A smart park parking space proactive recommendation system that combines user profiles with real-time popularity.

Abstract

This invention discloses a smart park parking space proactive recommendation system that combines user profiles and real-time popularity, belonging to the field of parking space recommendation technology. The key technical points include: acquiring historical parking space usage information of the smart park; obtaining parking space benchmarking elements based on the historical parking space usage information; wherein, the historical parking space usage information includes parking space feature parameters, recommendation classification results, and popularity evolution process; the parking space benchmarking elements include element number, element feature parameters, and element recommendation tags; determining the impact of parking space popularity fluctuations on the supply-demand matching relationship based on the degree of fit to obtain a supply-demand popularity deviation base; correcting the deviation between the rigidity of parking demand and the stability of parking space popularity based on the supply-demand popularity deviation base to obtain a profile popularity difference coefficient; and outputting target parking space recommendation results for users to be recommended based on the profile popularity difference coefficient, thereby improving the overall utilization rate of parking spaces in the park.

Description

Technical Field

[0001] This invention relates to the field of parking space recommendation technology, and more specifically, to a smart park parking space proactive recommendation system that combines user profiles with real-time popularity. Background Technology

[0002] Current smart park parking space recommendations use a fixed parameter matching method, deriving recommendation coefficients based on static hardware information such as parking space size and location. This makes it difficult to adapt to the dynamically changing parking environment within the park. Existing parking space matching benchmarks only record the inherent hardware parameters of the parking spaces, failing to combine long-term parking space usage patterns and historical recommendation results to generate standardized reference data. This easily leads to recommendation errors where the parking space hardware meets the standards but the scenario is incompatible. It also ignores real-time changes in parking space popularity due to pedestrian and vehicle traffic within the park, and cannot synchronously collect real-time parking space turnover data. The recommendation results can only adapt to normal scenarios with stable traffic flow in the park, easily resulting in the problem of recommended parking spaces being temporarily occupied. Summary of the Invention

[0003] To address the shortcomings of existing technologies, the purpose of this invention is to provide a smart park parking space proactive recommendation system that combines user profiles with real-time popularity data.

[0004] To achieve the above objectives, the present invention provides the following technical solution: A smart park parking space proactive recommendation system that combines user profiles with real-time popularity includes: Acquisition Module: Acquires historical parking space usage information of the smart park, and obtains parking space benchmarking elements based on the historical parking space usage information; wherein, the historical parking space usage information includes parking space feature parameters, recommendation classification results, and popularity evolution process; the parking space benchmarking elements include element number, element feature parameters, and element recommendation tags; Analysis module: Extract user profile features and real-time parking space popularity features of the users to be recommended, and analyze user profile features, real-time parking space popularity features, parking space benchmarking elements and inherent attribute features to obtain a set of parking space resource matching elements; Processing module: The module performs benchmarking and matching screening on the set of parking space resource matching elements to obtain target benchmarking elements; it judges the degree of matching between the actual parking demand of users and the matching status of parking space supply based on the target benchmarking elements; and it determines the impact of parking space heat fluctuations on the supply and demand matching relationship based on the degree of matching to obtain the supply and demand heat deviation baseline. Correction module: Based on the supply and demand heat deviation baseline, the module corrects the rigidity of parking demand and the stability of parking space heat to obtain the portrait heat difference coefficient; Output module: Based on the user profile popularity difference coefficient, output the target parking space recommendation results for the user to be recommended.

[0005] Preferably, the benchmarking elements for parking spaces are obtained based on historical parking space usage information, specifically including the following steps: Based on historical parking space usage information, the inherent usage characteristics of parking spaces, historical recommendation and adaptation results, and the evolution of parking space popularity throughout the time period are integrated to obtain historical parking space usage information. By hierarchically decomposing historical parking space usage information, we can obtain stable usage characteristics and scenario adaptation features of parking spaces. Based on the stable use characteristics and scenario adaptation features of parking spaces, parking space adaptation service scenarios are divided, and the parking space service capabilities of different parking space adaptation service scenarios are determined. Based on the parking space service capabilities, the basic elements for parking space benchmarking are obtained. Based on the historical recommendation adaptation results, the basic elements for parking space benchmarking are verified in the scenario to obtain the parking space benchmarking elements.

[0006] Preferably, the user profile features of the user to be recommended and the real-time parking space popularity features are extracted, specifically including the following steps: Based on historical parking behavior data and identity attribute data, we can obtain parking habit preferences and fixed parking needs; User profiles are created based on parking habits, preferences, and fixed parking needs. Collect real-time parking space usage status and dynamic pedestrian and vehicle flow data within the park; The real-time parking space activity level is determined by the real-time parking space usage status and dynamic pedestrian and vehicle flow data in the park, thus obtaining the real-time parking space popularity characteristics.

[0007] Preferably, the analysis of user profile features, real-time parking space popularity features, parking space benchmarking elements, and inherent attribute features yields a set of parking space resource adaptation elements, specifically including the following steps: Generate corresponding parking space recommendation elements based on user profile characteristics and real-time parking space popularity characteristics; Based on the parking space recommendation elements and parking space benchmarking elements, determine the matching benchmarking elements and extract the inherent attribute features of the parking needs of the users to be recommended. A profile of inherent parking demand is created based on matching benchmarking elements and inherent attribute characteristics; Based on the existing parking demand profile, determine the set of parking resource adaptation elements for various types of parking resources in the park; wherein, the set of parking resource adaptation elements includes service adaptation attributes and dynamic circulation service attributes.

[0008] Preferably, the set of matching elements for parking space resources is subjected to benchmarking and matching screening to obtain target benchmarking elements, which specifically includes the following steps: Verify the feature matching consistency of the parking space resource adaptation element set to obtain a preliminary matching element set; Based on the initial set of matching elements and the characteristics of parking demand adaptation, the target benchmarking elements are obtained.

[0009] Preferably, determining the degree of fit between users' actual parking needs and parking space supply based on target benchmarking factors specifically includes the following steps: The target benchmarking elements are processed to obtain an overall adaptation representation; The overall adaptation representation is divided into multiple fitting gradients, so that each fitting gradient corresponds to the supply and demand matching attribute. Based on the static parking space inventory, adjust the supply and demand matching attributes corresponding to each matching gradient to obtain the degree of matching between users' actual parking needs and the parking space supply.

[0010] Preferably, the target benchmarking elements are processed to obtain an overall adaptation representation, specifically including the following steps: The target elements are broken down item by item to obtain the user parking demand sub-items and the parking space supply sub-items; By comparing the user parking demand sub-items and the parking space supply sub-items, we can determine the overlap and gaps in the content. Determine the basic matching volume of demand and supply based on the scope of content overlap, and determine the basic deviation volume of demand and supply based on the scope of content gaps. The basic fit volume and the basic deviation volume are integrated to form an overall fit representation.

[0011] Preferably, the impact of parking space demand fluctuations on the supply-demand matching relationship is determined based on the degree of fit, resulting in a supply-demand demand deviation baseline. This specifically includes the following steps: The overall baseline status of the supply and demand matching of parking spaces in the park is defined based on the degree of fit. Capture the fluctuations in parking space popularity caused by dynamic changes in vehicle and pedestrian traffic within the park; The positive and negative impacts of parking space demand fluctuations on the supply-demand matching relationship can be determined based on the fluctuations in parking space demand and the overall baseline status. Based on the effects of positive gain and negative interference, the supply and demand mismatch amplitude of heat fluctuations with different degrees of fit is analyzed; The overall deviation of the supply and demand matching offset is integrated to obtain the base of supply and demand heat deviation.

[0012] Preferably, the profile heat difference coefficient is obtained by correcting the rigidity of parking demand and the stability of parking space heat based on the supply and demand heat deviation baseline, specifically including the following steps: Based on the supply-demand temperature deviation baseline, the matching offset of the rigidity of user parking demand is corrected to obtain the demand correction attribute; Based on the supply and demand heat deviation baseline, the resource offset of stable parking space heat is corrected to obtain the parking space correction attribute; The portrait heat difference coefficient is obtained by integrating and correcting the demand-adjusted attributes and parking space-adjusted attributes.

[0013] Preferably, the method of outputting target parking space recommendations for users based on the profile heat difference coefficient includes the following steps: Candidate parking spaces that match user parking preferences and are compatible with the park's real-time supply and demand status are selected based on the user profile popularity difference coefficient. The candidate parking spaces are ranked according to their suitability to obtain the target parking space recommendation result.

[0014] Compared with the prior art, the present invention has the following beneficial effects: This invention generates standardized parking space benchmarking elements based on full-cycle historical parking space usage information. It integrates user profile features with real-time parking space popularity characteristics, taking into account both long-term user parking habits and instantaneous traffic flow changes within the park. It not only locks in inherent attributes such as user vehicle specifications, charging needs, and monthly parking privileges, but also simultaneously incorporates real-time fluctuations in pedestrian and vehicle traffic and parking space popularity. The generated parking space resource matching element set considers both static and dynamic attributes, achieving bidirectional dynamic matching between user needs and parking space resources. By filtering to obtain target benchmarking elements, the degree of fit is determined based on element breakdown, and a supply-demand popularity deviation base is generated by combining popularity fluctuations. This quantifies the supply-demand offset caused by instantaneous passenger flow surges, compensating for the inability of static matching to quantify the impact of popularity disturbances. Based on the supply-demand popularity deviation base, a deviation correction is applied to the rigidity of parking demand and the stability of parking space popularity to obtain a profile popularity difference coefficient. This eliminates calculation errors caused by static original scores not considering real-time popularity. Based on the profile popularity difference coefficient, target parking space recommendations are output for users to be recommended, prioritizing parking spaces with higher matching degrees, thereby improving the overall utilization rate of parking spaces in the park. Attached Figure Description

[0015] Figure 1 A schematic diagram of a smart park parking space proactive recommendation system that combines user profiles and real-time popularity is provided for an embodiment of the present invention; Figure 2 This is a schematic diagram illustrating the parking space matching elements obtained in a smart park proactive parking space recommendation system that combines user profiles and real-time popularity, as provided in an embodiment of the present invention. Detailed Implementation

[0016] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0017] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0018] Secondly, the term "an embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places throughout this specification does not necessarily refer to the same embodiment, nor is it a single embodiment or an embodiment selectively excluded from other embodiments.

[0019] Reference Figures 1-2 As shown.

[0020] The embodiments further illustrate the smart park parking space proactive recommendation system that combines user profiles and real-time popularity proposed in this invention.

[0021] A smart park parking space proactive recommendation system that combines user profiles with real-time popularity includes: Acquisition Module: Acquires historical parking space usage information of the smart park, and obtains parking space benchmarking elements based on the historical parking space usage information; among which, the historical parking space usage information includes parking space characteristic parameters, recommended classification results and popularity evolution process; parking space benchmarking elements include element number, element characteristic parameters and element recommended tags; Analysis module: Extract user profile features and real-time parking space popularity features of the users to be recommended, and analyze user profile features, real-time parking space popularity features, parking space benchmarking elements and inherent attribute features to obtain a set of parking space resource matching elements; Processing module: The module performs benchmarking and matching screening on the set of parking space resource matching elements to obtain target benchmarking elements; it judges the degree of matching between the actual parking demand of users and the matching status of parking space supply based on the target benchmarking elements; and it determines the impact of parking space heat fluctuations on the supply and demand matching relationship based on the degree of matching to obtain the supply and demand heat deviation baseline. Correction module: Based on the supply and demand heat deviation baseline, the module corrects the rigidity of parking demand and the stability of parking space heat to obtain the portrait heat difference coefficient; Output module: Based on the user profile popularity difference coefficient, output the target parking space recommendation results for the user to be recommended.

[0022] Based on historical parking space usage information, the benchmarking elements for parking spaces are obtained, specifically including the following steps: Based on historical parking space usage information, the inherent usage characteristics of parking spaces, historical recommendation and adaptation results, and the evolution of parking space popularity throughout the time period are integrated to obtain historical parking space usage information. The inherent usage characteristics of a parking space represent its hardware and location-related parameters, including the length and width of the space, the installation status of the supporting charging equipment, the straight-line distance from the main entrance of the park, and the upper limit of the vehicle specifications that can be parked on the floor of the building. For example, parking space No. 05 in Area A on the first basement floor of the park has fixed dimensions of 5.3 meters long and 2.5 meters wide, is equipped with a DC fast charging pile, is 120 meters away from the main entrance of the park, and is only suitable for parking small passenger vehicles. Historical recommendation and adaptation results include the total number of times users actually parked in the corresponding parking space after it was recommended, the frequency of users receiving the parking space recommendation but abandoning it and choosing other parking spaces, and the statistics of adaptation and compliance after different user groups selected parking spaces. The evolution of parking space popularity throughout the day is divided into the morning commuting peak from 7:00 to 9:00, the midday visitor period from 12:00 to 14:00, and the evening commuting peak from 18:00 to 21:00. The real-time occupancy time of parking spaces is recorded for each period, and the average occupancy rate of a single parking space in each period is summarized monthly to obtain the fluctuation pattern of parking space popularity over the entire year.

[0023] The total amount of historical parking space usage information = the total number of parameters of the inherent usage characteristics of parking spaces + the total amount of statistical data of historical recommendation and adaptation results + the total amount of time-series data of the evolution trend of parking space popularity throughout the time period. After summarizing and converting, the historical parking space usage information of a single parking space is obtained.

[0024] By hierarchically decomposing historical parking space usage information, we can obtain stable usage characteristics and scenario adaptation features of parking spaces. Stable parking space usage characteristics are inherent attributes of parking spaces, including space size, supporting facilities information, whether a monthly long-term rental agreement has been signed, whether the space is located near the park, whether it is a freight loading / unloading lane space, and ground load-bearing weight limits. For example, if parking space No. 12 in Zone B of the park is locked by an internal company for a monthly rental, this monthly occupancy attribute is directly classified as a stable parking space usage characteristic. Scene adaptation features are extracted from historical recommendation adaptation results and the evolution of parking space popularity throughout the day to reflect the adaptation performance of parking spaces in different parking environments. Historical recommendation adaptation results are used to extract the adaptation feedback after users with different identities select parking spaces, and the evolution of parking space popularity throughout the day is used to extract the changes in parking space occupancy under fluctuations in pedestrian flow at different times. For example, during peak visitor periods on statutory holidays, the parking space occupancy rate reaches 92%, while during weekdays, the occupancy rate is only 28%. Content that shows significant numerical fluctuations with scene changes is classified as a scene adaptation feature. The total number of parameters for stable usage characteristics of a single parking space = the number of fixed parameters selected from inherent usage characteristics; the total number of parameters for scene adaptation characteristics of a single parking space = the number of scene parameters extracted from the recommendation adaptation results + the number of time period parameters extracted from the trend of popularity evolution.

[0025] Based on the stable use characteristics and scenario adaptation features of parking spaces, parking space adaptation service scenarios are divided, and the parking space service capabilities of different parking space adaptation service scenarios are determined. Based on stable usage characteristics, the hardware specifications of vehicles that a parking space can accommodate are determined. Based on scenario adaptability characteristics, the user groups and usage periods suitable for the parking space are determined. This allows for the classification of basic service scenarios such as fixed commuter parking, temporary visitor parking, new energy vehicle charging parking, and temporary parking for large trucks. For example, a basement parking space equipped with a fast-charging station has stable usage characteristics including hardware adaptability for new energy vehicles. Its scenario adaptability characteristic shows that the parking space occupancy rate exceeds 80% during weekday morning and evening commuting hours; therefore, this parking space is classified as a new energy commuter parking scenario. An open-air parking space near the main entrance of the park without charging equipment has scenario adaptability characteristics showing a significant surge in occupancy during statutory holidays; therefore, this parking space is classified as a temporary visitor parking scenario.

[0026] After the scene division is completed, the service capacity of each parking space in its respective scene is calculated one by one. The service capacity value of a single parking space in a single scene = the maximum number of vehicles that can be parked in the parking space at one time × the total available idle time per day. The total available idle time per day = the total natural time of the day - the historical average daily occupancy time. The total natural time of the day is taken as 24 hours. Both the occupancy time and the idle time are measured in hours. The overall service capacity of all parking spaces in the same service scene = the sum of the service capacities of each parking space in the scene.

[0027] Based on the parking space service capabilities, the basic elements for parking space benchmarking are obtained. Based on the historical recommendation adaptation results, the basic elements for parking space benchmarking are verified in the scenario to obtain the parking space benchmarking elements.

[0028] Parking spaces are linked to basic elements, including their corresponding parking space number, basic feature parameters, and initial service capacity values. The total number of parameters for a single basic element equals the number of element number items plus the number of basic feature parameter items plus the number of service capacity value items for the corresponding scenario. The scenario verification process uses historical recommendation adaptation results as a benchmark, comparing the adapted service scenarios labeled with the basic elements with the actual adaptation performance after historical recommendations were implemented. If a basic element labels a parking space as suitable for temporary visitor parking scenarios, but historical recommendation data shows that the actual parking success rate after allocating the space to visitor users is lower than the preset standard value, then the service capacity value within that basic element needs to be corrected and lowered. The corrected service capacity value equals the original basic element service capacity value minus the service loss value calculated from historical adaptation failures. The service loss value calculated from historical adaptation failures represents the equivalent of invalid resource loss caused by recommendation matching errors. This value is converted to hourly units and is a quantitative result of the parking space resource vacancy loss caused by users abandoning parking after receiving parking space recommendations. After completing scene verification and parameter correction for all basic elements, invalid parameters with mismatched scene labels are removed, and corrected feature information and recommended label content are added. Finally, parking space benchmarking elements are generated, consisting of element number, element feature parameters, and element recommended labels.

[0029] Extracting user profile features and real-time parking space popularity features from the users to be recommended, specifically including the following steps: Based on historical parking behavior data and identity attribute data, we can obtain parking habit preferences and fixed parking needs; Historical parking behavior data includes the user's specific time of entering the park, vehicle departure time, location information of the parking space occupied during each parking session, whether the parking space charging facility was used, and the total number of actual parking sessions per month. For example, if a user chooses to enter the park at 7:40 am every day a total of 238 times, and the parking locations are concentrated in the parking spaces near the park's elevator lobby on the first basement level, and the DC charging device in the parking space was used during 162 of those parking sessions, these records are included in the user's historical parking behavior data.

[0030] Identity attribute data includes the user's identity category (e.g., employee of a company in the park or temporary visitor), the length and width specifications of the registered vehicle, whether the vehicle is a new energy vehicle, and whether the user has activated a monthly parking pass for the park. Historical parking behavior data and identity attribute data are broken down and analyzed to obtain parking habit preferences and fixed parking needs. Parking habit preferences are the user's self-selected parking tendencies, including preferred parking time periods, preferred parking space floors, and preferred charging facilities. Fixed parking needs are long-term, unchanging parking requirements; for example, new energy vehicle users must be matched with parking spaces with charging devices, and monthly pass users need to prioritize matching with parking spaces in their contracted fixed areas. The reference frequency of a single user's monthly parking habits = actual number of parking days on weekdays + actual number of parking days on weekends, with the unit of frequency uniformly set as times. For example, if a user completes 21 parking days on weekdays and 4 parking days on weekends, the reference frequency of a single user's monthly parking habits = 21 + 4 = 25 times.

[0031] User profiles are created based on parking habits, preferences, and fixed parking needs. User profile features are a standardized set of all parking-related attributes of a user. The parameters consist of various parameters corresponding to parking habit preferences and fixed parking needs. The parameters corresponding to parking habit preferences include preferred daily parking time, preferred parking space location, and frequency of charging facility usage. The parameters corresponding to fixed parking needs include vehicle size restrictions, vehicle energy type, and monthly parking privileges. The total number of user profile feature parameters = number of parking habit preference parameters + number of fixed parking need parameters. The statistical unit for parameters is uniformly set as an item. For example, a single user's parking habit preferences include 3 parameters: preference for 8 am, preference for parking on the ground floor, and frequent use of charging piles. Fixed parking needs include 2 parameters: new energy small passenger vehicles and possession of monthly parking privileges in the park. The total number of user profile feature parameters = 3 + 2 = 5 items.

[0032] Collect real-time parking space usage status and dynamic pedestrian and vehicle flow data within the park; Based on two types of data captured by the park's parking space geomagnetic induction equipment, the park's entrance and exit gate counting device, and the park's gate passenger flow capture equipment, the real-time usage status data of parking spaces provides feedback on the instantaneous state of whether a single parking space is occupied or vacant. The system records the specific time when the parking space status changes. For example, if a vehicle leaves parking space No. 09 on the first basement level at 11:10 a.m. on the same day, the parking space status changes to vacant, and the corresponding time data is entered into the system database.

[0033] The park's dynamic pedestrian and vehicle flow data is divided into vehicle flow data and pedestrian flow data. The vehicle flow data is based on the total number of vehicles entering and leaving the park per hour according to the entrance and exit gates. The pedestrian flow data is based on the total number of people walking into and leaving the park per unit time period according to the entrance and exit passenger flow equipment. The data collection cycle is 15 minutes.

[0034] The real-time parking space activity level is determined by the real-time parking space usage status and dynamic pedestrian and vehicle flow data in the park, thus obtaining the real-time parking space popularity characteristics.

[0035] Real-time parking space turnover activity measures the frequency with which parking spaces switch between vacant and occupied states within a fixed time period. A higher switching frequency indicates more active parking space turnover, resulting in a higher real-time parking space popularity score. The number of turnovers per parking space per cycle equals the number of times a parking space changes from vacant to occupied within the statistical period plus the number of times a parking space changes from occupied to vacant within the statistical period. For example, if a parking space sees 4 vehicles enter and occupy it, and 4 vehicles leave it within a single data collection period, the number of turnovers per parking space per cycle is 4 + 4 = 8. The fluctuations in real-time pedestrian and vehicle traffic within the park are then added to the turnover count. The higher the total number of vehicles entering the park at the same time, the higher the parking space popularity score for the same number of turnovers. Finally, by integrating the parking space's own turnover data and the surrounding area's passenger flow data, a standardized real-time parking space popularity feature is generated.

[0036] The analysis of user profile features, real-time parking space popularity features, parking space benchmarking elements, and inherent attribute features yields a set of parking space resource adaptation elements, specifically including the following steps: Generate corresponding parking space recommendation elements based on user profile characteristics and real-time parking space popularity characteristics; User profile features are divided into parking habit preference parameters and fixed parking demand parameters. Parking habit preference parameters are the user's long-term parking choices, including daily fixed parking time periods, preferred parking space floors, and charging facility selection habits. Fixed parking demand parameters are the hard parking conditions, including the registered vehicle model and specifications, vehicle energy type, and monthly parking privileges in the park.

[0037] Real-time parking space popularity features are generated based on parking space turnover frequency and park pedestrian and vehicle flow data. These features include the number of parking space status changes within the statistical period, the real-time available parking space inventory in the area where the parking space is located, and the fluctuation value of popularity brought about by the traffic flow in the area. This is used to characterize the current hot / cold and vacant status of parking spaces. The total number of parameters for parking space recommendation elements = the number of parameters included in user profile features + the number of parameters included in real-time parking space popularity features. For example, a user's user profile features include 5 valid parameters: parking at 8 am on weekdays, preferring underground parking spaces, frequently using charging piles, using small new energy vehicles, and having monthly parking privileges. The real-time parking space popularity features include 3 valid parameters: high turnover frequency of underground parking spaces during the morning peak, large real-time vacancy of open-air parking spaces, and high current traffic flow in the park. The total number of parameters for parking space recommendation elements = 5 + 3 = 8.

[0038] Based on the parking space recommendation elements and parking space benchmarking elements, determine the matching benchmarking elements and extract the inherent attribute features of the parking needs of the users to be recommended. Parking space benchmarking elements are baseline data validated using historical data. Each element is accompanied by a number, fixed feature parameters, and a scenario recommendation tag, including the hardware conditions and service limits of parking spaces in different scenarios. Each parameter of the parking space recommendation element is compared with the tag and feature parameters of the parking space benchmarking element. Items with identical parameter content are grouped into matching benchmarking elements. The number of parameter items in matching benchmarking elements is equal to the statistical number of overlapping parameters between the two types of elements.

[0039] The inherent attribute features focus on the hard conditions of the user's vehicle and identity, and are not affected by single parking selection or real-time changes in parking space popularity. These include the vehicle's length and width dimensions, vehicle power type, and eligibility for monthly parking service. For example, 6 out of the 8 parameters in the user's parking space recommendation elements overlap with the corresponding parameters of the new energy monthly parking space on the first basement level. That is, the matching benchmark elements include 6 parameters. The user's inherent attribute features extract 3 fixed parameters, namely, the total length of the vehicle is 4.7 meters, it is a pure electric vehicle, and it has activated the park's annual monthly parking service.

[0040] A profile of inherent parking demand is created based on matching benchmarking elements and inherent attribute characteristics; Matching benchmark elements define the parking service scenarios and hardware standards suitable for users from the perspective of parking space baseline data. The total number of parameters in the inherent parking demand profile = the number of parameters in the matching benchmark elements + the number of parameters in the user's inherent attribute characteristics. For example, if there are a total of 6 parameters in the matching benchmark elements and a total of 3 parameters in the user's inherent attribute characteristics, the total number of parameters in the inherent parking demand profile is 6 + 3 = 9. After the demand profile is formed, it is clearly marked that the user can only be matched with small new energy dedicated parking spaces on the first basement level equipped with charging facilities and open to monthly users, thereby excluding resources that do not meet the hard conditions, such as open-air parking spaces and large truck parking spaces.

[0041] Based on the existing parking demand profile, determine the set of parking resource adaptation elements for various types of parking resources in the park; among them, the set of parking resource adaptation elements includes service adaptation attributes and dynamic circulation service attributes.

[0042] Service adaptation attributes characterize whether the parking space's hardware and the surrounding environment match the user's inherent parking needs. These attributes include parking space dimensions, charging equipment availability, target user group classification, and the parking scenario, emphasizing the parking space's static inherent adaptability. Dynamic flow service attributes characterize the dynamic usage status of parking spaces affected by real-time pedestrian and vehicle traffic within the park. These attributes include the average number of parking space turnovers over a fixed period, real-time available vacancy time, and the impact coefficient of area popularity fluctuations, emphasizing the instantaneous dynamic resource changes of parking spaces. The system calculates the total number of adaptation element parameters for a single selected parking space using a formula: Total Adaptation Element Parameters for a Single Parking Space = Number of Service Adaptation Attribute Parameters + Number of Dynamic Flow Service Attribute Parameters. For example, a qualified parking space may have 4 service adaptation attributes and 3 dynamic flow service attributes, resulting in a total of 4 + 3 = 7 parameters. All parking space adaptation parameters are aggregated to form a complete set of parking space resource adaptation elements, including both static and dynamic popularity data.

[0043] The set of parking space resource matching elements is benchmarked and screened to obtain target matching elements, which includes the following steps: Verify the feature matching consistency of the parking space resource adaptation element set to obtain a preliminary matching element set; The parking space resource adaptation element set is fixedly divided into service adaptation attributes and dynamic circulation service attributes. Service adaptation attributes are static characteristics of parking spaces that remain unchanged over a long period of time, including physical length and width dimensions of the parking space, on-board charging equipment installation parameters, usage scenario parameters for parking space planning adaptation, and vehicle specification parameters for parking space restrictions. Dynamic circulation service attributes are dynamic characteristics that change in real time with the park's on-site environment, including average parking space vacancy time parameters within the statistical period, number of parking space occupancy switching parameters per unit time period, and real-time popularity fluctuation parameters for the area to which the parking space belongs. Each parking space entry is checked for logical conflicts between the parameters of the two types of attributes. If the service adaptation parameters and dynamic circulation parameters of the same parking space contradict each other, the corresponding contradictory parameters are directly removed. The remaining valid parameters without conflicts are then aggregated to generate a preliminary matching element set. The changes in parameters for a single parking space are calculated using a quantitative formula. The total number of original elements for a single parking space = the number of service adaptation attribute parameters + the number of dynamic transfer service attribute parameters. The number of initially matching element parameters for a single parking space = the total number of original elements for a single parking space - the number of contradictory feature items removed during verification for the same parking space. For example, parking space No. 03 on the first basement floor of the park has 4 service adaptation attributes: 5.3 meters long, 2.5 meters wide, equipped with DC fast charging pile, suitable for new energy commuting parking scenarios, and limited to small passenger vehicles. The dynamic transfer service attributes have 3 parameters: an average of 3 hours of idle time during weekday morning peak hours, an average of 6 parking space transfers per hour, and high traffic volume in the area during morning peak hours. There is no logical contradiction between the two types of attribute parameters. The total number of original elements for a single parking space = 4 + 3 = 7 items. There are no contradictory parameters to be removed. The number of initially matching element parameters for a single parking space = 7 - 0 = 7 items. The service compatibility attribute of open-air parking space No. 15 in the same area is marked as suitable for large trucks. The dynamic flow parameter records that the actual size of the parking space can accommodate small vehicles. The two parameters constitute a content conflict. The total number of original elements for a single parking space is 4 + 3 = 7. After removing 1 contradictory parameter, the number of preliminary matching element parameters for a single parking space is 7 - 1 = 6. After collecting all compliant parameters, a complete set of preliminary matching elements is formed.

[0044] Based on the initial set of matching elements and the characteristics of parking demand adaptation, the target benchmarking elements are obtained.

[0045] Parking demand matching features are generated by deconstructing and extracting the user's inherent parking demand profile. They include the user's objective hard constraints and daily parking preferences, namely the user's registered vehicle's exterior dimensions (length and width), vehicle power type, the user's high-frequency parking time periods, and whether the user has activated monthly parking access. Each parameter item in the initial matching element set is compared with the parking demand matching features. Items whose parameters do not match the user's needs are directly eliminated. The remaining effective elements with matching parameters are summarized and used to produce target benchmarking elements. The effective parameter volume after screening a single parking space is quantified by a calculation formula. The number of target benchmark parameters for a single parking space = the number of initially matching parameters for the corresponding parking space - the number of parameters that do not match the user's parking needs. For example, the parking needs of the user to be recommended include 4 parameters: total vehicle length 4.6 meters, pure electric new energy vehicle, parking before or after 8 am on weekdays, and holding a monthly parking pass for the underground area of ​​the park. Parking space No. 03 on the first basement floor has 7 initially matching parameters, all of which match the user's 4 requirements. There are no mismatched parameters, so the number of target benchmark parameters for a single parking space = 7 - 0 = 7, and it is successfully included in the target benchmark elements. Parking space No. 15 in the open air has 6 initially matching parameters, but it lacks configuration parameters related to charging equipment, which cannot meet the user's hard parking requirements for new energy vehicles. The corresponding mismatched parameters are removed. The remaining parameters cannot meet all the user's constraints, so this parking space is removed from the target benchmark elements. All effective elements retained after screening are integrated and archived to form target benchmark elements.

[0046] Determining the degree of match between users' actual parking needs and parking space supply based on target benchmarking factors includes the following steps: The target benchmark elements are processed to obtain the overall adaptation representation, which includes the following steps: The target elements are broken down item by item to obtain the user parking demand sub-items and the parking space supply sub-items; The target benchmarking elements are composite data integrating user constraints and parking space resource parameters. Based on the user's inherent parking demand profile, they are broken down into sub-items of user parking demand, including vehicle size constraints, vehicle energy type constraints, preferred daily parking time slots, charging equipment usage needs, and monthly parking access permissions. All parameters are determined by the user's own vehicle conditions and long-term parking habits. The parking space supply sub-items originate from the service adaptability and dynamic circulation service attributes of parking spaces, including actual length and width dimensions, onboard charging pile installation status, park location, suitable usage scenarios, and real-time idle time. All parameters are determined by the parking space hardware configuration and the real-time usage status of the park. The total number of parameters after a single split is calculated using a formula. The total number of parameters for the target benchmark element is equal to the number of parameters for user parking demand plus the number of parameters for parking space supply. For example, a target benchmark element for a parking space to be matched has a total of 7 parameters. After splitting, the user parking demand sub-item has 4 parameters, corresponding to 4.7-meter small new energy vehicles, the need for DC fast charging, parking at 8 am on weekdays, and monthly access to the underground park. The parking space supply sub-item has 3 parameters, corresponding to 5.3 meters and 2.5 meters in length and width, the installation of DC fast charging piles, and the location on the first floor of the underground area. Substituting these into the calculation formula, the total number of parameters for the target benchmark element is 4 + 3 = 7.

[0047] By comparing the user parking demand sub-items and the parking space supply sub-items, we can determine the overlap and gaps in the content. The user parking demand items are compared item by item with the parking space supply items. Based on the parameter matching results, overlapping and missing items are identified. The overlapping item refers to the set of parameter entries where user demand parameters and parking space supply parameters can be matched. The missing item is divided into two categories: those where the user has corresponding parking demand and the parking space supply has available resources, and those where the user has corresponding parking demand and the parking space supply cannot be matched but are still categorized as missing. For example, if four demand parameters in the user item can be matched with parking space supply parameters, the overlapping item contains four parameters. If the user has no nighttime parking demand, the missing item contains one parameter. The specific number of overlapping and missing parameter items is then counted.

[0048] Determine the basic matching volume of demand and supply based on the scope of content overlap, and determine the basic deviation volume of demand and supply based on the scope of content gaps. The basic fit volume and basic deviation volume between demand and supply are calculated separately based on the overlapping and vacant areas. The total number of parameters in the overlapping area directly determines the value of the basic fit volume, which is used to quantify the scale of content where supply and demand are successfully matched. The total number of parameters in the vacant area directly determines the value of the basic deviation volume, which is used to quantify the scale of content where supply and demand cannot be matched. Basic fit volume = total number of parameters in the overlapping area, and basic deviation volume = total number of parameters in the vacant area. For example, if there are 4 parameters in the overlapping area and 1 parameter in the vacant area, substituting them into the formula, we get the basic fit volume = 4 and the basic deviation volume = 1.

[0049] The basic fit volume and the basic deviation volume are integrated to form an overall fit representation; Overall fit representation is a comprehensive quantitative data that takes into account both matching advantages and disadvantages. The overall fit representation value = basic fit volume - basic deviation volume. For example, the overall fit representation value = 4 - 1 = 3. The larger the value, the better the basic matching effect.

[0050] The overall adaptation representation is divided into multiple fitting gradients, so that each fitting gradient corresponds to the supply and demand matching attribute. The overall fit representation is divided into multiple levels of matching gradients, with each gradient level bound to a specific supply-demand matching attribute. The system pre-sets five progressively higher matching gradients, from 1 to 5. A higher gradient value indicates a better supply-demand matching effect. Level 1 gradients are bound to low-matching supply-demand attributes, indicating less matching content; Level 2 gradients are bound to slightly lower-matching attributes; Level 3 gradients are bound to moderately matched attributes; Level 4 gradients are bound to slightly higher-matching attributes; and Level 5 gradients are bound to the optimal matching attributes. The overall value above is 3, corresponding to a Level 3 matching gradient, and is simultaneously bound to a moderately matched supply-demand matching attribute.

[0051] Based on the static parking space inventory, adjust the supply and demand matching attributes corresponding to each matching gradient to obtain the degree of matching between users' actual parking needs and the parking space supply.

[0052] The static parking space inventory refers to the fixed total number of registered parking spaces of the same specifications and usage scenarios within the park. When there is a surplus of space, the matching attribute level is increased; when there is a shortage of space, the matching attribute level is decreased. The corrected matching reference value = original gradient baseline value + surplus adjustment value - shortage deduction value. For example, if the static inventory of underground new energy monthly parking spaces for this user is 32, and the inventory of this type of parking space in the park is in a surplus state, the surplus adjustment value is 1, and there is no deduction value due to shortage. The original gradient baseline value is 3. Substituting into the calculation formula, the corrected matching reference value is 3 + 1 - 0 = 4. The original level 3 medium matching attribute is increased to level 4 high matching attribute, and the user's fit with the corresponding parking space is finally determined to be level 4 high fit.

[0053] The impact of parking space demand fluctuations on the supply-demand matching relationship is determined based on the degree of fit, resulting in a supply-demand demand deviation baseline. This involves the following steps: The overall baseline status of the supply and demand matching of parking spaces in the park is defined based on the degree of fit. The overall baseline status represents the inherent matching benchmark between user parking demand and park parking resources. The degree of fit is pre-divided into five tiers from 1 to 5, each corresponding to a fixed baseline fit score. Level 1 corresponds to a low fit baseline score, and level 5 corresponds to an optimal fit baseline score. The baseline fit score is a quantitative representation of the overall baseline status. A single user's baseline fit score equals the baseline value corresponding to their fit level. For example, if a user's fit level is 4, the corresponding baseline value is 4. Substituting this into the calculation formula, the single user's baseline fit score is 4. This score corresponds to a relatively high fit baseline level, meaning that in the absence of temporary passenger flow disruptions, a suitable available parking space can be easily matched. The baseline statuses corresponding to different fit levels are independent of each other. At level 1, there is a scarcity of available parking spaces for the user, while at level 5, there is an abundance of available parking resources.

[0054] Capture the fluctuations in parking space popularity caused by dynamic changes in vehicle and pedestrian traffic within the park; The system collects real-time vehicle flow data from the entrance and exit gate counting devices, and counts the number of vehicles entering and leaving the park within a fixed time period. It also collects the total number of people entering and leaving the park per unit time period from the park gate passenger flow capture device. The system calculates the change in parking space popularity in a single time period with a fixed statistical cycle of 1 hour. The change in parking space popularity in a single hour = the total number of newly added vehicles entering the park in the statistical cycle - the total number of vehicles leaving the park in the statistical cycle. The vehicle count is uniformly measured in vehicles. If the calculated result is positive, it means that the parking space popularity is in a fluctuating state of rising. If the calculated result is negative, it means that the parking space popularity is in a fluctuating state of falling. If the calculated result is close to zero, it means that the parking space popularity remains stable without significant fluctuations.

[0055] For example, during a weekday morning rush hour, 35 new vehicles entered the park and 12 vehicles left during the same period. Substituting these values ​​into the calculation formula, we get the change in parking space popularity per hour as 35 - 12 = 23 vehicles. This indicates that the parking space popularity during that period showed a significant upward fluctuation.

[0056] The positive and negative impacts of parking space demand fluctuations on the supply-demand matching relationship can be determined based on the fluctuations in parking space demand and the overall baseline status. Positive gains occur when parking space popularity increases slightly and moderately. At this time, idle parking spaces in the park are gradually put into use, and previously wasted parking resources are used to meet the parking needs of more users with similar needs, thus improving the actual effect of supply and demand matching. Positive gains are quantified by a matching score. Negative interference occurs when parking space popularity surges rapidly. At this time, a large number of suitable parking spaces of the same type are occupied in advance by outside vehicles. Previously available parking spaces that could match users become occupied, compressing available resources and causing a decline in supply and demand matching effectiveness. Negative interference is quantified by a matching score deducted. For example, if a user's baseline matching score is 4 points, a slight increase in popularity results in a positive gain of 1 point, indicating improved matching effectiveness. However, if a sudden surge in visitors causes a sharp spike in popularity, the corresponding negative interference score is 2 points, indicating that the matching effectiveness is negatively affected by resource congestion.

[0057] Based on the effects of positive gain and negative interference, the supply and demand mismatch amplitude of heat fluctuations with different degrees of fit is analyzed; The supply-demand matching offset reflects the change in matching values ​​caused by real-time heat fluctuations compared to the static baseline state. The single-parking-space supply-demand matching offset is calculated as: Positive Gain Score - Negative Interference Score. For example, under conditions of moderate heat increase without sudden visitor interference, the positive gain score is 1 and the negative interference score is 0, resulting in a single-parking-space supply-demand matching offset of 1 - 0 = 1 point. When a sudden influx of visitors causes a surge in heat, the positive gain score is 1 and the negative interference score is 2, resulting in a single-parking-space supply-demand matching offset of 1 - 2 = -1 point. A negative number indicates that the current heat fluctuation has caused the supply-demand matching level to decline compared to the baseline state. Level 1 users, with lower matching levels, already have fewer baseline parking spaces; a similar surge in heat will generate a higher negative interference score, resulting in a greater decline in the corresponding offset value.

[0058] The overall deviation of the supply and demand matching offset is integrated to obtain the base of supply and demand heat deviation.

[0059] The supply-demand heat deviation base is equal to the sum of the supply-demand mismatch deviations of all candidate parking spaces. For example, if the deviations of the three candidate parking spaces are 1 point, -1 point, and 0 points respectively, the supply-demand heat deviation base = 1 + -1 + 0 = 0 points. This records the overall impact of the heat fluctuation on the supply-demand matching of the entire group of candidate parking spaces.

[0060] Based on the supply-demand temperature deviation baseline, the rigidity of parking demand and the stability of parking space demand are corrected to obtain the profile temperature difference coefficient. The specific steps include: Based on the supply-demand temperature deviation baseline, the matching offset of the rigidity of user parking demand is corrected to obtain the demand correction attribute; The rigidity of user parking demand is used to quantify the urgency level of a user's need to match a corresponding parking space. The rigidity level is pre-divided into 1 to 5. The original rigidity score is statically generated based on the user's identity and historical parking data. Monthly fixed commuter users have a higher rigidity of demand, with an original score in the range of 4 to 5 points. Random temporary visitor users have a lower rigidity of demand, with an original score in the range of 1 to 2 points.

[0061] From the inherent attribute data of user identity, we extracted user monthly parking contract features, user identity classification features, and dedicated fixed parking space contract features. The monthly parking contract feature determines whether a user has signed a monthly or annual fixed parking contract with the park. Users with a full-year monthly contract receive a base score of 3 points, users with a short-term quarterly monthly contract receive a base score of 2 points, and users without any monthly subscriptions receive a base score of 0 points. The user identity classification feature distinguishes between resident employees and temporary visitors. Registered and on-duty employees receive an additional 1 point, while temporary visitors with only one visit receive an additional 0 points. The dedicated fixed parking space contract feature determines whether a user has signed a contract for a dedicated fixed parking space. Users who have signed a fixed parking space contract receive an additional 1 point, while users without a dedicated parking space contract receive a score of 0. The total score for each identity attribute item is calculated as follows: Monthly contract base score + Identity classification additional score + Dedicated parking space contract additional score. For example, a full-time employee of a company in the park applies for a yearly monthly parking pass and signs a contract for a dedicated underground parking space. The basic score for the monthly pass contract is 3 points, the additional score for the identity category is 1 point, and the additional score for the dedicated parking space contract is 1 point. Substituting these into the formula, the score for each identity attribute item is 3 + 1 + 1 = 5 points.

[0062] We retrieved all historical parking behavior data retained by users over the past 12 months, breaking it down into monthly parking frequency characteristics and historical temporary parking change percentage characteristics. The monthly parking frequency characteristic is tiered based on the actual number of park visits per calendar month: a base score of 2 points for 22 or more parking visits in a single month, 1 point for 10 to 21 visits, and 0 points for less than 10 visits. The historical temporary parking change percentage characteristic is used to calculate the percentage of times a user, after receiving a parking space recommendation, abandoned their parking space without a valid reason and changed to another space out of the total number of recommendations. A change percentage of less than 10% indicates high parking plan stability and awards an additional 1 point; a change percentage of 10% or higher awards no additional points. The total score for historical parking behavior is calculated as follows: Monthly Parking Frequency Base Score + Change Percentage Bonus Score. For example, if a user parks an average of 25 times per month over the past 12 months, the monthly parking frequency baseline score is 2 points. If the percentage of parking violations is 6% which is less than 10% throughout the year, the additional score for parking violations is 1 point. Substituting these values ​​into the formula, the historical parking behavior score is 2 + 1 = 3 points.

[0063] The final original rigid score is obtained by adjusting the values ​​according to the capping and minimum score rules. For example, if the identity attribute score is 5 points and the behavior attribute score is 3 points, the total score before the capping is 5 + 3 = 8 points. This value exceeds the upper limit of 5 points, so the user's final original rigid score is 5 points. For example, if a visitor has no monthly pass and no fixed parking space, the identity attribute score is 0 points, and the average monthly parking is 2 times, the behavior attribute score is 0 points, and the total score is 0 + 0 = 0 points, which is 1 point lower than the minimum score. The visitor's final original rigid score is 1 point.

[0064] The demand correction attribute score = user's original parking rigidity score + supply-demand heat deviation baseline value. For example, a monthly commuter user in a certain park has an original parking rigidity score of 5 points and a supply-demand heat deviation baseline value of -2 points. Substituting these values ​​into the calculation formula, the demand correction attribute score is 5 + -2 = 3 points. The demand correction attribute includes the corrected user's actual rigidity level parameter, the corrected range of alternative parking spaces acceptable to the user, and the priority parking time period parameter adjusted due to the influence of heat.

[0065] Based on the supply and demand heat deviation baseline, the resource offset of stable parking space heat is corrected to obtain the parking space correction attribute; The stability of parking space popularity represents the steady level of parking space vacancy on a daily basis. The original stability score is derived from historical parking space turnover data.

[0066] The original stable total score = the score of the inherent attributes of the parking space + the score of the historical circulation statistics. When the total score exceeds 5 points, the original stable score is uniformly set to 5 points. When the total score is less than 1 point, the original stable score is uniformly set to 1 point.

[0067] Based on the inherent attribute data of parking space registration, the attributes are broken down into three categories: parking space location attribute, supporting hardware attribute, and planned usage scenario attribute. The parking space location attribute distinguishes the degree to which the parking space's environment is affected by temporary traffic flow and weather. Parking spaces located in the underground enclosed area of ​​the park, far from the main entrance and main road, receive 2 points; parking spaces in open areas near the main access road receive 1 point; and parking spaces in remote corners of the park receive 0 points. The supporting hardware attribute focuses on the fixed installation of equipment in the parking space. Parking spaces equipped with DC or AC charging devices receive 1 point; ordinary parking spaces without any additional equipment receive 0 points. The planned usage scenario attribute is determined based on the park's previous parking space planning and registration content. Scene parking spaces specifically allocated to monthly pass users receive 1 point; general temporary scene parking spaces open to all visitors receive 0 points. The total score for each inherent attribute of a parking space = location attribute score + supporting hardware attribute score + planned usage scenario attribute score. For example, a monthly parking space for new energy vehicles located on the underground floor of the park, far from the main road, has a location attribute score of 2 points, a supporting hardware attribute score of 1 point, and a planning scenario attribute score of 1 point. Substituting these into the calculation formula, the inherent attribute score of the parking space is 2 + 1 + 1 = 4 points.

[0068] The complete historical usage log for the parking space over the past 365 days was retrieved, and the characteristics of average daily idle time and monthly occupancy fluctuation were broken down to calculate the historical turnover statistics score. The average daily idle time is scored based on the total idle time accumulated over each calendar day. An average daily idle time greater than 6 hours within the statistical period indicates ample occupancy and scores 1 point; an average daily idle time between 3 and 6 hours scores 0.5 points; and an average daily idle time less than 3 hours indicates consistently high occupancy and scores 0 points. Monthly occupancy fluctuation measures the magnitude of fluctuations in the number of occupied days per month. Small fluctuations in the number of occupied days between months indicate stable usage and earn an additional 1 point; significant fluctuations in the number of occupied days per month result in no additional points. The historical turnover statistics score = average daily idle time score + monthly occupancy fluctuation bonus score. For example, if the average daily vacancy time of a monthly parking space is 7 hours throughout the year, the average daily vacancy score is 1 point. The number of days the parking space is occupied varies little throughout the year, and the fluctuation score is 1 point. Substituting these values ​​into the calculation formula, we get the historical turnover statistics score = 1 + 1 = 2 points.

[0069] For example, if the inherent attribute score of a parking space is 4 points, the historical turnover statistics score is 2 points, and the original stable total score is 4 + 2 = 6 points, exceeding the upper limit of 5 points, the final original stable score for this parking space is 5 points. For another example, if the parking space's location attribute score is 1 point, the score for no supporting facilities is 0 points, the score for a general visitor scenario is 0 points, the inherent attribute score is 1 + 0 + 0 = 1 point, the average daily idle time is 2 hours (corresponding to 0 points), the monthly occupancy fluctuation due to visitor influence has no additional points, the historical turnover statistics score is 0 + 0 = 0 points, and the total score is 1 + 0 = 1 point, the final original stable score is 1 point. In the extreme scenario where an open-air parking space without supporting facilities on a street has a total score of 0 points, due to the lower limit of the score, the original stable score is uniformly guaranteed to be 1 point.

[0070] The initial stability score ranges from 1 to 5 points. A higher score indicates more stable weekday vacancy times and less fluctuation in occupancy due to traffic flow. Open-air temporary parking spaces are more susceptible to visitor fluctuations, resulting in generally lower initial stability scores. The parking space correction attribute score equals the initial parking space popularity stability score plus the base value of the supply-demand deviation. For example, if the initial popularity stability score of the parking space to be matched is 4 points and the base value of the supply-demand deviation is -2 points, the correction attribute score is 4 + -2 = 2 points. The parking space correction attribute includes the corrected parking space vacancy stability parameter, the short-term available parking space quantity parameter after adjusting for popularity fluctuations, and the parking space turnover fluctuation parameter updated due to passenger flow.

[0071] The portrait heat difference coefficient is obtained by integrating and correcting the demand-adjusted attributes and parking space-adjusted attributes.

[0072] The profile popularity difference coefficient is calculated as follows: Demand-corrected attribute score - Parking space-corrected attribute score. For example, if the demand-corrected attribute score is 3 points and the parking space-corrected attribute score is 2 points, the profile popularity difference coefficient is 3 - 2 = 1 point. The profile popularity difference coefficient is positively correlated with recommendation priority. A higher coefficient indicates a better match between the user's corrected parking demand and the corrected parking space supply, resulting in the parking space appearing higher in the candidate list. A negative profile popularity difference coefficient indicates that the corrected parking space resources are insufficient to meet the user's current needs, and the corresponding parking space will be relegated to a later position or eliminated entirely during the candidate selection process.

[0073] Based on the user profile's popularity difference coefficient, the system outputs target parking space recommendations for the user to be recommended, specifically including the following steps: Candidate parking spaces that match user parking preferences and are compatible with the park's real-time supply and demand status are selected based on the user profile popularity difference coefficient. The profile heat index difference coefficient is a quantitative indicator derived by integrating user correction demand attributes and parking space correction attributes. It intuitively reflects the matching level between parking spaces and users' actual parking needs after heat index deviation correction. User parking preferences include the floor level where users habitually park, whether they need onboard charging facilities, preferred distance from entrances / exits, and fixed parking time periods. The park's real-time supply and demand status includes the current instantaneous vacancy status of parking spaces, the real-time traffic congestion value of the area where the parking space is located, and the remaining stock of matching parking spaces of the same size. All parameters are updated based on real-time pedestrian and vehicle flow and parking space occupancy data collected in the park. The candidate admission benchmark score = the system's preset minimum admission score. If the profile heat index difference coefficient of a single parking space is greater than the candidate admission benchmark score, it will be filtered through the basic value. For example, if the set admission benchmark score is 0, and there are 3 candidate parking spaces within the initial selection range for a user to be recommended, with a user profile popularity difference coefficient of 2 for parking space 1, 1 for parking space 2, and -1 for parking space 3, after applying the judgment rules, if the coefficient values ​​of parking spaces 1 and 2 exceed the admission benchmark, they will pass the numerical screening; if the coefficient value of parking space 3 is lower than the admission benchmark, it will be directly eliminated. Additional constraints are then applied to the two remaining parking spaces: parking space 1 is located in the user's preferred basement area and is equipped with the user's required fast charging station; both additional conditions are met. Parking space 2 matches the user's preferences and the real-time supply and demand environment, ultimately forming a candidate parking space set containing 2 spaces.

[0074] The candidate parking spaces are ranked according to their suitability to obtain the target parking space recommendation result.

[0075] Based on the user profile popularity difference coefficient for each parking space, the spaces are prioritized for adaptation. After sorting, the entries are aggregated to generate target parking space recommendations. A higher user profile popularity difference coefficient indicates a better match between the parking space supply (adjusted for popularity fluctuations) and the user's actual parking needs, resulting in a higher recommendation ranking. The parking space priority reference value equals the single parking space user profile popularity difference coefficient. For example, if parking space 1 has a priority reference value of 2 points and parking space 2 has a priority reference value of 1 point, after sorting by value from highest to lowest, parking space 1 will be ranked first in the recommendation list, and parking space 2 will be ranked second. The parking space information is then organized according to this predetermined order to form an ordered list of target parking space recommendations.

[0076] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0077] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0078] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A smart park parking space proactive recommendation system that combines user profiles and real-time popularity, characterized in that: include: Acquisition Module: Acquires historical parking space usage information of the smart park, and obtains parking space benchmarking elements based on the historical parking space usage information; wherein, the historical parking space usage information includes parking space feature parameters, recommendation classification results, and popularity evolution process; the parking space benchmarking elements include element number, element feature parameters, and element recommendation tags; Analysis module: Extract user profile features and real-time parking space popularity features of the users to be recommended, and analyze user profile features, real-time parking space popularity features, parking space benchmarking elements and inherent attribute features to obtain a set of parking space resource matching elements; Processing module: The module performs benchmarking and matching screening on the set of parking space resource matching elements to obtain target benchmarking elements; it judges the degree of matching between the actual parking demand of users and the matching status of parking space supply based on the target benchmarking elements; and it determines the impact of parking space heat fluctuations on the supply and demand matching relationship based on the degree of matching to obtain the supply and demand heat deviation baseline. Correction module: Based on the supply and demand heat deviation baseline, the module corrects the rigidity of parking demand and the stability of parking space heat to obtain the portrait heat difference coefficient; Output module: Based on the user profile popularity difference coefficient, output the target parking space recommendation results for the user to be recommended.

2. The smart park parking space proactive recommendation system combining user profiles and real-time popularity as described in claim 1, characterized in that, Based on historical parking space usage information, the benchmarking elements for parking spaces are obtained, specifically including the following steps: Based on historical parking space usage information, the inherent usage characteristics of parking spaces, historical recommendation and adaptation results, and the evolution of parking space popularity throughout the time period are integrated to obtain historical parking space usage information. By hierarchically decomposing historical parking space usage information, we can obtain stable usage characteristics and scenario adaptation features of parking spaces. Based on the stable use characteristics and scenario adaptation features of parking spaces, parking space adaptation service scenarios are divided, and the parking space service capabilities of different parking space adaptation service scenarios are determined. Based on the parking space service capabilities, the basic elements for parking space benchmarking are obtained. Based on the historical recommendation adaptation results, the basic elements for parking space benchmarking are verified in the scenario to obtain the parking space benchmarking elements.

3. The smart park parking space proactive recommendation system combining user profiles and real-time popularity as described in claim 1, characterized in that, Extracting user profile features and real-time parking space popularity features from the users to be recommended, specifically including the following steps: Based on historical parking behavior data and identity attribute data, we can obtain parking habit preferences and fixed parking needs; User profiles are created based on parking habits, preferences, and fixed parking needs. Collect real-time parking space usage status and dynamic pedestrian and vehicle flow data within the park; The real-time parking space activity level is determined by the real-time parking space usage status and dynamic pedestrian and vehicle flow data in the park, thus obtaining the real-time parking space popularity characteristics.

4. The smart park parking space proactive recommendation system combining user profiles and real-time popularity as described in claim 1, characterized in that, The analysis of user profile features, real-time parking space popularity features, parking space benchmarking elements, and inherent attribute features yields a set of parking space resource adaptation elements, specifically including the following steps: Generate corresponding parking space recommendation elements based on user profile characteristics and real-time parking space popularity characteristics; Based on the parking space recommendation elements and parking space benchmarking elements, determine the matching benchmarking elements and extract the inherent attribute features of the parking needs of the users to be recommended. A profile of inherent parking demand is created based on matching benchmarking elements and inherent attribute characteristics; Based on the existing parking demand profile, determine the set of parking resource adaptation elements for various types of parking resources in the park; wherein, the set of parking resource adaptation elements includes service adaptation attributes and dynamic circulation service attributes.

5. The smart park parking space proactive recommendation system combining user profiles and real-time popularity as described in claim 1, characterized in that, The set of parking space resource matching elements is benchmarked and screened to obtain target matching elements, which includes the following steps: Verify the feature matching consistency of the parking space resource adaptation element set to obtain a preliminary matching element set; Based on the initial set of matching elements and the characteristics of parking demand adaptation, the target benchmarking elements are obtained.

6. The smart park parking space proactive recommendation system combining user profiles and real-time popularity as described in claim 1, characterized in that, Determining the degree of match between users' actual parking needs and parking space supply based on target benchmarking factors includes the following steps: The target benchmarking elements are processed to obtain an overall adaptation representation; The overall adaptation representation is divided into multiple fitting gradients, so that each fitting gradient corresponds to the supply and demand matching attribute. Based on the static parking space inventory, adjust the supply and demand matching attributes corresponding to each matching gradient to obtain the degree of matching between users' actual parking needs and the parking space supply.

7. The smart park parking space proactive recommendation system combining user profiles and real-time popularity as described in claim 6, characterized in that, The target benchmark elements are processed to obtain the overall adaptation representation, which includes the following steps: The target elements are broken down item by item to obtain the user parking demand sub-items and the parking space supply sub-items; By comparing the user parking demand sub-items and the parking space supply sub-items, we can determine the overlap and gaps in the content. Determine the basic matching volume of demand and supply based on the scope of content overlap, and determine the basic deviation volume of demand and supply based on the scope of content gaps. The basic fit volume and the basic deviation volume are integrated to form an overall fit representation.

8. The smart park parking space proactive recommendation system combining user profiles and real-time popularity as described in claim 1, characterized in that, The impact of parking space demand fluctuations on the supply-demand matching relationship is determined based on the degree of fit, resulting in a supply-demand demand deviation baseline. This involves the following steps: The overall baseline status of the supply and demand matching of parking spaces in the park is defined based on the degree of fit. Capture the fluctuations in parking space popularity caused by dynamic changes in vehicle and pedestrian traffic within the park; The positive and negative impacts of parking space demand fluctuations on the supply-demand matching relationship can be determined based on the fluctuations in parking space demand and the overall baseline status. Based on the effects of positive gain and negative interference, the supply and demand mismatch amplitude of heat fluctuations with different degrees of fit is analyzed; The overall deviation of the supply and demand matching offset is integrated to obtain the base of supply and demand heat deviation.

9. The smart park parking space proactive recommendation system combining user profiles and real-time popularity as described in claim 8, characterized in that, Based on the supply-demand temperature deviation baseline, the rigidity of parking demand and the stability of parking space demand are corrected to obtain the profile temperature difference coefficient. The specific steps include: Based on the supply-demand temperature deviation baseline, the matching offset of the rigidity of user parking demand is corrected to obtain the demand correction attribute; Based on the supply and demand heat deviation baseline, the resource offset of stable parking space heat is corrected to obtain the parking space correction attribute; The portrait heat difference coefficient is obtained by integrating and correcting the demand-adjusted attributes and parking space-adjusted attributes.

10. The smart park parking space proactive recommendation system combining user profiles and real-time popularity as described in claim 9, characterized in that, Based on the user profile's popularity difference coefficient, the system outputs target parking space recommendations for the user to be recommended, specifically including the following steps: Candidate parking spaces that match user parking preferences and are compatible with the park's real-time supply and demand status are selected based on the user profile popularity difference coefficient. The candidate parking spaces are ranked according to their suitability to obtain the target parking space recommendation result.

Images