Parking space sharing matching system and method based on space-time constraint

Abstract

The invention provides a space-time constraint-based parking space sharing matching system and method, and belongs to the technical field of computers, the system comprises a user terminal module, a parking space information acquisition module and a data processing server, and parking space sharing matching is realized by establishing a space-time constraint index and calculating a matching degree. The system supports a user to customize a space radius threshold value, corrects a linear distance into an actual driving distance according to an urban road network, and sets a buffer time parameter extension matching time window at the same time. In addition, a weight coefficient is dynamically optimized through a machine learning algorithm to maximize a user reservation success rate, and hierarchical encryption protection is performed on user coordinates and parking time windows by adopting a homomorphic encryption technology. The problems that an existing parking space sharing system is insufficient in space-time matching precision, poor in algorithm adaptability and weak in privacy protection are solved, the accuracy, flexibility and data safety of parking space matching are improved, and the parking resource utilization rate and the user experience are effectively improved.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of computer, and particularly relates to a parking space sharing matching system and method based on space-time constraints. BACKGROUND

[0002] With the continuous growth of urban motor vehicle ownership, the problem of parking resource shortage is increasingly prominent, and the traditional parking space management mode has defects such as information asymmetry between supply and demand, low space-time resource utilization rate. The existing parking space sharing system generally has the following technical problems: first, the spatial matching only uses straight-line distance calculation, without combining the characteristics of urban road network to correct the actual driving distance, resulting in a large deviation of the accessibility of the recommended parking space and the user's destination; second, the time constraint lacks a flexible buffer mechanism, and the user's expected time window and the parking space's available time need to completely coincide to match, which is difficult to adapt to the time fluctuation demand in the parking scene; third, the matching algorithm weight coefficient is fixed and cannot be dynamically optimized according to user behavior preferences, resulting in a low reservation success rate; fourth, the user privacy protection mechanism is insufficient, and there is a risk of leakage of coordinate and time data in the transmission and processing process, which limits the user's enthusiasm for participating in parking space sharing.

[0003] In view of the above problems, the present application provides a parking space sharing matching system and method based on space-time constraints, which corrects the actual driving distance by introducing the road traffic coefficient, sets a buffer time to expand the time window overlap condition, dynamically optimizes the matching weight by combining machine learning, and realizes user privacy protection by using homomorphic encryption technology, thereby improving the accuracy, flexibility and data security of parking space matching, and effectively solving the technical problems of low space-time resource matching efficiency and insufficient privacy protection in the prior art. SUMMARY

[0004] The present application aims to provide a parking space sharing matching system and method based on space-time constraints, which aims to solve the problem that the spatial matching in the prior art only uses straight-line distance calculation, without combining the characteristics of urban road network to correct the actual driving distance, resulting in a large deviation of the accessibility of the recommended parking space and the user's destination.

[0005] To achieve the above-mentioned purpose, the present application provides the following technical scheme:

[0006] A parking space sharing matching system based on space-time constraints, characterized in that it comprises: a user terminal module for inputting user parking demand parameters, the parameters including: departure coordinates (x1, y1), destination coordinates (x2, y2), expected parking time T, expected parking time window [tstart, tend], and collecting user real-time position coordinates (xu, yu);

[0007] Parking space information collection module, for collecting parking space state parameters, the parameters include: parking space position coordinates (xp, yp), available time window [tbegin, tend], parking space type (indoor / outdoor), parking space size, charging standard P;

[0008] Data processing server, comprising: space-time constraint modeling module, for establishing space constraint index D≤R (wherein D is the straight-line distance between parking space and destination, and R is the user-set space radius threshold) and time constraint index

[0009] Matching algorithm module, for calculating matching degree S=ω1·(1-RD)+ω2·(1-tduration∣tstart-tbegin∣)+ω3·E, wherein ω1+ω2+ω3=1 is weight coefficient, tduration is parking space available time length, and E is historical evaluation score (0-5 points).

[0010] Privacy protection module, for homomorphic encryption processing on user coordinates (xu, yu) and parking time window [tstart, tend].

[0011] As a preferred scheme of the application, the space radius threshold R supports user self-defined setting, and the default value is 500 meters, and the system supports correcting the straight-line distance D to actual driving distance D' according to the urban road network, and the correction formula is D'=k·D, wherein k is a road traffic coefficient (1≤k≤1.5).

[0012] As a preferred scheme of the application, the time constraint index further includes a buffer time parameter Δt, for expanding Δt (default 15 minutes) at both ends of the user's expected parking time window to form a matching time window [tstart-Δt, tend+Δt], and requiring that the parking space available time window and the matching time window overlap for a length of ≥T.

[0013] As a preferred scheme of the application, the time constraint index further includes a buffer time parameter Δt, for expanding Δt (default 15 minutes) at both ends of the user's expected parking time window to form a matching time window [tstart-Δt, tend+Δt], and requiring that the parking space available time window and the matching time window overlap for a length of ≥T.

[0014] As a preferred scheme of the application, according to historical matching data, the weight coefficients ω1, ω2 and ω3 are optimized by a machine learning algorithm, and the optimization target is to maximize the user reservation success rate = (matching result display number)×(successful reservation number)×100%, and the success rate threshold is set to ≥70%.

[0015] As a preferred scheme of the present application, the parameter input step further comprises user privacy parameter setting, supporting user to select encryption level (primary / intermediate / advanced), corresponding to different homomorphic encryption calculation complexity, primary encryption processing time≤200 milliseconds, and advanced encryption processing time≤500 milliseconds.

[0016] Compared with the prior art, the present application has the following beneficial effects:

[0017] 1. In this scheme, the space constraint introduces the road passing coefficient k (1≤k≤1.5), and the straight-line distance D is corrected to the actual driving distance D'=k·D, for example, k=1.3 in the city center and k=1.1 in the suburbs, to ensure the actual accessibility of the recommended parking space.

[0018] The time constraint adds a buffer time Δt (default 15 minutes), which expands the user's expected window to [tstart-Δt, tend+Δt], and requires that the overlapping length of the parking space available time and the expanded window≥T, allowing partial overlap of the time window.

[0019] The matching accuracy is improved by more than 30%, avoiding invalid recommendations caused by straight-line distance misjudgment; the time elastic buffer makes the matching success rate increase by about 25%, adapting to the time fluctuation in the driving process.

[0020] 2. In this scheme, the space constraint introduces the road passing coefficient k (1≤k≤1.5), and the straight-line distance D is corrected to the actual driving distance D'=k·D, for example, k=1.3 in the city center and k=1.1 in the suburbs, to ensure the actual accessibility of the recommended parking space.

[0021] The time constraint adds a buffer time Δt (default 15 minutes), which expands the user's expected window to [tstart-Δt, tend+Δt], and requires that the overlapping length of the parking space available time and the expanded window≥T, allowing partial overlap of the time window.

[0022] The matching accuracy is improved by more than 30%, avoiding invalid recommendations caused by straight-line distance misjudgment; the time elastic buffer makes the matching success rate increase by about 25%, adapting to the time fluctuation in the driving process. BRIEF DESCRIPTION OF DRAWINGS

[0023] The accompanying drawings are included to provide a further understanding of the present application, and constitute a part of the specification, together with the embodiments of the present application, to explain the present application, and do not constitute a limitation on the present application. In the drawings:

[0024] Figure 1 The present application is a space-time constraint-based parking space sharing matching system module cooperation flowchart.

[0025] Figure 2 The present application is a dynamic parameter adjustment and privacy protection flowchart. DETAILED DESCRIPTION

[0026] With reference to the drawings of the embodiments of the present application, the technical solutions in the embodiments of the present application will be described clearly and completely. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all the other embodiments obtained by those skilled in the art without creative work are within the protection scope of the present application.

[0027] Embodiment 1

[0028] Please refer to Figures 1-2 , the present application provides the following technical solutions:

[0029] A parking space sharing matching system based on space-time constraints, characterized in that it comprises: a user terminal module for inputting user parking demand parameters, the parameters including: departure coordinates (x1, y1), destination coordinates (x2, y2), expected parking duration T, expected parking time window [tstart

[0030] , tend], and collecting real-time position coordinates (xu, yu) of the user;

[0031] A parking space information collection module for collecting parking space state parameters, the parameters including: parking space position coordinates (xp, yp), available time window [tbegin, tend], parking space type (indoor / outdoor), parking space size, and charging standard P;

[0032] A data processing server comprising: a space-time constraint modeling module for establishing a space constraint index D≤R (where D is the straight-line distance between the parking space and the destination, and R is the user-set space radius threshold) and a time constraint index

[0033] A matching algorithm module for calculating a matching degree S=ω1·(1-RD)+ω2·(1-tduration∣tstart-tbegin∣)+ω3·E, where ω1+ω2+ω3

[0034] =1 is a weight coefficient, tduration is the available time length of the parking space, and E is the historical evaluation score (0-5 points);

[0035] A privacy protection module for homomorphic encryption processing of the user coordinates (xu, yu) and the parking time window [tstart

[0036] , tend].

[0037] In the specific embodiments of the present application, I. Module cooperation process and data flow

[0038] 1. Initial data interaction between user terminal module and parking space information collection module

[0039] User demand input stage:

[0040] The user inputs the departure coordinate (x1, y1) and destination coordinate (x2, y2) through the mobile phone APP (user terminal module), and the system automatically associates the map service to parse the address and allows the user to set the spatial radius threshold R (such as the default 500 meters).

[0041] The user selects the expected parking duration T and desired time window [tstart, tend] (such as "14:00-16:00"), and the APP real-time obtains the mobile phone positioning as the user's current coordinate (xu, yu) for subsequent path planning.

[0042] Parking space information synchronization stage:

[0043] The parking space information collection module obtains the parking space status in real time through Internet of Things sensors (such as geomagnetic sensors, cameras) or parking lot management system API, including location coordinates (xp, yp), available time window [tbegin, tend], type (indoor / outdoor), size (such as small parking space / SUV parking space), and charging standard P, and stores the encrypted data in the server database.

[0044] 2. Spatio-temporal constraint filtering and matching calculation of data processing server

[0045] Spatio-temporal constraint modeling module processing logic:

[0046] Spatial filtering: Calculate the straight-line distance D = (xp-x2

[0047] )2+(yp-y2)2 between each parking space and the destination (x2, y2), and filter out the parking space set that satisfies D≤R.

[0048] Time filtering: Check whether the user's desired time window [tstart, tend] is completely contained in the parking space's available time window [tbegin, tend], i.e. tstart≥tbegin and tend≤tend

[0049] , and filter out the time-matched parking spaces.

[0050] Matching algorithm module calculation process:

[0051] For the filtered parking spaces, calculate the spatial constraint score: (1-RD), when D=0, the score is 1, and when D=R, the score is 0.

[0052] The calculation time constraint score: tduration = tend-tbegin (the duration of the available parking space), and the time difference between the user's desired start time and the start time of the available parking space is |tstart-tbegin|, and the score formula is (1-tduration |tstart-tbegin|), and the smaller the time difference, the higher the score.

[0053] The historical evaluation score E (0-5 points, summarized by the user who has used the parking space) is combined with the weight coefficients ω1, ω2, ω3 (such as ω1=0.4, ω2=0.3, ω3=0.3) to calculate the total matching degree S=ω1·(1-RD)+ω2·(1-tduration |tstart-tbegin

[0054] ∣)+ω3·E.

[0055] According to S from high to low, a recommended parking space list is generated.

[0056] 3. Encryption processing and secure interaction of the privacy protection module

[0057] Data encryption scenario:

[0058] Before the user terminal sends the coordinates (xu, yu) and the time window [tstart, tend], the data is encrypted using a homomorphic encryption algorithm (such as Paillier encryption), ensuring that the server can only process data in an encrypted state (such as when calculating the distance, the ciphertext operation is used to compare D and R, without decrypting the original coordinates).

[0059] The parking space coordinates (xp, yp) uploaded by the parking space information collection module can also be encrypted and stored, and the server can complete the time and space constraint filtering in the ciphertext state through secure multi-party computation (MPC) technology, avoiding the leakage of plaintext data.

[0060] Decryption and result return:

[0061] After the server calculates the matching degree, the encrypted recommended result is returned to the user terminal, and the terminal uses the private key to decrypt, only displaying the matching degree ranking and the desensitized parking space information (such as the location only showing XX meters away from the destination, without exposing the accurate coordinates).

[0062] II. Module coordination example in actual application scenarios

[0063] Scenario: User drives to the mall for shopping and looks for nearby parking spaces

[0064] User input requirements (user terminal module):

[0065] Open APP, input destination mall coordinates (x2, y2), set space radius R = 300 meters, expected parking time [15:00, 17:00], and expected parking time 2 hours. APP real-time acquisition of user's current location (xu, yu), display remaining driving distance and time.

[0066] Parking space information synchronization (parking space information collection module):

[0067] The parking lot uploads the parking space status in real time through the sensor: A parking space (xp, yp) available time [14:30, 17:30], charge 10 yuan / hour; B parking space (xp, yp) available time [15:10, 18:00], charge 8 yuan / hour, both meet the space constraint D≤300 meters.

[0068] Space-time filtering and matching calculation (data processing server):

[0069] The space-time constraint modeling module filters out the parking spaces whose time window contains [15:00, 17:00] (both A and B meet).

[0070] Matching algorithm module calculation:

[0071] A parking space: D = 200 meters, |15:00-14:30| = 30 minutes, tduration = 3 hours (180 minutes), historical evaluation E = 4.5 points.

[0072] Space score: 1-300200≈0.33; Time score: 1-18030≈0.83; Total matching degree S = 0.4×0.33+0.3×0.83+0.3×4.5≈1.73.

[0073] B parking space: D = 150 meters, |15:00-15:10| = 10 minutes, tduration = 2 hours 50 minutes (170 minutes), historical evaluation E = 4 points.

[0074] Space score: 1-300150 = 0.5; Time score: 1-17010≈0.94; Total matching degree S = 0.4×0.5+0.3×0.94+0.3×4≈1.88.

[0075] B parking space has a higher matching degree, and the recommendation priority is higher than that of A parking space.

[0076] Privacy protection and result display:

[0077] The server returns the encrypted matching results (including S value, parking space type, and charging standard) to the user terminal, and the terminal displays: "Recommended B parking space, 150 meters away from the destination, available time 15:10-18:00, charging 8 yuan / hour, rating 4 points" after decryption. The user's coordinates are encrypted throughout the journey, and the server cannot obtain the real location.

[0078] Subsequent process:

[0079] The user selects the B parking space, the APP generates a navigation path (calculates the route based on encrypted coordinates to avoid leaking the driving track), scans the code to unlock the parking space after arriving, and the system automatically deducts the fee after parking and prompts the user to evaluate the parking space, updating the historical evaluation score E of the parking space.

[0080] III. Implementation of key technologies and optimization points

[0081] Real-time guarantee:

[0082] The parking space information collection module uses the MQTT protocol to push state changes in real time, and the data processing server uses Redis to cache hot area parking space data, reducing database query delay.

[0083] The user terminal updates the real-time position (xu, yu) regularly (such as every 30 seconds), and the server dynamically adjusts the recommended parking space according to the user's moving track (such as reducing the R threshold when approaching the destination).

[0084] Dynamic adjustment of weight coefficients:

[0085] The system adaptively adjusts ω1, ω2, ω3 according to the user's historical behavior (such as whether to prefer low-price parking spaces or whether to care about walking distance), for example, high-frequency users can set personalized weights.

[0086] Homomorphic encryption efficiency optimization:

[0087] For the square root operation in distance calculation, use approximate calculation (such as Taylor expansion) or convert to Euclidean distance square comparison (D2≤R2) to avoid complex ciphertext operations and improve server processing speed.

[0088] Abnormal processing mechanism:

[0089] If the matching degree S is lower than the threshold (such as S<0.5), the system automatically expands the R threshold or relaxes the time constraint (such as allowing the user's time window to partially overlap with the parking space's time window, adjusting the constraint condition to And prompts the user that the nearby parking space is tight.

[0090] IV. Module interface and data interaction protocol

[0091] User terminal→server interface:

[0092] POST / api / parking / request: Transmit encrypted (x1, y1), (x2, y2), T, [tstart, tend], (xu, yu), request body in JSON format, use homomorphic encryption key to encrypt numerical fields.

[0093] Parking space information collection -> server interface:

[0094] POST / api / parking / spots: Transmit encrypted (xp, yp), [tbegin, tend], parking type, size, P, with sensor timestamp to ensure data real-time.

[0095] Server -> user terminal interface:

[0096] GET / api / parking / recommend: Return encrypted matching parking space list, including (xp, yp) (encrypted), S, parking type, P, decrypted by user terminal and displayed.

[0097] For details, please refer to Figures 1-2 , the space radius threshold R supports user-defined setting, the default value is 500 meters, and the system supports correcting the straight line distance D to the actual driving distance D' according to the urban road network, the correction formula is D' = k·D, where k is the road traffic coefficient (1 ≤ k ≤ 1.5).

[0098] In this embodiment: 1. Actual distance correction of spatial constraints

[0099] User terminal module: Allow users to customize the space radius threshold R (default 500 meters), and synchronize the default configuration of the road traffic coefficient k (such as k = 1.3 in the city center and k = 1.1 in the suburbs).

[0100] Data processing server:

[0101] After the time-space constraint modeling module calculates the straight line distance D, it is corrected to the actual driving distance according to the formula D' = k·D, and the screening condition becomes D' ≤ R. For example, the straight line distance between the destination and the parking space is D = 400 meters, k = 1.3, D' = 520 meters, and if R = 500 meters, the constraint is not met, and the parking space needs to be excluded.

[0102] 2. Buffer expansion mechanism of time constraints

[0103] User terminal module: Default to add buffer time Δt = 15 minutes, expand the user's expected time window to [tstart-Δt, tend+Δt] (such as expanding the original window [14:00, 16:00] to [13:45, 16:15]).

[0104] Data processing server:

[0105] Check the overlap length of the parking space available time window [tbegin, tend] and the extended matching time window, and require the overlap length ≥ T (estimated parking time). For example, the parking space available time is [13:30, 17:00], the user T = 2 hours, the extended window is [13:45, 16:15], and the overlap length is 13:45-16:00 (2.25 hours), which meets the condition.

[0106] For details, please refer to Figures 1-2 , the time constraint index further includes a buffer time parameter Δt, which is used to extend Δt (default 15 minutes) at both ends of the user's expected parking time window to form a matching time window [tstart-Δt, tend+Δt], and requires that the parking space available time window overlaps with the matching time window for a length ≥ T.

[0107] In this embodiment: user terminal module: add buffer time Δt = 15 minutes by default, and expand the user's expected time window to [tstart-Δt, tend+Δt] (such as expanding the original window [14:00, 16:00] to [13:45, 16:15]).

[0108] Data processing server:

[0109] Check the overlap length of the parking space available time window [tbegin, tend] and the extended matching time window, and require the overlap length ≥ T (estimated parking time). For example, the parking space available time is [13:30, 17:00], the user T = 2 hours, the extended window is [13:45, 16:15], and the overlap length is 13:45-16:00 (2.25 hours), which meets the condition.

[0110] For details, please refer to Figure 2 , including the following steps: parameter input step: receiving the user input of the departure coordinates (x1, y1), the destination coordinates (x2, y2), the expected parking time T, the expected parking time window [tstart, tend], and the parking space owner input of the parking space location coordinates (xp

[0111] , yp), available time window [tbegin, tend], and charging standard P;

[0112] Space-time constraint screening step: calculate the distance D = (xp-x2)2+(yp-y2

[0113] )2 between the parking space and the destination, and screen the candidate parking spaces that meet D≤R and ;

[0114] Matching degree calculation step: calculate the matching degree S of the candidate parking space = ω1·(1-RD) + ω2·(1-tduration∣tstart-tbegin∣) + ω3·E, where tduration=tend-tbegin, weight coefficients ω1∈[0.4, 0.6], ω2∈[0.3, 0.5], ω3∈[0.1, 0.2];

[0115] Result generation step: arrange the candidate parking spaces in descending order of matching degree S, generate a matching result list and push it to the user terminal.

[0116] In this embodiment: 1. User-defined encryption level

[0117] User terminal module: provide primary / intermediate / advanced encryption options:

[0118] Primary encryption: use lightweight homomorphic encryption (such as Paillier simplified version), processing time ≤200ms, only encrypt coordinate numerical part;

[0119] Advanced encryption: use full homomorphic encryption (such as BGV scheme), processing time ≤500ms, encrypt all parameters (including time window, coordinate).

[0120] 2. Server-side hierarchical processing logic

[0121] Privacy protection module: dynamically adjust the ciphertext calculation complexity according to the user's selected encryption level. For example, under primary encryption, directly process the distance square sum of coordinate ciphertext (avoid square root), under advanced encryption, complete complex distance correction operations through secure multi-party computation (MPC), ensure privacy and performance balance.

[0122] For details, please refer to Figures 1-2 It also includes a dynamic parameter adjustment step: according to historical matching data, optimize weight coefficients ω1, ω2, ω3 through machine learning algorithm, optimization goal is to maximize user reservation success rate = matching result display times successful reservation times × 100%, and success rate threshold is set to ≥70%.

[0123] In this embodiment: parameter input and correction: user input R=300 meters, k default 1.2, expected time window [15:00, 17:00] (Δt=15 minutes extended to [14:45, 17:15]), T=2 hours; Parking space A available time [14:30, 18:00], straight-line distance D=250 meters, D'=1.2×250=300 meters (just meet D'≤R).

[0124] Space-time screening and matching degree calculation: the time overlap length of parking space A is 14:45-17:00 (2.25 hours ≥ 2 hours), which is included in the candidate; when calculating the matching degree, ω1, ω2, and ω3 are dynamically adjusted by machine learning (for example, the current weights are 0.5, 0.3, and 0.2).

[0125] Privacy processing and result feedback: the user selects intermediate encryption, the server calculates D' for the coordinate ciphertext, returns the encrypted matching result, and the terminal displays the distance of parking space A as 300 meters, the available time as 14:30-18:00, and the matching degree S = 0.85 after decryption.

[0126] For details, please refer to Figures 1-2 The parameter input step also includes user privacy parameter setting, supports user selection of encryption level (primary / intermediate / advanced), corresponds to different homomorphic encryption calculation complexity, and the primary encryption processing time is ≤200 milliseconds and the advanced encryption processing time is ≤500 milliseconds.

[0127] In this embodiment: real-time correction of distance and time constraints: dynamically adapt urban road characteristics and user time flexibility through k and Δt parameters to improve matching accuracy;

[0128] Data-driven weight optimization: continuously iterate matching strategies based on reservation success rate to adapt to different user group preferences;

[0129] Hierarchical privacy protection: balance user privacy needs and system response speed to meet data security requirements in different scenarios.

[0130] Finally, it should be noted that: the above only describes the preferred embodiments of the present application and is not intended to limit the present application. Although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent replacements to some technical features. Any modification, equivalent replacement, improvement, etc. within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A time-space constraint based parking space sharing matching system, characterized in that, Comprise: User terminal module for inputting user parking demand parameters, including: departure coordinates (x1, y1), destination coordinates (x2, y2), expected parking duration T, expected parking time window [tstart, tend], and collecting user real-time position coordinates (xu, yu); Parking space information collection module for collecting parking space state parameters, including: parking space position coordinates (xp, yp), available time window [tbegin, tend], parking space type (indoor / outdoor), parking space size, and charging standard P; The data processing server includes: a spatiotemporal constraint modeling module, used to establish spatial constraint index D≤R (where D is the straight-line distance between the parking space and the destination, and R is the user-defined spatial radius threshold) and time constraint index. Matching algorithm module for calculating matching degree S = ω1·(1-RD) + ω2·(1-tduration∣tstart-tbegin∣) + ω3·E, wherein ω1+ω2+ω3=1 is the weight coefficient, tduration is the parking space available time length, and E is the historical evaluation score (0-5 points); Privacy protection module for homomorphic encryption processing of user coordinates (xu, yu) and parking time window [tstart, tend].

2. The system of claim 1, wherein, The spatial radius threshold R supports user self-defined setting, and the default value is 500 meters, and the system supports correcting the straight-line distance D to the actual driving distance D' according to the urban road network, and the correction formula is D' = k·D, wherein k is the road traffic coefficient (1≤k≤1.5).

3. The system of claim 1, wherein, The time constraint index further comprises a buffer time parameter Δt, which is used to expand Δt (default 15 minutes) at both ends of the user's expected parking time window to form a matching time window [tstart-Δt, tend+Δt], and requires that the parking space available time window and the matching time window overlap for a length of ≥T.

4. A parking space sharing matching method based on space-time constraints, characterized in that, Comprise the following steps: parameter Input step: receiving the departure coordinates (x1, y1), destination coordinates (x2, y2), expected parking duration T, and expected parking time window [tstart, tend] input by the user, and the parking space position coordinates (xp, yp), available time window [tbegin, tend], and charging standard P input by the parking space owner; The spatio-temporal constraint screening step: calculate the distance D between the parking space and the destination = (xp-x2)2+(yp-y2)2, screen the candidate parking spaces that satisfy D≤R and ​ Matching degree calculation step: calculating the matching degree S = ω1·(1-RD) + ω2·(1-tduration∣tstart-tbegin∣) + ω3·E of the candidate parking space, wherein tduration = tend-tbegin, the weight coefficients ω1∈[0.4, 0.6], ω2∈[0.3, 0.5], and ω3∈[0.1, 0.2]; Result generation step: arranging the candidate parking spaces in descending order of matching degree S, generating a matching result list and pushing it to the user terminal.

5. The method of claim 4, wherein, Further comprising a dynamic parameter adjustment step: optimizing the weight coefficients ω1, ω2, and ω3 through machine learning algorithm according to historical matching data, and the optimization goal is to maximize the user reservation success rate = matching result display times × successful reservation times × 100%, and the success rate threshold is set to ≥70%.

6. The method of claim 4, wherein, The parameter input step further includes user privacy parameter setting, supporting user selection of encryption level (primary / intermediate / advanced), corresponding to different homomorphic encryption calculation complexity, primary encryption processing time ≤ 200 milliseconds, and advanced encryption processing time ≤ 500 milliseconds.

Images