Festival and holiday flow fluctuation-oriented adaptive parking space recommendation method

Abstract

The invention discloses a self-adaptive parking space recommendation method for flow fluctuation in festivals and holidays, and relates to the technical field related to intelligent traffic, and the method comprises the steps: obtaining a target travel request of a user side, at least including a destination and travel time; through festival and holiday flow prediction and parking space demand pre-judgment, in combination with a target travel request, a festival and holiday travel scene is identified, and a scene rule base including recommendation priority rules and path planning preferences is loaded; generating a parking space recommendation scheme based on a multi-dimensional decision model by fusing the real-time parking space state, the road condition information and the user portrait; and planning a full-link travel path, and displaying the full-link travel path on a terminal interface. The technical problems that in the prior art, due to the fact that the contradiction between supply and demand of parking spaces in holidays and festivals is sharp and the adaptability of parking space recommendation is poor, users are difficult to find parking spaces, and surrounding traffic jam is aggravated are solved, intelligent and efficient matching recommendation of parking space resources is achieved through scene pre-judgment, and the user experience is improved. And the user travel experience and the regional traffic operation efficiency are improved.

Description

An Adaptive Parking Space Recommendation Method for Holiday Traffic Fluctuations Technical Field

[0001] This invention relates to the field of intelligent transportation technology, and in particular to an adaptive parking space recommendation method for holiday traffic fluctuations. Background Technology

[0002] Urban parking difficulties are a prominent social problem affecting residents' travel experience and urban traffic efficiency. This problem is exacerbated during holidays, exhibiting typical periodicity, clustering, and high volatility. During holidays, the supply and demand imbalance for parking spaces in specific areas such as shopping malls, tourist attractions, and transportation hubs is particularly acute, often resulting in a shortage of spaces. This not only causes users to spend long periods searching for parking spaces, wasting time and energy, but also triggers chain congestion on surrounding roads, seriously affecting the overall operational efficiency of the urban transportation system. Traditional parking space recommendation methods are mostly based on static or quasi-static data models. Their recommendation logic often relies on real-time parking space occupancy information and distance priority principles. While these methods alleviate daily parking pressure to some extent, they show significant limitations in extreme scenarios such as holidays. On the one hand, holiday traffic is sudden and tidal, and static models cannot accurately predict the spatiotemporal distribution changes in parking space demand. On the other hand, the lack of in-depth identification of travel scenarios and dynamic consideration of user preferences leads to simplistic and poorly adaptable recommendation schemes, failing to provide users with optimal travel decision support in complex traffic environments.

[0003] At present, there are technical problems in the relevant technologies, such as the sharp contradiction between the supply and demand of parking spaces during holidays, poor adaptability of parking space recommendations, difficulty for users to find parking spaces, and exacerbation of traffic congestion in the surrounding area. Summary of the Invention

[0004] This application provides an adaptive parking space recommendation method for holiday traffic fluctuations, which solves the technical problems in the prior art, such as the sharp contradiction between parking space supply and demand during holidays, poor adaptability of parking space recommendations leading to difficulties for users in finding parking spaces and exacerbating traffic congestion in the surrounding area. It achieves the technical effect of intelligent and efficient matching and recommendation of parking space resources through scenario prediction, thereby improving the user travel experience and regional traffic operation efficiency.

[0005] This application provides an adaptive parking space recommendation method for holiday traffic fluctuations. The method includes: obtaining a user's target travel request, wherein the target travel request includes at least a destination and a travel time; identifying holiday travel scenarios by combining holiday traffic prediction and parking demand forecast with the target travel request, and loading a scenario rule base, wherein the scenario rule base includes recommendation priority rules and route planning preferences; reading the holiday travel scenarios and the scenario rule base, and generating a parking space recommendation scheme based on a multi-dimensional decision model by integrating real-time parking space status, traffic information, and user profiles; planning a full-link travel route according to the parking space recommendation scheme, and displaying it on the terminal interface.

[0006] In possible implementations, holiday traffic flow prediction and parking demand forecasting include: retrieving historical holiday traffic data, collecting real-time dynamic data, and expanding related event information; and generating, through predictive methods, road traffic flow, regional parking demand, and congestion nodes for a preset time period based on the historical holiday traffic data, real-time dynamic data, and related event information, as prediction data.

[0007] In possible implementations, identifying holiday travel scenarios includes: using a scenario recognition engine to identify holiday travel scenarios, wherein the scenario recognition engine takes predicted data and target travel requests as input, and uses destination geographical attributes, overlap of peak hours, and spatial distribution characteristics of traffic flow as scenario matching criteria.

[0008] In one possible implementation, a parking space recommendation scheme is generated based on a multi-dimensional decision model, including: calculating the comprehensive suitability of each candidate parking lot through the multi-dimensional decision model to generate a parking space recommendation list; determining a parking space recommendation scheme based on the parking space recommendation list; wherein, when the parking space saturation in the core area exceeds a preset threshold, cross-regional collaborative scheduling is initiated, and secondary area parking lots are recommended and a connection scheme is matched.

[0009] In a possible implementation, the multi-dimensional decision-making model includes: the multi-dimensional decision-making model adopts a weighted scoring logic, wherein the weighted scoring logic includes at least parking space dimension score, path dimension score, preference dimension score based on user profile, and scene dimension score based on scene rule base; wherein the scene rule base is a set of recommendation priorities preset for multiple scenes.

[0010] In a possible implementation, the adaptive parking space recommendation method for holiday traffic fluctuations further includes: the entire travel path includes the main driving path from the origin to the parking lot, the parking space guidance path based on the internal layout planning of the parking lot, and the connection path based on the connection resource planning.

[0011] In a possible implementation, the adaptive parking space recommendation method for holiday traffic fluctuations further includes: introducing a conflict coordinator, wherein the conflict coordinator is optionally triggered in the parking space recommendation stage and the route planning stage; when a planning conflict exists, the conflict coordinator is triggered, and the weight allocation is dynamically adjusted according to the scenario type and priority label.

[0012] In possible implementations, after displaying the information on the terminal interface, the following steps are taken: continuously monitoring the status of recommended parking spaces and planned routes through parallel data streams; and triggering a replanning of the scheme when at least one of the recommended parking space status and the planned route status deviates and meets a preset threshold.

[0013] This application proposes an adaptive parking space recommendation method for holiday traffic fluctuations. The method obtains the user's target travel request, including at least the destination and travel time. By predicting holiday traffic flow and parking demand, and combining this with the target travel request, it identifies holiday travel scenarios and loads a scenario rule base, including recommendation priority rules and route planning preferences. By integrating real-time parking space status, traffic information, and user profiles, it generates a parking space recommendation scheme based on a multi-dimensional decision model. Finally, it plans the entire travel route and displays it on the terminal interface. This method solves the technical problems of existing technologies, such as the acute contradiction between parking space supply and demand during holidays, poor adaptability of parking space recommendations leading to difficulties in finding parking spaces for users and exacerbating surrounding traffic congestion. It achieves the technical effect of intelligent and efficient matching and recommendation of parking space resources through scenario prediction, improving user travel experience and regional traffic efficiency. Attached Figure Description

[0014] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings of the embodiments of the present invention will be briefly described below. Flowcharts are used in this application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed precisely in sequence. Instead, various steps can be processed in reverse order or simultaneously as needed. Furthermore, other operations can be added to these processes, or one or more steps can be removed from these processes.

[0015] Figure 1 is a flowchart illustrating an adaptive parking space recommendation method for holiday traffic fluctuations provided in an embodiment of this application. Detailed Implementation

[0016] To further illustrate the technical means and effects adopted by the present invention in order to achieve the intended purpose, the following detailed description is provided in conjunction with the accompanying drawings and preferred embodiments, based on the specific implementation methods, structures, features and effects of the present invention.

[0017] This application provides an adaptive parking space recommendation method for holiday traffic fluctuations, as shown in Figure 1. The method includes:

[0018] Step S100: Obtain the target travel request from the user's end, wherein the target travel request includes at least the destination and travel time.

[0019] Preferably, the system acquires request objects containing structured data actively submitted by users through terminal interfaces such as mobile applications and in-vehicle navigation systems to determine the target travel request. This request includes at least the destination and travel time. The destination refers to the final location information that the user plans to go to, usually in the form of geographical coordinates, structured addresses, such as No. xx, xx Road, xx District, xx City, or a point name with clear geographical significance, such as "South Gate of xx Museum," which is used to determine the final spatial direction of the parking demand. The travel time refers to the specific time point or time period that the user plans to arrive at the destination or start their journey, such as 14:30 on January 1, 2026, or 3 pm today. This time is used to combine with holiday traffic prediction models to determine the traffic state stage that the user's journey will be in, such as off-peak period, peak period, or congestion accumulation period, and is a key input for achieving time-dimensional scenario matching.

[0020] Step S200: By predicting holiday traffic flow and parking space demand, and combining the target travel request, identify holiday travel scenarios and load the scenario rule base, wherein the scenario rule base includes recommendation priority rules and route planning preferences.

[0021] Preferably, by predicting holiday traffic flow and parking demand, historical big data, real-time data, weather, and event information are used to predict traffic flow during the planned travel period and the availability of parking spaces at the destination. During special periods such as holidays with surges in traffic flow, parking shortages, and variable road conditions, more intelligent, personalized, and holiday-appropriate travel planning suggestions are provided to users' target travel requests. Specifically, traffic patterns, bottlenecks, and optimization goals vary significantly across different scenarios. Commute scenarios prioritize efficiency, while holiday scenarios may focus more on avoiding congestion and ensuring parking availability. The system then determines whether the current or planned time falls within a holiday period and, based on the user's travel request and destination (e.g., popular tourist attractions or transportation hubs), automatically identifies and determines the holiday travel scenario. When a holiday travel scenario is identified, the system invokes a scenario rule base designed for that scenario. This rule base is a pre-defined set of strategies and knowledge, containing two key rule categories: recommendation priority rules and route planning preferences. The recommendation priority rules determine the sorting logic when recommending routes to users. For example, holiday priority rules might be set such that routes with reserved parking spaces are prioritized over routes with the shortest travel time, routes with low congestion probability are prioritized over routes with the shortest distance, and public transportation connections are prioritized over pure driving options when parking spaces at scenic spots are predicted to be saturated. Route planning preferences refer to adjusting the weight parameters in route planning, such as avoiding increasing the weight of toll stations during peak hours, increasing the real-time congestion penalty value for roads around scenic spots, and increasing buffer time to cope with unpredictable delays during holidays.

[0022] Furthermore, step S200 also includes retrieving historical holiday traffic data, collecting real-time dynamic data, and expanding related event information; based on the historical holiday traffic data, real-time dynamic data, and related event information, predictively generating road traffic flow, regional parking space demand, and congestion nodes for a preset time period as predictive data.

[0023] Preferably, historical holiday traffic data is retrieved, i.e., traffic data from the same holiday or similar weekends over the past few years, including historical traffic flow, average speed, congested road sections, and parking lot occupancy rate curves. Real-time dynamic traffic data is then collected, including real-time traffic conditions obtained through map apps, current parking space occupancy information obtained using parking lot IoT sensors, and traffic camera footage and highway toll station passage speeds, reflecting the current initial state. Simultaneously, extended related event information is acquired as external interference factors, i.e., planned events or emergencies that may affect traffic, such as large-scale events like concerts and sporting events, road construction announcements, temporary traffic control information, weather forecasts, geological disaster warnings, and sudden traffic accidents. Based on historical holiday traffic data, real-time dynamic data, and related event information, a spatiotemporal prediction model learns patterns from historical data and corrects them by combining real-time status and event information to generate road traffic flow for a preset time period, such as the number of vehicles passing through key road sections in a future time period, regional parking space demand, such as the total number of parking spaces required in a target area in a future time period, and congestion nodes, i.e., predicting specific locations of potential congestion, the expected start / end time of congestion, and its severity.

[0024] Furthermore, step S200 also includes using a scene recognition engine to identify holiday travel scenarios, wherein the scene recognition engine takes predicted data and target travel requests as input, and uses destination geographical attributes, overlap of peak hours, and spatial distribution characteristics of traffic flow as scene matching criteria.

[0025] Preferably, a scene recognition engine is used to accurately determine whether a user's target travel request belongs to a holiday travel scenario. This scene recognition engine is a trained machine learning classifier that takes predicted data and the target travel request as input, and uses destination geographical attributes, peak-hour overlap, and traffic flow spatial distribution characteristics as scene matching criteria for comprehensive analysis, outputting scene recognition results. Destination geographical attributes are the most basic scene classification labels, including the destination's point of interest type and functional area attributes. For example, a destination that is a theme park / 5A-level scenic spot strongly points to a holiday leisure tourism scenario; a destination that is a transportation hub airport or high-speed rail station points to a holiday commuting / connection scenario; and a destination that is a large shopping mall / business district points to a holiday shopping and entertainment scenario. The user's planned travel time is compared with the regional peak traffic hours shown in the predicted data to calculate the overlap of peak hours and determine whether the user has entered a congested area. For example, if a user plans to go to a scenic spot at 10:00 AM, and the predicted data shows that roads around the scenic spot are congested from 9:30 AM to 11:30 AM, the overlap is extremely high, further reinforcing the judgment of a holiday leisure tourism scenario, and indicating that it belongs to the peak congestion sub-scenario within that scenario. Analyzing the spatial pattern of traffic flow along the predicted path from origin to destination or within a wider area determines the spatial distribution characteristics of traffic flow, such as smooth flow, point-like congestion, corridor-like congestion, or area-wide congestion. This information is used to determine the scope and form of congestion in order to match more refined response strategies. For example, if the destination is a scenic spot and the prediction shows area-wide congestion, extreme strategies such as strong traffic control, strong diversion, and recommended connecting routes may be triggered. If the destination is a commercial area and the prediction shows point-like congestion only at the parking lot entrance, strategies such as prioritizing reservation parking and optimizing the last 500 meters of the route may be triggered.

[0026] Step S300: Read the holiday travel scenario and scenario rule base, and generate a parking space recommendation scheme based on a multi-dimensional decision model by integrating real-time parking space status, road condition information and user profile.

[0027] Preferably, the system acquires traffic scene recognition results and corresponding scene rule bases, along with real-time parking space status (current available parking spaces, reservation status, real-time prices, and entry / exit queue times for all available parking lots within the target area); road condition information (real-time travel time, congestion status, and traffic events from the origin to each alternative parking lot and from the parking lot to the final destination); and user profiles (users' personal preferences and historical behavior data, such as price sensitivity, convenience preference, walking tolerance, and vehicle type restrictions). This data, along with holiday travel scenarios and the scene rule base, is integrated to generate parking space recommendation schemes through a multi-dimensional decision model. The decision dimensions may include multiple objectives such as total time cost, economic cost, convenience, reliability, and congestion level. The multi-dimensional decision model seeks the optimal balance among these objectives. Specifically, it evaluates all candidate parking spaces within the area that meet basic conditions, scores each candidate scheme across various dimensions, and weights and combines the scores according to the recommendation priority rules in the scene rule base and the weight preferences of the user profiles. Finally, it outputs parking space recommendation schemes ranked by overall utility.

[0028] Furthermore, step S300 also includes calculating the comprehensive suitability of each candidate parking lot through a multi-dimensional decision model to generate a parking space recommendation list; determining a parking space recommendation scheme based on the parking space recommendation list; wherein, when the parking space saturation in the core area exceeds a preset threshold, cross-regional collaborative scheduling is initiated to recommend secondary area parking lots and match connection schemes.

[0029] Preferably, the multi-dimensional decision-making model weights and scores candidate parking lots within the region based on multiple dimensions such as time cost, economic cost, convenience, reliability, and congestion pressure. It calculates the comprehensive suitability of each candidate parking lot, assigning a comprehensive score to each parking lot representing its matching degree to the current user request. Then, parking spaces are sorted from highest to lowest comprehensive suitability to generate a parking space recommendation list and determine the preferred parking space recommendation scheme within the core area. When the parking space saturation in the core area exceeds a preset threshold, which is set based on historical data and experience—for example, if the real-time occupancy rate of all parking lots in the core area is >95% or the reservation occupancy rate for the next 30 minutes is 100%—it indicates that the parking supply in the core area is nearing or has reached its limit. Continuing to divert vehicles to the core area will lead to search-based congestion on surrounding roads. If vehicles are forced to circle around looking for parking spaces, causing traffic congestion and a poor user experience, cross-regional collaborative scheduling will be initiated to recommend parking lots in secondary areas. These secondary areas refer to areas that are some distance from the core area but can be easily connected via expressways or public transportation. Examples include parking lots attached to subway stations outside scenic spots, vacant parking spaces at large transportation hubs, and surplus parking lots in less popular commercial areas. The calculation scope of the multi-dimensional decision-making model will be expanded from the core area to the secondary area to assess the suitability of peripheral parking lots, which are usually cheaper and have more parking spaces. Then, based on real-time information, the optimal connection method from the secondary parking lot to the final destination will be matched for the user, and the total travel time and total cost will be calculated. Connection options will be matched, such as recommending rail transit connections, shuttle bus connections, shared bicycle / electric vehicle connections, or short-distance ride-hailing connections.

[0030] Furthermore, step S300 also includes the multi-dimensional decision model employing a weighted scoring logic, wherein the weighted scoring logic includes at least parking space dimension score, path dimension score, preference dimension score based on user profile, and scene dimension score based on scene rule base; wherein the scene rule base is a set of recommendation priorities preset for multiple scenes.

[0031] Preferably, the multi-dimensional decision-making model employs a weighted scoring logic, which includes at least parking space dimension scores, path dimension scores, user profile-based preference dimension scores, and scenario dimension scores based on a scenario rule base. The weights of each dimension are dynamically adjusted by the scenario rule base. The parking space dimension score is used to evaluate the attributes and status of the parking lot itself, with scoring indicators including the number / saturation of parking spaces, reservation success rate, parking fee per unit, parking space size, facility condition, safety rating, and walking distance to the final destination. The path dimension score is used to evaluate the driving path from the origin to the parking lot, with scoring indicators including predicted travel time, path congestion level, number of traffic lights, path complexity, and road condition risk. The user profile's preference dimension score assesses the match between the proposed solution and the user's personal preferences. Scoring metrics include price preference, convenience preference, and time preference. The scenario dimension score, based on a scenario rule base, assesses the contribution of the proposed solution to the macro-level goals of a specific scenario. Scoring metrics are determined by priority rules defined in the scenario rule base. For example, in holiday tourist scenarios, solutions with high arrival probability and those alleviating congestion in the core area may receive extremely high positive weights, while solutions that are slightly further away but guarantee parking and utilize shuttle services will receive very high bonuses. In daily commuting scenarios, solutions with certain time and lowest cost may receive the highest weights. In rainy weather scenarios, indoor parking and shortest walking distance may receive significantly increased weights to ensure that personal recommendations align with overall traffic management goals. The scenario rule base is used to pre-set recommendation priorities for different scenarios, including peak tourist seasons, large event exits, commuting in inclement weather, daily commuting, and nighttime travel. It clearly defines the recommendation dimensions that should be prioritized in specific scenarios, thereby ensuring accurate, flexible, and strategic parking space recommendations.

[0032] Step S400: Based on the parking space recommendation scheme, plan the entire travel route and display it on the terminal interface.

[0033] Step S400 further includes the following: the full-link travel path includes the main driving path from the departure point to the parking lot, the parking space guidance path based on the internal layout planning of the parking lot, and the connection path based on the connection resource planning.

[0034] Preferably, the structured parking space recommendation scheme is transformed into a comprehensive, step-by-step, immersive main driving route from the origin to the parking lot, a parking space guidance route based on the internal layout planning of the parking lot, and a connection route based on the planning of connecting resources. This determines the entire travel path and guides users in complex travel scenarios. Specifically, the main driving route from the origin to the parking lot refers to the driving route from the user's current origin to the entrance of the recommended parking lot or a specific gate. The destination is a specific parking space resource with status attributes. The route planning may be strongly bound to the real-time status of the parking space. For example, if 15 minutes have been reserved for you, the route may choose a faster route to arrive before the reservation expires. During the navigation process, the interface not only displays road information but also continuously updates the status of the recommended parking space, such as 8 minutes remaining and the current parking space still being available.

[0035] Preferably, the parking guidance route refers to the route that guides users to a recommended or reserved parking space after they drive into the parking lot. It relies on a high-precision indoor map of the parking lot that is accurate to the parking space, and takes into account micro-elements such as one-way streets, speed bumps, pillars, occupied parking spaces, and pedestrian walkways inside the parking lot to plan the most efficient driving route. During the guidance process, it combines parking lot camera or vehicle sensor data to adjust the route in real time to avoid temporary obstacles or oncoming vehicles. Furthermore, on the terminal interface, the camera combined with AR technology overlays arrows on the real image to intuitively guide users to "turn left, your parking space is in area A023".

[0036] Preferably, the connecting route is the last leg of the journey from the user's parked car to the final destination. Based on the matching connecting resources in the plan, corresponding navigation is generated. Specifically, if the connecting route is on foot, walking navigation is generated from the parking space to the bus stop / subway station / scenic spot entrance, taking into account pedestrian crossings, overpasses, and stairs; if the connecting route is a shuttle bus, the user is guided to the correct pick-up point and the estimated arrival time of the next bus is displayed; if the connecting route is a shared bicycle, the user is guided to the nearest available bicycle point and the riding route is planned; if the connecting route is a short-distance ride-hailing service, a ride-hailing interface may pop up directly or the user may be guided to the designated ride-hailing pick-up area. Ultimately, the three routes are seamlessly integrated on the user's navigation app or in-vehicle screen, achieving a seamless transition. When the user confirms the route, the interface displays a continuous, color-coded line winding from the starting point to the final destination, with different icons or colors distinguishing driving, walking, and public transportation segments. In the first stage, the user follows the main driving route. When the vehicle enters the parking lot entrance, the navigation automatically and seamlessly switches from city road navigation to the parking lot's internal parking guidance interface. In the second stage, the user follows AR or 3D overhead views to a specific parking space. Once the vehicle has stopped or the user manually clicks the "parked" button, the interface automatically switches to the navigation guidance for the third connecting route. Throughout the entire process, an information card may always be displayed in a corner of the interface, dynamically showing key information such as the current destination, connecting vehicle status, and total time updates.

[0037] Furthermore, step S400 also includes continuously monitoring the status of recommended parking spaces and planned routes through parallel data streams; when at least one of the status of recommended parking spaces and planned routes deviates and meets a preset threshold, a replanning of the scheme is triggered.

[0038] Preferably, the system continuously monitors the real-time status of the specific parking space recommended to the user through parking lot IoT sensors, parking lot management API, and status feedback from the reservation platform. This includes monitoring the occupancy status, reservation availability status, and attribute status to determine the status of the recommended parking space. Real-time traffic data streams and traffic incident reports are used to monitor the real-time traffic conditions of the entire planned route from the origin to the parking space and subsequent connections. This includes monitoring the current actual or predicted travel time, whether new traffic accidents, traffic controls, or severe congestion have occurred on the route, and whether any road sections are completely closed due to unforeseen circumstances to determine the status of the planned route. The monitored real-time status is then continuously compared with the predicted / expected status. If at least one of the recommended parking space status and the planned route status deviates and meets a preset threshold, the deviation may be due to the original plan's expectation of "parking space available and reserved," but the monitoring shows "parking space occupied" or "reservation canceled for some reason." Similarly, a deviation in the route status may be due to the original plan predicting "main driving route reachable in 25 minutes," but the monitoring shows "due to an accident ahead, at least 50 minutes are required." If the travel time deviation exceeds M minutes, the status of key resources changes fundamentally, or the probability of arriving on time and successfully using a parking space based on new data falls below a certain threshold, it indicates that the preset threshold has been met. In this case, a replanning of the route is triggered. This means that the user's current location is used as a new starting point, the latest parking space status and real-time traffic conditions are integrated, and the multi-dimensional decision model is run again to find an alternative parking space in the same parking lot or switch to another more feasible parking lot and route combination. On the terminal interface, a new recommended route is provided in a way with minimal interference, such as a voice prompt "Severe congestion detected ahead. A better solution has been found for you. Please confirm the switch." This also achieves seamless updates of the navigation route.

[0039] Furthermore, the adaptive parking space recommendation method for holiday traffic fluctuations also includes the introduction of a conflict coordinator, wherein the conflict coordinator is optionally triggered during the parking space recommendation stage and the route planning stage; when a planning conflict exists, the conflict coordinator is triggered, and the weight allocation is dynamically adjusted according to the scenario type and priority label.

[0040] Preferably, planning conflicts typically occur between two phases or between different user requests. These conflicts may include logical conflicts between phases, such as in the parking space recommendation phase, where parking space 001 in parking lot P is recommended to user A based on a high reliability dimension, but the route planning phase reveals that the only internal passage leading to parking space 001 is temporarily closed for construction, causing the planning to fail; resource competition conflicts, such as recommending the same parking space 002 in parking lot P to two users B and C almost simultaneously based on a high arrival probability; and multi-objective optimization conflicts, such as in extreme congestion scenarios where the scenario rules require priority diversion, recommending a more distant secondary parking lot to user D, but the user profile shows that the user is pregnant and has a strong personal preference dimension that requires the shortest walking distance, thus the diversion objective conflicts with the user's urgent personal needs.

[0041] Preferably, continuous monitoring allows for the optional activation of the conflict coordinator when a conflict is detected. Information such as the conflict type, involved users / solutions, and conflict points is input. Based on the scenario type and priority tags, the weight allocation is dynamically adjusted. This involves adjusting the weights of each dimension in the multi-dimensional decision-making model or directly adjusting the candidate solution list. For example, for inter-stage logical conflicts, the system might command a return to the parking space recommendation stage, temporarily and significantly increasing the weight of the path accessibility dimension to recalculate and avoid construction areas, selecting another accessible parking space. For resource competition conflicts, resources are assigned to one party based on priority tags, and a new recommendation calculation is immediately triggered for the other party, excluding the occupied resource, while potentially granting them some compensatory weight. For multi-objective optimization conflicts, in holiday emergency rescue scenarios, the system might determine that global diversion has a higher priority, maintaining the original solution, but simultaneously activating privileges for pregnant users—notifying parking lot administrators to provide electric vehicle shuttles to compensate for their walking distance loss. In daily commuting scenarios, the system might prioritize individual needs, increasing the weight of the preference dimension to find a closer parking space for them.

[0042] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. An adaptive parking space recommendation method for holiday traffic fluctuations, characterized in that, The method includes: acquiring a user's target travel request, wherein the target travel request includes at least a destination and a travel time; identifying holiday travel scenarios by combining holiday traffic prediction and parking space demand forecast with the target travel request, and loading a scenario rule base, wherein the scenario rule base includes recommendation priority rules and route planning preferences; reading the holiday travel scenarios and the scenario rule base, and generating a parking space recommendation scheme based on a multi-dimensional decision model by integrating real-time parking space status, road condition information, and user profiles; planning a full-link travel route according to the parking space recommendation scheme, and displaying it on the terminal interface.

2. The adaptive parking space recommendation method for holiday traffic fluctuations as described in claim 1, characterized in that, The method involves predicting traffic flow and parking demand during holidays, including: retrieving historical holiday traffic data, collecting real-time dynamic data, and expanding related event information; and generating predictive data such as road traffic flow, regional parking demand, and congestion nodes for a preset time period based on the historical holiday traffic data, real-time dynamic data, and related event information.

3. The adaptive parking space recommendation method for holiday traffic fluctuations as described in claim 2, characterized in that, Identifying holiday travel scenarios includes: using a scene recognition engine to identify holiday travel scenarios, wherein the scene recognition engine takes predicted data and target travel requests as input, and uses destination geographical attributes, overlap of peak hours, and spatial distribution characteristics of traffic flow as scene matching criteria.

4. The adaptive parking space recommendation method for holiday traffic fluctuations as described in claim 1, characterized in that, The method for generating parking space recommendation schemes based on a multi-dimensional decision model includes: calculating the comprehensive suitability of each candidate parking lot using the multi-dimensional decision model to generate a parking space recommendation list; determining a parking space recommendation scheme based on the parking space recommendation list; wherein, when the parking space saturation in the core area exceeds a preset threshold, cross-regional collaborative scheduling is initiated, and secondary area parking lots are recommended and matched with connection schemes.

5. The adaptive parking space recommendation method for holiday traffic fluctuations as described in claim 1, characterized in that, The multi-dimensional decision-making model includes: the multi-dimensional decision-making model adopts a weighted scoring logic, wherein the weighted scoring logic includes at least parking space dimension score, path dimension score, preference dimension score based on user profile, and scene dimension score based on scene rule base; wherein the scene rule base is a set of recommendation priorities preset for multiple scenes.

6. The adaptive parking space recommendation method for holiday traffic fluctuations as described in claim 1, characterized in that, The entire travel route includes the main driving route from the departure point to the parking lot, the parking space guidance route based on the internal layout planning of the parking lot, and the connecting route based on the connecting resources planning.

7. The adaptive parking space recommendation method for holiday traffic fluctuations as described in claim 1, characterized in that, The method further includes: introducing a conflict coordinator, wherein the conflict coordinator is optionally triggered in the parking space recommendation stage and the route planning stage; when there is a planning conflict, the conflict coordinator is triggered, and the weight allocation is dynamically adjusted according to the scenario type and priority label.

8. The adaptive parking space recommendation method for holiday traffic fluctuations as described in claim 1, characterized in that, After being displayed on the terminal interface, the process includes: continuously monitoring the status of recommended parking spaces and planned routes through parallel data streams; and triggering a replanning of the plan when at least one of the recommended parking space status and the planned route status deviates and meets a preset threshold.

Images