Intelligent shopping mall-based advanced reservation garage parking guidance method and system

Abstract

The present application relates to parking management technical field, especially intelligent shopping mall based on advanced reservation garage parking guidance method and system, wherein the method comprises: obtaining information; screening remaining parking spaces, obtaining distribution map; calculating parking difficulty value; selecting target parking space and planning parking path; collecting driving person's operation habit, parking difficulty value and parking time in parking process, dividing proficiency according to driving person's operation habit, and respectively storing into novice data set and proficiency data set; training parking time model; obtaining the predicted parking time corresponding to each parking space in the parking lot; according to the predicted parking time, the parking path is corrected and guided. The present application reduces the channel retention caused by the mismatch of parking spaces or the long waiting time of the previous vehicle, thereby reducing the waiting time in the parking lot and further reducing the congestion probability.

Description

Technical Field

[0001] This invention relates to the field of parking management technology, and in particular to a method and system for guiding parking in advance reservation garages based on smart shopping malls. Background Technology

[0002] With the acceleration of urbanization and the continuous growth of car ownership, the service quality of parking lots in smart shopping malls, which are comprehensive commercial spaces integrating consumption, leisure and services, has become a core element in improving user experience and mall competitiveness. Traditional shopping mall parking lots can no longer meet the needs of intelligent and refined services, and upgrading parking guidance technology has become an important issue in the construction of smart shopping malls.

[0003] However, existing parking guidance technologies still have many drawbacks: On the one hand, parking space screening and matching lack specificity. Existing systems mostly recommend spaces based solely on their availability, without adapting them to the specific characteristics of vehicles. On the other hand, parking difficulty assessment and scheduling do not consider individual differences. Existing technologies rely heavily on subjective experience (such as distance from the entrance) to assess "parking difficulty," failing to quantify the impact of physical characteristics of parking spaces (aisle width, turning angle, etc.) on parking operations, leading to excessively long parking times and congestion in access lanes.

[0004] Therefore, existing technologies have shortcomings and urgently need improvement. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for guiding parking in advance reservation garages based on smart shopping malls, in order to solve the problems in the existing technology where parking space screening and matching lacks specificity and parking difficulty assessment and scheduling does not take into account individual differences, resulting in excessively long parking times and causing congestion in the passageways.

[0006] This invention provides a method for guiding parking in advance reservation garages based on smart shopping malls, comprising:

[0007] In response to a parking reservation, obtain the current vehicle information;

[0008] Based on the current vehicle information, the remaining parking spaces are filtered, and image information of the parking lot is obtained. The image information of the parking lot is then identified to obtain a distribution map of the candidate parking spaces.

[0009] The parking difficulty value of each of the candidate parking spaces is calculated based on the distribution map and the current vehicle information;

[0010] Select the candidate parking space with the lowest parking difficulty value as the target parking space and plan the parking route;

[0011] The system collects drivers' operating habits, parking difficulty values, and parking time during the parking process. It categorizes drivers' proficiency levels based on their operating habits and stores them in separate datasets for beginners and proficient drivers.

[0012] Train parking time models based on the novice dataset and the proficient dataset respectively;

[0013] In response to a car entering the parking lot, obtain the predicted parking time for each parking space currently occupied.

[0014] Based on the predicted parking time, the parking route is corrected and parking guidance is provided.

[0015] As a preferred technical solution for the advanced reservation parking guidance method based on smart shopping malls, the current vehicle information includes: vehicle model, vehicle length, vehicle width, maximum steering angle, and whether charging is required.

[0016] As a preferred technical solution for a smart shopping mall-based advance reservation parking guidance method, the step of calculating the parking difficulty value of each of the candidate parking spaces based on the distribution map and the current vehicle information includes:

[0017] Acquire images of each candidate parking space and identify obstacles in the parking lot;

[0018] The actual width of the passageway between parking spaces for vehicles is obtained based on the distribution map and obstacles in the parking lot.

[0019] Based on the vehicle model, length, width, actual channel width, and kinematic model, the current vehicle's parking trajectory is simulated.

[0020] Select the trajectories with the largest and smallest radii, calculate their corresponding steering angles, and use them as the maximum and minimum values ​​to form the steering angle intervals;

[0021] The percentage of the steering angle range in which the steering angle is greater than the preset steering angle is calculated and recorded as the parking difficulty value.

[0022] As a preferred technical solution for a smart shopping mall-based advance reservation parking guidance method, the method involves collecting drivers' operating habits, parking difficulty values, and parking time during the parking process. Based on drivers' operating habits, proficiency levels are categorized and stored separately in a novice dataset and a proficiency dataset, including:

[0023] For each parking space, data on parking time, parking habits, and parking difficulty are collected from the moment a vehicle starts reversing until it enters the garage, and stored in the form of data groups.

[0024] The data set is cleaned to remove data sets with parking times less than the first-level threshold or greater than the second-level threshold, thus obtaining a standard data set.

[0025] Based on the aforementioned operating habits, each of the standard data groups is divided into a proficient standard data group and a novice standard data group;

[0026] Store the proficient standard data set into the proficient dataset, and the novice standard data set into the novice dataset.

[0027] As a preferred technical solution for the advanced reservation parking guidance method for smart shopping malls, the operating habits are one of the following: the total number of steering wheel turns required to complete parking, the maximum steering angle deviation during parking, or the interval time from starting to reverse to the first adjustment of the steering wheel.

[0028] As a preferred technical solution for the advanced reservation parking guidance method based on smart shopping malls, the step of dividing the standard data groups into skilled standard data groups and novice standard data groups according to the operating habits includes:

[0029] In response to the total number of steering wheel rotations exceeding a preset threshold, or the maximum steering angle deviation exceeding a preset angle threshold, or the interval time being less than a preset time threshold, the standard data group is divided into the novice standard data group; otherwise, it is divided into the proficient standard data group.

[0030] The parking time model is trained using the novice standard data set and the expert standard data set, respectively.

[0031] As a preferred technical solution for the advance reservation parking guidance method based on smart shopping malls, the parking time model is built based on the correspondence between the parking difficulty value, current vehicle information and parking time as input data. If the parking difficulty value and current vehicle information are input, the predicted parking time is output.

[0032] As a preferred technical solution for a smart shopping mall-based advance reservation parking guidance method, the method of predicting the parking time corresponding to each parking space currently occupied in the parking lot in response to a car entering the entrance includes:

[0033] Acquire images of the parking lot, identify each parking space in the image where a parking operation is underway, and obtain its parking difficulty value;

[0034] Collect the parking operation characteristics of the driver who is currently parking and assess their proficiency.

[0035] Select the corresponding parking time model based on the proficiency assessment results;

[0036] Input the parking difficulty value and current vehicle information into the corresponding parking time model, and output the predicted parking time.

[0037] As a preferred technical solution for a smart shopping mall-based advance reservation parking guidance method, the step of correcting the parking path based on the predicted parking time and guiding vehicles according to the corrected parking path includes:

[0038] Calculate the total time to reach the target parking space, select the path with the shortest total time as the parking path, and guide the vehicle accordingly;

[0039] The total time to reach the target parking space is the sum of the driving time in the passage and the waiting time at each parking space.

[0040] This invention also provides a smart shopping mall-based advance reservation parking guidance system, comprising:

[0041] The reservation module, in response to the existence of a parking reservation, retrieves the current vehicle information;

[0042] The parking space distribution module is used to filter the remaining parking spaces based on the current vehicle information and obtain a distribution map of the candidate parking spaces;

[0043] The difficulty value calculation module is used to calculate the parking difficulty value of each of the candidate parking spaces based on the distribution map and the current vehicle information.

[0044] The route planning module selects the candidate parking space with the lowest parking difficulty value as the target parking space and plans the parking route.

[0045] The data acquisition module is used to collect drivers' operating habits, parking difficulty values, and parking time during the parking process. It classifies drivers' proficiency levels according to their operating habits and stores them in novice datasets and proficiency datasets respectively.

[0046] The model building module is used to train parking time models based on the novice dataset and the proficient dataset, respectively.

[0047] The result prediction module responds to the presence of a car entering the parking lot by obtaining the predicted parking time for each parking space currently occupied.

[0048] The route correction module is used to correct the parking route and provide parking guidance based on the predicted parking time.

[0049] Compared with the prior art, the beneficial effects of the present invention are as follows: the present invention obtains vehicle information based on parking reservation and accurately filters candidate parking spaces, calculates parking difficulty values ​​by combining kinematic models to achieve reasonable allocation of parking spaces, divides novice / experienced driver datasets by collecting driver operating habits and trains corresponding parking time models, and finally dynamically corrects the path based on real-time predicted parking time, thereby ensuring more accurate parking space matching and more reliable parking time prediction, effectively reducing vehicle congestion caused by mismatched parking spaces or excessive waiting time for the vehicle in front, thereby reducing waiting time in the parking lot and further reducing the probability of congestion. Attached Figure Description

[0050] Figure 1 This is a flowchart illustrating the steps of the advance reservation parking guidance method for smart shopping malls according to an embodiment of the present invention.

[0051] Figure 2 This is a structural block diagram of a smart shopping mall-based advance reservation parking guidance system according to an embodiment of the present invention. Detailed Implementation

[0052] The features and exemplary embodiments of various aspects of this application will be described in detail below. To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain this application and not to limit it. For those skilled in the art, this application can be implemented without some of these specific details. The following description of the embodiments is merely to provide a better understanding of this application by illustrating examples.

[0053] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes said element.

[0054] Please see Figure 1 As shown, it is a flowchart of the advance reservation parking guidance method for smart shopping malls according to an embodiment of the present invention, including:

[0055] Step S1: In response to the existence of a parking reservation, obtain the current vehicle information;

[0056] Step S2: Filter the remaining parking spaces based on the current vehicle information, obtain image information of the parking lot, and identify the image information of the parking lot to obtain the distribution map of the candidate parking spaces.

[0057] Step S3: Calculate the parking difficulty value of each candidate parking space based on the distribution map and current vehicle information;

[0058] Step S4: Select the candidate parking space with the lowest parking difficulty value as the target parking space and plan the parking route;

[0059] Step S5: Collect the driver's operating habits, parking difficulty value and parking time during the parking process, classify the proficiency level according to the driver's operating habits, and store them into novice dataset and proficiency dataset respectively.

[0060] Step S6: Train the parking time model based on the novice dataset and the proficient dataset respectively;

[0061] Step S7: In response to a car entering the entrance, obtain the predicted parking time for each parking space currently occupied in the parking lot.

[0062] Step S8: Based on the predicted parking time, correct the parking route and provide parking guidance.

[0063] In detail, this invention accurately filters parking spaces based on vehicle information obtained through parking reservations, calculates parking difficulty values ​​using kinematic models, and then rationally allocates parking spaces and vehicles according to the parking difficulty values. By collecting driver operating habits to divide the data into novice / experienced datasets and training corresponding parking time models, the invention finally dynamically corrects the path based on real-time predicted parking time, thereby reducing vehicle congestion caused by mismatched parking spaces and excessive waiting time for the vehicle in front, thus reducing waiting time in parking lots and further reducing the probability of congestion.

[0064] Furthermore, the current vehicle information includes: vehicle model, vehicle length, vehicle width, maximum steering angle, and whether charging is required.

[0065] In practice, users can initiate parking reservations through the mall's app / mini-program, triggering vehicle information collection, which supports two acquisition methods:

[0066] Active input: Users manually fill in information such as vehicle model, length, width, maximum steering angle, and charging requirements;

[0067] Automatic recognition: Scan license plates through parking lot entrance cameras and link them to vehicle parameter databases (such as identifying vehicle models by license plates and calling vehicle registration data from the vehicle management office), or read real-time parameters through the vehicle's OBD interface or Bluetooth to automatically identify vehicle model, length, width, maximum steering angle, and charging requirements.

[0068] Furthermore, this invention obtains full-dimensional vehicle information (vehicle type, size, steering angle, special needs, etc.) by combining active input and automatic recognition, ensuring that the vehicle characteristics can be accurately matched when selecting parking spaces (such as matching new energy vehicles with charging piles and large vehicles avoiding narrow parking spaces), avoiding parking failures or secondary adjustments due to missing information, reducing the ineffective dwell time of vehicles at parking spaces, thereby reducing waiting time in parking lots and further reducing the probability of congestion.

[0069] Specifically, the parking difficulty value for each candidate parking space is calculated based on the distribution map and current vehicle information, including:

[0070] Acquire images of each candidate parking space and identify obstacles in the parking lot;

[0071] The actual width of the passageway between parking spaces for vehicles is obtained based on the distribution map and obstacles in the parking lot.

[0072] Based on the vehicle model, length, width, actual channel width, and kinematic model, the current vehicle's parking trajectory is simulated.

[0073] Select the trajectories with the largest and smallest radii, calculate their corresponding steering angles, and use them as the maximum and minimum values ​​to form the steering angle intervals;

[0074] The percentage of the steering angle range in which the steering angle is greater than the preset steering angle is calculated and recorded as the parking difficulty value.

[0075] In practice, the parking space screening and distribution map obtains the list of remaining parking spaces by connecting with the parking lot's geomagnetic sensors and camera real-time data; based on vehicle information (such as filtering micro parking spaces with a width of <2.2m when the width of an SUV is >1.9m, and prioritizing charging pile parking spaces for new energy vehicles), suitable parking spaces are screened out, and a distribution map of candidate parking spaces is generated through 3D modeling (marking the location of the parking space, the width of the passage, and the status of adjacent parking spaces).

[0076] Parking difficulty calculation: It is understood that the trajectory formed by a vehicle during reversing will correspond to a unique turning angle. The turning angle can be obtained based on the angle between the trajectory tangent and the vehicle direction, or through kinematic models, etc. Simulate the parking trajectory of each candidate parking space and calculate the proportion of the trajectory with a turning angle > 30° (i.e., the preset angle in this invention. The selection of the preset angle value is determined according to the actual situation. For example, if the simulation trajectory finds that the distance between the current vehicle and the surrounding environment is less than 10cm, there may be a risk of scratching during the actual parking process. Therefore, the turning angle corresponding to this simulation trajectory is recorded as the preset angle value, or other methods that conform to reality can be used). This proportion is used as the parking difficulty value (e.g., if the turning angle range of a certain parking space is 10° to 35°, and the preset turning angle is 30°, then the parking difficulty value is 5 / 30 = 16.7%).

[0077] Understandably, during the process of reversing into a parking space, the larger the angle required for the vehicle to turn, the smaller the available space during reversing, and the more difficult it is to park. For example, in extreme cases, the vehicle can only park in the garage from the furthest position with the maximum turning angle, leaving only two parking methods. The error tolerance during parking is extremely small, making parking difficult. Calculating the parking difficulty value can provide a data-driven representation of the difficulty of parking in a garage. On the one hand, it provides a foundation for the establishment of subsequent parking time models, and on the other hand, it makes it possible to dynamically adjust the path based on predicted parking time, reduce waiting time in the parking lot, and thus further reduce the probability of congestion.

[0078] In detail, the system collects data on drivers' operating habits, parking difficulty levels, and parking time during the parking process. Based on these habits, drivers are categorized into proficiency levels, and the data is stored in separate datasets for beginners and experienced drivers.

[0079] For each parking space, data on parking time, parking habits, and parking difficulty are collected from the moment a vehicle starts reversing until it enters the garage, and stored in the form of data groups.

[0080] The data sets are cleaned by filtering out data sets whose parking time is less than the first-level threshold or greater than the second-level threshold, thus obtaining the standard data sets.

[0081] Based on operating habits, the standard data groups are divided into the proficient standard data group and the novice standard data group;

[0082] Store the proficient standard data set into the proficient dataset, and the novice standard data set into the novice dataset.

[0083] In implementation, route planning aims for "lowest parking difficulty + closest to the entrance," planning the optimal route from the garage entrance to the target parking space and simultaneously generating route navigation instructions (e.g., "Go straight for 50 meters, turn left, the target parking space is the third one from the right"). Data collection is achieved through cameras near the parking spaces, steering wheel sensors, and onboard timers. The purpose of primary and secondary thresholds is to filter data and remove problematic data sets. For example, data sets generated in cases where parking time is short due to violations or excessively long due to novice drivers practicing reversing into parking spaces are not representative and should be filtered out. The values ​​of primary and secondary thresholds should be selected based on the actual situation. Preferably, the primary threshold is the average parking time of drivers with more than 15 years of driving experience, and the secondary threshold is the average parking time of drivers with less than one year of driving experience.

[0084] Furthermore, this invention collects data in the form of data sets, cleans the data sets to obtain representative standard data sets, and then divides the standard data sets according to operating habits. This provides a basis for training parking time models based on the expert standard data sets and the novice standard data sets respectively. This allows for the prediction of parking time during the driver's reversing process, and the correction of the parking path based on the prediction results, thereby reducing the waiting time in the parking lot and further reducing the probability of congestion.

[0085] Furthermore, the operating practice can be any of the following:

[0086] The total number of steering wheel turns required to complete parking is recorded by a steering wheel angle sensor.

[0087] The maximum steering angle deviation during parking, which is the difference between the actual steering angle and the optimal trajectory steering angle;

[0088] The time interval from starting to reverse to the first adjustment of the steering wheel, that is, the time from starting to reverse to the first turn of the steering wheel.

[0089] In detail, based on operating habits, the standard data groups are divided into proficient standard data groups and novice standard data groups, including:

[0090] If the total number of steering wheel rotations exceeds a preset threshold, or the maximum steering angle deviation exceeds a preset threshold, or the interval time is less than a preset time threshold, the standard data group is divided into the novice standard data group; otherwise, it is divided into the proficient standard data group.

[0091] The parking time model was trained using both novice and expert standard datasets.

[0092] In implementation, the preset lap count threshold, preset angle threshold, and preset time threshold are set to classify the driver's proficiency level; the values ​​should be reasonable and reflect the actual situation. Preferably, in this embodiment of the invention, the preset lap count threshold, preset angle threshold, and preset time threshold are determined by collecting the average value of drivers with more than 15 years of driving experience, while the secondary threshold is selected from the average value of drivers with less than one year of driving experience.

[0093] Specifically, by classifying drivers by their proficiency level, and training corresponding relational models based on proficiency levels, the impact of driver proficiency on parking time can be reduced. This allows the two types of models to specifically fit the parking patterns of the corresponding groups (novice drivers face high difficulty and long waiting times, while proficient drivers face low difficulty and short waiting times), improving the accuracy of parking time prediction, reducing scheduling errors and vehicle waiting times caused by ambiguity in the classification, thereby reducing waiting time in parking lots and further reducing the probability of congestion.

[0094] Furthermore, based on the parking time model, it is built using the correspondence between parking difficulty value, current vehicle information, and parking time as input data. If the parking difficulty value and current vehicle information are input, the predicted parking time is output.

[0095] In practice, the parking time model is trained using the random forest regression algorithm, with novice / experienced datasets as samples. The input features are parking difficulty value and vehicle information (vehicle length, steering angle), and the output is parking time. The model is optimized through cross-validation (e.g., the novice model focuses on fitting the association between "high-difficulty parking spaces - long time consumption", while the experienced model focuses on the pattern of "low-difficulty parking spaces - short time consumption").

[0096] Furthermore, in response to a vehicle entering the parking lot, the parking time for each parking space currently occupied is predicted, including:

[0097] Acquire images of the parking lot, identify each parking space in the image where a parking operation is underway, and obtain its parking difficulty value;

[0098] Collect the parking operation characteristics of the driver who is currently parking and assess their proficiency.

[0099] Select the corresponding parking time model based on the proficiency assessment results;

[0100] Input the parking difficulty value and current vehicle information into the corresponding parking time model, and output the predicted parking time.

[0101] Furthermore, the system collects driver operation characteristics in real time to assess proficiency, calls the corresponding parking time model to predict the time spent in the parking space, and dynamically updates the prediction results to match the actual parking progress. This allows the system to detect potential congestion points in the lane in advance (such as a parking space with a predicted time that is too long), providing timely reference for subsequent vehicle route planning, reducing lane congestion caused by prediction lag, thereby reducing waiting time in the parking lot and further reducing the probability of congestion.

[0102] Furthermore, based on the predicted parking time, the parking route is corrected, and vehicles are guided according to the corrected parking route, including:

[0103] Calculate the total time to reach the target parking space, select the path with the shortest total time as the parking path, and guide the vehicle accordingly;

[0104] The total time to reach the target parking space is the sum of the driving time in the passageway and the waiting time at each parking space.

[0105] Furthermore, by combining the travel time in the integrated lane with the predicted parking time in the parking space to calculate the total time, the system prioritizes and dynamically adjusts the route with the shortest total time. This can proactively avoid lane congestion caused by excessive parking time in a certain area, reduce unnecessary waiting and detours in congested sections, improve the turnover efficiency of traffic flow in the parking lot, thereby reducing waiting time in the parking lot and further reducing the probability of congestion.

[0106] Please see Figure 2 As shown, it is a structural block diagram of the advanced reservation parking guidance system for smart shopping malls based on an embodiment of the present invention, including:

[0107] The reservation module, in response to the existence of a parking reservation, retrieves the current vehicle information;

[0108] The parking space distribution module is used to filter the remaining parking spaces based on the current vehicle information and obtain a distribution map of the parking spaces to be selected.

[0109] The difficulty value calculation module is used to calculate the parking difficulty value of each candidate parking space based on the distribution map and current vehicle information.

[0110] The route planning module selects the candidate parking space with the lowest parking difficulty value as the target parking space and plans the parking route.

[0111] The data acquisition module is used to collect drivers' operating habits, parking difficulty values, and parking time during the parking process. It classifies drivers' proficiency levels according to their operating habits and stores them in novice datasets and proficiency datasets respectively.

[0112] The model building module is used to train parking time models based on novice and proficient datasets respectively.

[0113] The result prediction module responds to the presence of a car entering the parking lot by obtaining the predicted parking time for each parking space currently occupied.

[0114] The route correction module is used to correct the parking route and provide parking guidance based on the predicted parking time.

[0115] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the implementation of the present invention. Those skilled in the art can make other variations or modifications based on the above description. It is neither necessary nor possible to exhaustively describe all embodiments here. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the claims of the present invention.

Claims

1. A method for guiding parking in advance reservation garages based on smart shopping malls, characterized in that: include: In response to a parking reservation, obtain the current vehicle information; Based on the current vehicle information, the remaining parking spaces are filtered, and image information of the parking lot is obtained. The image information of the parking lot is then identified to obtain a distribution map of the candidate parking spaces. The parking difficulty value of each of the candidate parking spaces is calculated based on the distribution map and the current vehicle information; Select the candidate parking space with the lowest parking difficulty value as the target parking space and plan the parking route; The system collects drivers' operating habits, parking difficulty values, and parking time during the parking process. It categorizes drivers' proficiency levels based on their operating habits and stores them in separate datasets for beginners and proficient drivers. Train parking time models based on the novice dataset and the proficient dataset respectively; In response to a car entering the parking lot, obtain the predicted parking time for each parking space currently occupied. Based on the predicted parking time, the parking route is corrected and parking guidance is provided; The step of calculating the parking difficulty value of each of the candidate parking spaces based on the distribution map and the current vehicle information includes: Acquire images of each candidate parking space and identify obstacles in the parking lot; The actual width of the passageway between parking spaces for vehicles is obtained based on the distribution map and obstacles in the parking lot. Based on the vehicle model, length, width, actual channel width, and kinematic model, the current vehicle's parking trajectory is simulated. Select the trajectories with the largest and smallest radii, calculate their corresponding steering angles, and use them as the maximum and minimum values ​​to form the steering angle ranges; The percentage of the steering angle range in the steering angle range that is greater than the preset steering angle is calculated and recorded as the parking difficulty value.

2. The method for guiding parking in a smart shopping mall-based pre-booking garage according to claim 1, characterized in that, The current vehicle information includes: vehicle model, vehicle length, vehicle width, maximum steering angle, and whether charging is required.

3. The method for guiding parking in a smart shopping mall-based pre-booking garage according to claim 1, characterized in that, The process of collecting data on drivers' operating habits, parking difficulty, and parking time during parking is used to categorize drivers' proficiency levels based on their operating habits, and these levels are stored in separate datasets for beginners and experienced drivers. For each parking space, data on parking time, parking habits, and parking difficulty are collected from the moment a vehicle starts reversing until it enters the garage, and stored in the form of data groups. The data set is cleaned to remove data sets with parking times less than the first-level threshold or greater than the second-level threshold, thus obtaining a standard data set. Based on the aforementioned operating habits, each of the standard data groups is divided into a proficient standard data group and a novice standard data group; Store the proficient standard data set into the proficient dataset, and the novice standard data set into the novice dataset.

4. The method for guiding parking in a smart shopping mall-based pre-booking garage according to claim 3, characterized in that, The operating habits are one of the following: the total number of steering wheel turns required to complete parking, the maximum steering angle deviation during parking, or the interval from starting to reverse to the first adjustment of the steering wheel.

5. The method for guiding parking in a smart shopping mall-based pre-booking garage according to claim 4, characterized in that, The step of dividing the standard data groups into skilled standard data groups and novice standard data groups according to the operating habits includes: In response to the total number of steering wheel rotations exceeding a preset threshold, or the maximum steering angle deviation exceeding a preset angle threshold, or the interval time being less than a preset time threshold, the standard data group is divided into the novice standard data group; otherwise, it is divided into the proficient standard data group. The parking time model is trained using the novice standard data set and the expert standard data set, respectively.

6. The method for guiding parking in a smart shopping mall-based pre-booking garage according to claim 5, characterized in that, The parking time model is built based on the correspondence between the parking difficulty value, current vehicle information, and parking time as input data. If the parking difficulty value and current vehicle information are input, the predicted parking time is output.

7. The method for guiding parking in a smart shopping mall-based pre-booking garage according to claim 5, characterized in that, The response to a vehicle entering the parking lot includes predicting the parking time for each parking space currently occupied, including: Acquire images of the parking lot, identify each parking space in the image where a parking operation is underway, and obtain its parking difficulty value; Collect the parking operation characteristics of the driver who is currently parking and assess their proficiency. Select the corresponding parking time model based on the proficiency assessment results; Input the parking difficulty value and current vehicle information into the corresponding parking time model, and output the predicted parking time.

8. The method for guiding parking in a smart shopping mall-based pre-booking garage according to claim 1, characterized in that, The step of correcting the parking path based on the predicted parking time and guiding vehicles according to the corrected parking path includes: Calculate the total time to reach the target parking space, select the path with the shortest total time as the parking path, and guide the vehicle accordingly; The total time to reach the target parking space is the sum of the driving time in the passage and the waiting time at each parking space.

9. A smart shopping mall-based advance reservation parking guidance system, used to implement the smart shopping mall-based advance reservation parking guidance method according to any one of claims 1 to 8, characterized in that, include: The reservation module, in response to the existence of a parking reservation, retrieves the current vehicle information; The parking space distribution module is used to filter the remaining parking spaces based on the current vehicle information and obtain a distribution map of the candidate parking spaces; The difficulty value calculation module is used to calculate the parking difficulty value of each of the candidate parking spaces based on the distribution map and the current vehicle information. The route planning module selects the candidate parking space with the lowest parking difficulty value as the target parking space and plans the parking route. The data acquisition module is used to collect drivers' operating habits, parking difficulty values, and parking time during the parking process. It classifies drivers' proficiency levels according to their operating habits and stores them in novice datasets and proficiency datasets respectively. The model building module is used to train parking time models based on the novice dataset and the proficient dataset, respectively. The result prediction module responds to the presence of a car entering the parking lot by obtaining the predicted parking time for each parking space currently occupied. The route correction module is used to correct the parking route and provide parking guidance based on the predicted parking time.

Images