Personalized parking recommendation and path planning method based on user behavior portrait

Abstract

The invention provides a personalized parking recommendation and path planning method based on user behavior portraits, and relates to the technical field of smart traffic, and the method comprises the steps: constructing multi-user behavior portraits; identifying user travel scene features, and matching current scene rules; performing parking resource recommendation and conflict coordination according to the multi-user behavior portraits and the user travel scene features; and based on the real-time position of the user and the target parking lot, performing full-link path planning in combination with the real-time traffic road condition data, performing path optimization according to the current scene rule, obtaining a driving guide path, and pushing the driving guide path to the user side. Through the parking recommendation method and device, the technical problem that in the prior art, due to the fact that parking recommendation lacks personalized consideration, the recommendation result is disjointed with the real demand of the user, and then the travel efficiency is affected is solved, personalized parking recommendation is carried out by constructing the multi-user behavior portraits, full-link path planning is carried out, the utilization rate of parking resources is increased, and the user experience is improved. And the overall travel experience of the user is optimized.

Description

Technical Field

[0001] This application relates to the field of intelligent transportation technology, specifically to a personalized parking recommendation and route planning method based on user behavior profiles. Background Technology

[0002] With accelerated urbanization and continuous growth in car ownership, parking difficulties and complex route planning have become core pain points restricting urban travel efficiency. Parking and route needs under different travel scenarios exhibit diverse characteristics. Currently, mainstream parking recommendation services rely on user location and destination, using single dimensions such as distance, base price, or real-time parking availability for information filtering and route planning. However, such technical solutions generally ignore users' individual preferences, historical behavioral habits, and the essential differences in needs under different travel scenarios, resulting in homogenized recommendation results. Users still need to manually compare and filter repeatedly, which not only places a heavy decision-making burden on them but also makes it difficult to obtain the optimal solution that truly meets their needs, leading to a poor user travel experience and low utilization of urban transportation resources.

[0003] In summary, existing technologies suffer from a lack of personalized consideration in parking recommendations, leading to a disconnect between the recommendations and users' actual needs, which in turn affects travel efficiency. Summary of the Invention

[0004] The purpose of this application is to provide a personalized parking recommendation and route planning method based on user behavior profiles, in order to solve the technical problem in the prior art that the parking recommendation results are out of touch with the user's actual needs due to the lack of personalized consideration, which in turn affects travel efficiency.

[0005] To achieve the above objectives, this application provides a personalized parking recommendation and route planning method based on user behavior profiles. The method includes: collecting multi-dimensional historical travel datasets and multi-user identity feature information to construct multi-user behavior profiles; obtaining the user's real-time travel destination, identifying user travel scenario features, and matching current scenario rules in a scenario rule base; recommending parking resources and coordinating conflicts based on the multi-user behavior profiles and user travel scenario features to determine target parking lots; performing full-link route planning based on the user's real-time location and the target parking lot, combined with real-time traffic data, and optimizing the route according to the current scenario rules to obtain a driving guidance route and push it to the user's terminal.

[0006] Optionally, a multi-dimensional historical travel dataset of users is collected, including historical parking data, travel trajectory data, and preference setting data; the multi-dimensional historical travel dataset of users is preprocessed to obtain a standard dataset; based on the standard dataset, users are tagged to obtain user identification tags, wherein the user identification tags include at least travel mode tags, cost sensitivity tags, and facility preference tags; according to the user identity feature information, weight values ​​are assigned to the user identification tags, and the weight values ​​are mapped one-to-one with the user identification tags to construct the user behavior profile.

[0007] Optionally, based on the user's real-time travel destination, the destination type is determined; the user's travel time is simultaneously acquired, and the user's travel scenario is determined in combination with the destination type; the user's travel scenario features are obtained through feature extraction; similarity matching is performed in the scenario rule base according to the user's travel scenario features, and a scenario probability distribution is output; the top P scenario rules of the scenario probability distribution are sent to the user terminal, user confirmation information is obtained, and the current scenario rule is obtained, wherein the current scenario rule includes at least priority rules, path planning rules, service adaptation rules, and space tolerance rules.

[0008] Optionally, historical travel data, historical traffic route data, and historical parking resource data are acquired; the historical travel data, historical traffic route data, and historical parking resource data are aligned and associated according to their spatiotemporal locations to obtain a historical scene dataset; multiple travel scenarios are predefined, and the historical scene datasets are clustered to obtain multiple scene historical datasets; based on the multiple scene historical datasets, a set of scene rules corresponding to multiple travel scenarios is generated to construct the scene rule base; user decision feedback data is continuously collected to update the scene rule base.

[0009] Optionally, taking the user's real-time travel destination as the center, a geographical search range is determined according to the spatial tolerance rules of the current scenario rules; based on the multi-user behavior profile and the user's travel scenario characteristics, multi-condition parallel filtering is performed within the geographical search range to determine an initial parking lot candidate set; the quantified utility value of each dimension of the initial parking lot candidate set is calculated, and a parking lot adaptability score is obtained through weighted calculation, and the initial parking lot candidate set is sorted to obtain an initial parking lot sequence; the variance of the quantified utility values ​​of each dimension of the first N parking lots in the initial parking lot sequence is calculated, and if the variance of the quantified utility value of any dimension is greater than a preset threshold, a conflict is determined, where N is a positive integer; the lowest dimension of the quantified utility value of each dimension is obtained, the corresponding weight is reduced, and the weighted calculation and sorting are performed again to obtain the parking lot sequence; the parking lot ranked first in the parking lot sequence is selected as the target parking lot.

[0010] Optionally, based on the user's real-time location and the target parking lot, and combined with real-time traffic data, route planning is performed to obtain multiple alternative routes and multiple estimated arrival times; a first route is determined based on the shortest estimated arrival time among the multiple estimated arrival times; the real-time status of the target parking lot is obtained, and the user's parking preferences are retrieved based on the multi-user behavior profile to determine the target parking space; an in-park route is planned based on the target parking space and the parking lot entrance to determine a second route; a walking route is planned based on the target parking space and the user's real-time travel destination to determine a third route; the first, second, and third routes are integrated to obtain the full-link path; the full-link path is optimized according to the current scenario rules to obtain the driving guidance path.

[0011] Optionally, based on the current scenario rules, a path optimization objective is determined; according to the path optimization objective, multi-objective optimization is performed on the entire link path to obtain the driving guidance path.

[0012] Optionally, the system monitors changes in the user's travel status in real time, determines the user's travel update scenario, and matches the updated travel rules in the scenario rule base. Based on the user's travel update scenario, it determines the target updated parking lot (the target updated parking lot refers to the parking lot that is re-determined based on real-time changes during the travel process). Based on the target updated parking lot and the user's real-time updated location, it performs full-link path planning in conjunction with current traffic condition data, and optimizes the path according to the updated travel rules to obtain the updated driving guidance path and pushes it to the user's terminal.

[0013] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0014] By collecting multi-dimensional historical travel datasets and multi-user identity feature information, a multi-user behavior profile is constructed. The system obtains the user's real-time travel destination, identifies user travel scenario characteristics, and matches the current scenario rules in a scenario rule base. Based on the multi-user behavior profile and the user travel scenario characteristics, parking resource recommendations and conflict coordination are performed to determine the target parking lot. Based on the user's real-time location and the target parking lot, combined with real-time traffic data, a full-link path planning is performed, and the path is optimized according to the current scenario rules to obtain a driving guidance path, which is then pushed to the user's end. In other words, by collecting multi-dimensional historical travel datasets and multi-user identity feature information, a comprehensive multi-user behavior profile is constructed. The system obtains the user's real-time travel destination, combines user travel habits and current time factors to identify the user's travel scenario, thereby providing parking recommendations and path planning that better meet user needs. It matches the current scenario rules in a scenario rule base and optimizes the path according to these rules, ensuring that the path planning meets the user's actual needs, improving the utilization rate of parking resources, reducing parking search time and traffic congestion, and optimizing the overall travel experience.

[0015] The above description is merely an overview of the technical solution of this application. To better understand the technical means of this application and to facilitate its implementation according to the description, and to make the above and other objects, features, and advantages of this application more apparent, specific embodiments of this application are described below. It should be understood that the content described in this section is not intended to identify key or important features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent through the following description. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in 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 merely exemplary. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0017] Figure 1 This is a flowchart illustrating a personalized parking recommendation and route planning method based on user behavior profiles, as described in this application.

[0018] Figure 2 This is a schematic diagram illustrating the process of determining the target parking lot in a personalized parking recommendation and route planning method based on user behavior profiles according to this application. Detailed Implementation

[0019] This application provides a personalized parking recommendation and route planning method based on user behavior profiles. This addresses the technical problem in existing technologies where parking recommendations lack personalization, leading to a disconnect between the recommendations and the user's actual needs, thus impacting travel efficiency. By collecting multi-dimensional historical travel datasets and multi-user identity feature information, a comprehensive multi-user behavior profile is constructed. This profile obtains the user's real-time travel destination and, combined with factors such as travel habits and current time, identifies the user's travel scenario. This allows for the provision of parking recommendations and route planning that better meet the user's needs. The method matches the current scenario rules in a scenario rule base and optimizes the route based on these rules, ensuring that the route planning aligns with the user's actual needs. This improves the utilization rate of parking resources, reduces parking search time and traffic congestion, and optimizes the overall travel experience.

[0020] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. It should be understood that this application is not limited to the exemplary embodiments described herein. 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. It should also be noted that, for ease of description, only the parts related to this application are shown in the accompanying drawings, not all of them.

[0021] For examples, please refer to the appendix. Figure 1 This application provides a personalized parking recommendation and route planning method based on user behavior profiles, wherein the personalized parking recommendation and route planning method based on user behavior profiles specifically includes the following steps:

[0022] S100: Collect multi-dimensional historical travel datasets and multi-user identity feature information to construct multi-user behavior profiles.

[0023] Furthermore, S100 of this application includes: collecting a multi-dimensional historical travel dataset of users, including historical parking data, travel trajectory data, and preference setting data; performing data preprocessing on the multi-dimensional historical travel dataset of users to obtain a standard dataset; labeling users based on the standard dataset to obtain user identification tags, wherein the user identification tags include at least travel mode tags, cost sensitivity tags, and facility preference tags; assigning weight values ​​to the user identification tags according to the user identity feature information, mapping the weight values ​​to the user identification tags one-to-one, and constructing the user behavior profile.

[0024] Specifically, multi-dimensional historical travel datasets are collected from users through applications, vehicle GPS, or other sensors. These datasets include historical parking data (selected parking lot type, location, pricing preferences, parking duration), travel trajectory data (origin, destination, frequently traveled routes, travel time patterns), and preference setting data (manually set price caps, parking lot type preferences, whether accessibility facilities are needed, etc.). The raw data often contains missing values, outliers, or noisy data. The first step in preprocessing is data cleaning, removing irrelevant or erroneous records. For example, if a parking record has a negative parking duration, that data is automatically removed. Data standardization ensures that all types of data are represented in a uniform format and unit. For example, parking fees can be uniformly converted to hourly rates, and travel times can be standardized to minutes.

[0025] Based on standardized data, each user is tagged. Their travel patterns are determined by analyzing their historical travel routes. For example, a user who mostly avoids peak areas and typically chooses faster routes might be tagged with a preference for fast routes. Users' cost sensitivity is assessed by analyzing the price and frequency of their parking lot choices. If a user frequently chooses low-priced parking lots, they are more sensitive to parking costs, and their cost sensitivity tag might be set to high. Facility preference tags are constructed by analyzing whether users tend to choose certain facilities when selecting parking lots, such as charging stations, security, and parking space width. For example, if a user frequently chooses parking lots with charging stations, they are assigned a charging station preference tag.

[0026] Based on user identity data—including basic information such as age, occupation, and disability status obtained through real-name authentication—different weights are assigned to tags. This weighting takes into account behavioral differences among different user groups. For example, older users may prioritize parking lot security, while younger users may prefer cheaper parking spaces. Combining user identification tags with weighted values ​​ultimately creates a personalized user behavior profile, reflecting users' parking habits, cost awareness, and preferences.

[0027] User identification tags include travel type tags, such as commuting, temporary, tourism, and emergency; price sensitivity tags, such as highly sensitive, moderately sensitive, and low sensitive; parking type tags, such as ground-level preference, underground preference, and acceptable multi-level parking; convenience need tags, such as walking distance sensitivity, traffic efficiency sensitivity, and service facility sensitivity; and special need tags, such as accessibility needs, elderly-friendly needs, and electric vehicle charging needs.

[0028] The system collects users' latest parking behavior and feedback information in real time, dynamically adjusting the weight of user profile tags. For example, if a user frequently chooses parking lots with charging facilities recently, the weight of the electric vehicle charging demand tag is automatically increased to ensure that the user profile matches their actual needs. By capturing each user's parking needs and preferences, the system determines each user's behavioral patterns, preferences, and sensitivities, thereby providing personalized services.

[0029] S200: Obtain the user's real-time travel destination, identify the user's travel scenario characteristics, and match the current scenario rules in the scenario rule base.

[0030] Furthermore, S200 of this application includes: determining the destination type based on the user's real-time travel destination; synchronously acquiring the user's travel time, determining the user's travel scenario in combination with the destination type, and obtaining the user's travel scenario features through feature extraction; performing similarity matching in the scenario rule base according to the user's travel scenario features, and outputting a scenario probability distribution; sending the top P scenario rules of the scenario probability distribution to the user terminal, obtaining user confirmation information, and obtaining the current scenario rule, wherein the current scenario rule includes at least priority rules, path planning rules, service adaptation rules, and space tolerance rules.

[0031] Specifically, the system first obtains the user's real-time travel destination, either manually entered or input through other means. This real-time destination is the user's intended destination at the current or scheduled time, typically a place the user wants to reach, such as a shopping mall, office building, or parking lot. Assuming the user's destination is a commercial area, it is categorized as a commercial area, considering that such destinations generally have high traffic density and parking demand is concentrated during peak hours. Destinations are then classified based on their specific locations; different destination types will influence the user's parking needs and travel routes.

[0032] At the same time, it obtains users' travel time, that is, the time when users plan to travel or actually travel. Travel time affects traffic flow, road selection, and parking lot availability, and is usually used to analyze travel patterns during peak and off-peak hours.

[0033] By combining the type of travel destination and the user's travel time, the system identifies the user's travel scenario. For example, if a user goes to a business district at 8 a.m., it infers that this is a commuting scenario. In this case, the system extracts travel scenario features to analyze how to deal with traffic congestion during the morning rush hour, whether it is necessary to reserve a parking space in advance, or choose a parking lot with more available spaces; if the destination is a hospital and the travel time is in the early morning, it is automatically identified as an emergency medical treatment scenario.

[0034] The system searches for rules similar to the current travel scenario in the scenario rule base. By calculating the similarity between the current travel scenario and all scenario rules in the base, the system selects the most closely matching scenario rules. The scenario rule base is a database containing rules for different travel scenarios, used to define how to perform route planning, parking recommendations, etc., based on different travel scenarios. It includes parking recommendation priority rules for each scenario, such as: commuting scenario priority: parking space stability → distance → price; emergency medical treatment scenario priority: traffic efficiency → walking distance → parking space type; route planning preference rules, such as commuting scenario preference: route stability → time; tourism scenario preference: scenic route → time; and service adaptation rules, such as automatically associating parking lot fast-pass information in emergency scenarios and automatically associating scenic area navigation information in tourism scenarios. By calculating the similarity between the user's current travel scenario and existing scenarios in the rule base, the system selects the closest scenario rules for matching. Common similarity calculation methods include Euclidean distance and cosine similarity.

[0035] Based on the similarity matching results, probability distributions for multiple scenario rules are generated. For example, scenario rule 1 (prioritizing parking lots with more available spaces during peak hours) has a matching degree of 80%, while scenario rule 2 (avoiding main roads during peak hours) has a matching degree of 60%. The probability distributions for these two rules are 0.8 and 0.6, respectively. The scenario probability distribution represents the probability distribution of different scenario rules matching the current user's travel scenario, indicating the degree to which each scenario rule adapts to the user's current needs.

[0036] The system sends the top P scenario rules with the highest probability to the user's device, requiring user confirmation. For example, the first two rules might be: prioritize recommending parking lots with more spaces and avoiding congested roads. After user confirmation, the system generates detailed route planning based on the confirmed rules. Based on the user's confirmation, the system finalizes the current scenario rules, which include a series of rules guiding the user's travel and parking decisions, at least prioritization rules, route planning rules, service adaptation rules, and space tolerance rules. Priority rules determine which travel options are prioritized based on the urgency or importance of the user's needs, such as prioritizing available parking spaces in a particular parking lot. Route planning rules plan the user's travel route, which may include avoiding peak hours and choosing faster routes. Service adaptation rules ensure that the provided services (such as parking lots and transportation facilities) meet the user's specific needs (such as charging stations and special parking facilities). Space tolerance rules determine the user's adaptability to parking space, such as whether they are willing to park far from their destination or can accept slightly inconvenient parking conditions.

[0037] The current scenario rules include at least priority rules, path planning rules, service adaptation rules, and space tolerance rules. For example, parking recommendation priority rules prioritize parking space stability over distance over price in commuting scenarios, and traffic efficiency over walking distance over parking space type in emergency medical treatment scenarios. Path planning preference rules prioritize route stability over time in commuting scenarios, and scenic routes over time in tourism scenarios. Service adaptation rules automatically associate parking lot fast passage information in emergency scenarios and scenic area navigation information in tourism scenarios.

[0038] Real-time monitoring of user travel status changes, such as destination changes or extended travel times, automatically triggers scenario re-identification and rule adjustments to ensure continuous service adaptability. By combining travel destination, time, and real-time user needs, it provides more personalized scenario recommendations, avoiding the uniform recommendations of traditional methods. Through precise scenario matching and route optimization, it can provide users with optimal travel plans based on different peak periods and traffic conditions.

[0039] Furthermore, this application also includes the following steps: acquiring historical travel data, historical traffic route data, and historical parking resource data; aligning and associating the historical travel data, historical traffic route data, and historical parking resource data according to their spatiotemporal locations to obtain a historical scene dataset; predefining multiple travel scenarios and clustering the historical scene dataset to obtain multiple scene historical datasets; generating a set of scene rules corresponding to multiple travel scenarios based on the multiple scene historical datasets to construct the scene rule library; and continuously collecting user decision feedback data to update the scene rule library.

[0040] Specifically, this involves acquiring historical travel data, historical traffic route data, and historical parking resource data. Historical travel data refers to a user's past travel records, including travel routes, times, and destinations. For example, it might show a user's route and time from home to the office, and whether they chose peak hours. Historical traffic route data is historical data related to the transportation network, describing traffic flow, speed, and congestion at different times and locations. It is typically acquired through traffic sensors or GPS data and used to analyze traffic conditions and route selection. Historical parking resource data is data related to a user's past parking experiences, including available parking spaces, parking fees, parking lot locations, and parking duration. This data helps understand users' parking selection habits and the distribution of parking resources.

[0041] Historical travel data, historical traffic route data, and historical parking resource data typically come from traffic management systems, GPS devices, or third-party applications. For analysis, this data needs to be aligned according to time and space. For example, when a user leaves home at a specific time (such as 8:00 AM during rush hour), the relevant data needs to be integrated into a complete historical scenario dataset, taking into account traffic flow and parking availability at that time. Aligning and associating historical travel data, historical traffic route data, and historical parking resource data according to spatiotemporal location means synchronizing different types of historical data according to time and space, ensuring that all data points match at the same time and location.

[0042] By aligning historical travel data, traffic route data, and parking resource data according to their spatiotemporal locations, a comprehensive dataset is formed, which is the historical scene dataset. It contains information such as the user's travel trajectory, the traffic routes taken, the status of parking lots, traffic flow, and parking resources in a specific spatiotemporal scenario.

[0043] Several travel scenario types are predefined, such as weekday commuting, weekend shopping, and holiday travel. Each scenario represents different user travel needs and behavioral patterns. A travel scenario is a travel characteristic under specific time, location, and user needs. For example, a weekday morning rush hour commuting scenario, a weekend shopping scenario, etc. Each scenario has different transportation and resource requirements.

[0044] Clustering is performed on historical scenario datasets. Clustering algorithms group data points with similar characteristics into multiple scenario-based historical datasets. For example, all users' travel data from the city center to the commercial area during the morning rush hour might be clustered into one group, while travel data from home to the shopping mall on weekends would be clustered into another. Clustering is an unsupervised learning method used to group data points in a dataset based on their similarity. The historical scenario dataset is divided into multiple groups using clustering algorithms, each group corresponding to a specific travel scenario type.

[0045] Based on each clustering result, a set of scenario rules is defined for each scenario type. For example, for the weekday commuting scenario, the following rules are defined: prioritize expressways; parking lot recommendations should consider available spaces and pricing; choose roads with lower traffic volume to avoid peak-hour congestion. Scenario rules are generated through analysis of historical data and pattern learning, aiming to optimize the user's travel experience. The scenario rule set is a set of rules defined for each travel scenario based on clustering results. Each rule describes the best behavior or strategy to be taken in that scenario, such as which road to choose and which parking lots to recommend. All scenarios and their corresponding scenario rules are stored in a scenario rule library, a database containing multiple scenarios and their corresponding rules, used for dynamic matching and recommendations based on real-time user needs.

[0046] Continuously collect user decision feedback data, such as user satisfaction with recommended routes and parking lots, and whether the selected parking lot meets expectations, to adjust existing scenario rules and improve the recommendation algorithm. User decision feedback data refers to user feedback information during actual travel, including data on user feedback on recommended routes, parking lot selection results, and satisfaction levels, which is used to improve the system's recommendation algorithm. Based on new user behavior data, regularly update the scenario rule base to make it more in line with users' actual needs and travel habits.

[0047] By clustering and analyzing historical travel and traffic route data, the system accurately identifies and defines different travel scenarios, thereby providing users with travel solutions that better meet their needs. Through the construction of a scenario rule base, the system can quickly match the most suitable recommended routes and parking lots based on real-time travel scenarios, improving users' travel efficiency.

[0048] S300: Based on the multi-user behavior profile and the user travel scenario characteristics, recommend parking resources and coordinate conflicts to determine the target parking lot.

[0049] Further details are attached. Figure 2 As shown, S300 of this application includes: determining a geographical search range based on the user's real-time travel destination and the spatial tolerance rule of the current scenario rules; performing multi-condition parallel filtering within the geographical search range based on the multi-user behavior profile and the user's travel scenario features to determine an initial parking lot candidate set; calculating the quantitative utility value of each dimension of the initial parking lot candidate set, obtaining a parking lot adaptability score through weighted calculation, and sorting the initial parking lot candidate set to obtain an initial parking lot sequence; calculating the variance of the quantitative utility values ​​of each dimension of the first N parking lots in the initial parking lot sequence, and determining a conflict if the variance of any dimension's quantitative utility value is greater than a preset threshold, where N is a positive integer; obtaining the lowest dimension of each dimension's quantitative utility value, reducing the corresponding weight, re-calculating the weighted value, and sorting to obtain a parking lot sequence; and selecting the parking lot ranked first in the parking lot sequence as the target parking lot.

[0050] Specifically, the system obtains the user's real-time travel destination and, based on the spatial tolerance rules in the current scenario, determines the geographical range of parking lots the user is willing to accept. For example, if the user's destination is a commercial area, and the spatial tolerance rule is set to allow parking within a maximum 10-minute walking distance, then all parking lots within a 10-minute walking distance of the commercial area are searched. The user's real-time travel destination is their current or planned destination, obtained through GPS or user input, such as a commercial area, residential area, or office building.

[0051] The spatial tolerance rule in the current scenario is a spatial tolerance set based on the user's travel scenario, specifically the user's spatial tolerance for parking locations. The spatial tolerance rule describes the user's tolerance for the distance between the parking lot and their destination. For example, a user might prefer a parking lot farther from their destination, but still prefer it to be within a specific range.

[0052] After obtaining the geographic search scope, parallel filtering is performed based on multi-user behavior profiles and user travel scenario characteristics. For example, the following conditions are considered simultaneously: real-time parking resource data (parking space availability, traffic efficiency, charging standards, open status, service facilities), real-time traffic condition data (road congestion, travel time, construction control information), user profile data (preference tags, behavioral characteristics), and scenario characteristic data (scenario priority, related needs). Based on these conditions, parking lots within the scope are filtered to select an initial candidate set of parking lots that meet user needs.

[0053] The geographic search scope is a geographical area defined based on the user's travel destination and spatial tolerance rules, encompassing all possible parking lots. Parking lots within the geographic search scope are filtered based on multiple criteria, such as location, price, and availability. Parallel filtering means these criteria can be processed simultaneously to quickly identify parking lots that meet the requirements. The initial parking lot candidate set is a preliminary set of parking lots that meets the user's needs, obtained based on the geographic search scope and the results of multi-criteria parallel filtering.

[0054] For each selected parking lot, its quantified utility value is calculated based on multiple dimensions. For example, a parking lot might have a low cost (high cost utility value), a close proximity (high distance utility value), and many available parking spaces (high parking space utility value), but no charging stations (low facility utility value). These utility values ​​are calculated for each parking lot and weighted according to user preferences to obtain a parking lot suitability score. The quantified utility value measures the suitability of each parking lot across different dimensions (such as price, available parking spaces, distance, etc.). The quantified utility value is calculated based on multiple factors and is used to quantify the parking lot's attractiveness to users.

[0055] The quantitative utility value of parking lots is weighted according to the weight values ​​of each dimension to obtain a comprehensive parking lot adaptability score. The weights of different dimensions reflect the degree of importance users attach to these features; for example, users may value parking fees more and have a lower preference for parking lot facilities. The parking lot adaptability score is a score for each parking lot after comprehensively considering all influencing factors, used to measure whether each parking lot meets the user's needs and preferences. In short, by combining real-time traffic conditions to calculate the entire journey time from the origin to the parking lot and then to the destination, the overall cost-effectiveness of parking resources is comprehensively evaluated, including time cost, economic cost, and convenience, and finally a ranked recommendation list is generated.

[0056] The adaptability scores of all parking lots are sorted from most suitable to least suitable, resulting in a preliminary parking lot sequence. Simultaneously, the variance of the utility values ​​of these parking lots is calculated. If the utility values ​​of some parking lots vary excessively across certain dimensions (i.e., high variance), these parking lots are considered conflicting, meaning they may not consistently meet user needs in some aspects. In other words, if the variance exceeds a preset threshold, the weight of that parking lot on the conflicting dimension is reduced, and the ranking is recalculated and adjusted to ensure more consistent and stable recommendation results. Variance is an indicator of data volatility, used to detect the consistency of parking lots across different dimensions. If a parking lot's utility value fluctuates significantly across certain dimensions, it indicates that the parking lot's adaptability is unstable in those dimensions.

[0057] If a parking lot has a large variance in utility value across certain dimensions, it indicates that the parking lot may have inconsistent recommendation features and may not be able to meet the consistent needs of users. Such parking lots should be adjusted or excluded from the candidate set.

[0058] After sorting and conflict resolution, the parking lot with the highest adaptability score is selected as the target parking lot and recommended to the user. The target parking lot is the optimal parking lot selected based on sorting and conflict detection; it is considered the most suitable for the user's needs, and the user will be guided to this parking lot.

[0059] When multiple objectives conflict, such as the nearest parking lot price exceeding a user's budget, a balance is struck based on user profile tag weights. For example, price-sensitive users are prioritized for more cost-effective, slightly farther parking lots, while time-sensitive users are prioritized for closer, slightly more expensive lots. By comprehensively considering the user's real-time travel destination, travel scenario characteristics, and behavioral profile, personalized parking recommendations are provided to better meet user needs. Variance detection avoids recommending unstable parking lots, ensuring consistent and reliable parking suggestions for users. Spatial tolerance rules and multi-dimensional filtering quickly select the optimal parking lot, and adaptive scoring ensures the efficiency of the recommendation results and user satisfaction.

[0060] S400: Based on the user's real-time location and the target parking lot, and combined with real-time traffic data, perform full-link path planning, and optimize the path according to the current scenario rules to obtain a driving guidance path and push it to the user terminal.

[0061] Furthermore, S400 of this application includes: performing route planning based on the user's real-time location and the target parking lot, combined with real-time traffic data, to obtain multiple alternative routes and multiple estimated arrival times; determining a first route based on the shortest estimated arrival time among the multiple estimated arrival times; obtaining the real-time status of the target parking lot, retrieving the user's parking preferences based on the multi-user behavior profile, and determining a target parking space; performing in-park route planning based on the target parking space and the parking lot entrance to determine a second route; performing walking route planning based on the target parking space and the user's real-time travel destination to determine a third route; integrating the first route, the second route, and the third route to obtain a full-link route; and optimizing the full-link route according to the current scenario rules to obtain the driving guidance route.

[0062] Specifically, by combining the user's real-time location and target parking lot with real-time traffic data, multiple alternative routes from the user's location to the target parking lot are calculated. Each route is dynamically adjusted based on traffic conditions (such as traffic congestion, road construction, etc.). For each route, an estimated arrival time is calculated. The user's real-time location is the user's current geographical location, usually obtained through positioning technologies such as GPS. The target parking lot is the parking lot the user plans to park in, usually specified by a parking recommendation system or the user. Real-time traffic data reflects the current road traffic conditions, including road congestion, traffic flow, accident information, etc., and is usually obtained from traffic sensors, road surveillance cameras, GPS devices, etc. The estimated arrival time is usually calculated based on route planning and traffic conditions, taking into account factors such as traffic conditions and road congestion.

[0063] Based on the user's origin, destination, and traffic data, one or more driving routes from the origin to the destination are calculated. Based on the estimated arrival times of multiple alternative routes, the route with the shortest estimated arrival time is selected as the primary route. Among multiple alternative routes, the route with the shortest estimated arrival time is usually the optimal route.

[0064] The system acquires the real-time status of the target parking lot, including information such as the current number of available parking spaces, the parking lot's open status, and payment details. This includes the number of available spaces, the availability of special parking spots, and current traffic flow. By combining this information with multi-user behavior profiles to retrieve user parking preferences, the system determines the target parking space. User preferences for parking spaces include their choice of location (e.g., near the entrance, near the exit) and preferred parking space type (e.g., with charging stations, accessible for disabled users). Within the target parking lot, the system ultimately determines the user's parking location and target parking space, taking into account both user preferences and the parking lot's real-time status.

[0065] After a user enters the parking lot, based on the target parking space and the location of the parking lot entrance, in-park route planning is performed to calculate the optimal driving route from the parking lot entrance to the target parking space. Factors such as the parking lot's lane layout and parking space distribution are considered to ensure the user can easily find a parking space. In other words, within the parking lot, route planning is performed based on the location of the entrance and parking spaces to determine the optimal path for the user from the parking lot entrance to the designated parking space.

[0066] Based on the user's target parking space and final destination, plan the user's walking route from the parking lot to the destination, consider the shortest or most convenient walking route from the parking lot to the destination, and try to avoid congested areas or interference factors such as traffic signals.

[0067] The entire journey, from the user's real-time location to the target parking lot, the driving route within the parking lot, and the walking route from the parking lot to the destination, is integrated into a complete end-to-end path. Users will receive a complete travel plan before departure, encompassing all aspects of driving, parking, and walking. The end-to-end path includes the driving route from the user's real-time location to the target parking lot, the driving route within the parking lot, and the walking route from the parking lot to the destination.

[0068] The integrated end-to-end path is optimized based on current scenario rules. Scenario rules may involve peak-hour optimization, weather factors, and changes in user preferences. For example, during peak hours, the system might avoid main streets and prioritize secondary roads to avoid traffic congestion. Scenario rules are strategies for optimizing route planning based on different travel scenarios, such as peak hours, holidays, and ordinary daily travel. Scenario rules may include selecting the optimal route and adapting to traffic conditions.

[0069] The driving guidance route is the final travel path provided to the user, considering driving routes from the departure point to the parking lot, driving routes within the parking lot, and walking routes, providing users with comprehensive guidance. It not only plans the driving route from the departure point to the parking lot but also the walking / connecting routes from the parking lot to the final destination, providing end-to-end navigation; it monitors real-time road conditions and parking lot status, such as decreasing parking space availability or sudden traffic congestion, dynamically adjusting driving routes and recommended parking lots to ensure the timeliness of recommendations; it provides internal parking lot navigation, such as parking space locations in underground garages and elevator location guidance, as well as access prompts, such as parking lot entrance queue status and the best entry route, helping users quickly find parking spaces and reach their destination.

[0070] By combining the user's real-time location, target parking lot, traffic conditions, and parking preferences, personalized and optimal driving routes and parking lot recommendations are provided, significantly improving users' travel efficiency. The comprehensive integration of driving routes, parking lot routes, and walking routes offers complete travel guidance, resolving common user issues such as unclear routes and parking difficulties.

[0071] Furthermore, this application also includes the following steps: determining the path optimization objective based on the current scenario rules; and performing multi-objective optimization on the entire link path according to the path optimization objective to obtain the driving guidance path.

[0072] Specifically, route optimization goals are determined based on current scenario rules. For example, during peak hours, the goal is to minimize traffic congestion and reduce travel time; in severe weather, the preference is to choose safer routes, such as avoiding areas with flooded or snow-covered roads; during holidays, priority is given to parking availability and parking fees, rather than simply optimizing travel time. Current scenario rules are optimization rules set based on the specific circumstances of the user's travel (such as traffic congestion, time constraints, weather, etc.). For example, during peak traffic hours, major roads should be avoided; in severe weather, safer routes are preferred. Route optimization goals are determined by analyzing the current travel scenario to identify the priorities for route optimization. For example, optimization goals could be the shortest time, lowest cost, lowest risk, or maximum safety.

[0073] Based on the determined path optimization objectives, the system performs multi-objective optimization of the entire route. This optimization considers more than just travel time; it may also involve shortest time, lowest cost, safety, and parking convenience. During path planning, multiple optimization objectives are considered simultaneously to address the balance between them. For example, users might want to reduce travel time while avoiding high-toll sections. Based on user preferences and scenario rules, multiple objectives are comprehensively considered to select an optimal path. After multi-objective optimization, the resulting driving guidance path is a final path that comprehensively considers factors such as time, cost, and safety, avoiding unsatisfactory path selections caused by users focusing on only a single factor (such as time or cost).

[0074] Based on current travel scenarios, such as peak hours and holidays, we flexibly adjust and optimize our routes to ensure users always find the most suitable travel path in ever-changing traffic environments. Through route optimization, we guide users to their target parking lot in the shortest possible time, while avoiding congested and unsafe sections, thus improving travel efficiency and safety.

[0075] Furthermore, this application also includes the following steps: real-time monitoring of changes in user travel status, determining user travel update scenarios, and matching updated travel rules in the scenario rule base; determining target updated parking lots based on the user travel update scenarios; performing full-link path planning based on the target updated parking lots and the user's real-time updated location, combined with current traffic data, and optimizing the path according to the updated travel rules to obtain a driving update guidance path and push it to the user terminal.

[0076] Specifically, the system continuously monitors users' real-time travel status, including updates to their location, the status of the destination parking lot, and traffic conditions. For example, assuming a user's starting point is a residential area and their destination is a parking lot in a commercial area, the system periodically acquires GPS data of the user's current location, road traffic conditions, and parking lot availability. In other words, it continuously tracks and analyzes changes during the user's journey, including real-time updates to their location, traffic conditions, and parking lot status. The system obtains this information in real time to dynamically adjust travel recommendations for the user.

[0077] As a user's journey progresses, changes may occur, such as alterations in traffic conditions or temporary changes to the destination. After monitoring these changes, the user's current travel context is reassessed to determine if adjustments to routes or parking suggestions are necessary. Based on the identified updated travel scenario, suitable rules are retrieved from the scenario rule base, and optimizations to travel routes and parking lots are made based on these rules.

[0078] Based on changes in the user's travel status and real-time traffic information, a new parking lot may need to be recommended. The target updated parking lot refers to the most suitable parking lot determined based on the latest information during the trip. Based on the user's real-time location, the target updated parking lot, and traffic data, end-to-end route planning is performed, including: the driving route from the user's current location to the target parking lot, the route within the parking lot from the entrance to the designated parking space, and the walking route from the parking lot to the final destination after parking.

[0079] Based on updated travel rules, the entire route is optimized. Optimization goals include minimizing time, cost, and risk, ensuring users reach their destination in the most suitable way. After optimization, a new travel route is generated to guide users smoothly through changing traffic conditions.

[0080] By monitoring changes in user travel status in real time, travel plans can be flexibly adjusted to ensure users always receive the most suitable routes and parking recommendations in a dynamically changing environment. Real-time route acquisition and optimization avoid traffic accidents and peak-hour congestion, significantly improving user travel efficiency. Timely feedback and route optimization enhance the overall user travel experience and reduce inconvenience caused by traffic or parking issues.

[0081] By leveraging IoT devices in parking lots and traffic monitoring data, the system monitors parking resources and route status in real time, promptly identifying anomalies such as temporary parking lot closures, sold-out parking spaces, and road closures. It supports proactive user feedback, such as incorrect parking information or unreasonable recommendations. Simultaneously, it automatically collects service process data, such as whether the user accepted the recommendation, whether parking was successful, and whether the travel time met expectations, as indirect feedback. Based on feedback data and user behavior data, iteratively optimizes user profile tag weights, scenario rule bases, and recommendation algorithm parameters, such as adjusting the weights of various recommendation factors in different scenarios to improve recommendation accuracy and route planning efficiency.

[0082] User behavior data and identity information are encrypted and stored to prevent data leakage; a tiered data access permission system is established, allowing only authorized personnel to access relevant data. User data is collected following the principle of minimum necessity, and users are clearly informed of the scope of data use, with the option for users to disable unnecessary data collection functions.

[0083] In a specific example, the scenario is a commuting trip, focusing on personalized parking recommendations and route planning for office workers. A commuting scenario rule base is constructed, prioritizing parking space stability over walking distance to the workplace, over monthly payment discounts, and over traffic efficiency. Route preferences prioritize frequently used routes and avoid congested areas during morning rush hour. Urban parking resource data is integrated, including parking lot location, total number of parking spaces, monthly payment price, opening hours, and service facilities. Real-time traffic data is also integrated. User data collection and privacy protection rules are configured. User profile initialization and user A's first system use involve real-name authentication, identification as a company employee, and manual preference settings: a price cap of 500 yuan / month, preference for surface parking, and a walking distance of no more than 10 minutes. User A's historical travel data is collected: Over the past three months, user A departs from home around 8:30 AM every Monday to Friday, heading to an office building in the city center. They have repeatedly chosen two surface parking lots near the office building, namely Parking Lot A and Parking Lot B, both with monthly passes, and parking for approximately 8 hours / day. Construct a user behavior profile for user A: travel type tag (commuter, weight 95%), price sensitivity tag (moderately sensitive, weight 80%), parking type tag (ground preference, weight 90%), convenience need tag (walking distance sensitive, weight 85%), and special need tag (none), and update the profile data in real time.

[0084] Scene recognition and rule loading: At 7:30 a.m. on Monday, user A opened the system and entered the company office building as the destination. The system automatically recognized that the destination type was office building and the travel time was weekday morning rush hour. Combined with the commuting tag in the user profile, the system determined that the current scene was a commuting scene. The system automatically loaded the commuting scene rule library and clarified the recommendation priority and route planning preference. Real-time collection and fusion of multi-dimensional data: Parking resource data: Parking lot A (surface, 5-minute walk from the company, monthly payment 450 yuan, currently 3 remaining monthly parking spaces, high entrance efficiency during morning rush hour), Parking lot B (surface, 8-minute walk from the company, monthly payment 480 yuan, currently 1 remaining monthly parking space, occasional entrance congestion during morning rush hour), Parking lot C (underground, 3-minute walk from the company, monthly payment 550 yuan, ample remaining monthly parking spaces); Traffic condition data: The morning rush hour congestion index from user A's home to parking lot A is 30% (smooth), the congestion index to parking lot B is 45% (lightly congested), and the congestion index to parking lot C is 25% (smooth); User profile data: User A is a commuter, prefers surface parking, is moderately price-sensitive, and sensitive to walking distance. Excluding parking lot C (which exceeds the budget), comparing parking lots A and B: Parking lot A is superior in walking distance, traffic efficiency, and price, and has ample remaining monthly parking spaces, making it the most cost-effective overall. Parking lot A is determined to be the optimal recommendation, and parking lot B is the alternative recommendation. The system plans the entire route, including the driving route: starting from User A's home, via XX Road - XX Avenue - XX Branch Road to Parking Lot A, estimated driving time of 25 minutes, with real-time avoidance of congested sections; and the walking route: starting from the exit of Parking Lot A, via the pedestrian walkway on the north side of the office building to the company's main entrance, a 5-minute walk, with reference points such as traffic lights and convenience stores marked along the way. After User A confirms acceptance of the recommendation, real-time navigation begins: during the journey, real-time traffic information is updated, indicating congestion at the intersection 500 meters ahead, and the route has been adjusted, with an estimated delay of 2 minutes; upon approaching Parking Lot A, it indicates that the entrance to Parking Lot A is in the right lane, with 2 vehicles currently in the queue, indicating smooth traffic flow; upon arrival at the parking lot, internal navigation is provided: please enter Zone 2, the remaining parking spaces are located at B1 floor, spaces 123, and there is an elevator nearby that goes directly to the office building. After User A successfully parks, a feedback questionnaire pops up, User A selects "satisfied," "convenient walking distance," and the system automatically records service data, with a 100% recommendation acceptance rate, a 100% parking success rate, and the total travel time meeting expectations. Increase the weight of the walking distance sensitive tag in User A's profile, and record high satisfaction with Parking Lot A, prioritizing it in future recommendations. If an anomaly occurs midway, such as the system detecting that Parking Lot A is temporarily closed during the journey, immediately trigger a re-decision, recommending Parking Lot B as an alternative and adjusting the route: Parking Lot A is temporarily closed due to equipment maintenance; you have been switched to Parking Lot B. The new route is estimated to take 28 minutes. Confirm? After User A confirms, the navigation information is updated in real time.

[0085] In summary, the personalized parking recommendation and route planning method based on user behavior profiles provided in this application has the following technical effects:

[0086] By collecting multi-dimensional historical travel datasets and multi-user identity feature information, a multi-user behavior profile is constructed. The system obtains the user's real-time travel destination, identifies user travel scenario characteristics, and matches the current scenario rules in a scenario rule base. Based on the multi-user behavior profile and the user travel scenario characteristics, parking resource recommendations and conflict coordination are performed to determine the target parking lot. Based on the user's real-time location and the target parking lot, combined with real-time traffic data, a full-link path planning is performed, and the path is optimized according to the current scenario rules to obtain a driving guidance path, which is then pushed to the user's end. In other words, by collecting multi-dimensional historical travel datasets and multi-user identity feature information, a comprehensive multi-user behavior profile is constructed. The system obtains the user's real-time travel destination, combines user travel habits and current time factors to identify the user's travel scenario, thereby providing parking recommendations and path planning that better meet user needs. It matches the current scenario rules in a scenario rule base and optimizes the path according to these rules, ensuring that the path planning meets the user's actual needs, improving the utilization rate of parking resources, reducing parking search time and traffic congestion, and optimizing the overall travel experience.

[0087] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0088] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application also intends to include such modifications and variations.

Claims

1. A personalized parking recommendation and route planning method based on user behavior profiles, characterized in that, include: Collect multi-dimensional historical travel datasets and user identity feature information to construct multi-user behavior profiles; Obtain the user's real-time travel destination, identify the user's travel scenario characteristics, and match the current scenario rules in the scenario rule base; Based on the multi-user behavior profiles and user travel scenario characteristics, parking resources are recommended and conflicts are coordinated to determine target parking lots; Based on the user's real-time location and the target parking lot, and combined with real-time traffic data, a full-link path planning is performed, and the path is optimized according to the current scenario rules to obtain a driving guidance path and push it to the user's terminal.

2. The personalized parking recommendation and route planning method based on user behavior profiles as described in claim 1, characterized in that, Collect multi-dimensional historical travel data and user identity feature information to construct user behavior profiles, including: Collect multi-dimensional historical travel datasets from users, including historical parking data, travel trajectory data, and preference settings data; The user's multi-dimensional historical travel dataset is preprocessed to obtain a standard dataset; Based on the standard dataset, users are tagged to obtain user identification tags, wherein the user identification tags include at least travel mode tags, cost sensitivity tags, and facility preference tags; Based on the user identity feature information, weight values ​​are assigned to the user identification tags, and the weight values ​​are mapped one-to-one with the user identification tags to construct the user behavior profile.

3. The personalized parking recommendation and route planning method based on user behavior profiles as described in claim 1, characterized in that, Obtain the user's real-time travel destination, identify the user's travel scenario characteristics, and match the current scenario rules in the scenario rule base, including: Determine the destination type based on the user's real-time travel destination; Simultaneously acquire user travel time, determine user travel scenario based on destination type, and obtain user travel scenario features through feature extraction; Based on the user's travel scenario characteristics, similarity matching is performed in the scenario rule base, and the scenario probability distribution is output. The first P scenario rules of the scenario probability distribution are sent to the user terminal to obtain user confirmation information, and the current scenario rule is obtained. The current scenario rule includes at least priority rules, path planning rules, service adaptation rules, and space tolerance rules.

4. The personalized parking recommendation and route planning method based on user behavior profiles as described in claim 3, characterized in that, Constructing the scenario rule base includes: Obtain historical travel data, historical traffic route data, and historical parking resource data; The historical travel data, historical traffic route data, and historical parking resource data are aligned and correlated according to their spatiotemporal locations to obtain a historical scene dataset; Multiple travel scenarios are predefined, and the historical scenario datasets are clustered to obtain multiple scenario historical datasets; Based on the historical datasets of the multiple scenarios, a set of scenario rules corresponding to multiple travel scenarios is generated, and the scenario rule library is constructed. Continuously collect user decision feedback data and update the scenario rule base accordingly.

5. The personalized parking recommendation and route planning method based on user behavior profiles as described in claim 3, characterized in that, Based on the multi-user behavior profiles and user travel scenario characteristics, parking resource recommendations and conflict coordination are performed to determine target parking lots, including: Centered on the user's real-time travel destination, the geographical search range is determined according to the spatial tolerance rules of the current scenario rules; Based on the multi-user behavior profile and the user travel scenario characteristics, multi-condition parallel filtering is performed within the geographical retrieval range to determine the initial parking lot candidate set; Calculate the quantified utility values ​​of each dimension of the initial parking lot candidate set, obtain the parking lot adaptability score through weighted calculation, and sort the initial parking lot candidate set to obtain the initial parking lot sequence; Calculate the variance of the quantified utility values ​​of each dimension for the first N parking lots in the initial parking lot sequence. If the variance of the quantified utility value of any dimension is greater than a preset threshold, it is determined that there is a conflict, where N is a positive integer. The lowest dimension of the quantified utility value of each dimension is obtained, the corresponding weight is reduced, the weighted calculation is re-performed and sorted to obtain the parking lot sequence; The parking lot ranked first in the parking lot sequence is selected as the target parking lot.

6. The personalized parking recommendation and route planning method based on user behavior profiles as described in claim 1, characterized in that, Based on the user's real-time location and the target parking lot, and combined with real-time traffic data, a full-link path planning is performed, and the path is optimized according to the current scenario rules to obtain a driving guidance path, including: Based on the user's real-time location and target parking lot, combined with real-time traffic data, route planning is performed to obtain multiple alternative routes and multiple estimated arrival times; The first path is determined based on the shortest estimated arrival time among the plurality of estimated arrival times; Obtain the real-time status of the target parking lot, retrieve user parking preferences based on the multi-user behavior profile, and determine the target parking space; Based on the target parking space and the parking lot entrance, perform on-site path planning to determine the second path; Based on the target parking space and the user's real-time travel destination, a walking route is planned to determine a third route; By integrating the first path, the second path, and the third path, a complete link path is obtained. The driving guidance path is obtained by optimizing the entire link path according to the current scenario rules.

7. The personalized parking recommendation and route planning method based on user behavior profiles as described in claim 6, characterized in that, The entire link path is optimized according to the current scenario rules to obtain the driving guidance path, including: Based on the current scenario rules, determine the path optimization objective; Based on the path optimization objective, the entire link path is optimized using a multi-objective method to obtain the driving guidance path.

8. The personalized parking recommendation and route planning method based on user behavior profiles as described in claim 1, characterized in that, Also includes: Real-time monitoring of changes in user travel status, determination of user travel update scenarios, and matching and updating travel rules in the scenario rule base; Based on the user travel update scenario, the target parking lot for update is determined; Based on the target updated parking lot and the user's real-time updated location, and combined with the current traffic condition data, the entire link path is planned, and the path is optimized according to the updated travel rules to obtain the updated driving guidance path and push it to the user terminal.

Images