally, the user browsing preferences and expected duration of stay assigned to each candidate tourist spot are integrated to generate travel demand features.

[0125] This solution uses natural language processing to extract keywords and time descriptions from travel guides, ensuring that user intent is accurately captured and represented. Based on keywords, it matches POI data within the target area to identify candidate tourist attractions, mapping user interests in the guide to actual geographical locations and providing a candidate set for attraction selection. It analyzes POI data to determine historical travel patterns, providing data support for predicting dwell time and ensuring that patterns are driven by historical data. Based on historical travel patterns and time descriptions, it determines the expected dwell time for each candidate tourist attraction, ensuring that the dwell time reflects historical patterns while meeting user time constraints, providing time parameters for demand characteristics. It acquires historical user behavior data and, based on this data, determines user travel preferences, personalizing demand characteristics and reflecting historical behavior patterns. Finally, it generates travel demand characteristics based on user travel preferences and expected dwell time, ensuring complete expression of demand information.

[0126] In some embodiments, the target dining location is analyzed to determine the characteristics of the food consumed; based on the characteristics of the food consumed and the density of people, the expected queuing time at the target dining location is predicted.

[0127] The characteristics of food consumption can be classified according to the way the food is eaten at the target restaurant.

[0128] Specifically, through a database query interface, the unique identifier of the target restaurant is input, and the "Food Characteristics" field of the POI data entry is retrieved to determine the food characteristics. A pre-stored mapping table based on historical third-party review data is queried to map the food characteristics to the corresponding baseline queue time. Based on the crowd density value, an adjustment factor is calculated using predefined adjustment rules based on historical crowd density and queue time. Finally, based on the baseline queue time and the adjustment factor, the expected queue time for the target restaurant is generated.

[0129] This solution analyzes target dining locations and identifies the characteristics of their food offerings, ensuring accurate identification of these characteristics to provide a basis for predicting queue times. Based on the food characteristics and pedestrian density, it predicts the expected queue times for target dining locations, reflecting the impact of pedestrian density on actual queue conditions and providing time parameters for optimizing guided tour routes.

Claims

1. A smart electronic bus stop signage system based on dynamic interaction and dynamic information fusion, characterized in that, include: Obtain passenger flow data for the target area; Analyze the passenger flow data to determine the pedestrian distribution characteristics of the target area; Based on the aforementioned pedestrian flow distribution characteristics, a dynamic traffic guide is generated, which includes route planning and route accessibility. Obtain travel guides actively imported by users; analyze the travel guides to determine the characteristics of travel needs; Based on the aforementioned tourism demand characteristics and the aforementioned dynamic transportation guide, generate personalized tour routes; The analysis of the passenger flow data to determine the pedestrian distribution characteristics of the target area includes: Based on the time series, the passenger flow data is analyzed to determine the passenger data and the number of passengers alighting; Obtain traffic information within the target area, and determine several available modes of transportation and the operating hours of each available mode of transportation within the target area; Analyze the operating period, determine the period characteristics, and based on the period characteristics, perform time decay weighted calculation on the passenger data and alighting data to obtain the calculation results; The target area is spatially gridded to obtain a unit grid, and a pedestrian density prediction model is established based on the unit grid. Based on the pedestrian density prediction model and the calculation results, the pedestrian distribution characteristics of the target area are determined; The process of generating a dynamic traffic guide that includes route planning and route accessibility based on the pedestrian flow distribution characteristics includes: Get the user's current location and initial target coordinates; Based on the user's current location and the initial target coordinates, several candidate routes are generated according to a path planning algorithm; Based on the pedestrian flow distribution characteristics, calculate the travel time correction coefficient for each candidate route; Based on the travel time correction factor, the route and travel time of each candidate route are adjusted to obtain a dynamic traffic guide; The process of generating personalized tour routes based on the tourism demand characteristics and the dynamic transportation guide includes: Based on the aforementioned tourism demand characteristics, determine the preferred duration of stay at each target tourist destination; The order of sightseeing spots is dynamically adjusted based on the population density and the preferred length of stay. A personalized guided tour route will be generated based on the order of visits to the attractions and the operating hours. The analysis of the travel guide determines the characteristics of travel demand, including: Natural language processing was performed on the travel guide to extract keywords and time descriptions; Based on the keywords, match the POI data within the target area to determine candidate tourist spots; Analyze the POI data to determine historical visit patterns; Based on the historical travel patterns and the time descriptions, the expected duration of stay for each candidate tourist destination is determined. Obtain user historical behavior data, and determine user browsing preferences based on the user historical behavior data; Based on the user's travel preferences and expected length of stay, tourism demand characteristics are generated.

2. The system according to claim 1, characterized in that, The acquisition of passenger flow data in the target area includes: Obtain basic passenger flow data within the target area through the interface of a third-party platform; Analyze the basic passenger flow data to determine the data type; The data type is compared with the preset data item type to determine whether there are any missing data items; If it exists, determine the associated data source based on the missing data item; Based on the associated data source, customer data is estimated to generate predicted data for the missing data items; The predicted data and the basic passenger flow data are used as the passenger flow data for the target area.

3. The system according to claim 2, characterized in that, The process of establishing a pedestrian density prediction model based on the cell grid includes: Obtain historical pedestrian traffic data for the target area; Analyze the historical pedestrian flow data to determine the pedestrian flow transmission patterns among several available modes of transportation; Acquire meteorological data, analyze the meteorological data and the available modes of transportation, and determine weather interference factors; Based on the spatial division of the cell grid, the historical pedestrian flow data is associated with the corresponding cell grid; Based on the historical pedestrian flow data associated with each grid cell, and according to the pedestrian flow transmission pattern and the weather interference factor, a regression algorithm is used to train the model, so that the model outputs the predicted pedestrian flow density value for each grid cell, thus obtaining the pedestrian flow density prediction model.

4. The system according to claim 1, characterized in that, The process of generating personalized guided tour routes based on the order of visits to attractions and the operating hours includes: Based on the order of visits to the attractions and the preferred length of stay, predict the dining locations; Obtain restaurant data within the preset range of the dining attractions; Analyze the restaurant data to determine the characteristics of the restaurants; Based on the characteristics of the restaurants, a recommendation list is generated, and recommendation feedback is received; Based on the recommended feedback, determine the target dining location; Based on the pedestrian density, predict the expected queuing time at the target dining location; A personalized tour route is generated based on the order of visits to the attractions, the operating hours, and the expected queuing time.

5. The system according to claim 3, characterized in that, The step of determining the associated data source based on the missing data item includes: Analyze the missing data items to determine the missing attribute monitoring time; Based on the missing attributes and the pattern of pedestrian flow, the associated transportation modes are determined; Based on the missing monitoring time, the operational data of the associated transportation mode is obtained as the associated data source.

6. The system according to claim 4, characterized in that, The step of predicting the expected queuing time at the target dining location based on the pedestrian density includes: Analyze the target dining locations to determine the characteristics of the food consumed; Based on the characteristics of the food and the population density, the expected queuing time at the target dining location is predicted.

Images