Intelligent parking management system and method based on data analysis

Abstract

The invention discloses an intelligent parking management system and method based on data analysis, relates to the technical field of intelligent parking management, and aims to solve the problems that a parking lot matched with a vehicle model is difficult to obtain and parking spaces are pre-judged by depending on static data in an existing system. A three-dimensional association database is constructed by collecting physical data of a parking lot and vehicle type data of vehicle owners; screening surrounding parking lots according to destinations input by vehicle owners to generate candidate lists; preliminarily judging a scratching risk by combining an entrance and exit slope correction size comparison result; calling the real-time channel occupancy rate and the entrance and exit congestion data to secondarily verify the passing suitability and update the risk level; calculating a parking space release rate and an availability index of an adaptive parking lot, and generating a recommendation list by combining a distance and a risk level calculation comprehensive score; when the number of adaptive parking lots is insufficient, the range is expanded for supplementary selection, and alternative schemes are generated; and collecting the feedback of the vehicle owner to update the database, and optimizing the adaptation judgment logic.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent parking management, and particularly relates to an intelligent parking management system and method based on data analysis. BACKGROUND

[0002] With the continuous growth of the number of cars, the intelligent parking management system becomes an important means to alleviate the urban parking problem, but the existing technology still has key defects, which is difficult to meet the core parking needs of car owners.

[0003] On the one hand, the existing system is difficult to help the car owner to obtain the parking lot which is suitable for the car model and can ensure that there is an available parking space. Such system mostly does not deeply correlate the parking lot physical data (such as the entrance height, the passage width, and the entrance and exit slope) with the car owner's car model data (such as the maximum car height and whether the roof device is installed), only judges the adaptability through simple size comparison, neither considers the actual size change when the car body is inclined due to the entrance and exit slope, such as the steep slope may cause scratching, nor combines the real-time traffic state to check the adaptability, resulting in that the recommended parking lot may not be normally used due to the mismatch between the car model and the parking lot, or although it is suitable, there is no available parking space, so as to realize the double protection of "adaptation" and "availability".

[0004] On the other hand, the existing system depends on the real-time parking space to recommend the parking lot, and cannot predict the parking space release rule. The system only takes the remaining parking space number at the recommendation time as the basis, does not combine the historical parking lot release rate (such as the average number of released parking spaces per unit time) to predict the availability of parking space when the car owner arrives, so that the car owner often has no parking space when he arrives at the recommended parking lot, not only wastes time, but also greatly reduces the parking efficiency.

[0005] To solve the above problems, the present application provides an intelligent parking management system and method based on data analysis. SUMMARY

[0006] The purpose of the present application is to provide an intelligent parking management system and method based on data analysis to solve the problems in the prior art.

[0007] To achieve the above purpose, the present application provides the following technical scheme:

[0008] An intelligent parking management method based on data analysis, comprising the following steps:

[0009] S1, collecting the physical data of all existing parking lots, correlating it with the geographic information system of the navigation software, and synchronously receiving the car model data, and constructing the correlation database of the two types of data;

[0010] S2, according to the destination input by the owner, in combination with geographic information, the parking lots within the preset range are screened from the associated database to form a candidate list;

[0011] S3, the owner's vehicle model data is compared with the parking lot data, and the entry and exit slope characteristics are combined to determine whether the vehicle type is prone to scratching;

[0012] S4, the real-time channel occupancy rate and the entry and exit congestion data of the candidate parking lot which is preliminarily adapted to the vehicle type are called to verify the traffic adaptation and update the parking lot risk level;

[0013] S5, the historical parking space data of the same period in the preset historical period of the adaptive parking lot is called to calculate the average parking space release rate per unit time, and the parking space availability index is constructed; then, the distance between the adaptive parking lot and the destination, the parking space availability index and the risk level are combined to calculate the comprehensive score, and the parking lot recommendation list is generated according to the priority;

[0014] S6, if the number of adaptive parking lots in the recommendation list does not reach the preset value, the screening range is expanded to repeat the adaptive analysis, the route adjustment time is calculated, and the low-risk alternative scheme is generated;

[0015] S7, the actual traffic feedback of the owner and the parameter change data of the parking lot are collected, the associated database is updated synchronously, and the subsequent vehicle type-parking lot adaptive judgment logic is optimized.

[0016] S1 further includes the following contents:

[0017] S1.1: Collect all existing physical data of various parking lots, said physical data including parking lot entrance height, channel width and entry and exit slope; and the collected physical data is standardized to form a structured parking lot physical data set, and the data set is associated with the geographic information system of the navigation software, so that the physical data of each parking lot corresponds to its geographic coordinates;

[0018] S1.2: Receive the vehicle type data entered by the owner when registering in the system, said vehicle type data including vehicle type name, whether to install roof additional device; then call the built-in motor vehicle parameter database, match and extract the original factory maximum vehicle height and maximum vehicle width according to the received vehicle type name, if the owner marks to install roof additional device, then add height value to update the maximum vehicle height, form a vehicle type parameter set containing the individual characteristics of the owner's vehicle type;

[0019] S1.3: The parking lot physical data set and the vehicle type parameter set are associated with the parking lot geographic coordinates as the index to construct a three-dimensional associated database of parking lot geographic coordinates-parking lot physical data set-vehicle type parameter set.

[0020] S2 further includes the following contents:

[0021] S2.1: According to the destination geographic coordinates input by the vehicle owner, set a preset space screening range centered on the coordinates; determine the geographic coordinates of all parking lots within the range through the spatial distance calculation function of the geographic information system; then retrieve all parking lot data with geographic coordinates within the above defined range from the associated database, extract the parking lot data and preliminarily sort it according to the distance from the destination from near to far, forming an original list containing basic information and physical data of the parking lot;

[0022] S2.2: Verify the effectiveness of the parking lots in the above original list, eliminate parking lots marked as temporary closure, permanent closure, marked for use only by specific vehicles, and this vehicle model and vehicle owner vehicle model parameter set do not match; organize the remaining parking lot data after verification into a structured candidate list, each parking lot entry in the list is associated with its geographic coordinates, physical data and vehicle owner vehicle model preliminary matching status.

[0023] S3 further comprises the following contents:

[0024] S3.1: Retrieve the physical data of each parking lot in the candidate list, and retrieve the vehicle model parameter set bound to the vehicle owner's account, then compare the maximum height and maximum vehicle width in the vehicle owner's vehicle model parameter set with the entrance height and passage width of the corresponding parking lot physical data; record the difference between the vehicle model parameters and the parking lot parameters in each comparison result to form a structured size comparison table;

[0025] S3.2: Retrieve the vehicle tilt-size offset model built-in the system, according to the entrance slope value of the candidate parking lot physical data, match the corresponding slope coefficient K slope from the model, and use the formula to get the corrected size difference; the formula is as follows:

[0026] Corrected size difference = original size difference × (1 / K slope );

[0027] Wherein, the original size difference is the difference between the parking lot and the vehicle model parameter, and the corrected size difference quantitatively reflects the actual size of the vehicle model and the real allowance of the parking lot parameter after the body is inclined due to slope; update the corrected size difference to the structured size comparison table at the same time;

[0028] S3.3: Set two thresholds: a safety threshold T1 representing the minimum size margin that meets the safety of passing; a critical threshold T2 representing the critical value of the size margin that reaches the risk, and T2 < T1; compare the modified size difference of each candidate parking lot with T1 and T2. If any one of the modified size difference, the vehicle height difference and the vehicle width difference < T2, it is determined as mismatch; if any one of the modified size difference, the vehicle height difference and the vehicle width difference meets T2 ≤ difference < T1, it is determined as risk to be verified; if the modified size difference all meets ≥ T1, it is determined as preliminary adaptation; generate a structured preliminary risk table containing parking lot identification, modified size difference, threshold comparison result and preliminary risk level.

[0029] S4 further comprises the following contents:

[0030] S4.1: According to the structured preliminary risk table, filter out the parking lots determined as preliminary adaptation and risk to be verified; call the system real-time data interface to obtain the real-time channel occupancy rate and the entrance and exit congestion data of these two types of parking lots; wherein the real-time channel occupancy rate is the proportion data of the actual occupied width of the channel to the total width of the channel, and the entrance and exit congestion data is the average passing time data of the vehicles currently waiting to pass through the entrance and exit; then the above two types of data are associated with the modified size difference of the corresponding parking lot to form a verification data set;

[0031] According to the real-time channel occupancy rate and the entrance and exit congestion data in the verification data set, the mean and standard deviation of the two types of data are calculated respectively; the outliers beyond ±3 times the standard deviation of the mean are removed, and the mean is recalculated with the remaining data to obtain the effective channel occupancy rate and the effective entrance and exit congestion time; update the effective data to the verification data set;

[0032] S4.2: Calculate the actual available channel width W of the parking lot based on the effective channel occupancy rate act The formula is as follows:

[0033] W act =W ori ×(1-C occ );

[0034] Wherein, W act is the actual available channel width; W ori is the original value of the channel width of the parking lot; C occ is the effective channel occupancy rate after processing;

[0035] Compare the calculated W act with the modified vehicle width difference, and according to the formula, the actual width margin = W act - the maximum vehicle width of the vehicle type;

[0036] At the same time, set the entrance and exit congestion time threshold T congIf the congestion duration of the effective entrance and exit of a parking lot exceeds T cong , it is determined that the parking lot has a risk of traffic delay, which needs to be added to the risk assessment dimension.

[0037] S4.3: According to the actual width margin and the congestion duration judgment result, update the risk level in combination with the safety threshold T1 and the critical threshold T2:

[0038] If the actual width margin is greater than or equal to T1 and the congestion duration of the effective entrance and exit is less than or equal to T cong , the level is determined as adaptive.

[0039] If the actual width margin satisfies T2≤actual width margin cong and the actual width margin is greater than T1, and any one of the above two conditions is satisfied, the level is determined as needing attention.

[0040] If the actual width margin is less than T2, the level is determined as not adaptive.

[0041] Finally, a structured final risk table containing the parking lot identifier, secondary verification result, and final risk level is generated.

[0042] S5 further includes the following content:

[0043] S5.1: Collect the historical parking space data of the same period in the preset historical period and the current real-time remaining parking space number of the parking lot determined as the adaptive level in the final risk table; the historical parking space data includes the remaining parking space number at each collection time node in the period, and the current real-time remaining parking space number is the instantaneous remaining parking space number of the parking lot at the collection time; bind the collected historical parking space data and the current real-time remaining parking space number with the three-dimensional association database constructed in S1.3 to form an expanded association database containing parking lot geographic coordinates-parking lot physical data-vehicle type parameters-historical parking space data-real-time remaining parking space number;

[0044] S5.2: From the expanded association database, retrieve the historical parking space data of the same period in the preset historical period of the parking lot of the adaptive level; calculate the total released parking space number ΔN in the period, ΔN is the absolute value of the difference between the remaining parking space number at the start of the period and the remaining parking space number at the end of the period; divide the total released parking space number ΔN by the total duration ΔT of the period to obtain the average parking space release rate R avg per unit time, the formula is as follows:

[0045] R avg = ΔN / ΔT;

[0046] Where R avg is the average parking space release rate per unit time, ΔN is the total released parking space number in the preset historical period, and ΔT is the total duration of the preset historical period.

[0047] retrieve the current real-time remaining parking spaces N of the adaptive level parking lot from the extended association database real ; obtain the estimated driving time T from the current location of the vehicle owner to the parking lot through the geographic information system of the navigation software arrive ; based on the obtained average parking space release rate R avg , construct the parking space availability index I using formula I ava = N real + (R avg × T arrive ), where I ava is the parking space availability index, N real is the current real-time remaining parking spaces of the parking lot, and T arrive is the estimated arrival time of the vehicle owner at the parking lot

[0048] S5.3: Extract the distance data between the adaptive level parking lot and the destination from the original list generated in S2.1 and perform standardization processing to obtain D norm ; set the distance weight W dist , the parking space availability index weight W ava , and the risk level weight W risk , and the sum of the three is 1; wherein the risk level weight W risk corresponds to the fixed weight value of the adaptive level in S4.3; calculate the comprehensive score of each adaptive parking lot using the formula, which is as follows:

[0049] S total = W dist × D norm + W ava × I ava + W risk × R score ;

[0050] where S total is the comprehensive score; R score is the risk score corresponding to the adaptive level, and R score is based on the system preset risk parameter-score mapping rule, which assigns a fixed risk score R score to the parameter combination with a corrected size difference ≥ T1 and an actual width allowance ≥ T1, this mapping rule is built-in the system, and R score takes a value matching the risk level of the adaptive level;

[0051] S5.4: Sort all adaptive level parking lots in descending order of comprehensive score S total , if there are cases with the same comprehensive score, then sort them according to the distance D normThe parking information is sorted twice from nearest to furthest. The sorted parking information is then organized into a structured recommendation list. Each item in the list is associated with a parking lot identifier, a comprehensive score, a parking space availability index, a distance from the destination, and a risk level, generating a priority-ranked parking recommendation list.

[0052] S6 further includes the following:

[0053] S6.1: Retrieve the priority-sorted parking lot recommendation list and count the total number N of parking lots in the list that are suitable for each level. 适配 And the system's preset threshold number of adapters N 预设 , will N 适配 With N 预设 For comparison, if N 适配 ≥N 预设 If N 适配 <N 预设 This will trigger an operation to expand the filtering range;

[0054] S6.2: Based on the geographical coordinates of the destination entered by the car owner, filter the radius according to the system's preset range expansion rules to form a new expanded filtering range; repeat the adaptation analysis process of S2.1-S4.3 to filter out parking lots with adaptation levels within the expanded range to form a supplementary list of adapted parking lots.

[0055] Extract the nearest suitable parking lot from the recommendation list described in S5.4, and calculate the estimated travel time T to that parking lot using the navigation software's geographic information system. 原 And for each parking lot in the supplementary matching parking list, the estimated travel time T is also calculated using the navigation software's geographic information system. 新 The route adjustment time ΔT is calculated using a formula. 调 The formula is as follows:

[0056] ΔT 调 =T 新 -T 原 ;

[0057] Where ΔT 调 For the time spent adjusting the route, T 新 To supplement the estimated travel time for parking lots, T 原 The estimated travel time is the closest suitable parking lot in the original recommendation list;

[0058] S6.3: For parking lots in the supplementary adaptation list, repeat the calculation logic of S5.2-S5.3 to obtain the comprehensive score S of the supplementary parking lots. total补 Then sort the Total scores from highest to lowest. If the overall scores are the same, sort them according to ΔT. 调Sort the parking lots from smallest to largest; then merge the sorted supplementary parking lots with the original recommended list in S5.4, and label the ΔT for each parking lot. 调 The results were compiled into a structured low-risk alternative plan that includes parking lot signage, overall score, parking space availability index, route adjustment time, and risk level.

[0059] The S7 further includes the following:

[0060] S7.1: Receive actual passage feedback submitted by the car owner after completing passage through any parking lot in the recommended list or low-risk alternative plan. The feedback includes the actual compatibility between the vehicle model and the parking lot, and whether there is a risk of scratches during the passage. Then, the feedback data is synchronously updated to the three-dimensional association database and the extended association database, replacing the old data of the corresponding parking lot in the database.

[0061] S7.2: Based on the updated associated database, extract records from vehicle owner traffic feedback where the feedback adaptation differs from the system's predicted adaptation, and analyze the reasons for the discrepancies:

[0062] If the discrepancy stems from an adaptation bias caused by the slope, then retrieve the slope values ​​of the parking lot entrance and exit, along with the corresponding K value used in the prediction. slope By comparing the actual scratches reported by car owners with the calculated and corrected size difference, when the actual risk level on the slope is higher than the model's prediction, the corresponding slope value K is adjusted. slope The value is gradually increased according to the preset range so that the corrected size difference meets the size allowance for actual passage.

[0063] If the discrepancy stems from a mismatch between the overall score and the actual needs of car owners, then the frequency of car owners' concerns regarding distance, parking availability, and traffic risks in such inconsistent feedback will be statistically analyzed, and the weight preset values ​​in S5.3 will be adjusted: when the frequency of feedback regarding parking availability is higher than the current W... ava The percentage will be gradually increased according to the preset range. ava When the frequency of distance-related feedback is lower than the current W dist The percentage will be gradually reduced according to the preset range. dist This ensures that the weight settings match the actual traffic needs of car owners.

[0064] A data-based intelligent parking management system includes a data acquisition and association database construction module, a geographic information interaction module, a parking lot screening module, an adaptation risk assessment module, a parking space availability calculation and recommendation list generation module, an alternative solution generation module, and a feedback collection and adaptation logic optimization module.

[0065] The data collection and associated database construction module is responsible for collecting parking lot physical data and vehicle owner vehicle type data, associating navigation geographic information to construct a three-dimensional associated database; supplementary collection of parking space historical data and real-time remaining parking space number, binding to the three-dimensional database to form an expanded associated database; the geographic information interaction module is used for receiving the destination input by the vehicle owner and converting it into standardized geographic coordinates, providing spatial distance calculation and estimated driving time of the vehicle owner to the parking lot, supporting the definition of the screening range and the time-consuming calculation of route adjustment; the parking lot screening module is responsible for setting the screening range according to the destination coordinates, calling the parking lot data within the range from the associated database to form an original list; verifying and removing invalid parking lots to generate a structured candidate list; the adaptive risk judgment module is used to call the candidate parking lot data and vehicle owner vehicle type parameters, and through size comparison, slope correction and real-time data verification, the actual width margin is calculated and combined with the congestion time to determine the risk level, and the final risk table is generated; the parking space availability calculation and recommended list generation module is used to call the adaptive parking lot space data to calculate the average release rate per unit time, and combined with the real-time remaining parking space number and the estimated arrival time to construct the parking space availability index; according to the distance, index, risk level weight, the comprehensive score is calculated, and the priority recommendation list is generated; the alternative scheme generation module is responsible for expanding the screening range to supplement the adaptive parking lot when the number of adaptive parking lots in the recommendation list does not reach the preset threshold; calculate the time-consuming of route adjustment and sort the supplementary parking lots, and merge the original list to generate a low-risk alternative scheme; the feedback collection and adaptive logic optimization module is responsible for receiving the vehicle owner's passing feedback and parking lot parameter change data, updating the associated database; analyze the adaptive difference between the feedback and the prediction, adjust the parameters, and optimize the adaptive judgment logic.

[0066] Compared with the prior art, the beneficial effects of the present application are:

[0067] 1、The present application can accurately select the parking lot completely adapted to the vehicle type by constructing the associated database of "parking lot physical data-vehicle owner vehicle type data-navigation geographic information", and combining the size comparison results of the entrance and exit slope correction, the real-time channel occupancy rate and the congestion data secondary verification adaptability, effectively avoiding the problems of vehicle body inclination caused by slope, too narrow channel occupancy and other problems, solving the core pain point of "difficult to obtain the parking lot adapted to the vehicle type" of the prior art.

[0068] 2、The present application is not dependent on a single real-time parking space recommendation, but through calling the parking space data at the same period in the preset historical period of the parking lot, calculating the average parking space release rate per unit time, combining the vehicle owner's estimated arrival time to construct the "parking space availability index", dynamically predicting the parking space state when the vehicle owner arrives, solving the problem of "depending on static data recommendation, no parking space available after the vehicle owner arrives" of the prior art, improving the parking efficiency and reducing the waste of travel time.

[0069] 3、The application collects the actual passing feedback of the car owner, analyzes the difference between the prediction of the system and the actual adaptation, adjusts the key parameters such as the slope coefficient mapping rule and the comprehensive score weight, synchronously updates the database, so that the system adaptation judgment logic can be dynamically optimized according to the actual use scene, long-term high accuracy is maintained, and the defects of the prior art such as 'no optimization mechanism and long-term use accuracy decline' are avoided. BRIEF DESCRIPTION OF DRAWINGS

[0070] Figure 1 The method flowchart of the intelligent parking management method based on data analysis. DETAILED DESCRIPTION

[0071] The technical solutions in the embodiments of the application will be clearly and completely described below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the application.

[0072] Embodiment: As shown in the figure, the application provides a technical solution, Figure 1

[0073] An intelligent parking management method based on data analysis comprises the following steps:

[0074] S1, collect the physical data of all existing parking lots, associate it with the geographic information system of the navigation software, and synchronously receive the car type data of the car owner, and construct the associated database of the two types of data;

[0075] S2, according to the destination input by the car owner, combine the geographic information to filter the parking lots within the preset range from the associated database, and form a candidate list;

[0076] S3, compare the car type data of the car owner with the parking lot data, and judge whether there is a scratch risk when the car type passes through the entrance and exit according to the slope characteristics;

[0077] S4, call the real-time channel occupancy rate and the entrance and exit congestion data of the preliminary adapted car type in the candidate parking lot, verify the passing adaptation again and update the risk level of the parking lot;

[0078] S5, call the historical data of the parking space in the same period within the preset historical period of the adapted parking lot, calculate the average parking space release rate per unit time, construct the parking space availability index, and then calculate the comprehensive score according to the distance between the adapted parking lot and the destination, the parking space availability index and the risk level, and generate a parking lot recommendation list sorted by priority;

[0079] ​S6, if the number of adapted parking lots in the recommendation list does not reach the preset value, expand the screening range and repeat the adaptive analysis, calculate the time consumption of route adjustment and generate low-risk alternative solutions;

[0080] S7, collect the actual traffic feedback of the vehicle owner and the parking lot parameter change data, update the associated database synchronously, and optimize the subsequent vehicle-parking lot adaptive judgment logic.

[0081] S1 further comprises the following contents:

[0082] S1.1: Collect all existing physical data of each parking lot, including the height of the parking lot entrance, the width of the passage and the slope of the entrance and exit; and standardize the collected physical data to form a structured parking lot physical data set, associate the data set with the geographic information system of the navigation software, so that the physical data of each parking lot corresponds to its geographic coordinates;

[0083] S1.2: Receive the vehicle type data entered by the vehicle owner during system registration, including the vehicle type name, whether the roof-mounted device is installed; then call the built-in motor vehicle parameter database, match and extract the original factory maximum vehicle height and maximum vehicle width according to the received vehicle type name, if the vehicle owner marks the installation of roof-mounted device, then superimpose the height value to update the maximum vehicle height, form a vehicle type parameter set containing the individual characteristics of the vehicle owner's vehicle type;

[0084] S1.3: Take the parking lot geographic coordinates as the index, associate the parking lot physical data set and the vehicle type parameter set, and construct a three-dimensional associated database of parking lot geographic coordinates-parking lot physical data set-vehicle type parameter set.

[0085] S2 further comprises the following contents:

[0086] S2.1: According to the destination geographic coordinates input by the vehicle owner, set a preset spatial screening range with the coordinates as the center; determine the geographic coordinates of all parking lots within the range through the spatial distance calculation function of the geographic spatial information system; then call all parking lot data with geographic coordinates within the above defined range from the associated database, extract the parking lot data and preliminarily sort it according to the distance from the geographic coordinates to the destination from near to far, forming an original list containing the basic information and physical data of the parking lot;

[0087] S2.2: Verify the effectiveness of the parking lots in the above original list, eliminate the parking lots marked as temporary closure, permanent closure, marked for use only by specific vehicles, and this vehicle type and the vehicle owner's vehicle type parameter set do not match; organize the remaining parking lot data after verification into a structured candidate list, each parking lot entry in the list is associated with its geographic coordinates, physical data and vehicle owner's vehicle type preliminary matching state.

[0088] S3 further comprises the following:

[0089] S3.1: retrieve the physical data of each parking lot in the candidate list, and retrieve the vehicle model parameter set bound to the vehicle owner account, then compare the maximum height and maximum width in the vehicle owner vehicle model parameter set with the entrance height and passage width of the corresponding parking lot physical data; and record the difference between the vehicle model parameters and the parking lot parameters in each comparison result to form a structured size comparison table;

[0090] S3.2: retrieve the vehicle tilt-size offset model built-in the system, and according to the entrance and exit slope value of the candidate parking lot physical data, match the corresponding slope coefficient K slope from the model; the formula is as follows:

[0091] The corrected size difference = original size difference x (1 / K slope );

[0092] Wherein, the original size difference is the difference between the parking lot and the vehicle model parameter, and the corrected size difference quantitatively reflects the actual size of the vehicle model and the real allowance of the parking lot parameter after the vehicle body is inclined due to slope; the corrected size difference is updated to the structured size comparison table at the same time;

[0093] S3.3: set two threshold values: a safety threshold T1 representing the minimum size allowance that meets traffic safety; a critical threshold T2 representing the critical value of size allowance that reaches risk, and T2 < T1; compare the corrected size difference of each candidate parking lot with T1 and T2. If any one of the corrected size difference, the vehicle height difference and the vehicle width difference is < T2, it is determined as mismatched; if any one of the corrected size difference, the vehicle height difference and the vehicle width difference satisfies T2 < difference < T1, it is determined as risk to be checked; if the corrected size difference all satisfies ≥ T1, it is determined as preliminary adaptation; generate a structured preliminary risk table containing the parking lot identifier, the corrected size difference, the threshold comparison result and the preliminary risk level.

[0094] S4 further comprises the following:

[0095] S4.1: according to the structured preliminary risk table, screen out the parking lots determined as preliminary adaptation and risk to be checked; call the system real-time data interface to obtain the real-time passage occupancy rate and the entrance and exit congestion data of these two types of parking lots; wherein the real-time passage occupancy rate is the proportion data of the actual occupied width of the passage to the total width of the passage, and the entrance and exit congestion data is the average passage time data of the vehicles currently waiting to pass through the entrance and exit; then the above two types of data are associated with the corrected size difference of the corresponding parking lot to form a verification data set;

[0096] According to the real-time lane occupancy rate and the entrance and exit congestion data in the verification data set, the mean and standard deviation of the two types of data are calculated respectively; the abnormal values exceeding ±3 times the standard deviation of the mean are removed, and then the mean is recalculated using the remaining data to obtain the effective lane occupancy rate and the effective entrance and exit congestion duration; the effective data is updated to the verification data set;

[0097] S4.2: Calculate the current actual available lane width W of the parking lot based on the effective lane occupancy rate act The formula is as follows:

[0098] W act =W ori ×(1-C occ );

[0099] Wherein, W act is the actual available lane width; W ori is the original value of the lane width of the parking lot; C occ is the effective lane occupancy rate after processing; the formula is to quantify the influence of real-time lane occupancy on the actual available width;

[0100] Compare the calculated W act with the corrected vehicle width difference, and according to the formula, the actual width margin = W act - the maximum vehicle width of the vehicle type;

[0101] At the same time, set the entrance and exit congestion duration threshold T cong ; if the effective entrance and exit congestion duration of a parking lot exceeds T cong , it is determined that the parking lot has a traffic delay risk and needs to be added to the risk assessment dimension; through the dual judgment of the actual width margin and the congestion duration, the secondary verification result is formed;

[0102] S4.3: According to the judgment result of the actual width margin and the congestion duration, update the risk level combined with the safety threshold T1 and the critical threshold T2:

[0103] The actual width margin is greater than or equal to T1 and the effective entrance and exit congestion duration is less than or equal to T cong , and the level is determined as adaptive;

[0104] The actual width margin satisfies T2≤ actual width margin < T1, the effective entrance and exit congestion duration > T cong and the actual width margin is greater than or equal to T1, and any one of the above two conditions is satisfied, and the level is determined as needing attention;

[0105] The actual width margin is less than T2, and the level is not adaptive;

[0106] Finally, a structured final risk table containing the parking lot identifier, the secondary verification result and the final risk level is generated.

[0107] S5 further comprises the following contents:

[0108] S5.1: Collect the historical parking space data in the same period and the current real-time remaining parking space number of the parking lot determined as the adaptive level in the final risk table in the preset historical period; the historical parking space data includes the remaining parking space number at each collection time node in the period, and the current real-time remaining parking space number is the real-time remaining parking space number of the parking lot at the collection time; taking the parking lot identifier as the association item, the collected historical parking space data and the current real-time remaining parking space number are bound with the three-dimensional association database constructed in S1.3 to form an extended association database containing parking lot geographic coordinates-parking lot physical data-vehicle type parameters-parking space historical data-real-time remaining parking space number;

[0109] S5.2: From the extended association database, the historical parking space data in the same period in the preset historical period of the adaptive level parking lot is called; the total released parking space number ΔN in the period is calculated, ΔN is the absolute value of the difference between the remaining parking space number at the beginning of the period and the remaining parking space number at the end of the period; the total released parking space number ΔN is divided by the total length ΔT of the period to obtain the average parking space release rate R avg , the formula is as follows:

[0110] R avg = ΔN / ΔT;

[0111] Wherein R avg is the average parking space release rate per unit time, ΔN is the total released parking space number in the preset historical period, and ΔT is the total length of the preset historical period;

[0112] The current real-time remaining parking space number N real of the adaptive level parking lot is called from the extended association database; the estimated driving time T arrive of the vehicle owner from the current position to the parking lot is obtained through the geographic information system of the navigation software; based on the obtained average parking space release rate R avg , the formula I ava = N real +(R avg ×T arrive ) is used to construct the parking space availability index, wherein I ava is the parking space availability index, N real is the current real-time remaining parking space number of the parking lot, and T arrive is the estimated arrival time of the parking lot; the formula is used to combine the current parking space amount and the parking space release amount in the estimated arrival time to predict the parking space availability when the vehicle owner arrives, and to convert the static parking space data into a dynamic availability quantitative index;

[0113] S5.3: The distance data of the adaptive level parking lot and the destination is extracted from the original list generated in S2.1 and standardized to obtain Dnorm Set distance weight W dist Parking space availability index weight W ava Risk level weight W risk The sum of the three is 1; where the risk level weight W risk The fixed weight values ​​correspond to the adaptation levels in S4.3; the comprehensive score for each adapted parking lot is calculated using the following formula:

[0114] S total =W dist ×D norm +W ava ×I ava +W risk ×R score ;

[0115] Where S total For the overall score; R score To adapt to the risk score corresponding to the level, R score Based on the system's preset risk parameter-score mapping rules, the parameter combination where the corrected size difference is ≥ T1 and the actual width margin is ≥ T1 is assigned a fixed risk score R. score This mapping rule is built into the system, and R score The value is matched with the risk level of the suitability level; the function of this formula is to quantify the comprehensive suitability value of each parking lot by comprehensively considering three dimensions: the suitability of the parking lot to the destination, the availability of parking spaces, and the level of traffic risk.

[0116] S5.4: All parking lots of varying suitability levels will be ranked according to their overall score S. total Sort the data from highest to lowest. If there are cases with the same overall score, sort them according to the distance D between the parking lot and the destination. norm The parking information is sorted twice from nearest to furthest. The sorted parking information is then organized into a structured recommendation list. Each item in the list is associated with a parking lot identifier, a comprehensive score, a parking space availability index, a distance from the destination, and a risk level, generating a priority-ranked parking recommendation list.

[0117] S6 further includes the following:

[0118] S6.1: Retrieve the priority-sorted parking lot recommendation list and count the total number N of parking lots in the list that are suitable for each level. 适配 And the system's preset threshold number of adapters N 预设 , will N 适配 With N 预设 For comparison, if N 适配 ≥N 预设 If N 适配 <N 预设 This will trigger an operation to expand the filtering range;

[0119] S6.2: Based on the geographical coordinates of the destination entered by the car owner, filter the radius according to the system's preset range expansion rules to form a new expanded filtering range; repeat the adaptation analysis process of S2.1-S4.3 to filter out parking lots with adaptation levels within the expanded range to form a supplementary list of adapted parking lots.

[0120] Extract the nearest suitable parking lot from the recommendation list described in S5.4, and calculate the estimated travel time T to that parking lot using the navigation software's geographic information system. 原 And for each parking lot in the supplementary matching parking list, the estimated travel time T is also calculated using the navigation software's geographic information system. 新 The route adjustment time ΔT is calculated using a formula. 调 The formula is as follows:

[0121] ΔT 调 =T 新 -T 原 ;

[0122] Where ΔT 调 For the time spent adjusting the route, T 新 To supplement the estimated travel time for parking lots, T 原 This is the estimated travel time to the nearest suitable parking lot in the original recommendation list; the function of this formula is to quantify the additional time spent adjusting from the original planned route to the supplementary suitable parking lot, providing a time dimension basis for judging the priority of alternative solutions;

[0123] S6.3: For parking lots in the supplementary adaptation list, repeat the calculation logic of S5.2-S5.3 to obtain the comprehensive score S of the supplementary parking lots. total补 Then sort the Total scores from highest to lowest. If the overall scores are the same, sort them according to ΔT. 调 Sort the parking lots from smallest to largest; then merge the sorted supplementary parking lots with the original recommended list in S5.4, and label the ΔT for each parking lot. 调 The results were compiled into a structured low-risk alternative plan that includes parking lot signage, overall score, parking space availability index, route adjustment time, and risk level.

[0124] S7 further includes the following:

[0125] S7.1: Receive actual passage feedback submitted by the car owner after completing passage through any parking lot in the recommended list or low-risk alternative plan. The feedback includes the actual compatibility between the vehicle model and the parking lot, and whether there is a risk of scratches during the passage. Then, the feedback data is synchronously updated to the three-dimensional association database and the extended association database, replacing the old data of the corresponding parking lot in the database.

[0126] S7.2: Based on the updated association database, extract records in the vehicle owner's traffic feedback that are inconsistent with the system's predicted adaptation, and analyze the reasons for the differences:

[0127] If the difference is caused by adaptation deviation triggered by slope, retrieve the slope value of the parking lot entrance and the corresponding K slope value when the risk level under the actual slope is higher than the model prediction result, then adjust the K slope value according to the preset range, so that the corrected size difference meets the size margin during actual traffic;

[0128] If the difference is caused by the mismatch between the comprehensive score and the actual demand of the vehicle owner, then count the attention frequency of the vehicle owner to the distance, parking space availability, and traffic risk in such inconsistent feedback, and adjust the weight preset value in S5.3: when the feedback attention frequency of the parking space availability is higher than the current W ava , then gradually increase W ava according to the preset range; when the feedback attention frequency of the distance is lower than the current W dist , then gradually decrease W dist according to the preset range, to ensure that the weight setting is consistent with the actual traffic demand of the vehicle owner.

[0129] An intelligent parking management system based on data analysis, including a data collection and association database construction module, a geographic information interaction module, a parking lot screening module, an adaptation risk judgment module, a parking space availability calculation and recommendation list generation module, a backup solution generation module, and a feedback collection and adaptation logic optimization module.

[0130] The data collection and association database construction module is responsible for collecting parking lot physical data and vehicle owner vehicle type data, associating navigation geographic information to construct a three-dimensional association database; supplementary collection of parking space historical data and real-time remaining parking spaces, binding to the three-dimensional database to form an expanded association database; the geographic information interaction module is used to receive the destination input by the vehicle owner and convert it into standardized geographic coordinates, provide spatial distance calculation and estimated driving time of the vehicle owner to the parking lot, support range definition and route adjustment time consumption calculation; the parking lot screening module is responsible for setting the screening range according to the destination coordinates, retrieving parking lot data within the range from the association database to form an original list; verifying and removing invalid parking lots to generate a structured candidate list; the adaptive risk judgment module is used to retrieve candidate parking lot data and vehicle owner vehicle type parameters, compare sizes, correct slopes, and verify real-time data to calculate the actual width margin and determine the risk level combined with congestion duration to generate a final risk table; the parking space availability calculation and recommended list generation module is used to retrieve adaptive parking lot data to calculate the average release rate per unit time, combined with real-time remaining parking spaces and estimated arrival time to construct a parking space availability index; calculate the comprehensive score according to the distance, index, risk level weight, and generate a priority recommendation list; the alternative scheme generation module is responsible for expanding the screening range to supplement adaptive parking lots when the number of adaptive parking lots in the recommendation list does not reach the preset threshold; calculate the route adjustment time and sort the supplementary parking lots, and merge them with the original list to generate a low-risk alternative scheme; the feedback collection and adaptive logic optimization module is responsible for receiving vehicle owner traffic feedback and parking lot parameter change data, updating the association database; analyze the feedback and the adaptive difference of the prediction, adjust the parameters, and optimize the adaptive judgment logic.

[0131] A vehicle owner Zhang San drives a SUV vehicle labeled "install roof rack" in a city, needs to go to the destination "Wangfujing Department Store" (geographic coordinates 116.41°E, 39.92°N), and finds an adaptive parking lot through the intelligent parking management system of the application. The specific operation process of the system is as follows:

[0132] Firstly, the system starts the data collection and association database construction step. In the parking lot physical data collection link (S1.1), the system collects the physical data of the existing parking lots in the target area, wherein the entrance height of parking lot A (geographic coordinates 116.40°E, 39.90°N) is 2.2 m, the passage width is 3.5 m, and the entrance slope is 5°; the entrance height of parking lot B (geographic coordinates 116.42°E, 39.91°N) is 2.0 m, the passage width is 3.2 m, and the entrance slope is 8°; the entrance height of parking lot C (geographic coordinates 116.39°E, 39.89°N) is 2.1 m, the passage width is 3.3 m, and the entrance slope is 10°. The system standardizes these data, unifies the units as "meters" and "degrees", forms a structured parking lot physical data set, and associates it with the navigation software geographic information system, so that the physical data of each parking lot accurately corresponds to its geographic coordinates. In the vehicle model data collection link (S1.2), the system receives the vehicle model data entered by Zhang San when registering - vehicle model name "a brand SUV, install roof rack", then calls the built-in motor vehicle parameter database, matches and extracts the original factory maximum vehicle height 1.8 m and maximum vehicle width 1.9 m of the SUV, and because of the label "install roof rack", the roof rack height is added 0.1 m, the maximum vehicle height is updated to 1.9 m, forming a vehicle model parameter set containing the individual characteristics of Zhang San's vehicle model. Finally (S1.3), the system indexes the parking lot geographic coordinates, associates the parking lot physical data set with Zhang San's vehicle model parameter set, and constructs a three-dimensional association database of parking lot geographic coordinates-parking lot physical data set-vehicle model parameter set, providing a data basis for subsequent steps.

[0133] After the three-dimensional association database is constructed, the system enters the parking lot screening link (S2). In the first step (S2.1), the system sets a preset spatial screening range with the destination Wangfujing Department Store geographic coordinates as the center and a radius of 1 km, determines the parking lots within the range as A (distance 0.8 km) and B (distance 0.6 km) through the spatial distance calculation function of the geographic spatial information system, retrieves the basic information and physical data of these two parking lots from the three-dimensional association database, sorts them according to "distance from destination from near to far", and forms an original list: parking lot B (0.6 km), parking lot A (0.8 km). In the second step (S2.2), the system verifies the effectiveness of the original list and finds that parking lot B is marked as "temporary closed" in the database, has no "limited to specific vehicles" label, and the vehicle model matches Zhang San's vehicle model parameter set, but is still excluded due to temporary closure; parking lot A has no closure label, no vehicle type restriction, and preliminary matches Zhang San's vehicle model, and finally arranges parking lot A as a structured candidate list, which associates the geographic coordinates, physical data of parking lot A, and the status of preliminary matching with Zhang San's vehicle model.

[0134] After the candidate list is generated, the system starts the adaptive risk judgment step (S3). First (S3.1), the system retrieves the physical data of parking lot A from the candidate list, and retrieves the vehicle model parameter set bound to Zhang San's account at the same time. The maximum vehicle height of Zhang San's vehicle model is 1.9m, which is compared with the entrance height of parking lot A, 2.2m, to obtain a vehicle height difference of 0.3m; the maximum vehicle width is 1.9m, which is compared with the channel width of parking lot A, 3.5m, to obtain a vehicle width difference of 1.6m; these differences are recorded to form a structured size comparison table. Next (S3.2), the system retrieves the built-in vehicle inclination-size offset model, and according to the slope of the entrance and exit of parking lot A, 5°, matches the corresponding slope coefficient K slope =1.05 (model preset slope 5° corresponds to K slope =1.05), calculates according to the formula "corrected size difference = original size difference x (1 / K slope )": corrected vehicle height difference = 0.3 x (1 / 1.05) ≈ 0.286m, corrected vehicle width difference = 1.6 x (1 / 1.05) ≈ 1.524m, and the corrected difference is updated to the size comparison table at the same time. Finally (S3.3), the system sets a safety threshold T1 = 0.2m (representing the minimum size margin that meets traffic safety) and a critical threshold T2 = 0.1m (representing the critical value of the size margin that reaches risk), and compares the corrected difference of parking lot A with the threshold: corrected vehicle height difference 0.286m ≥ T1, corrected vehicle width difference 1.524m ≥ T1, and determines "initial adaptation", generating a structured initial risk table containing "parking lot A identification, corrected difference, threshold comparison result, initial adaptation level".

[0135] After the initial risk judgment is completed, the system enters the real-time data secondary verification link (S4). First (S4.1), the system selects "initial adaptation" parking lot A according to the initial risk table, and calls the real-time data interface to obtain its real-time channel occupancy rate and entrance and exit congestion data: the continuous 5 times collected values of real-time channel occupancy rate are 15%, 16%, 14%, 50%, and 15%, and the continuous 5 times collected values of entrance and exit congestion data are 2min, 3min, 2.5min, 12min, and 2.2min. The system calculates the mean μ occ =(15%+16%+14%+50%+15%) / 5=22%, the standard deviation σ occ ≈15.6%, and the 50% exceeding the range of "μ occ ±3σ occ "(22%-46.8% to 22%+46.8%) is removed, and the remaining 4 times of data are used to recalculate the mean, obtaining the effective channel occupancy rate C occ =15%; similarly, the mean μ cong =(2+3+2.5+12+2.2) / 5=4.34min, and the standard deviation σcong ≈4.1min, remove data beyond "μ cong ±3σ cong " (4.34-12.3min to 4.34+12.3min) range, recalculate mean value with remaining 4 data, get effective entrance-exit congestion duration 2.425min, then associate effective data with difference of modified size of parking lot A, form verification data set. Second step (S4.2), system calculates actual available lane width according to formula "W act = W ori ×(1-C occ )", where W ori is original value of lane width of parking lot A 3.5m, C occ is effective lane occupancy rate 15%, substitute to get W act =3.5×(1-15%) =2.975m; then calculate according to formula "actual width margin = W act -maximum vehicle width of vehicle type" to get actual width margin =2.975-1.9 =1.075m. At the same time, system sets entrance-exit congestion duration threshold T cong =5min, compare effective entrance-exit congestion duration of parking lot A 2.425min with T cong , determine no traffic delay risk. Third step (S4.3), system updates risk level in combination with safety threshold T1=0.2m and critical threshold T2=0.1m: actual width margin of parking lot A 1.075m≥T1, and effective entrance-exit congestion duration≤T cong , finally determine risk level as "adaptation", generate structured final risk table containing "identification of parking lot A, secondary verification result, adaptation level".

[0136] After completing adaptation determination, system enters step of parking space availability calculation and recommended list generation (S5). First step (S5.1), system collects preset historical period data of parking lot A with "adaptation" level in final risk table - parking space historical data (remaining parking space number at each time node: 14:00 for 20 vehicles, 14:15 for 18 vehicles, 14:30 for 15 vehicles, 14:45 for 10 vehicles, 15:00 for 8 vehicles) in the past 7 days 14:00-15:00 (consistent with current time period), and current real-time remaining parking space number N real=15 vehicles; using the parking lot A identifier as the associated item, these parking space data are bound to the three-dimensional associated database constructed in S1.3, forming an extended associated database containing "parking lot geographic coordinates - physical data - vehicle parameters - historical parking space data - real-time remaining parking spaces". In the second step (S5.2), the system retrieves the historical parking space data of parking lot A from the extended associated database, calculates the total number of released parking spaces ΔN = |20 - 8| = 12 vehicles within the preset historical period (60 minutes), and then calculates the number of vehicles released according to the formula "R". avg =ΔN / ΔT (ΔT = 60min) Calculate the average parking space release rate R per unit time avg =12 / 60=0.2 vehicles / min; simultaneously, using the navigation software's geographic information system, calculate the estimated travel time T from Zhang San's current location (116.405°E, 39.905°N) to parking lot A. arrive =10min, according to formula "I ava =N real +(R avg ×T arrive Construct a parking space availability index, and substitute it into the equation to obtain I. ava =15 + (0.2 × 10) = 17. In the third step (S5.3), the system extracts the distance of parking lot A from the original list generated in S2.1 (0.8 km) to the destination, and presses "D". norm =1 - (distance / maximum screening radius)" (maximum screening radius 1km) is standardized to obtain D norm =1-(0.8 / 1)=0.2; Set distance weight W dist =0.3, Parking space availability index weight W ava =0.5, Risk Level Weight W risk =0.2 (the sum of the three is 1), where the risk level weight W risk The fixed weight value corresponding to the "adaptation" level is 0.2; at the same time, according to the system's preset risk parameter-score mapping rules, the parameter combination of "corrected size difference ≥ T1 and actual width allowance ≥ T1" corresponds to a fixed risk score R. score =90, according to the formula "S total =W dist ×D norm +W ava ×I ava +W risk ×R score Calculate the overall score and substitute it to get S. total= 0.3×0.2 + 0.5×17 + 0.2×90 = 0.06 + 8.5 + 18 = 26.56. In the fourth step (S5.4), the system sorts the "suitable" level parking lots in descending order according to the comprehensive score. Currently, only Parking Lot A meets the criteria. Therefore, it is sorted into a structured recommendation list, and the identification of Parking Lot A, the comprehensive score of 26.56, the parking space availability index of 17, the distance of 0.8 km from the destination, and the "suitable" level are associated in the list to generate a prioritized parking lot recommendation list.

[0137] After the recommendation list is generated, the system checks whether the number of suitable parking lots meets the standard (S6). In the first step (S6.1), the system counts the number N of "suitable" level parking lots in the recommendation list 适配 = 1, while the preset threshold N for the number of suitable ones 预设 = 2. Since 1 < 2, an operation to expand the screening range is triggered. In the second step (S6.2), the system expands the screening radius from 1 km to 2 km according to the preset range expansion rule, and determines the newly added Parking Lot C (1.2 km away from the destination) within the expanded range through the Geographic Information System; then repeats the adaptation analysis process of S2.1 - S4.3: screens Parking Lot C into the candidate list, compares its physical data with Zhang San's vehicle model parameters (entrance height 2.1 m - vehicle height 1.9 m = 0.2 m, aisle width 3.3 m - vehicle width 1.9 m = 1.4 m), and matches K according to a slope of 10° slope＝ 1.15, the corrected vehicle height difference = 0.2×(1 / 1.15) ≈ 0.174 m, vehicle width difference = 1.4×(1 / 1.15) ≈ 1.217 m (both ≥ T1 = 0.2 m? Here 0.174 m < T1, judged as "risk to be verified"); then retrieves the real-time data of Parking Lot C, the effective aisle occupancy rate is 20%, calculates W act = 3.3×(1 - 20%) = 2.64 m, the actual width margin = 2.64 - 1.9 = 0.74 m ≥ T1, the effective entrance and exit congestion duration is 4 min ≤ T cong = 5 min, and finally determines that Parking Lot C is "suitable", forming a supplementary list of suitable parking lots. At the same time, the system extracts the estimated travel time T of the nearest suitable Parking Lot A from the original recommendation list 原 = 10 min, calculates the estimated travel time T of Parking Lot C 新 = 15 min, according to the formula "ΔT 调 = T 新 - T 原 " to get ΔT 调 = 15 - 10 = 5 min, quantifying the route adjustment time. In the third step (S6.3), the system repeats the calculation logic of S5.2 - S5.3 for Parking Lot C: its historical parking space data ΔN = 10 vehicles (within 60 min), R avg = 10 / 60 ≈ 0.167 vehicles / min, Nreal = 12, T arrive = 15 min, I ava = 12 + (0.167 * 15) = 14.5; distance 1.2 km, D norm = 1 - (1.2 / 2) = 0.4; R score = 90, comprehensive score S total = 0.3 * 0.4 + 0.5 * 14.5 + 0.2 * 90 = 0.12 + 7.25 + 18 = 25.37. The system ranks parking lot C by S total suppl from high to low (second to parking lot A), and if the comprehensive score is the same, it is ranked by AT 调 from small to large, and then it is merged with the original recommended list, and the AT 调 = 5 min, and it is arranged as a structured low-risk alternative solution containing "parking lot identification, comprehensive score, parking space availability index, route adjustment time consumption, risk level".

[0138] Zhang San finally chooses parking lot A to complete parking, and the system enters the feedback collection and adaptive logic optimization step (S7). The first step (S7.1), the system receives the actual traffic feedback submitted by Zhang San - "no risk of scratching, vehicle type and parking lot are matched", and at the same time collects the parameter change data of parking lot A (no parameter change), updates these feedback data to the three-dimensional correlation database and the extended correlation database, and replaces the old data of parking lot A in the database. The second step (S7.2), based on the updated correlation database, the system extracts the traffic feedback of other vehicle owners, and finds that some vehicle owners feedback "Parking lot B (has been restored) slope 8°, the system judges as'matched' but the actual traffic has a risk of scratching", analyze the difference reason: the K slope of parking lot B slope 8° corresponds to 1.1 originally preset, and the corrected vehicle height difference value = 0.3 * (1 / 1.1) = 0.273 m > T1, while the actual scratching shows that K slope The value is too low, resulting in overestimation of the safety margin of the corrected difference. The system immediately adjusts the K slope matching rule of the vehicle tilt-size offset model to K slope from 1.1 to 1.2, so that the corrected vehicle height difference value = 0.3 * (1 / 1.2) = 0.25 m (still > T1, but closer to the actual safety margin); At the same time, the frequency of "vehicle owners' attention to parking space availability" in recent feedback is higher than the current W ava proportion", the weight W ava of the parking space availability index in S5.3 is adjusted from 0.5 to 0.55, and the distance weight W distDown from 0.3 to 0.25 (keep the sum of the three as 1), ensure that the weight setting is consistent with the actual demand of the owner, complete the optimization of the subsequent vehicle-parking lot adaptation judgment logic, form a "data-feedback-optimization" closed loop.

[0139] Through the above process, the present application realizes the whole process management from data collection, parking lot screening, adaptation judgment, parking space pre-judgment to feedback optimization, accurately solves the core problems of "vehicle type adaptation" and "parking space availability", and provides efficient and reliable parking recommendation services for the owner.

[0140] It is apparent for those skilled in the art that the present application is not limited to the details of the above exemplary embodiments, but can be implemented in other concrete forms without departing from the spirit or essential characteristics of the present application. Therefore, the embodiments should be considered in all aspects as illustrative and not restrictive, the scope of the present application being defined by the appended claims rather than the above description, and all changes falling within the meaning and range of the equivalent elements of the claims are intended to be embraced by the present application. Any reference signs in the claims should not be considered as limiting the claims involved.

Claims

1. A data analysis based intelligent parking management method, characterized in that: Comprise the following steps: S1, collect all existing parking lot physical data, associated with the navigation software geographic information system, and receive the car owner car type data, build the association database of two kinds of data; S2, according to the destination input by the car owner, combined with the geographic information from the associated database in the preset range of parking lot screening, form a candidate list; S3, the car owner car type data and parking lot data are compared, combined with the entry slope characteristics to judge whether the same line of car type exists scratch risk; S4, call the real-time channel occupancy rate of the preliminary adaptation of the candidate parking lot, the entrance congestion data, the second verification of the traffic adaptation and the update of the parking lot risk level; S5, call the historical data of parking space in the same period in the preset historical period of the adaptive parking lot, calculate the average release rate of parking space per unit time, and build the parking space availability index; then combined with the distance between the adaptive parking lot and the destination, the parking space availability index and the risk level to calculate the comprehensive score, generate the priority sorting of the parking lot recommendation list; S6, if the number of adaptive parking lot in the recommendation list does not reach the preset value, expand the screening range and repeat the adaptive analysis, calculate the time consumption of route adjustment and generate low risk alternative scheme; S7, collect the actual traffic feedback of the car owner and the parameter change data of the parking lot, update the associated database synchronously, and optimize the subsequent car type-parking lot adaptive judgment logic.

2. The intelligent parking management method based on data analysis according to claim 1, characterized in that: S1 further comprises the following contents: S1.1: collect all existing physical data of each parking lot, the physical data including the entrance height, channel width and entrance slope of the parking lot; and standardize the collected physical data to form a structured parking lot physical data set, associate the data set with the geographic information system of the navigation software, so that the physical data of each parking lot corresponds to its geographic coordinates; S1.2: receive the car type data input by the car owner when registering in the system, the car type data including the car type name, whether the roof additional device is installed; then call the built-in motor vehicle parameter database in the system, match and extract the original factory maximum height and maximum width of the car type according to the received car type name, if the car owner marks the installation of roof additional device, then superimpose the height value to update the maximum height, form the car type parameter set containing the individual characteristics of the car type of the car owner; S1.3: taking the geographic coordinates of the parking lot as the index, the parking lot physical data set and the car type parameter set are associated to construct the three-dimensional associated database of parking lot geographic coordinates-parking lot physical data set-car type parameter set.

3. The intelligent parking management method based on data analysis according to claim 1, characterized in that: S2 further comprises the following contents: S2.1: according to the destination geographic coordinates input by the car owner, set the preset spatial screening range with the coordinates as the center; determine the geographic coordinates of all parking lots in the range through the spatial distance calculation function of the geographic spatial information system; then call all parking lot data from the associated database whose geographic coordinates are in the above defined range, extract the parking lot data and preliminarily sort them according to the distance from the destination from near to far to form the original list containing the basic information and physical data of the parking lot; S2.2: Verify the effectiveness of the parking lots in the original list, and exclude those marked as temporary closure, permanent closure, only for specific vehicles, and the current vehicle model does not match the owner's vehicle model parameter set. The remaining parking lot data after verification is organized into a structured candidate list, with each parking lot entry associated with its geographic coordinates, physical data, and owner's vehicle model preliminary matching status.

4. The intelligent parking management method based on data analysis according to claim 1, characterized in that: S3 further includes the following: S3.1: Retrieve the physical data of each parking lot in the candidate list, and retrieve the vehicle model parameter set bound to the owner's account. Then compare the maximum height and maximum width of the owner's vehicle model parameter set with the entrance height and passage width of the corresponding parking lot physical data. Record the difference between the vehicle model parameter and the parking lot parameter in each comparison result to form a structured size comparison table. S3.2: Call the vehicle tilt-size offset model built in the system, according to the entrance slope value of the candidate parking lot physical data, match the corresponding slope coefficient K from the model slope , adopt the formula to get the corrected size difference; The formula is as follows: Modified size difference = original size difference x (1 / K slope ); Wherein, the original size difference is the difference between the parking lot and the vehicle model parameter, and the corrected size difference reflects the actual size of the vehicle model and the true allowance of the parking lot parameter after the vehicle body is inclined due to the slope. Update the corrected size difference to the structured size comparison table. S3.3: Set two thresholds: a safety threshold T1 representing the minimum size allowance that meets traffic safety, and a critical threshold T2 representing the critical value of the size allowance that reaches the risk, and T2 < T1. Compare the corrected size difference of each candidate parking lot with T1 and T2. If any of the corrected size difference, height difference and width difference is < T2, it is determined as not matching. If any of the corrected size difference, height difference and width difference satisfies T2 ≤ difference < T1, it is determined as risk to be verified. If all the corrected size differences satisfy ≥ T1, it is determined as preliminary adaptation. Generate a structured preliminary risk table containing parking lot identification, corrected size difference, threshold comparison result and preliminary risk level.

5. The intelligent parking management method based on data analysis according to claim 1, characterized in that: S4 further includes the following: S4.1: According to the structured preliminary risk table, select the parking lots determined as preliminary adaptation and risk to be verified. Call the system real-time data interface to obtain the real-time passage occupancy rate and entrance congestion data of these two types of parking lots. Wherein, the real-time passage occupancy rate is the proportion data of the actual occupied width to the total width of the passage, and the entrance congestion data is the average passage time data of the vehicles currently waiting at the entrance. Then, the above two types of data are associated with the corrected size difference of the corresponding parking lot to form a verification data set. According to the real-time passage occupancy rate and entrance congestion data in the verification data set, calculate the mean and standard deviation of the two types of data respectively. Remove outliers that are more than ±3 times the standard deviation of the mean, and then recalculate the mean using the remaining data to obtain the effective passage occupancy rate and effective entrance congestion time. Update the effective data to the verification data set. S4.2: Calculate the current actual available lane width W of the parking lot based on the effective lane occupancy act The formula is as follows: W act = W ori × (1 - C occ ); Wherein, W act is the actual available lane width; W ori is the parking lot lane width original value; C occ is the processed effective lane occupancy rate; The calculated W act The actual width allowance = W act - The maximum vehicle width of the vehicle model; At the same time, the exit congestion duration threshold T is set cong If the effective exit congestion duration of a parking lot exceeds T cong , it is determined that the parking lot has a risk of traffic delay, which needs to be added to the risk assessment dimension S4.3: Update the risk level according to the actual width allowance and congestion time judgment results, combined with the safety threshold T1 and the critical threshold T2: Actual width allowance ≥ T1 and effective access congestion duration ≤ T cong , the level is determined as adaptive; The actual width allowance satisfies T2≤actual width allowance<T1, and the effective entrance and exit congestion duration>T cong And the actual width allowance≥T1, satisfying any one of the above two conditions, the grade is determined as attention needed; Actual width allowance < T2, level is not adapted. Finally, generate a structured final risk table containing parking lot identification, secondary verification result and final risk level.

6. The intelligent parking management method based on data analysis according to claim 1, characterized in that: S5 further includes the following: S5.1: Collect the historical parking space data of the same period in the preset historical period and the current real-time remaining parking space number of the parking lot determined as the adaptive level in the final risk table; the historical parking space data includes the remaining parking space number of each collection time node in the period, and the current real-time remaining parking space number is the instant remaining parking space number of the parking lot at the collection time; taking the parking lot identifier as the association item, the collected historical parking space data and the current real-time remaining parking space number are bound with the three-dimensional association database constructed in S1.3 to form an expanded association database containing parking lot geographic coordinates-parking lot physical data-vehicle type parameters-parking space historical data-real-time remaining parking space number; S5.2: From the extended association database, call the historical data of parking spaces in the same period within the preset historical period of the adaptive level parking lot; calculate the total released parking space number ΔN in the period, ΔN is the absolute value of the difference between the remaining parking space number at the beginning of the period and the remaining parking space number at the end of the period; divide the total released parking space number ΔN by the total time length ΔT of the period to obtain the average parking space release rate R per unit time avg , The formula is as follows: R avg = ΔN / ΔT; wherein R avg is the average release rate of parking spaces per unit of time, ΔN is the total number of released parking spaces in the preset historical period, and ΔT is the total duration of the preset historical period. Retrieve the current real-time number of remaining parking spaces N for the appropriate parking level from the extended relational database. real The estimated travel time T for the driver to reach the parking lot from their current location is obtained through the geographic information system of the navigation software. arrive Based on the obtained average parking space release rate R avg Formula I ava =N real +(R avg ×T arrive Construct a parking space availability index, where I ava N is the parking space availability index. real T represents the number of parking spaces currently available in the parking lot in real time. arrive Estimate the time it takes for drivers to arrive at the parking lot; S5.3: Extract the distance data between the adaptive level parking lot and the destination from the original list generated in S2.1 and process it to obtain D norm ; Set the distance weight W dist , the parking space availability index weight W ava , and the risk level weight W risk , and the sum of the three is 1; wherein the risk level weight W risk corresponds to the fixed weight value of the adaptive level in S4.3; the comprehensive score of each adaptive parking lot is calculated by using the formula as follows: S total = W dist x D norm + W ava x I ava + W risk x R score ; Wherein S total is the comprehensive score; R score is the risk score corresponding to the adaptation level, R score is a fixed risk score R score corresponding to the parameter combination of the modified size difference ≥T1 and the actual width allowance ≥T1 based on the system preset risk parameter-score mapping rule, the mapping rule has been built-in the system, and R score takes a value matching the risk level of the adaptation level; S5.4: Rank all the parking lots by the comprehensive score S total Sort from high to low, if there are the same comprehensive score, then sort by the distance D between the parking lot and the destination norm Sort from near to far, and arrange the sorted parking lot information into a structured recommendation list, each entry in the list is associated with the parking lot identifier, the comprehensive score, the parking space availability index, the distance to the destination, and the risk level, and generate a priority-ordered parking lot recommendation list.

7. The intelligent parking management method based on data analysis according to claim 1, characterized in that: S6 further comprises the following content: S6.1: Call the priority ranking parking lot recommendation list, count the total number of adaptive level parking lots in the list N 适配 and the system preset adaptive number threshold N 预设 , compare N 适配 with N 预设 , if N 适配 ≥ N 预设 , no need to execute the subsequent sub-steps; if N 适配 < N 预设 , trigger the operation of expanding the screening range; S6.2: According to the geographic coordinates of the destination input by the vehicle owner, a new expanded screening range is formed by screening the radius according to the system preset range expansion rule; the adaptive analysis process of S2.1-S4.3 is repeatedly executed to screen the parking lots of the adaptive level in the expanded range to form a supplementary adaptive parking lot list; extracting the closest adapted parking lot from the recommended list described in S5.4, calculating the estimated driving time T of the parking lot by the navigation software geographic information system 原 , and calculating the estimated driving time T of each parking lot in the supplementary adapted parking lot list by the navigation software geographic information system as well 新 ; calculating the route adjustment time consumption ΔT by the formula 调 , the formula is as follows: ΔT 调 = T 新 - T 原 ; where ΔT 调 is the route adjustment time, T 新 is the estimated travel time to the supplementary adaptation parking lot, T 原 is the estimated travel time to the nearest adaptation parking lot in the original recommendation list; S6.3: Repeat the calculation logic of S5.2-S5.3 for the supplementary parking lots in the supplementary parking lot list, and get the supplementary parking lot comprehensive score S total补 , and sort Stotal from high to low, and if the comprehensive scores are the same, sort them according to ΔT 调 from small to large; then merge the sorted supplementary parking lots with the original recommended list of S5.4, and mark the ΔT 调 of each parking lot, and arrange them into a structured low-risk alternative solution containing the parking lot identifier, comprehensive score, parking space availability index, route adjustment time consumption, and risk level.

8. The intelligent parking management method based on data analysis according to claim 1, characterized in that: The S7 further comprises the following content: S7.1: Receive the actual traffic feedback submitted by the vehicle owner after completing the traffic in any parking lot in the recommended list or the low-risk alternative scheme, the feedback including the actual adaptation of the vehicle type and the parking lot, and whether there is a scratch risk in the traffic process; Then update the feedback data to the three-dimensional association database and the expanded association database to replace the old data of the corresponding parking lot in the database; S7.2: Based on the updated association database, extract the records of the feedback adaptation inconsistent with the system prediction adaptation in the vehicle owner's traffic feedback, and analyze the difference reasons: If the difference is caused by the slope-induced adaptive deviation, the slope value of the parking lot entrance and the corresponding K at the prediction time are called slope The actual scratch situation of the owner feedback is compared with the calculated corrected size difference. When the risk level under the actual slope is higher than the model prediction result, the K value corresponding to the slope value is adjusted slope The value is gradually adjusted according to the preset range, so that the corrected size difference meets the size allowance during actual traffic; If the difference is due to the mismatch between the comprehensive score and the actual demand of the owner, the frequency of the owner's attention to the distance, parking availability, and traffic risk in such inconsistent feedback is counted, and the preset weight in S5.3 is adjusted: when the feedback attention frequency of the parking availability is higher than the current W ava , the preset range is gradually increased ava ; when the feedback attention frequency of the distance is lower than the current W dist , the preset range is gradually decreased dist , to ensure that the weight setting is consistent with the actual traffic demand of the owner.

9. An intelligent parking management system based on data analysis, applied to the intelligent parking management method based on data analysis in claims 1-8, characterized in that: The data collection and association database construction module, the geographic information interaction module, the parking lot screening module, the adaptive risk judgment module, the parking space availability calculation and recommended list generation module, the alternative scheme generation module, and the feedback collection and adaptive logic optimization module are included. The data collection and association database construction module is responsible for collecting parking lot physical data and vehicle owner vehicle type data, and constructing a three-dimensional association database by associating navigation geographic information; supplementary collection of parking space historical data and real-time remaining parking space number, binding to the three-dimensional database to form an expanded association database; the geographic information interaction module is used to receive the destination input by the vehicle owner and convert it into standardized geographic coordinates, provide spatial distance calculation and estimated driving time from the vehicle owner to the parking lot, support screening range definition and route adjustment time consumption calculation; the parking lot screening module is responsible for setting the screening range according to the destination coordinates, and forming an original list from the parking lot data in the range; Verify and exclude invalid parking lots to generate a structured candidate list; The adaptive risk judgment module is used to call the candidate parking lot data and vehicle owner vehicle type parameters, perform size comparison, slope correction and real-time data verification, calculate the actual width margin, and determine the risk level combined with the congestion duration to generate a final risk table; The parking space availability calculation and recommendation list generation module is configured to call the average release rate of parking space data of the adaptive parking lot per unit time, combine the real-time remaining parking space number and the estimated arrival time to construct a parking space availability index, calculate a comprehensive score according to the distance, the index, the risk level weight, sort and generate a priority recommendation list; The alternative solution generation module is responsible for expanding the screening range to select the adaptive parking lot when the number of the adaptive parking lot in the recommendation list does not reach the preset threshold; The time consumption of route adjustment is calculated, the supplementary parking lot is sorted, and the low-risk alternative solution is generated by combining the original list; The feedback collection and adaptive logic optimization module is responsible for receiving the driver's passing feedback and the parking lot parameter change data, updating the associated database; The adaptive difference between the analysis feedback and the prediction is adjusted, the parameters are adjusted, and the adaptive judgment logic is optimized.

Images