A parking space state detection method and system

Abstract

The present application relates to the technical field of parking space state detection, and provides a parking space state detection method and system, the method comprising: collecting background magnetic field data by using a magnetic field sensor preset on the ground; collecting first original magnetic field data by using the magnetic field sensor, and generating a reference vehicle magnetic field profile of an authorized vehicle based on the background magnetic field data; collecting second original magnetic field data by using the magnetic field sensor, and generating a to-be-detected vehicle magnetic field profile of a to-be-detected vehicle based on the background magnetic field data; comparing the to-be-detected vehicle magnetic field profile with the reference vehicle magnetic field profile, and determining the identity of the to-be-detected vehicle according to the comparison result, so as to detect the occupancy state of the parking space corresponding to the bearing platform. The present application has the effect of overcoming the complex background magnetic field environment in the mechanical lifting type parking equipment, and realizing accurate detection of the parking space occupancy state and preliminary identification of the vehicle identity.

Description

Technical Field

[0001] This invention relates to the technical field of parking space status detection, and specifically to a parking space status detection method and system. Background Technology

[0002] In modern urban parking management, geomagnetic detectors are widely used to automatically detect whether parking spaces are occupied due to their advantages such as concealed installation and immunity to weather conditions. With the promotion of the sharing economy, many private parking spaces in commercial buildings or residential areas are opened as shared parking spaces during off-peak hours. This requires parking space status detection systems to not only be accurate but also support automated billing and management. However, when this technology is applied to special parking facilities designed to maximize space utilization, such as double-layer mechanical lift parking systems, traditional geomagnetic detection methods face significant challenges.

[0003] Specifically, in scenarios employing double-layer mechanical lift parking systems, geomagnetic detection systems face multiple technical challenges. When the lift is in operation, its drive motor and the large, moving metal platform collectively generate strong, dynamically changing magnetic field interference. The intensity of this interference's magnetic field variations far exceeds the weak signals generated by stationary vehicles on the upper level, easily leading to erroneous system judgments, such as misinterpreting the lift's operation as vehicle entry or exit, resulting in billing or management errors. Furthermore, when an authorized vehicle is replaced by another vehicle, traditional systems cannot recognize this change in vehicle identity, potentially leading to inaccurate billing or service abuse.

[0004] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0005] The purpose of this application is to provide a parking space status detection method and system, which can determine the vehicle identity and parking space occupancy status by collecting and processing background magnetic field data and comparing vehicle magnetic field profiles. This effectively overcomes the problems of complex background magnetic field environment, equipment operation interference and weak vehicle signals in mechanical lifting parking equipment, and achieves the advantages of accurate detection of parking space occupancy status and preliminary identification of vehicle identity.

[0006] This application provides a parking space status detection method, the technical solution of which is as follows:

[0007] When the platform of the mechanical lifting parking equipment is unloaded, the background magnetic field data corresponding to the lifting movement of the platform is collected by a magnetic field sensor pre-installed on the ground.

[0008] When the platform needs to carry the authorized vehicle and perform lifting movements, it uses a magnetic field sensor to collect the first raw magnetic field data, and processes the first raw magnetic field data based on the background magnetic field data to generate a reference vehicle magnetic field profile of the authorized vehicle.

[0009] When the carrying platform needs to carry the vehicle to be inspected and perform lifting and lowering movements, a second original magnetic field data is collected using a magnetic field sensor, and the second original magnetic field data is processed based on the background magnetic field data to generate the magnetic field profile of the vehicle to be inspected.

[0010] The magnetic field profile of the vehicle to be inspected is compared with that of a reference vehicle. Based on the comparison results, the identity of the vehicle to be inspected is determined, so as to detect the occupancy status of the parking space corresponding to the carrying platform.

[0011] The above solution solves the problems of complex background magnetic field environment, weak vehicle magnetic field signal and dynamic magnetic field interference generated by equipment operation in mechanical lifting parking equipment, which lead to misjudgment by traditional geomagnetic detection methods. It realizes accurate detection of parking space occupancy status and can identify vehicle identity.

[0012] Furthermore, this application also proposes a step for processing the second original magnetic field data based on the background magnetic field data to generate a magnetic field profile of the vehicle under inspection, including:

[0013] The magnetic field characteristics of the carrying platform in the initial stage of lifting and lowering motion are extracted from the background magnetic field data and used as the benchmark start-up characteristics.

[0014] From the second original magnetic field data, the magnetic field characteristics of the carrying platform in the initial stage of the lifting and lowering motion are extracted as the current start-up characteristics;

[0015] The correction value is determined based on the baseline startup characteristics and the current startup characteristics;

[0016] The background magnetic field data is adjusted based on the correction value to generate dynamic background magnetic field data.

[0017] Dynamic background magnetic field data is removed from the second original magnetic field data to generate a magnetic field profile of the vehicle to be inspected.

[0018] The above scheme effectively eliminates dynamic magnetic field interference caused by the operation of lifting equipment by generating dynamic background magnetic field data, thus improving the accuracy of magnetic field profile generation.

[0019] Furthermore, this application also proposes a step of comparing the magnetic field profile of the vehicle to be inspected with that of a reference vehicle, and determining the identity of the vehicle to be inspected based on the comparison results, so as to detect the occupancy status of the parking space corresponding to the carrying platform, including:

[0020] Calculate the similarity between the magnetic field profile of the vehicle under test and the magnetic field profile of the reference vehicle, and generate preliminary comparison results;

[0021] When the initial comparison result is lower than the preset identity confirmation threshold, a differential profile is generated based on the difference between the magnetic field profile of the vehicle under test and the magnetic field profile of the reference vehicle.

[0022] Analyze the data distribution characteristics of the differential profile, and determine whether the differential profile exhibits a locally concentrated or globally distributed change pattern based on the data distribution characteristics, thus obtaining the change pattern determination result.

[0023] If the change pattern determination result is a locally concentrated change pattern, then the vehicle to be inspected is identified as an authorized vehicle with structural changes, and the occupancy status of the parking space corresponding to the carrying platform is detected as an authorized occupancy status; if the change pattern determination result is a globally distributed change pattern, then the vehicle to be inspected is identified as an unauthorized vehicle, and the occupancy status of the parking space corresponding to the carrying platform is detected as a temporary occupancy status.

[0024] The above scheme can further distinguish vehicle identities based on the distribution characteristics of magnetic field profile differences, enabling the identification of authorized vehicle structural changes and unauthorized vehicles, thus meeting the needs of refined management.

[0025] Furthermore, this application also proposes a step for generating a differential profile based on the difference between the magnetic field profile of the vehicle under test and the magnetic field profile of a reference vehicle, including:

[0026] A time offset is applied to the magnetic field profile of the vehicle to be inspected to generate a set of candidate magnetic field profiles;

[0027] Calculate the matching degree between each candidate magnetic field profile and the reference vehicle magnetic field profile in a set of candidate magnetic field profiles;

[0028] Based on the matching degree, the candidate magnetic field profile with the highest matching degree with the reference vehicle magnetic field profile is selected as the aligned magnetic field profile of the vehicle to be inspected.

[0029] A differential profile is generated based on the aligned magnetic field profile of the vehicle under inspection and the magnetic field profile of the reference vehicle.

[0030] The above scheme, through time offset and matching degree calculation, achieves precise alignment between the magnetic field profile of the vehicle under inspection and the magnetic field profile of the reference vehicle, ensuring the accuracy of differential profile generation.

[0031] Furthermore, this application also proposes a step for analyzing the data distribution characteristics of the differential profile, determining whether the differential profile exhibits a locally concentrated or globally distributed change pattern based on the data distribution characteristics, and obtaining the change pattern determination result, including:

[0032] The differential profile is divided into several data segments;

[0033] Calculate the amount of data change for each data segment;

[0034] Identify data segments whose changes exceed a preset segment change threshold and designate them as target segments;

[0035] The number of target segments is counted, and it is determined whether the target segments are in a separated state on the differential profile, thus obtaining the separation state determination result;

[0036] If the number of target segments is less than the preset number threshold and the separation state judgment result is yes, then the differential profile is determined to present a locally concentrated change pattern; if the number of target segments is greater than or equal to the preset number threshold or the separation state judgment result is no, then the differential profile is determined to present a globally distributed change pattern.

[0037] By employing the above approach and conducting quantitative analysis of the distribution characteristics of differential profile data, it is possible to accurately determine whether the change pattern is locally concentrated or globally distributed, providing a reliable basis for the refined identification of vehicle identities.

[0038] Furthermore, this application also proposes a step for determining whether the target segment is in a separated state on the differential profile, and obtaining the separation state determination result, including:

[0039] Identify adjacent target segments from all data segments to form merged segments;

[0040] Analyze the internal data of the fusion section to determine whether the number of data peaks in the fusion section is greater than one, and obtain the data peak judgment result;

[0041] If the data peak value judgment result is yes, then obtain the valley value between adjacent data peak values ​​within the fusion segment, and determine whether the valley value is higher than the preset segment change threshold to obtain the valley value judgment result;

[0042] If the valley value judgment result is yes, then the separation state judgment result is determined to be that the target segment is in a separation state;

[0043] If the data peak value judgment result is negative or the valley value judgment result is negative, then the separation status judgment result will be determined as the target segment being in a non-separated state.

[0044] The above scheme further refines the judgment logic of local concentrated change patterns, and improves the recognition accuracy of local structural changes by analyzing the data peaks and valleys of the fusion section.

[0045] Furthermore, this application also proposes a step for calculating the matching degree between each candidate magnetic field profile in a set of candidate magnetic field profiles and the reference vehicle magnetic field profile, including:

[0046] The benchmark vehicle magnetic field profile and the candidate magnetic field profile are synchronously divided into several profile segments according to the time axis.

[0047] Calculate the segment similarity for each set of profile segments;

[0048] Aggregate the similarity of all segments to generate a matching score.

[0049] The above scheme improves the accuracy and robustness of magnetic field profile matching by calculating and aggregating similarity in segments.

[0050] Furthermore, this application also proposes a step for determining the correction value based on the baseline startup characteristics and the current startup characteristics, including:

[0051] The baseline startup characteristics and the current startup characteristics are synchronously divided into several groups of characteristic segments;

[0052] Calculate the segment deviation for each group of characteristic segments;

[0053] Identify segment deviations that deviate numerically from the corresponding segment deviation center trend;

[0054] From all the segment deviations, exclude the segment deviations that deviate numerically from the corresponding segment deviation central trend to obtain the remaining segment deviations;

[0055] Based on the remaining segment deviation, determine the correction value.

[0056] By eliminating deviations in abnormal sections, the accuracy of determining correction values ​​is improved, thereby allowing for more precise adjustment of the background magnetic field data.

[0057] Furthermore, this application proposes a step for aggregating the similarity of all segments to generate a matching score, including:

[0058] Based on the benchmark vehicle magnetic field profile, the feature saliency of the profile segment is determined, and the weight corresponding to each profile segment is generated.

[0059] The similarity of segments is weighted and aggregated to generate a matching score.

[0060] By introducing feature saliency as a weight, the above scheme makes the matching degree calculation more prominent in terms of the similarity of key feature regions, thus improving the accuracy of matching.

[0061] Furthermore, this application also proposes a parking space status detection system for performing parking space status detection, including:

[0062] The background magnetic field acquisition module is used to collect background magnetic field data corresponding to the lifting and lowering movement of the mechanical lifting parking equipment when the carrying platform is unloaded, using a magnetic field sensor pre-installed on the ground.

[0063] The reference profile generation module is used to collect first raw magnetic field data using a magnetic field sensor when the carrying platform needs to carry an authorized vehicle and perform lifting movements, and to process the first raw magnetic field data based on the background magnetic field data to generate a reference vehicle magnetic field profile of the authorized vehicle.

[0064] The inspection data generation module is used to collect second original magnetic field data using a magnetic field sensor when the carrying platform needs to carry the vehicle to be inspected and perform lifting movements, and to process the second original magnetic field data based on the background magnetic field data to generate the vehicle magnetic field profile of the vehicle to be inspected.

[0065] The identity status detection module is used to compare the magnetic field profile of the vehicle to be inspected with the magnetic field profile of a reference vehicle, and determine the identity of the vehicle to be inspected based on the comparison results, so as to detect the occupancy status of the parking space corresponding to the carrying platform.

[0066] The above scheme provides a system for implementing the parking space status detection method, which has the advantages of modularity and clear functions, and is easy to deploy and apply in practice.

[0067] As can be seen from the above, the parking space status detection method and system provided in this application, by collecting and processing background magnetic field data and determining the vehicle identity and parking space occupancy status based on vehicle magnetic field profile comparison, effectively overcomes the problems of complex background magnetic field environment, equipment operation interference and weak vehicle signal in mechanical lifting parking equipment, and realizes accurate detection of parking space occupancy status and preliminary identification of vehicle identity, which has significant practical value and technological progress. Attached Figure Description

[0068] Figure 1 This is a flowchart of a parking space status detection method according to one embodiment of the present invention;

[0069] Figure 2 This is one of the flowcharts of a parking space status detection method according to another embodiment of the present invention;

[0070] Figure 3 This is a second flowchart of a parking space status detection method according to another embodiment of the present invention;

[0071] Figure 4 This is a third flowchart of a parking space status detection method according to another embodiment of the present invention;

[0072] Figure 5 This is the fourth flowchart of a parking space status detection method according to another embodiment of the present invention;

[0073] Figure 6 This is the fifth flowchart of a parking space status detection method according to another embodiment of the present invention;

[0074] Figure 7 This is a flowchart of a parking space status detection method according to another embodiment of the present invention, number six.

[0075] Figure 8 This is the seventh flowchart of a parking space status detection method according to another embodiment of the present invention;

[0076] Figure 9 This is the eighth flowchart of a parking space status detection method according to another embodiment of the present invention;

[0077] Figure 10 This is a system block diagram of a parking space status detection system according to another embodiment of the present invention;

[0078] Explanation of reference numerals in the attached figures:

[0079] 1. Parking space status detection system; 11. Background magnetic field acquisition module; 12. Reference profile generation module; 13. Data to be inspected generation module; 14. Identity status detection module. Detailed Implementation

[0080] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0081] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, the terms "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0082] Traditional geomagnetic detection systems face multiple technical challenges when applied to mechanical lift parking systems. When the geomagnetic detector is positioned on the ground and the vehicle is parked on the upper platform, the vertical distance between the vehicle and the detector increases significantly, causing a sharp attenuation of the magnetic field signal generated by the vehicle and a decrease in the signal-to-noise ratio. Simultaneously, the operation of the mechanical lift, especially the movement of high-power motors and large metal structures, generates strong, dynamically changing magnetic field interference. The intensity of this interference signal far exceeds the weak signal generated by the vehicle, making it difficult for the system to accurately distinguish between vehicle signals and environmental noise and equipment operational interference. Furthermore, existing systems primarily focus on determining whether a parking space is occupied, failing to identify the specific identity of the vehicle. This is insufficient to meet the needs of identifying unauthorized vehicle changes in scenarios requiring refined management, such as shared parking.

[0083] For example, suppose an underground parking garage with a double-layer mechanical lift system has a geomagnetic sensor installed at the center of the ground for each parking space. This garage is open to the public as a shared parking resource during the day. When a small car is lifted by the lift to the upper platform and parked, the vertical distance between the vehicle's chassis and the ground geomagnetic sensor may be more than two meters. In this situation, the magnetic field signal generated by the vehicle itself is significantly attenuated by the time it reaches the sensor. Simultaneously, when the lift system performs its lifting action, its high-powered motor starts and moves the large metal platform, generating instantaneous and strong magnetic field disturbances. The amplitude of these disturbance signals may be much higher than the weak magnetic field signal generated by the vehicle parked on the upper level. Furthermore, if a user registers a small car and obtains parking authorization, but then drives the small car away and returns in a larger SUV and parks in the same space, a traditional system that only determines vehicle presence based on magnetic field strength thresholds will not be able to recognize the actual change in vehicle presence and will still consider it the original authorized vehicle.

[0084] In response, this application proposes a parking space status detection method, combining... Figure 1 As shown, it includes:

[0085] S1, When the carrying platform of the mechanical lifting parking equipment is unloaded, the background magnetic field data corresponding to the lifting movement of the carrying platform is collected by a magnetic field sensor pre-installed on the ground.

[0086] S2, when the carrying platform needs to carry the authorized vehicle and perform lifting motion, the first original magnetic field data is collected by the magnetic field sensor, and the first original magnetic field data is processed based on the background magnetic field data to generate the reference vehicle magnetic field profile of the authorized vehicle.

[0087] S3, when the carrying platform needs to carry the vehicle to be inspected and perform lifting and lowering movements, the magnetic field sensor is used to collect the second original magnetic field data, and the second original magnetic field data is processed based on the background magnetic field data to generate the magnetic field profile of the vehicle to be inspected.

[0088] S4. The magnetic field profile of the vehicle to be inspected is compared with the magnetic field profile of the reference vehicle. Based on the comparison results, the identity of the vehicle to be inspected is determined so as to detect the occupancy status of the parking space corresponding to the carrying platform.

[0089] Among them, the carrying platform of the mechanical lifting parking equipment refers to the structure used to carry vehicles and perform vertical lifting movements. It can be implemented using a steel structure platform, a hydraulic drive platform, or a chain drive platform, with the purpose of enabling multi-level vehicle parking within a limited space; the magnetic field sensor pre-installed on the ground is a device that can sense and measure changes in magnetic field strength. It can be implemented using Hall effect sensors, magnetoresistive sensors, or fluxgate sensors, with the purpose of collecting magnetic field data of the parking space area; background magnetic field data refers to the magnetic field information collected by the magnetic field sensor during the lifting movement of the carrying platform when it is unloaded. Its purpose is to characterize the inherent influence of the parking equipment's own operation and environmental factors on the magnetic field, serving as a benchmark for subsequent data processing; the benchmark vehicle magnetic field profile refers to the magnetic field information collected by the magnetic field sensor during the lifting movement of the authorized vehicle after processing the background magnetic field data. The unique magnetic field feature sequence formed during the descent motion is used to establish the magnetic field identity model of authorized vehicles. The magnetic field profile of the vehicle under inspection refers to the magnetic field feature sequence formed by the vehicle under inspection during the descent motion after background magnetic field data processing. Its purpose is to obtain the magnetic field identity information of the vehicle under inspection for comparison. Comparison refers to the process of comparing and analyzing the magnetic field profile of the vehicle under inspection with the magnetic field profile of a reference vehicle. Its purpose is to assess the similarity or difference between the two. Determining the identity of the vehicle under inspection refers to judging whether the vehicle under inspection is an authorized vehicle or its specific type based on the comparison results. Its purpose is to achieve accurate vehicle identification. Detecting the occupancy status of the parking space corresponding to the carrying platform refers to judging whether the parking space is authorized, temporarily occupied, or vacant based on the vehicle identity recognition results. Its purpose is to achieve refined management of parking spaces.

[0090] This application's solution achieves accurate detection of parking space status and vehicle identification in complex mechanical lift parking environments through a series of collaborative steps. First, when the platform is unloaded, a magnetic field sensor pre-installed on the ground collects background magnetic field data corresponding to the platform's lifting movement. This step aims to establish a magnetic field benchmark purely generated by the parking equipment's operation and environmental factors, providing a reference for subsequent data processing and effectively eliminating strong magnetic field interference from equipment operation. Subsequently, when the platform needs to carry an authorized vehicle and perform lifting movements, the magnetic field sensor collects first raw magnetic field data. This raw data is then processed based on the previously collected background magnetic field data, such as through background removal or differential processing, to generate a benchmark vehicle magnetic field profile for the authorized vehicle. This processing effectively isolates interference from the equipment and environment, highlighting the authorized vehicle's own magnetic field characteristics and forming its unique magnetic field fingerprint. Similarly, when the platform needs to carry a vehicle to be inspected and perform lifting movements, the magnetic field sensor collects second raw magnetic field data, which is also processed based on the background magnetic field data to generate a vehicle magnetic field profile for the vehicle to be inspected. This step ensures that the extraction method of the magnetic field characteristics of the vehicle to be inspected is consistent with that of the authorized vehicle, guaranteeing the effectiveness of subsequent comparisons. Finally, the magnetic field profile of the vehicle to be inspected is compared with that of a reference vehicle. The comparison result is directly used to determine the identity of the vehicle to be inspected, such as whether it is an authorized vehicle, and thus detect the occupancy status of the parking space corresponding to the platform. The entire process overcomes the signal attenuation and strong interference problems faced by traditional geomagnetic detection in mechanical lifting equipment through dynamic acquisition and background removal, and achieves vehicle identification functionality beyond simple presence or absence judgment through magnetic field profile comparison.

[0091] In some preferred embodiments, the above method can be implemented as follows: The magnetic field sensor can be a triaxial magnetoresistive sensor configured to continuously acquire magnetic field strength data at a fixed sampling rate (e.g., 100 times per second) during the lifting and lowering of the support platform. Background magnetic field data can be pre-stored in a database and associated with specific parking spaces. When the raw magnetic field data is acquired, a data processing unit can perform a background removal operation, for example, by subtracting the raw magnetic field data from the corresponding background magnetic field data point by point or by removing the background data based on model fitting, thereby generating a vehicle magnetic field profile. These profile data can be represented as magnetic field strength curves that vary over time. In the comparison phase, a pattern recognition algorithm can calculate the similarity between the magnetic field profile of the vehicle to be inspected and the magnetic field profile of a reference vehicle, for example, by calculating the Pearson correlation coefficient or Euclidean distance between them. Based on the relationship between the calculated similarity and a preset threshold, the system can determine whether the vehicle to be inspected is an authorized vehicle, and thus determine the occupancy status of the parking space.

[0092] Optional, combined Figure 2As shown, the steps in S3 to process the second original magnetic field data based on the background magnetic field data to generate the magnetic field profile of the vehicle under inspection include:

[0093] S31. Extract the magnetic field characteristics of the carrying platform in the initial stage of lifting and lowering motion from the background magnetic field data, and use them as the reference start-up characteristics.

[0094] S32, Extract the magnetic field characteristics of the carrying platform in the initial stage of lifting and lowering motion from the second original magnetic field data, and use them as the current start-up characteristics;

[0095] S33, determine the correction value based on the baseline startup characteristics and the current startup characteristics;

[0096] S34, Adjust the background magnetic field data based on the correction value to generate dynamic background magnetic field data;

[0097] S35, remove dynamic background magnetic field data from the second original magnetic field data to generate the magnetic field profile of the vehicle to be inspected.

[0098] The reference start-up feature refers to the set of background magnetic field data collected by the platform during the initial lifting and lowering phase under no-load conditions. This data reflects a specific magnetic field change pattern at the moment of equipment startup or in the early stages of startup. It can be represented by time-series data such as magnetic field strength, magnetic field direction, or magnetic field change rate. The current start-up feature refers to the magnetic field change pattern data extracted from the second original magnetic field data during the initial lifting and lowering phase of the platform carrying the vehicle to be inspected. This data corresponds to the reference start-up feature in terms of time or event. It can use the same data type and representation as the reference start-up feature. The correction value is a numerical value or function calculated based on the difference or correlation between the reference start-up feature and the current start-up feature to adjust the background magnetic field data. It can be represented by a difference, ratio, scaling factor, or linear / nonlinear mapping relationship. The dynamic background magnetic field data refers to the set of data obtained by adjusting the pre-collected background magnetic field data in real-time or near real-time under the influence of the correction value. This data more accurately reflects the magnetic field interference during the operation of the mechanical lifting equipment. It can be generated by translating, scaling, or overlaying the original background magnetic field data. Among them, the magnetic field profile of the vehicle under inspection refers to the time-series data that mainly reflects the magnetic field characteristics of the vehicle under inspection after removing the dynamic background magnetic field data from the second original magnetic field data. It can be represented by a curve of magnetic field intensity changing with time, multidimensional magnetic field component data or its derived characteristics.

[0099] In some preferred embodiments, specifically, when processing the second original magnetic field data based on the background magnetic field data to generate the magnetic field profile of the vehicle under inspection, the following steps can be taken: First, from the background magnetic field data, identify and extract the magnetic field intensity change curve or its spectral characteristics of the carrying platform in the initial stage of the lifting motion, such as the first 5 seconds after startup, as the baseline startup characteristics. Simultaneously, from the second original magnetic field data, extract the magnetic field intensity change curve or its spectral characteristics of the carrying platform in the current initial stage of the lifting motion, also such as the first 5 seconds after startup, as the current startup characteristics. Next, based on the baseline startup characteristics and the current startup characteristics, a correction value can be determined by calculating the average difference or scaling factor between the two at a specific time point or within a specific frequency range. For example, the root mean square error of the two characteristic curves can be calculated, or a linear or nonlinear function can be fitted using the least squares method to represent the relationship between the two, and the parameters of this function can be used as correction values. Then, based on the determined correction values, adjust the background magnetic field data to generate dynamic background magnetic field data. If the correction value is an average difference, this difference can be added to or subtracted from each data point in the background magnetic field data; if the correction value is a scaling factor, this factor can be multiplied by each data point in the background magnetic field data. In this way, the background magnetic field data can be dynamically adjusted according to the current actual magnetic field environment. Finally, the dynamic background magnetic field data is removed from the second original magnetic field data to generate the magnetic field profile of the vehicle under inspection. This can be achieved by subtracting the dynamic background magnetic field data at the corresponding time point from each data point in the second original magnetic field data. For example, if the magnetic field sensor samples at a frequency of 100Hz, then at each sampling moment, the currently acquired original magnetic field data is subtracted from the dynamic background magnetic field data at the corresponding time, thus obtaining the magnetic field profile of the vehicle under inspection after removing interference. In this way, magnetic field interference caused by equipment operation can be effectively eliminated, allowing the vehicle's own magnetic field characteristics to be highlighted.

[0100] Optional, combined Figure 3 As shown, S4 compares the magnetic field profile of the vehicle to be inspected with the magnetic field profile of a reference vehicle, and determines the identity of the vehicle to be inspected based on the comparison result, in order to detect the occupancy status of the parking space corresponding to the carrying platform. The steps include:

[0101] S41, calculate the similarity between the magnetic field profile of the vehicle under test and the magnetic field profile of the reference vehicle, and generate preliminary comparison results;

[0102] S42, when the preliminary comparison result is lower than the preset identity confirmation threshold, a differential profile is generated based on the difference between the magnetic field profile of the vehicle under test and the magnetic field profile of the reference vehicle;

[0103] S43, Analyze the data distribution characteristics of the differential profile, and determine whether the differential profile presents a locally concentrated change pattern or a globally distributed change pattern based on the data distribution characteristics, and obtain the change pattern determination result.

[0104] S44. If the change pattern determination result is a locally concentrated change pattern, then the vehicle to be inspected is determined to be an authorized vehicle with structural changes, and the occupancy status of the parking space corresponding to the carrying platform is detected as an authorized occupancy status; if the change pattern determination result is a globally distributed change pattern, then the vehicle to be inspected is determined to be an unauthorized vehicle, and the occupancy status of the parking space corresponding to the carrying platform is detected as a temporary occupancy status.

[0105] Similarity refers to the numerical value that measures the degree of matching between two magnetic field profiles. It can be quantified by calculating their correlation coefficient, Euclidean distance, or dynamic time-warped distance, aiming to make a preliminary quantitative judgment on vehicle identity. Differential profiles are data sets extracted and quantified by comparing the magnetic field profiles of the vehicle under test with those of a reference vehicle. This can be generated through point-by-point subtraction, normalized difference, or eigenvector difference, highlighting specific differences between the magnetic field profiles for subsequent refined analysis. Data distribution characteristics refer to the numerical and spatial arrangement or statistical properties of data points in the differential profile. Specifically, this includes peaks, valleys, fluctuation ranges, concentration, and dispersion, providing a basis for distinguishing different types of vehicle changes. Locally concentrated change patterns refer to data differences in the differential profile concentrated in a few specific areas, while differences are not significant or remain stable in other areas. This can manifest as a few significant, high-amplitude peaks or valleys in the differential profile, indicating possible local structural changes in the vehicle. The global distribution variation pattern refers to the data differences in the differential profile being widely distributed throughout the entire profile without obvious concentrated areas, or the differences exhibiting an overall, non-local variation state. Specifically, it can be manifested as a large deviation or frequent fluctuation in the overall values ​​of the differential profile. Its purpose is to indicate that there are fundamental differences between the vehicle under inspection and the reference vehicle, and that it may be an unauthorized vehicle.

[0106] In some preferred embodiments, this application is implemented as follows. When calculating the similarity between the magnetic field profile of the vehicle under inspection and the magnetic field profile of the reference vehicle, the Pearson correlation coefficient can be used as a similarity index. Specifically, the two magnetic field profiles are considered as two time series, and their linear correlation is calculated. The closer the absolute value of the correlation coefficient is to 1, the higher the similarity. When the calculated Pearson correlation coefficient is lower than a preset identity verification threshold, for example, it can be set to 0.8 or 0.75, the preliminary comparison result is considered not to meet the identity verification conditions, and further analysis is required. When the preliminary comparison result is lower than the identity verification threshold, a differential profile can be generated based on the difference between the magnetic field profile of the vehicle under inspection and the magnetic field profile of the reference vehicle. Specifically, each data point in the magnetic field profile of the vehicle under inspection can be subtracted point by point from the corresponding data point in the magnetic field profile of the reference vehicle to obtain a series of differences, which constitute the differential profile. For example, if the value of the reference profile at a certain time point is X, and the value of the profile under inspection at the same time point is Y, then the value of the differential profile at that time point is YX. Furthermore, when analyzing the data distribution characteristics of the differential profile, statistical methods can be used to determine whether it exhibits a locally concentrated or globally distributed pattern of variation. For example, the variance or standard deviation of all data points in the differential profile can be calculated and combined with a peak detection algorithm. If the differential profile contains only a few high-amplitude peaks or troughs with narrow widths, while the values ​​of most other data points are close to zero or within a small fluctuation range, it can be determined as a locally concentrated pattern of variation. This indicates that the magnetic field differences are mainly concentrated in a few discrete regions. Conversely, if the values ​​of data points in the differential profile generally deviate from zero and have a large fluctuation range, without obvious concentrated peaks, or if the peaks are widely distributed and irregular, it can be determined as a globally distributed pattern of variation. This indicates that the magnetic field differences are distributed throughout the entire profile, reflecting significant differences in the overall characteristics of the vehicle. If the change pattern is determined to be a locally concentrated change, for example, if the system identifies one or two significant magnetic field peaks at the corresponding position on the vehicle's top in the differential profile, while other parts show little change, then the vehicle under inspection can be identified as an authorized vehicle with structural modifications, such as an authorized vehicle with added roof racks or antennas. In this case, the occupancy status of the parking space corresponding to the support platform is detected as authorized occupancy. If the change pattern is determined to be a globally distributed change, for example, if the differential profile shows a general and significant deviation between the entire magnetic field curve and the baseline profile, then the vehicle under inspection can be identified as an unauthorized vehicle, such as a vehicle of a completely different model from an authorized vehicle. In this case, the occupancy status of the parking space corresponding to the support platform is detected as temporary occupancy.

[0107] Optional, combined Figure 4As shown, the step S42, which generates a differential profile based on the difference between the magnetic field profile of the vehicle under test and the magnetic field profile of the reference vehicle, includes:

[0108] S421, Apply time offset to the magnetic field profile of the vehicle to be inspected to generate a set of candidate magnetic field profiles;

[0109] S422, calculate the matching degree between each candidate magnetic field profile and the reference vehicle magnetic field profile in a set of candidate magnetic field profiles;

[0110] S423, Based on the matching degree, select the candidate magnetic field profile with the highest matching degree with the reference vehicle magnetic field profile as the aligned magnetic field profile of the vehicle to be inspected.

[0111] S424 generates a differential profile based on the aligned magnetic field profile of the vehicle under inspection and the magnetic field profile of the reference vehicle.

[0112] Here, time offset refers to the translation operation of the magnetic field profile data on the time axis, which can be achieved by increasing or decreasing the data point index. Its purpose is to simulate the small time deviations in the magnetic field data acquisition start point or process caused by the vehicle entering or leaving the detection area at different times. Candidate magnetic field profiles refer to a series of magnetic field profile data obtained by applying different time offsets to the magnetic field profile of the vehicle under test. Their purpose is to construct a dataset containing multiple time alignment possibilities in order to select the one that best matches the benchmark vehicle's magnetic field profile. Matching degree refers to a quantitative indicator that measures the similarity between two magnetic field profiles, which can be achieved by calculating the correlation coefficient, Euclidean distance, dynamic time warping (DTW) distance, or feature point matching. The method is as follows: The purpose is to evaluate the degree of similarity between the magnetic field profile of the vehicle under test after time offset and the magnetic field profile of the reference vehicle in terms of shape, trend, or characteristics. The aligned magnetic field profile of the vehicle under test refers to the magnetic field profile that is most similar to the reference vehicle magnetic field profile from a set of candidate magnetic field profiles based on the matching degree. Its purpose is to eliminate the time deviation between the original magnetic field profile of the vehicle under test and the reference vehicle magnetic field profile, and to ensure the accuracy of subsequent differential profile generation. The differential profile refers to the data sequence obtained by subtracting the aligned magnetic field profile of the vehicle under test and the reference vehicle magnetic field profile point by point or segment by segment. Its purpose is to highlight the actual differences in magnetic field characteristics between the two profiles and provide a clear basis for subsequent vehicle identification.

[0113] In some preferred embodiments, specifically, when performing the step of generating a differential profile based on the difference between the magnetic field profile of the vehicle under test and the reference vehicle magnetic field profile, the magnetic field profile of the vehicle under test can first be time-aligned. For example, the magnetic field profile of the vehicle under test can be shifted a series of small steps along the time axis, such as with a step size of 0.1 seconds, generating multiple time-off versions within a range of ±5 seconds, each version being a candidate magnetic field profile. Next, for each generated candidate magnetic field profile, the correlation coefficient method can be used to calculate its matching degree with the reference vehicle magnetic field profile. Specifically, the Pearson correlation coefficient between the two profile data sequences can be calculated; the closer the absolute value of the coefficient is to 1, the higher the matching degree. Then, the system compares the correlation coefficients between all candidate magnetic field profiles and the reference vehicle magnetic field profile, selecting the candidate magnetic field profile with the largest absolute value of the correlation coefficient as the aligned magnetic field profile of the vehicle under test. Finally, this aligned magnetic field profile of the vehicle under test is subtracted point by point from the reference vehicle magnetic field profile to generate the final differential profile. For example, if the value of the aligned magnetic field profile of the vehicle under test is X at a certain time point, and the value of the magnetic field profile of the reference vehicle at the corresponding time point is Y, then the value of the differential profile at that time point is XY. In this way, it can be ensured that the differential profile accurately reflects the differences in the magnetic field characteristics of the vehicle itself, rather than errors caused by time synchronization issues.

[0114] Optional, combined Figure 5 As shown, S43 analyzes the data distribution characteristics of the differential profile, and determines whether the differential profile exhibits a locally concentrated or globally distributed variation pattern based on these characteristics. The steps to obtain the variation pattern determination result include:

[0115] S431 divides the differential profile into several data segments;

[0116] S432, calculate the amount of data change in each data segment;

[0117] S433, identify data segments whose data changes exceed a preset segment change threshold, and use them as target segments;

[0118] S434, count the number of target segments and determine whether the target segments are in a separated state on the differential profile, and obtain the separation state judgment result;

[0119] S435, if the number of target segments is less than the preset number threshold and the separation state judgment result is yes, then the differential profile is determined to present a locally concentrated change pattern; if the number of target segments is greater than or equal to the preset number threshold or the separation state judgment result is no, then the differential profile is determined to present a globally distributed change pattern.

[0120] In this context, a data segment refers to a discrete data fragment formed by dividing continuous differential profile data according to time or spatial dimensions. This can be achieved using fixed-length windows or adaptive partitioning based on data characteristics (such as zero-crossing points or local extrema). Data variation refers to the quantitative indicator of the fluctuation or dispersion of magnetic field differences within a specific data segment. It can be calculated using statistical methods such as standard deviation, root mean square, the difference between the maximum and minimum values, or the sum of the absolute values ​​of differences between adjacent data points. The preset segment variation threshold is a numerical threshold used to distinguish between significant and insignificant changes. It can be set based on historical data analysis, empirical values, or trained using machine learning algorithms. The target segment refers to the data segment whose variation exceeds the preset segment variation threshold; these segments represent regions with significant magnetic field differences in the differential profile. Separation status refers to whether the target segments are independent and discontinuously distributed on the differential profile. Its purpose is to distinguish whether magnetic field changes are concentrated at a few discrete points or widely distributed over a continuous area. The preset quantity threshold refers to the numerical threshold used to determine whether the quantity of the target section is "too small" or "too large". It can be determined based on the actual application scenario, common characteristics of vehicle modification, or through experimental optimization.

[0121] In some preferred embodiments, a specific example is given below. Once the system obtains a differential profile, for example, containing 1000 sampling points, it can divide the differential profile into 100 data segments, each containing 10 consecutive sampling points. For each data segment, its data variation can be calculated, for example, by calculating the standard deviation of all sampling point values ​​within that segment. Subsequently, the system identifies data segments with standard deviations higher than a preset segment variation threshold and marks these segments as target segments. For example, if the preset segment variation threshold is set to 0.5, segments with standard deviations greater than 0.5 are identified as target segments.

[0122] Next, the system counts the total number of these target segments. Simultaneously, the system analyzes the distribution of these target segments on the differential profile to determine if they are separated. For example, if multiple non-target segments exist between two target segments, they can be considered separated. If the number of target segments, for example, is two, less than a preset threshold (e.g., the preset threshold is three), and these target segments are separated on the differential profile—for example, they are located at opposite ends of the profile with a large area of ​​non-target segments in the middle—then the system determines that the differential profile exhibits a locally concentrated change pattern. Conversely, if the number of target segments is five, greater than or equal to the preset threshold, or if these target segments are continuous and not separated—for example, they are closely connected to form a large area of ​​change—then the system determines that the differential profile exhibits a globally distributed change pattern. In this way, the system can accurately identify whether the vehicle has undergone local structural changes or a complete replacement based on the actual distribution of magnetic field differences in the differential profile.

[0123] Optional, combined Figure 6 As shown, the steps in S434 to determine whether the target segment is in a separated state on the differential profile and to obtain the separation state determination result include:

[0124] S4341, Identify adjacent target segments from all data segments to form merged segments;

[0125] S4342, Analyze the internal data of the fusion section, determine whether the number of data peaks in the fusion section is greater than one, and obtain the data peak judgment result;

[0126] S4343, if the data peak judgment result is yes, then obtain the valley value between adjacent data peaks in the fusion segment, and determine whether the valley value is higher than the preset segment change threshold to obtain the valley value judgment result;

[0127] S4344, if the valley value judgment result is yes, then the separation state judgment result is determined to be that the target segment is in a separation state;

[0128] S4345, if the data peak value judgment result is negative or the valley value judgment result is negative, then the separation state judgment result is determined to be that the target segment is in a non-separated state.

[0129] The fusion section refers to a continuous data region formed by merging multiple adjacent target sections in the differential profile. Its purpose is to analyze target sections that may be interrupted by noise or minor fluctuations but are essentially within the same change region as a whole. The number of data peaks refers to the number of points where the data value reaches a local maximum within the fusion section. It can be identified using peak detection algorithms in signal processing, such as threshold-based or derivative-based methods. Its purpose is to identify whether there are multiple change centers within the fusion section. The valley value refers to the local minimum value of data between adjacent data peaks within the fusion section. It can be obtained using valley detection algorithms in signal processing. Its purpose is to measure the degree of decline between adjacent peaks to determine whether they are truly separated. The preset section change threshold is a preset value used to determine whether the valley value is high enough to indicate that there is a separation between two peaks. It can be set according to the actual application scenario, the data characteristics of the magnetic field sensor, and the requirements for the accuracy of vehicle identification. Its purpose is to provide a criterion for judging the separation state.

[0130] In some preferred embodiments, the specific process of determining whether a target segment is in a separated state on the differential profile can be implemented as follows: First, the system can traverse all identified target segments in the differential profile. When a target segment is found, it checks whether its subsequent data segments are also target segments and whether the two are directly adjacent on the differential profile. If these conditions are met, these adjacent target segments are merged into a single fusion segment. For example, if the differential profile is divided into 100 data segments, and segments 10-15, 16-20, and 30-35 are target segments, then segment 10-20 will be identified as a fusion segment, while segment 30-35 may form a separate fusion segment.

[0131] Next, the magnetic field data within this fusion section is analyzed to identify data peaks. This can be achieved by applying a local maximum detection algorithm. For example, a sliding window can be set to find the maximum data value within the window, ensuring that this maximum value is higher than all data points within a certain range on both sides, and that its value is also higher than a preset peak identification threshold. If two or more such data peaks are detected within the fusion section, the number of data peaks is considered to be greater than one.

[0132] Furthermore, if multiple data peaks are identified within the fusion segment, the system will obtain the valley values ​​between these adjacent data peaks. For example, the point with the lowest data value between two adjacent peaks is the valley value. This valley value is then compared to a preset segment change threshold. This threshold can be determined empirically or through training data and is used to distinguish between separation and fluctuation. If the valley value is higher than this preset segment change threshold, it indicates that even if these target segments are spatially continuous, they exhibit a decrease in magnetic field intensity, sufficient to be considered two independent change regions. In this case, the separation state judgment result is determined that the target segment is in a separated state. Conversely, if there is only one data peak within the fusion segment, or if the valley value between adjacent peaks is lower than the preset segment change threshold, it indicates that these target segments actually represent a continuous, non-separated magnetic field change region. In this case, the separation state judgment result is determined that the target segment is in a non-separated state.

[0133] Optional, combined Figure 7 As shown in Figure S422, the steps for calculating the matching degree between each candidate magnetic field profile and the reference vehicle magnetic field profile in a set of candidate magnetic field profiles include:

[0134] S4221, the reference vehicle magnetic field profile and the candidate magnetic field profile are synchronously divided into several profile segments according to the time axis;

[0135] S4222, calculate the segment similarity of each group of profile segments;

[0136] S4223 aggregates the similarity of all segments to generate a matching score.

[0137] Among them, a profile segment refers to dividing continuous magnetic field profile data into sub-data segments with specific lengths or characteristics according to the time axis or sampling point sequence. This can be achieved by dividing at fixed time intervals, dividing based on magnetic field feature points, or dynamically adjusting the length. The purpose is to decompose the overall magnetic field profile into local units that are easier to analyze and compare. Segment similarity refers to a numerical value that quantifies the degree of similarity between two corresponding profile segments. It can be calculated using various algorithms such as Euclidean distance, cosine similarity, Pearson correlation coefficient, or dynamic time warping (DTW). Its purpose is to accurately measure the consistency of local magnetic field characteristics. Matching degree refers to a numerical value that comprehensively reflects the overall similarity between two complete magnetic field profiles. It can be generated by simply averaging, weighted averaging, summing, or aggregating all segment similarities according to specific rules. Its purpose is to provide a comprehensive and representative overall similarity assessment result.

[0138] In some preferred embodiments, this application is implemented as follows: First, the reference vehicle magnetic field profile and the candidate magnetic field profile are synchronously divided into several profile segments according to the time axis. For example, each magnetic field profile data stream can be sampled at a frequency of 10 sampling points per second, and divided into a profile segment of 50 sampling points each, so that each profile segment represents 5 seconds of magnetic field change data. This fixed-length division method can simplify the complexity of subsequent processing and ensure the comparability of each segment in the time dimension. Next, the segment similarity of each group of profile segments is calculated. Specifically, for each pair of synchronized profile segments, Euclidean distance can be used to measure their similarity, that is, the square root of the sum of the squares of the differences between the values ​​of all corresponding sampling points in the two segments is calculated. The smaller the Euclidean distance, the more similar the magnetic field characteristics of the two segments are. Finally, all segment similarities are aggregated to generate a matching degree. For example, the simple arithmetic mean of all calculated segment similarities can be used to obtain a final matching degree value. For example, if there are N profile segments, the matching degree can be defined as the sum of the similarities of all N segments divided by N. This aggregation method can provide an intuitive and easy-to-understand overall similarity assessment result, thus providing a reliable basis for subsequent vehicle identification.

[0139] Optional, combined Figure 8 As shown, the steps of S33 in determining the correction value based on the baseline startup characteristics and the current startup characteristics include:

[0140] S331, synchronously divide the baseline startup features and the current startup features into several groups of feature segments;

[0141] S332, calculate the segment deviation for each group of characteristic segments;

[0142] S333 identifies segment deviations that deviate numerically from the corresponding segment deviation center trend;

[0143] S334, from all the segment deviations, exclude the segment deviations that deviate numerically from the corresponding segment deviation central trend, and obtain the remaining segment deviations;

[0144] S335, determine the correction value based on the remaining segment deviation.

[0145] In this context, a characteristic segment refers to a subset of magnetic field characteristic data obtained by dividing the data along the time axis or sampling point sequence into equal or unequal lengths. Specifically, it can be the magnetic field data of the entire startup process divided according to fixed time intervals or the number of data points. The purpose is to decompose the overall magnetic field characteristic differences into multiple local differences for detailed analysis. Segment deviation refers to the numerical difference between the baseline startup characteristic and the current startup characteristic within the corresponding characteristic segment. This can be obtained by calculating the average difference, maximum difference, or root mean square difference of all data points within two corresponding characteristic segments. Its purpose is to quantify the magnetic field characteristic differences in each local region. The central tendency of segment deviation refers to the centrality or representativeness of a set of segment deviation data. This can be obtained by calculating the mean, median, or mode of all segment deviations. Its purpose is to provide a benchmark for determining which segment deviations fall within the normal fluctuation range. Identifying segment deviations that numerically deviate from the corresponding segment deviation central trend involves setting a threshold or using statistical methods to screen for those segment deviations that differ from the central trend. Specifically, this can be done by calculating the absolute difference between each segment deviation and the central trend and comparing it to a preset deviation threshold, or by using statistical methods such as box plots and Z-scores to identify outliers. The goal is to identify and mark abnormal deviations that may be caused by random interference. Exclusion refers to removing the identified segment deviations that deviate from the central trend from the entire set of segment deviations. This can be achieved through data filtering or deletion operations, aiming to eliminate outlier data and improve the accuracy of subsequent calculations. The remaining segment deviations refer to the representative and reliable set of segment deviations retained after outlier exclusion, providing a clean data foundation for calculating the correction value. Determining the correction value involves calculating a value for adjusting the background magnetic field data based on the remaining segment deviations. This can be obtained by calculating the average, median, or weighted average of the remaining segment deviations, providing a correction amount to eliminate magnetic field deviations during the startup process.

[0146] In some preferred embodiments, determining the correction value based on the baseline initiation feature and the current initiation feature can be implemented as follows: First, the magnetic field data of the two time series, the baseline initiation feature and the current initiation feature, are synchronously divided into several feature segments. For example, if each initiation feature contains 1000 sampling points, it can be divided into 100 feature segments, each containing 10 consecutive sampling points. Next, the segment deviation of each feature segment is calculated. This can be done by calculating the difference between the average values ​​of all sampling points within each corresponding segment. For example, for the i-th feature segment, the difference between the average value of the baseline initiation feature in that segment and the average value of the current initiation feature in that segment is calculated as the segment deviation of the i-th segment. Then, segment deviations that deviate numerically from the corresponding segment deviation central trend are identified. This can be achieved using statistical methods. For example, first, the average of all segment deviations can be calculated as the central trend. Then, the standard deviation of each segment deviation from this average can be calculated. If the difference between a segment deviation and the average exceeds a preset multiple of a certain standard deviation (e.g., 2 or 3 times the standard deviation), it is identified as a segment deviation deviating from the central trend. Subsequently, these identified deviating segment deviations are excluded from all segment deviations, resulting in the remaining segment deviations. Finally, a correction value is determined based on the remaining segment deviations. This can be achieved by calculating the arithmetic mean of the remaining segment deviations as the final correction value, or by using methods such as the median or weighted average. In this way, magnetic field deviations caused by random factors during startup can be effectively filtered out, ensuring the accuracy of the correction value.

[0147] Optional, combined Figure 9 As shown, the steps in S4223 to aggregate the similarity of all segments and generate the matching score include:

[0148] A1. Based on the benchmark vehicle magnetic field profile, determine the feature significance of the profile segment and generate the weight corresponding to each profile segment.

[0149] A2 generates a matching score by weighting and aggregating the segment similarity based on weights.

[0150] The saliency of a profile segment refers to the amount of information or uniqueness contained in a specific segment of the magnetic field profile that is capable of distinguishing vehicle identification. It can be determined using various methods, such as quantifying it by analyzing the variation amplitude, frequency characteristics, waveform complexity, or contrast with background noise of the magnetic field data within the segment. Alternatively, it can be quantified by training a machine learning model on historical data to identify segments that contribute significantly to the classification results. The aim is to identify key regions in the magnetic field profile that contribute more to vehicle identification. The weight refers to a numerical factor assigned to the segment similarity of each profile segment, reflecting the corresponding... The importance of segments in matching degree calculation can be dynamically adjusted based on the feature significance of the segment profile. For example, the higher the feature significance of a segment, the larger its corresponding weight value. The purpose is to distinguish the degree of contribution of different segments to the final matching degree. Weighted aggregation refers to multiplying the similarity of each segment by its corresponding weight, and then summing or averaging the weighted similarities to obtain the final matching degree. It can be implemented by linear weighted summation, nonlinear weighted averaging, or other statistical methods. Its purpose is to highlight the contribution of feature-significant segments while weakening the interference of feature-insignificant segments, thereby improving the accuracy of the matching degree.

[0151] In some preferred embodiments, this application is implemented as follows: When aggregating the similarity of all segments and generating the matching degree, the feature significance can first be determined by calculating the variance or standard deviation of the magnetic field data within each segment based on the benchmark vehicle magnetic field profile. The larger the variance or standard deviation, the more drastic the magnetic field change in that segment, and the more significant the feature. For example, if the magnetic field data of a certain segment shows obvious peaks or troughs in the benchmark profile, its variance will be relatively large, indicating that its feature is significant. Subsequently, weights corresponding to each segment can be generated based on these variance values. One way to generate weights is to normalize the variance of each segment or map it to a value between 0 and 1 as the weight of that segment. For example, the maximum value of the variance of all segments can be set as the benchmark, and the weight of other segments can be the ratio of their variance to the maximum variance. Then, when aggregating segment similarity, a weighted summation method can be used. Specifically, the segment similarity of each profile segment is multiplied by its corresponding weight, and then all weighted segment similarities are summed to obtain the final matching degree. For example, if the reference vehicle magnetic field profile is divided into N profile segments, and each segment i has its segment similarity Si and weight Wi, then the final matching degree M can be calculated as M = Σ(Si * Wi), where i ranges from 1 to N. In this way, segments with significant changes in magnetic field characteristics and a large contribution to vehicle identification will have a larger weight in the final matching degree, thereby effectively improving the accuracy of matching.

[0152] A parking space status detection system is used to perform parking space status detection, combined with Figure 10 As shown, the parking space status detection system 1 includes:

[0153] Background magnetic field acquisition module 11 is used to acquire background magnetic field data corresponding to the lifting and lowering movement of the carrying platform when the carrying platform of the mechanical lifting parking equipment is unloaded, by using a magnetic field sensor preset at the ground.

[0154] The reference profile generation module 12 is used to collect first raw magnetic field data using a magnetic field sensor when the carrying platform needs to carry the authorized vehicle and perform lifting movements, and to process the first raw magnetic field data based on the background magnetic field data to generate a reference vehicle magnetic field profile of the authorized vehicle.

[0155] The inspection data generation module 13 is used to collect second original magnetic field data using a magnetic field sensor when the carrying platform needs to carry the vehicle to be inspected and perform lifting movements, and to process the second original magnetic field data based on the background magnetic field data to generate the vehicle magnetic field profile of the vehicle to be inspected.

[0156] The identity status detection module 14 is used to compare the magnetic field profile of the vehicle to be inspected with the magnetic field profile of the reference vehicle, and determine the identity of the vehicle to be inspected based on the comparison result, so as to detect the occupancy status of the parking space corresponding to the carrying platform.

[0157] Among them, the background magnetic field acquisition module refers to the unit used to acquire magnetic field reference information under specific environment. It can be implemented by hardware combination of integrated magnetic field sensor, data acquisition device and preliminary data processor. Its purpose is to establish a magnetic field baseline without vehicle interference so as to identify the magnetic field changes introduced by vehicles in the future.

[0158] The baseline profile generation module is a unit used to extract and construct the unique magnetic field features of authorized vehicles from the original magnetic field data. It can be implemented using a computing unit that includes a signal processor, data storage and feature extraction algorithm. Its purpose is to provide a reference standard for subsequent vehicle identification comparison.

[0159] The data generation module is a unit used to extract and construct the magnetic field characteristics of the vehicle to be inspected from the collected raw magnetic field data. It can be implemented using a computing unit similar to the reference profile generation module. Its purpose is to provide magnetic field data to be compared for vehicle identification.

[0160] The identity and status detection module refers to the unit used to determine the vehicle identity and parking space occupancy status by comparing the vehicle's magnetic field characteristics. It can be implemented using a central processing unit that includes a comparison algorithm, decision logic, and status output interface. Its purpose is to determine the parking space occupancy status and identify the vehicle identity.

[0161] In some preferred embodiments, this application is implemented as follows:

[0162] The background magnetic field acquisition module can consist of a triaxial magnetoresistive sensor array, a data acquisition card, and a microcontroller. The magnetoresistive sensor array can be pre-positioned below the parking space floor, for example, below the centerline of the lifting path of the platform, to ensure magnetic field data is acquired during the platform's lifting process. The microcontroller controls the data acquisition card to sample data and transmits the acquired raw magnetic field data to the back-end processing unit via wired or wireless means. The reference profile generation module and the test data generation module can be integrated into a central processing unit, such as a computer or embedded processor. This processing unit can run signal processing software. After acquiring the raw magnetic field data, the software can first perform a background magnetic field data subtraction operation, such as using point-by-point subtraction or adaptive filtering algorithms to eliminate background magnetic field interference generated by the mechanical lifting equipment. Subsequently, feature extraction can be performed on the data after removing the background magnetic field, such as using Fourier transform, wavelet analysis, or principal component analysis to extract vehicle-specific magnetic field profile features and store them as a reference vehicle magnetic field profile or a test vehicle magnetic field profile. The identity status detection module can also be integrated into this central processing unit. This module can receive the magnetic field profiles of the vehicle to be inspected and a reference vehicle, and execute a comparison algorithm. The comparison algorithm can use methods such as correlation coefficient calculation, Euclidean distance measurement, or machine learning classifiers to quantify the similarity between the two magnetic field profiles. Based on the calculated similarity, the system can set a threshold. When the similarity is higher than the threshold, the vehicle to be inspected is determined to be an authorized vehicle, and the parking space status is authorized occupancy. When the similarity is lower than the threshold, further analysis of differences may be conducted to determine whether it is an unauthorized vehicle or an authorized vehicle with structural changes, thereby determining the temporary occupancy status or structural change occupancy status of the parking space. The entire system can interact with the parking management platform to update parking space status and vehicle identity information.

[0163] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for detecting the status of a parking space, characterized in that, include When the platform of the mechanical lifting parking equipment is unloaded, the background magnetic field data corresponding to the lifting movement of the platform is collected by a magnetic field sensor pre-installed on the ground. When the carrying platform needs to carry the authorized vehicle and perform lifting movements, it uses a magnetic field sensor to collect the first raw magnetic field data, and processes the first raw magnetic field data based on the background magnetic field data to generate a reference vehicle magnetic field profile of the authorized vehicle. When the carrying platform needs to carry the vehicle to be inspected and perform lifting and lowering movements, a second original magnetic field data is collected using a magnetic field sensor, and the second original magnetic field data is processed based on the background magnetic field data to generate a magnetic field profile of the vehicle to be inspected. The magnetic field profile of the vehicle under inspection is compared with the magnetic field profile of the reference vehicle, and the identity of the vehicle under inspection is determined based on the comparison result, so as to detect the occupancy status of the parking space corresponding to the carrying platform.

2. The parking space status detection method according to claim 1, characterized in that, The step of processing the second original magnetic field data based on the background magnetic field data to generate the magnetic field profile of the vehicle to be inspected includes: From the background magnetic field data, the magnetic field characteristics of the carrying platform in the initial stage of lifting and lowering motion are extracted as the reference start-up characteristics. From the second original magnetic field data, the magnetic field characteristics of the carrying platform in the initial stage of the lifting motion are extracted as the current start-up characteristics; Based on the baseline startup characteristics and the current startup characteristics, a correction value is determined; The background magnetic field data is adjusted based on the correction value to generate dynamic background magnetic field data; The dynamic background magnetic field data is removed from the second original magnetic field data to generate the magnetic field profile of the vehicle under inspection.

3. The parking space status detection method according to claim 2, characterized in that, The step of comparing the magnetic field profile of the vehicle to be inspected with the magnetic field profile of the reference vehicle, and determining the identity of the vehicle to be inspected based on the comparison result, so as to detect the occupancy status of the parking space corresponding to the carrying platform, includes: Calculate the similarity between the magnetic field profile of the vehicle under test and the magnetic field profile of the reference vehicle to generate preliminary comparison results; When the preliminary comparison result is lower than the preset identity confirmation threshold, a differential profile is generated based on the difference between the magnetic field profile of the vehicle under test and the magnetic field profile of the reference vehicle. Analyze the data distribution characteristics of the differential profile, and determine whether the differential profile presents a locally concentrated change pattern or a globally distributed change pattern based on the data distribution characteristics, and obtain the change pattern determination result. If the change pattern determination result is a locally concentrated change pattern, then the vehicle under inspection is identified as an authorized vehicle with structural changes, and the occupancy status of the parking space corresponding to the carrying platform is detected as an authorized occupancy status; if the change pattern determination result is a globally distributed change pattern, then the vehicle under inspection is identified as an unauthorized vehicle, and the occupancy status of the parking space corresponding to the carrying platform is detected as a temporary occupancy status.

4. The parking space status detection method according to claim 3, characterized in that, The step of generating a differential profile based on the difference between the magnetic field profile of the vehicle under test and the magnetic field profile of the reference vehicle includes: A time offset is applied to the magnetic field profile of the vehicle under test to generate a set of candidate magnetic field profiles; Calculate the matching degree between each candidate magnetic field profile in a set of candidate magnetic field profiles and the reference vehicle magnetic field profile; Based on the matching degree, the candidate magnetic field profile with the highest matching degree with the reference vehicle magnetic field profile is selected as the aligned vehicle magnetic field profile to be inspected. A differential profile is generated based on the aligned magnetic field profile of the vehicle under test and the magnetic field profile of the reference vehicle.

5. The parking space status detection method according to claim 4, characterized in that, The steps of analyzing the data distribution characteristics of the differential profile and determining whether the differential profile exhibits a locally concentrated or globally distributed variation pattern based on the data distribution characteristics to obtain the variation pattern determination result include: The differential profile is divided into several data segments; Calculate the amount of data change for each of the data segments; Identify data segments whose changes exceed a preset segment change threshold and designate them as target segments; The number of target segments is counted, and it is determined whether the target segments are in a separated state on the differential profile to obtain the separation state determination result; If the number of target segments is less than a preset threshold and the separation state judgment result is yes, then the differential profile is determined to present a locally concentrated change pattern; if the number of target segments is greater than or equal to the preset threshold or the separation state judgment result is no, then the differential profile is determined to present a globally distributed change pattern.

6. The parking space status detection method according to claim 5, characterized in that, The step of determining whether the target segment is in a separated state on the differential profile and obtaining the separation state determination result includes: Identify adjacent target segments from all the data segments to form fused segments; Analyze the internal data of the fusion section to determine whether the number of data peaks in the fusion section is greater than one, and obtain the data peak judgment result; If the data peak value judgment result is yes, then the valley value between adjacent data peak values ​​in the fusion segment is obtained, and it is determined whether the valley value is higher than the preset segment change threshold to obtain the valley value judgment result. If the valley value judgment result is yes, then the separation state judgment result is determined to be that the target segment is in a separation state; If the data peak value judgment result is negative or the valley value judgment result is negative, then the separation state judgment result is determined to be that the target segment is in a non-separated state.

7. The parking space status detection method according to claim 4, characterized in that, The step of calculating the matching degree between each candidate magnetic field profile in a set of candidate magnetic field profiles and the reference vehicle magnetic field profile includes: The reference vehicle magnetic field profile and the candidate magnetic field profile are synchronously divided into several profile segments according to the time axis. Calculate the segment similarity for each group of the profile segments; Aggregate the similarity of all the segments to generate a matching score.

8. The parking space status detection method according to claim 2, characterized in that, The step of determining the correction value based on the baseline startup characteristics and the current startup characteristics includes: The baseline startup feature and the current startup feature are synchronously divided into several feature segments; Calculate the segment deviation for each group of characteristic segments; Identify segment deviations that deviate numerically from the corresponding segment deviation center trend; From all the segment deviations mentioned, exclude the segment deviations that deviate numerically from the corresponding segment deviation central trend to obtain the remaining segment deviations; Based on the remaining segment deviation, determine the correction value.

9. A parking space status detection method according to claim 7, characterized in that, The step of aggregating the similarity of all the segments to generate a matching score includes: Based on the benchmark vehicle magnetic field profile, the feature saliency of the profile segment is determined, and the weights corresponding to each profile segment are generated. The similarity of the segments is weighted and aggregated to generate a matching score.

10. A parking space status detection system, used to perform parking space status detection, characterized in that, include: The background magnetic field acquisition module is used to collect background magnetic field data corresponding to the lifting and lowering movement of the mechanical lifting parking equipment when the carrying platform is unloaded, using a magnetic field sensor pre-installed on the ground. The reference profile generation module is used to collect first raw magnetic field data using a magnetic field sensor when the carrying platform needs to carry an authorized vehicle and perform lifting movements, and to process the first raw magnetic field data based on the background magnetic field data to generate a reference vehicle magnetic field profile of the authorized vehicle. The inspection data generation module is used to collect second original magnetic field data using a magnetic field sensor when the carrying platform needs to carry the vehicle to be inspected and perform lifting movements, and to process the second original magnetic field data based on the background magnetic field data to generate the vehicle magnetic field profile of the vehicle to be inspected. The identity status detection module is used to compare the magnetic field profile of the vehicle to be inspected with the magnetic field profile of the reference vehicle, and determine the identity of the vehicle to be inspected based on the comparison result, so as to detect the occupancy status of the parking space corresponding to the carrying platform.

Images