An intelligent garage vehicle re-allocation method and system based on a cell congestion risk field

Abstract

The application provides an intelligent garage vehicle re-distribution method and system based on a cell congestion risk field, and belongs to the technical field of garage management. The method comprises the following steps: determining a comprehensive congestion risk value of each cell based on attribute information of a plurality of cells in an intelligent garage, and further determining a congestion risk quantification feature of the intelligent garage; if the congestion risk quantification feature of the intelligent garage meets a preset triggering condition, screening, from a plurality of parked vehicles in the intelligent garage, a vehicle whose comprehensive congestion risk value of the occupied cell is greater than a preset first risk threshold, as a vehicle to be re-distributed; constructing a plurality of feasible pairs, each feasible pair comprising a vehicle to be re-distributed and a candidate target cell of the vehicle to be re-distributed; calculating a matching cost corresponding to each feasible pair, and determining a target parking cell corresponding to each vehicle to be re-distributed, so as to realize vehicle re-distribution. The application can reduce the local congestion probability and improve the overall vehicle access operation efficiency of the intelligent garage.

Description

Technical Field

[0001] This application belongs to the field of garage management technology, and more specifically, it relates to an intelligent garage vehicle redistribution method and system based on cellular congestion risk field. Background Technology

[0002] With the increasing application of multi-level parking garages and automated parking systems, the impact of vehicle layout within the garage on system operating efficiency is becoming more and more significant. Existing parking garage scheduling methods can typically achieve automatic vehicle storage, automatic vehicle retrieval, and shortest path-based movement control. However, in scenarios with high occupancy rates, frequent vehicle retrieval, or high load at entrances and exits, problems such as long-term congestion near entrances and exits, occupation of key passages, long-term occupation of buffer zones, and repeated obstruction of vehicle retrieval paths are still likely to occur. Summary of the Invention

[0003] The purpose of this application is to provide a method and system for intelligent parking garage vehicle redistribution based on cellular congestion risk field, in order to solve problems such as long-term congestion near entrances and exits, occupation of key passages, long-term occupation of buffer zones, and repeated obstruction of vehicle retrieval paths in existing parking garages.

[0004] A first aspect of this application provides a vehicle reallocation method for smart garages based on a cellular congestion risk field, applied to a smart garage. The smart garage includes a two-dimensional discrete space composed of M rows and N columns of minimum schedulable spatial units, each of which is a cell, comprising: The comprehensive congestion risk value of each cell is determined based on the attribute information of multiple cells within the intelligent parking garage. The congestion risk quantification characteristics of the intelligent parking garage are determined based on the comprehensive congestion risk value of each cell; If the congestion risk quantification characteristics of the smart garage meet the preset triggering conditions, then from the multiple parked vehicles in the smart garage, vehicles whose comprehensive congestion risk value of the occupied cells is greater than the preset first risk threshold are selected as vehicles to be reassigned. For each vehicle to be reassigned, based on the vehicle's pick-up frequency and the distance from each candidate cell to the entrance / exit, the predicted pick-up frequency risk of the vehicle moving to each candidate cell is calculated. Based on the predicted pick-up frequency risk and the comprehensive congestion risk value corresponding to each candidate cell, the predicted comprehensive congestion risk value corresponding to each candidate cell is determined. The risk difference between the comprehensive congestion risk value of the currently occupied cell of the vehicle to be reassigned and the predicted comprehensive congestion risk value corresponding to each candidate cell is calculated. From multiple candidate cells, cells whose corresponding predicted comprehensive congestion risk value is less than a preset second risk threshold and whose corresponding risk difference is greater than a preset difference threshold are selected as candidate target cells for the vehicle to be reassigned, and multiple feasible pairings are constructed. Each feasible pairing includes the vehicle to be reassigned and one candidate target cell of the vehicle to be reassigned. The candidate cell is a cell whose corresponding operating status information is idle and allows vehicle entry, and the second risk threshold is less than the first risk threshold. For each feasible pair, the cost required to move the vehicle to be reassigned in the feasible pair to the candidate target cell in the feasible pair is calculated to obtain the matching cost corresponding to the feasible pair; based on the matching cost corresponding to each feasible pair, the target parking cell corresponding to each vehicle to be reassigned is determined; based on the target parking cell corresponding to each vehicle to be reassigned, an executable movement path is generated for each vehicle to be reassigned, and each vehicle to be reassigned is moved to the corresponding target parking cell along the executable movement path to realize the reassignment of each vehicle to be reassigned.

[0005] A second aspect of this application provides a smart garage vehicle reallocation system based on a cellular congestion risk field, applied to a smart garage. The smart garage includes a two-dimensional discrete space composed of M rows and N columns of minimum schedulable space units, each of which is a cell, comprising: The risk perception module is used to determine the comprehensive congestion risk value of each cell based on the attribute information of multiple cells in the intelligent garage. The risk calculation module determines the quantification characteristics of the congestion risk of the intelligent garage based on the comprehensive congestion risk value of each cell; The first screening module is used to select vehicles whose comprehensive congestion risk value of the occupied cells is greater than a preset first risk threshold from multiple parked vehicles in the smart garage when the congestion risk quantification characteristics of the smart garage meet the preset triggering conditions, and then select them as vehicles to be reassigned. The second screening module, for each vehicle to be reassigned, calculates the predicted retrieval frequency risk of the vehicle moving to each candidate cell based on the vehicle's retrieval frequency and the distance from each candidate cell to the entrance / exit. It then determines the predicted comprehensive congestion risk value for each candidate cell based on the predicted retrieval frequency risk and the comprehensive congestion risk value. The module calculates the risk difference between the comprehensive congestion risk value of the currently occupied cell and the predicted comprehensive congestion risk value of each candidate cell. From multiple candidate cells, it selects cells whose predicted comprehensive congestion risk value is less than a preset second risk threshold and whose corresponding risk difference is greater than a preset difference threshold, as candidate target cells for the vehicle to be reassigned, and constructs multiple feasible pairs. Each feasible pair includes the vehicle to be reassigned and one candidate target cell for the vehicle. The candidate cell is a cell whose corresponding operating status information is idle and allows vehicle entry, and the second risk threshold is less than the first risk threshold. The vehicle reallocation module is used to calculate, for each feasible pair, the cost required to move the vehicle to be reallocated in the feasible pair to the candidate target cell in the feasible pair, and obtain the matching cost corresponding to the feasible pair; based on the matching cost corresponding to each feasible pair, determine the target parking cell corresponding to each vehicle to be reallocated; based on the target parking cell corresponding to each vehicle to be reallocated, generate an executable movement path for each vehicle to be reallocated, and move each vehicle to be reallocated to the corresponding target parking cell along the executable movement path, so as to realize the reallocation of each vehicle to be reallocated.

[0006] A third aspect of this application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the steps of the above-described intelligent garage vehicle reallocation method based on a cellular congestion risk field.

[0007] A fourth aspect of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the above-described intelligent garage vehicle reallocation method based on a cellular congestion risk field.

[0008] The beneficial effects of the intelligent garage vehicle reallocation method and system based on cellular congestion risk field provided in this application are as follows: This application embodiment models the space of a smart parking garage as a cellular network, incorporating the congestion risk distribution, key passage occupancy status, and multi-vehicle execution conflicts within the garage into garage management. The comprehensive congestion risk of each cell is calculated on the cellular network, forming a cellular congestion risk field. Then, using the comprehensive congestion risk value of each cell as a unified decision variable, the redistribution trigger, vehicle screening, target screening, vehicle-target matching, and path generation are sequentially completed. Ultimately, this achieves the relocation of vehicles parked in high-congestion-risk areas, thereby reducing the probability of localized congestion and improving the overall parking and retrieval efficiency of the smart parking garage. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 A flowchart illustrating a smart garage vehicle reallocation method based on a cellular congestion risk field, provided as an embodiment of this application; Figure 2 A flowchart illustrating a smart garage vehicle reallocation method based on a cellular congestion risk field, provided as another embodiment of this application; Figure 3 A structural block diagram of an intelligent garage vehicle redistribution system based on a cellular congestion risk field provided in one embodiment of this application; Figure 4 This is a schematic block diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0011] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0012] To make the objectives, technical solutions, and advantages of this application clearer, the following description will be provided in conjunction with the accompanying drawings and specific embodiments.

[0013] Please refer to Figure 1 , Figure 1A flowchart illustrating a vehicle reallocation method for a smart garage based on a cellular congestion risk field, provided in one embodiment of this application, is shown. This method is applied to a smart garage, which comprises a two-dimensional discrete space consisting of M rows and N columns of minimum schedulable spatial units, each of which is considered a cell. The method can be executed by an electronic device and includes: S101: Determine the comprehensive congestion risk value of each cell based on the attribute information of multiple cells in the intelligent parking garage.

[0014] In this embodiment, a cell, also called a cellular element, is the most basic and smallest unit in a cellular automaton. A large number of cells are arranged in a grid according to certain rules, commonly in one-dimensional linear or two-dimensional grid layouts. Each cell has an independent state (attribute information), such as alive, dead, stationary, or moving. The state is usually simply distinguished by numbers or colors. Each cell is affected by its neighboring cells and updates its own state synchronously according to fixed evolutionary rules. Complex overall phenomena can emerge through local interactions.

[0015] This embodiment primarily targets grid-type intelligent parking garages or puzzle-based grid-dense storage systems (PBS-type grid parking garages), where each location has a clearly defined adjacency relationship and passage function. Therefore, this embodiment draws upon the control concept of cellular automata mentioned above, uniformly modeling the intelligent parking garage space as a cellular network. Each cell corresponds to a minimum schedulable space unit, which can serve as a vehicle parking location, vehicle movement transition location, entrance / exit location, obstacle location, or temporary buffer location. During the movement of a vehicle or vehicle platform within the discrete grid, each movement is only permitted from the current cell to adjacent cells in the four directions of up, down, left, and right, without considering continuous vehicle steering, diagonal movement, changes in heading angle, or curved trajectory tracking.

[0016] The garage cell set C can be represented as: ; Each cell's attribute information includes at least the following: location coordinates, space type, operating status, occupied vehicles, passable identifier, adjacency set, risk value, functional zoning, and continuous occupancy duration. Among these, space type distinguishes between ordinary parking cells, passage cells, entrance / exit cells, buffer cells, and obstacle cells; operating status indicates whether it is idle, occupied, waiting, blocked, or impassable; continuous occupancy duration determines whether a buffer cell or a turnover cell is occupied for an extended period; for any cell c, its passable adjacency set is defined as: ; in, This represents the Manhattan distance between cells c and c′. The definition indicates that cell c′ allows vehicles to enter or pass through. This definition clarifies that vehicles can only move between adjacent cells in the vertical, horizontal, and vertical directions, which conforms to the motion characteristics of grid-type garages and PBS-type grid garages.

[0017] Furthermore, considering that in a grid-type smart parking garage or a PBS-type grid parking garage, whether a location is suitable for long-term parking depends not only on its vacancy status but also on its proximity to entrances / exits, the density of surrounding vehicles, its location in an area where multiple paths converge, whether it obstructs the movement of other vehicles, and whether the vehicle occupying that location is a frequently accessed vehicle. Therefore, relying solely on distance or vacancy status as the basis for judgment is insufficient to meet the overall operational efficiency optimization requirements of the parking garage.

[0018] To address the aforementioned issues, this embodiment incorporates the congestion risk distribution, critical passage occupancy status, and multi-vehicle execution conflicts within the garage into garage management, based on the aforementioned cell attribute information. The comprehensive congestion risk of each cell is calculated on the cell network, forming a cell congestion risk field. Subsequently, vehicles within the smart garage can be redistributed based on this cell risk field. That is, based on the cell congestion risk field, the positions of existing vehicles within the smart garage are adjusted, rearranged, and optimized. This process differs from single-stage vehicle allocation and single-vehicle retrieval path planning; its focus is on reorganizing, rebalancing, and optimizing the layout of already parked vehicles.

[0019] For example, for each cell, the comprehensive congestion risk value of that cell is determined based on the attribute information of multiple cells within the smart garage, including: (1) Based on the location information of the cell, determine the congestion risk caused by the cell being close to the entrance / exit, which is taken as the entrance / exit distance risk of the cell, and expressed by the following formula: ; in, This represents the risk of the entrance / exit distance of cell c at time t. This represents the Manhattan distance from cell c to the nearest entrance / exit. This represents the maximum Manhattan distance from the accessible cell to the entrance / exit in the smart garage. The location information of cell c is its row and column coordinates in the two-dimensional grid. Similarly, the location information of the nearest entrance / exit is its row and column coordinates in the two-dimensional grid.

[0020] (2) Based on the idle / occupied flag of each cell, count the number of occupied cells in the k-level neighborhood of that cell; based on the passable / impeded flag of each cell, count the number of passable cells in the k-level neighborhood of that cell; based on the number of occupied cells in the k-level neighborhood of that cell and the number of passable cells in the k-level neighborhood of that cell, determine the vehicle density around that cell, which is used as the neighborhood occupancy risk of that cell, expressed by the following formula: ; in, This represents the neighborhood occupancy risk of cell c at time t. This represents the number of occupied cells in the k-th layer neighborhood of cell c. This represents the number of traversable cells within the k-th layer neighborhood of cell c.

[0021] In this embodiment, the k-layer neighborhood of cell c refers to the set of all adjacent cells that fall within the range of k concentric circles centered on cell c, extending upwards, downwards, leftwards, and rightwards. For example: k=1 (1-layer neighborhood, first-order neighborhood): only the four adjacent parking spaces directly above, below, left, and right of the current parking space, i.e., the immediate adjacent parking spaces that can be reached in one step; k=2 (2-layer neighborhood, second-order neighborhood): based on the 1-layer neighborhood, extending outwards another circle, all parking spaces that can be reached in at most two steps from the central parking space; and so on. The larger k is, the larger the surrounding range is considered. Here, k can be set according to the garage size, occupancy rate, and scheduling accuracy. In this embodiment, k=2 is set.

[0022] (3) Based on the location information of the cell, determine the number of candidate access routes passing through the cell, which is used as the path criticality risk of the cell, expressed by the following formula: ; in, This represents the path criticality risk of cell c at time t. This represents the number of paths that pass through cell c in the candidate vehicle access path; This represents the total number of candidate access routes at time t.

[0023] It should be noted that the candidate parking and retrieval path is calculated in real time by the candidate parking and retrieval task. The candidate parking and retrieval task is a regular scheduling task in the garage management system that is independent of the vehicle reallocation task in this embodiment. It is usually initiated by the user. In contrast, the vehicle reallocation task in this embodiment is a globally optimized scheduling task. The reallocation scheduling task will not interfere with the normal user parking and retrieval business. The two tasks run in parallel and the paths are calculated separately to prevent path contention and instruction conflicts from occurring during the reallocation of vehicles and user parking / retrieval.

[0024] (4) Based on the cell's location information, determine the degree of impact on the movement of other vehicles after the cell is occupied, which is taken as the cell's blocking risk, expressed by the following formula: ; in, This represents the blocking risk of cell c at time t. This represents the number of vehicles affected by movement due to cell c being occupied. This refers to the collection of vehicles within a smart garage. This represents the total number of vehicles in the vehicle set.

[0025] For example, The method for determining L0 is as follows: For any vehicle v in the garage other than the vehicle occupying cell c, calculate the baseline path length L0(v,t) from vehicle v to the entrance / exit or target location when cell c is passable, and the updated path length Lc(v,t) from vehicle v to the entrance / exit or target location when cell c is occupied or impassable. If Lc(v,t) does not exist, or Lc(v,t) is not found... If L0(v,t) is greater than a preset step threshold (e.g., 4), or the path length increases by a certain percentage (Lc(v,t)). If L0(v,t) / L0(v,t) is greater than a preset increase threshold (e.g., 20%), then vehicle v is determined to be affected by the movement of cell c. Furthermore, the total number of vehicles affected by obstruction within the garage is counted to obtain... .

[0026] (5) Calculate the matching deviation between the cell's location information and the vehicle retrieval frequency, which is taken as the vehicle retrieval frequency risk of the cell, expressed by the following formula: ; in, Let represent the vehicle retrieval frequency risk of cell c at time t, and let represent the occupancy status of cell c at time t. This indicates that cell c is occupied at time t. This indicates that cell c is idle at time t; This indicates the historical vehicle retrieval frequency or vehicle retrieval priority of cell c.

[0027] (6) The comprehensive congestion risk value of the cell is obtained by weighting and summing the risks of entrance / exit distance, neighborhood occupancy, path criticality, obstruction, and vehicle retrieval frequency, as expressed by the following formula: ; in, This represents the overall congestion risk value of cell c at time t. , , , , For weighting coefficients that are greater than or equal to zero, and .

[0028] It should be noted that this embodiment does not propose several independent risk indicators, but rather maps the above five risk factors into a unified cellular comprehensive risk value. Subsequent steps will prioritize calling [the appropriate method / mechanism]. If necessary, then according to , , , and The specific percentages explain the sources of risk, thereby ensuring consistency in the decision-making logic throughout the text.

[0029] S102: Determine the quantitative characteristics of congestion risk in smart parking garages based on the comprehensive congestion risk value of each cell.

[0030] In this embodiment, vehicle reallocation should not be initiated unconditionally at any time. If the system frequently adjusts existing vehicles, it may increase the number of steps required, interfere with normal vehicle access tasks, and cause additional disturbances. Therefore, this embodiment uniformly bases the reallocation trigger judgment on whether the risk of a cell occupied by a vehicle is too high. Specifically, the congestion risk quantification characteristics of the smart garage can be determined based on the comprehensive congestion risk value of each cell, and then the judgment on whether the risk of a cell occupied by a vehicle is too high can be based on the congestion risk quantification characteristics.

[0031] For example, determining the quantification characteristics of congestion risk in a smart parking garage based on the comprehensive congestion risk value of each cell may include: From the multiple cells of the smart garage, select the cells occupied by vehicles and treat them as multiple used cells; The total congestion risk of the smart garage is obtained by summing the combined congestion risk values ​​of multiple used cells. The total amount of congestion risk is used as a quantitative characteristic of congestion risk in smart parking garages.

[0032] The calculation process for the total congestion risk mentioned above is expressed by the following formula: ; in, This represents the total congestion risk of the smart garage, which is also the total risk of cells currently occupied by vehicles in the smart garage. This represents the occupancy state of cell c at time t. This indicates that cell c is occupied at time t. This indicates that cell c is idle at time t.

[0033] The total congestion risk can characterize the total risk of cells occupied by vehicles in the current smart garage. A high number indicates that the vehicles occupy a significant amount of high-risk space, necessitating a redistribution of the garage layout.

[0034] Furthermore, in addition to calculating the total congestion risk, determining the quantitative characteristics of congestion risk for a smart parking garage based on the comprehensive congestion risk value of each cell may also include: Based on the comprehensive congestion risk value, the multiple cells of the smart garage are divided into risk levels, resulting in high-risk cells, medium-risk cells, low-risk cells, and stable low-risk cells. Calculate the number of high-risk cells, and the proportion of cells that are occupied for a long time among high-risk and medium-risk cells; cells that are occupied for a long time refer to cells that have been continuously occupied for a duration greater than a preset time threshold. The total congestion risk and the proportion of cells that are occupied for a long time in high-risk and medium-risk cells are used as quantitative characteristics of congestion risk in smart parking garages.

[0035] In this embodiment, cells can be dynamically functionally partitioned based on their comprehensive congestion risk value, allowing cells with different comprehensive congestion risk values ​​to undertake different scheduling functions. This transforms the risk field from an evaluation result into spatial usage rules. Specifically, a threshold can be set. Cell functional regions are divided according to the following formula:

[0036] Furthermore, based on the comprehensive congestion risk value, the multiple cells of the smart parking garage can be classified into risk levels, resulting in high-risk cells, medium-risk cells, low-risk cells, and stable low-risk cells, as shown in Table 1 below:

[0037] The key point of the above zoning rules is: the higher the risk, the less suitable it is for long-term parking; the lower the risk, the more suitable it is for... Final parking target. If a cell is due to... A cell that is relatively high-risk is considered high-risk because it is close to an entrance or exit and should not be targeted. Long-term occupation; if due to or A higher-risk cell indicates that it has a critical passage or obstruction effect. It is not suitable as a long-term parking target due to noise; if because The fact that a cell is at a relatively high risk level, thus becoming a medium-to-high risk cell, indicates that it is located in a localized area. In densely populated areas, the system should restrict the continued allocation of vehicles to that area.

[0038] Therefore, cellular functional zoning is no longer an independent rule, but a direct application of the risk field. When vehicles occupy turnover cells, buffer cells, or cells affected by entrances and exits for extended periods, the system can, according to... Include them in the vehicle set to be redistributed to free up passage and avoidance resources.

[0039] Based on this, the number of high-risk cells can be counted, expressed by the following formula: ; in, This indicates the number of high-risk cells.

[0040] Furthermore, the number of cells that are occupied for a long time in high-risk and medium-risk cells can be counted, and this number can be divided by the total number of high-risk and medium-risk cells to obtain the proportion of cells that are occupied for a long time in high-risk and medium-risk cells. .

[0041] Ultimately, the total congestion risk, the number of high-risk cells, and the proportion of cells that are occupied for a long time among high-risk and medium-risk cells will be used as the quantitative characteristics of congestion risk for smart parking garages.

[0042] S103: If the congestion risk quantification characteristics of the smart garage meet the preset triggering conditions, then select vehicles whose comprehensive congestion risk value of the occupied cells is greater than the preset first risk threshold from the multiple parked vehicles in the smart garage, and use them as vehicles to be reassigned.

[0043] In this embodiment, if the total congestion risk is used as the quantitative characteristic of the congestion risk of the smart garage, the preset triggering condition can be: the total congestion risk of the smart garage is greater than the preset total risk threshold.

[0044] If the total congestion risk, the number of high-risk cells, and the proportion of cells that are occupied for a long time among high-risk and medium-risk cells are used together as the quantitative characteristics of congestion risk in a smart parking garage, then the preset trigger condition can be any of the following conditions: ; or ; or ; in, This indicates the preset total risk threshold. This indicates a preset quantity threshold. This represents a preset ratio threshold, which can be flexibly designed by those skilled in the art according to actual needs. , and The specific value to be taken.

[0045] The triggering conditions mentioned above are all derived from the cellular risk field. For example, if the triggering cause mainly comes from... A high number of occupied cells indicates that the area affected by the entrance / exit needs to be released; if the triggering cause mainly comes from... or A high number of occupied cells indicates that critical paths or blocking positions need to be released; if the triggering cause mainly comes from... Higher concentrations in certain areas indicate that the local vehicle density needs to be alleviated.

[0046] It should be noted that after satisfying the necessity of reallocation, it is necessary to further determine whether a safe execution window is currently available. The execution window should at least meet the following conditions: there is no urgent vehicle retrieval task, the garage is in a low-load or task-interval state, and there are available idle cells. These conditions can be obtained from the vehicle storage and retrieval task calculation module.

[0047] After the risk field is generated and the reallocation process is triggered, vehicles to be reallocated can be selected from the parked vehicles. Unlike directly listing multiple independent selection criteria, this invention prioritizes the comprehensive risk value of the vehicle's current location as the basis for vehicle selection.

[0048] Current position risk of any vehicle v It can be represented as: ; in, This represents the cells occupied by vehicle v at time t.

[0049] A vehicle may be included in the set of vehicles to be reassigned if the risk of its current location meets the following condition: ; in, This indicates the preset first risk threshold.

[0050] Therefore, the vehicles to be reassigned It can be represented as: .

[0051] To improve the interpretability of the screening results, the system can also classify vehicles based on the main sources of risk at their current location: (1) If The high level indicates that vehicles are occupying the entrance / exit area and should be released first. (2) If A high reading indicates that the vehicle is in a densely populated area; removing the vehicle will help reduce local congestion. (3) If A high value indicates that the vehicle is occupying a critical path or path convergence cell, and long-term parking should be avoided. (4) If A high value indicates that the vehicle is obstructing other vehicles from moving or retrieving their vehicles, and should be adjusted first. (5) If A high number indicates that the frequency of vehicle retrieval is not matched with the current location. For example, vehicles that are frequently retrieved may be parked too deep and should be moved to a more suitable area.

[0052] Therefore, this invention does not reset an isolated set of indicators during the vehicle screening stage, but rather uses the vehicle's current location... A unified judgment is made, and then five risk factors are used to explain why a vehicle needs to be removed. This method can ensure that the selection of vehicles to be reassigned is consistent with the aforementioned risk field model.

[0053] S104: For each vehicle to be reassigned, based on the vehicle's pick-up frequency and the distance from each candidate cell to the entrance / exit, calculate the predicted pick-up frequency risk of the vehicle moving to each candidate cell. Based on the predicted pick-up frequency risk and the comprehensive congestion risk value corresponding to each candidate cell, determine the predicted comprehensive congestion risk value corresponding to each candidate cell. Calculate the risk difference between the comprehensive congestion risk value of the currently occupied cell of the vehicle to be reassigned and the predicted comprehensive congestion risk value corresponding to each candidate cell. From multiple candidate cells, select cells whose corresponding predicted comprehensive congestion risk value is less than a preset second risk threshold and whose corresponding risk difference is greater than a preset difference threshold, as candidate target cells for the vehicle to be reassigned, and construct multiple feasible pairs. Each feasible pair includes the vehicle to be reassigned and one candidate target cell of the vehicle to be reassigned. The candidate cell is a cell whose corresponding operating status information is idle and allows vehicle entry, and the second risk threshold is less than the first risk threshold.

[0054] In this embodiment, candidate target cells should not only meet the condition of being "idle", but also meet requirements such as being accessible and reachable. Specifically, the screening of target cells can be divided into two levels: hard constraints and risk constraints. Hard constraints are used to determine whether a cell possesses the basic qualifications to be the final parking target. The basic candidate set is determined based on hard constraints. It can be represented as: ; in, This indicates that the candidate cell is currently in an idle state. This indicates that candidate cells allow vehicles to enter or pass through. This indicates that there is an executable path from the vehicle's current location to the candidate cell.

[0055] Based on satisfying the hard constraints, for any vehicle to be redistributed and candidate cells Obtain vehicles to be redistributed At any moment Car pick-up frequency and obtain candidate cells Distance to entrance / exit Based on the vehicle retrieval frequency of the vehicles to be reassigned and the normalized distance from the candidate cells to the entrances / exits, the vehicles to be reassigned are calculated. Candidate cells to be transferred Risk of predicted car pick-up frequency afterward:

[0056] in, Indicates the hypothetical vehicle move in The resulting risk of predicted car pick-up frequency Indicates vehicle At any moment The frequency of car pick-up Indicates candidate cells Distance to entrance / exit E This represents the maximum distance from all valid parking cells in the garage to the entrance / exit.

[0057] While keeping other risk components such as entrance / exit impact risk, neighborhood occupancy risk, path criticality risk, and obstruction risk of candidate cells unchanged, the current vehicle retrieval frequency risk of candidate cells is replaced with the predicted vehicle retrieval frequency risk after the vehicles to be reallocated move in, resulting in the vehicle-cell predicted comprehensive congestion risk:

[0058] In the formula, Indicates vehicles to be reassigned Transferring into basic candidate cells The subsequent prediction of comprehensive congestion risk; Represents the basic candidate cell At any moment The current overall congestion risk; Represents the basic candidate cell Risk associated with current vehicle pick-up frequency; This indicates the weight of vehicle retrieval frequency risk in the overall congestion risk. This means first deducting the risk contribution from the old vehicle pickup frequency, and then adding the vehicle. The corresponding predicted contribution is used to avoid double counting.

[0059] Due to the basic candidate cells The idle state condition is met, therefore: , ; Therefore, regarding vehicles Any candidate cell The predicted comprehensive congestion risk value can be simplified as follows: ; Furthermore, it can be based on each candidate cell Predicted comprehensive congestion risk value to screen vehicles Candidate target cells:

[0060]

[0061] in, Indicates vehicle The overall congestion risk value of the currently occupied cells, Indicates the risk difference. This indicates the preset second risk threshold. This indicates the preset difference threshold.

[0062] Therefore, the same candidate cell can predict different comprehensive congestion risks for vehicles to be redistributed with different pick-up frequencies, thus distinguishing the problem of different vehicle pick-up frequencies.

[0063] This embodiment can pre-screen any vehicle to be reallocated by using basic constraints and predicting comprehensive congestion risk. The candidate target cells can be used to construct multiple feasible pairs; each feasible pair includes the vehicle to be reassigned and a candidate target cell of the vehicle to be reassigned.

[0064] If the number of vehicles to be reassigned is m, and each vehicle retains an average of n pre-candidate cells, then the maximum number of pairs predicted is m×n. During the calculation, only the actual pick-up frequency of each vehicle to be reassigned is substituted; it is not necessary to traverse all possible pick-up frequencies. Risks related to entrances / exits, neighborhood occupancy, path criticality, obstruction, and distance can be pre-calculated and cached.

[0065] In other implementations, each risk component can be recalculated after the vehicle moves in; when this method is used, the predicted overall congestion risk should be determined based on all the recalculated risk components.

[0066] S105: For each feasible pair, calculate the cost required to move the vehicle to be reassigned in the feasible pair to the candidate target cell in the feasible pair, and obtain the matching cost corresponding to the feasible pair; based on the matching cost corresponding to each feasible pair, determine the target parking cell corresponding to each vehicle to be reassigned; based on the target parking cell corresponding to each vehicle to be reassigned, generate an executable movement path for each vehicle to be reassigned, and move each vehicle to the corresponding target parking cell along the executable movement path to realize the reassignment of each vehicle to be reassigned.

[0067] In this embodiment, let V be the set of vehicles to be reassigned in the current batch. b (t)={v1,v2,…,v m The union of the final candidate target cells of each vehicle to be reassigned is C. T (t)={c1,c2,…,c n}. When n<m, no more than n vehicles can be selected to form the current batch of vehicles to be redistributed in descending order of the current comprehensive congestion risk value of the cell currently occupied by the vehicle, and the vehicles not entering the current batch will continue to be processed in the next safe execution window.

[0068] For each vehicle to be redistributed and its candidate target cell constituting a feasible pair, the normalized moving distance can be calculated according to the moving distance from the current position of the vehicle (i.e., the position of the currently occupied cell) to the candidate target cell:

[0069] In the formula, represents the vehicle the moving distance or number of moving steps required to move from the current position to the candidate target cell ; represents the maximum effective moving distance for normalization under the current garage layout; represents the normalized moving distance, with a value range of [0,1].

[0070] The non-negative matching cost of the feasible pair can be calculated according to the normalized moving distance and the predicted comprehensive congestion risk after the vehicle moves into the candidate target cell:

[0071]

[0072] In the formula, represents the vehicle and the candidate target cell the non-negative matching cost between them; represents the normalized moving distance; represents the vehicle moves into the subsequent predicted comprehensive congestion risk; and respectively represent the non-negative weights of the moving distance term and the predicted comprehensive risk term, and the sum of the two is 1. The smaller the value, the better the feasible pair.

[0073] As an example, when the original distance term weight is 0.40 and the original target risk term weight is 0.30, renormalization can be performed while maintaining the relative ratio of the two as:

[0074] In the formula, and These values, derived from 0.40 / (0.40+0.30) and 0.30 / (0.40+0.30) respectively, are used only to maintain the relative proportions of the original distance term and the target risk term; in actual implementation, they can be determined through simulation experiments, historical operating data, or parameter calibration. and .

[0075] Based on this, for each vehicle With the corresponding candidate target cell The feasible pairings are set with 0-1 decision variables:

[0076] In the formula, Indicates vehicle Whether to be assigned to a candidate target cell When the vehicle Assigned to If the value is 1, then 0; i = 1, 2, ..., m represents the vehicle number in this batch; j = 1, 2, ..., n represents the candidate target cell number.

[0077] The objective is to minimize the total matching cost of all selected pairs in this batch, among all feasible pairings:

[0078] In the formula, Indicates vehicle Assigned to candidate target cells Matching cost; The summation represents the cost of the selected pairing; the double summation represents the total cost of all selected vehicle-cell pairings in this batch; min represents the minimum total cost among all allocation schemes that satisfy the constraints.

[0079] Each vehicle to be redistributed in this batch is selected as exactly one target cell:

[0080] In the formula, for any vehicle The decision variables corresponding to the vehicle and all candidate target cells are summed and limited to 1, thereby ensuring that each vehicle to be reassigned in this batch is assigned to exactly one target cell.

[0081] Each candidate target cell can receive at most one vehicle to be redistributed:

[0082] In the formula, for any candidate target cell The decision variables corresponding to all vehicles and the cell are summed and limited to no more than 1, thereby ensuring that the same target cell in the same batch will not be assigned to two or more vehicles at the same time.

[0083] Vehicle-cell pairs that are not part of the final candidate vehicle set cannot be selected.

[0084] In the formula, Indicates vehicle The candidate target cell set is ultimately obtained by predicting risk and the amount of risk improvement; Represents cell Not a vehicle Feasible goals; implying the symbol " "This means that when the condition is met, it must be made" .

[0085] The above model can be solved using the Hungarian algorithm, the minimum-cost bipartite graph matching algorithm, or the 0-1 integer programming method. After obtaining the global matching result, each assigned target cell is immediately locked, and vehicle reallocation is performed based on conflict-free paths.

[0086] In this embodiment, after matching the vehicle with the candidate target cell, an executable movement path can be generated for each vehicle to be reassigned based on the target parking cell. Since the vehicle or vehicle platform only performs discrete translations between adjacent cells, the path generation is based on the number of movement steps, cell risk, and execution constraints, and does not involve constraints on vehicle turning radius, heading angle, or continuous trajectory curvature.

[0087] Tradition Pathfinding, typically based on movement cost and heuristic distance, can yield shorter paths but struggles to reflect issues in grid-type parking garages such as entrance / exit congestion, localized vehicle density, key lane obstruction, and multi-vehicle execution conflicts. Therefore, this embodiment proposes a risk perception mechanism. The path search algorithm, also known as the path generation algorithm, calls the aforementioned cell congestion risk field during the path generation phase to calculate the comprehensive congestion risk of the candidate next cell q. and potential conflict indicators By introducing a path evaluation function, path search can simultaneously consider movement cost, heuristic distance, cell risk, and execution safety.

[0088] Specifically, the vehicle redistribution path can be represented as: ; in, This represents the vehicle's current location (i.e., the starting point). The target cell must be an adjacent path node that satisfies the following conditions: ; Path generation should meet the following constraints: (1) At the same scheduling time, the same cell can be occupied by at most one vehicle; (2) The next cell must be in an idle, passable, or allowed short-term waiting state; (3) The path should avoid high-risk cells as much as possible, and the risk penalty should be increased when it must pass through them in a short time. (4) Buffer cells are only allowed to be occupied for a short time. When the threshold is exceeded, release, replanning or execution order adjustment should be triggered. (5) After the vehicle moves, it shall not cause any new obstruction exceeding the preset limit; (6) Two vehicles are prohibited from occupying the same cell at the same time, and two vehicles are also prohibited from exchanging in opposite directions at adjacent times.

[0089] Accordingly, at each stage of generating an executable movement path, a path evaluation function (risk perception) is used. The path evaluation function evaluates each candidate target cell in the next stage. For any candidate target cell q, the path evaluation function... for: ; in, This represents the cumulative number of steps the vehicle to be reassigned has taken from the starting point to the candidate target cell q. This represents the number of steps required for the vehicle to be reassigned to move from the candidate target cell q to the target parking cell. This represents the overall congestion risk value of candidate target cell q. ρ indicates whether the candidate cell q has the risk of occupancy conflict, reciprocal exchange conflict or short-term inaccessibility, and μ and μ are the risk penalty coefficient and conflict penalty coefficient, respectively.

[0090] In the above formula, The specific value can be determined based on the running status information of the candidate target cell q. When candidate cell q is occupied by another vehicle at time t or t+1, is already reserved by another vehicle's route, or when the current vehicle moves from current cell p to q and another vehicle moves from q to p, creating a reciprocal exchange conflict, or when q is a short-term inaccessible cell within the current scheduling window, the value is set to... Otherwise set In the path evaluation function This item is added to apply conflict penalties to candidate cells that have occupancy conflicts, reciprocal exchange conflicts, or short-term inaccessibility risks during the path search process.

[0091] in, The heuristic distance (Manhattan distance) can be used for calculation, and the specific calculation formula is as follows: ; in, Represents the row and column coordinates of the candidate target cell q. This represents the row and column coordinates of the target parking cell.

[0092] During the path generation phase, the system mainly calls... The risks directly related to traffic safety include entrance / exit distance risk, neighborhood occupancy risk, path criticality risk, and obstruction risk. When there are vehicles with pre-booked or frequently picked-up vehicles, the pick-up frequency risk can also be used as an auxiliary factor for path avoidance. Specifically, if the candidate cell q... A higher elevation indicates that the location is close to the entrance / exit, and frequent passage through this area may affect traffic flow at the entrance / exit; if A higher value indicates that it is located in a locally high-occupancy area, and the path passing through this area may increase the spread of congestion; if A higher value indicates that it is a critical channel or path convergence cell, and the path should reduce its occupation of this area; if A high value indicates that the cell is occupied or may affect the movement of other vehicles after a short period of occupation, and the passage penalty should be increased.

[0093] Using the above method, if a path has a shorter number of steps but passes through areas near entrances / exits, high-density areas, critical path convergence cells, or potentially conflicting cells, its path valuation will increase. Conversely, if another path has a slightly longer number of steps but passes through low-risk cells and has fewer execution conflicts, it may be prioritized. Thus, the vehicle redistribution path changes from simply choosing the "shortest number of steps" to a comprehensive path that is "low-risk, low-conflict, and executable."

[0094] It should be noted that this embodiment The pathfinding algorithm is not independent of the aforementioned risk field model, but rather a specific application of the cellular risk field in the path execution phase. This method enables... , , , and pass Continue to participate in the vehicle movement process to achieve seamless integration from risk identification, vehicle release, target matching to path execution.

[0095] As can be seen from the above, this embodiment models the space of the smart garage as a cellular network, and incorporates the distribution of congestion risks, the occupancy status of key passages, and multi-vehicle execution conflicts into the garage management. The comprehensive congestion risk of each cell is calculated on the cellular network to form a cellular congestion risk field. Then, using the comprehensive congestion risk value of the cells as a unified decision variable, the redistribution triggering, vehicle screening, target screening, vehicle target matching, and path generation are completed in sequence. Finally, the parked vehicles in high-congestion-risk areas are relocated, thereby reducing the probability of local congestion and improving the overall parking and retrieval efficiency of the smart garage.

[0096] The method flow of this embodiment is illustrated below with a specific example.

[0097] I. Implementation conditions and initial garage layout The garage has 6 rows and 6 columns, c 13 It is connected to entrance / exit E. Initially, there are 6 vehicles, with v1 parked at c. 34 v2 is parked in c 35 v3 is parked in c 32 v4 is parked in c 62 v5 is parked in c 52 v6 is parked in c 63 The remaining cells are idle. There is currently no vehicle retrieval task, and vehicles v1 and v2 still need to remain parked; the garage is in a task gap, providing a safe window for vehicle reassignment.

[0098] c in Table 2 34 c 35 For high-risk cell occupancy, c 32 For medium-risk occupied cells, c 52 c 62 c 63 For low-risk cell occupancy, c 13 The cells are designated as entry / exit cells, and the others (cells marked "empty") are designated as free cells. In this embodiment, Rc ≥ 0.50 is considered high risk, 0.30 ≤ Rc < 0.50 is considered medium risk, and Rc < 0.30 is considered low risk.

[0099] Table 2 - Initial Layout of 6×6 Garage

[0100]

[0101] Table 4 - Initial Vehicle Information

[0102] II. Calculate the current risk and determine whether to initiate reallocation. At time t, calculate the entrance / exit distance risk, neighborhood occupancy risk, path criticality risk, obstruction risk, and current vehicle retrieval frequency risk for each relevant cell, and then perform a weighted summation. The calculation formulas for each risk component and the overall risk are as follows:

[0103]

[0104]

[0105]

[0106] In the formula, Nocc is the number of occupied cells in the k-level neighborhood of cell c, Npass is the number of passable cells in that neighborhood, Npath is the number of candidate access routes passing through cell c, Npathtotal is the total number of candidate access routes, Naffected is the number of vehicles affected by movement due to cell c being occupied, and |V| is the total number of vehicles. Rc(c,t) represents the comprehensive congestion risk value. In this embodiment, k=2, Npathtotal=12, and |V|=6, with the five weights being 0.25, 0.20, 0.20, 0.20, and 0.15 respectively.

[0107] The risk associated with the distance between the entrance / exit and the cell decreases as the distance from the cell to the entrance / exit increases, and the calculation formula is as follows:

[0108] c 34 The normalized distance to the entrance / exit is 3 / 8, and its entrance / exit distance risk is 0.625; c 35 The normalized distance to the entrance / exit is 4 / 8, and its entrance / exit distance risk is 0.500; c 32 The normalized distance to the entrance / exit is 3 / 8, and its entrance / exit distance risk is 0.625.

[0109]

[0110] The current frequency term for each of the six occupied cells is calculated as "vehicle retrieval frequency × cell-to-entrance distance ratio". Idle cells currently have no vehicles, so their current frequency term is 0. The formula for the current frequency term is:

[0111] In the formula, Occ(c,t) represents the cell occupancy status, with 1 for occupied and 0 for idle; This represents the frequency at which vehicles are currently parked can be retrieved. Let be the normalized distance from the cell to the entrance. Therefore, the current frequency risk of the idle cell is 0, but its other risk components can still be non-zero.

[0112] Table 6 - Calculation of the current frequency term

[0113] Each score in Table 4 is obtained by dividing the original counts from the above formula. Substituting this into the comprehensive risk formula: c 34 (v1): 0.25×0.625+0.20×(2 / 12)+0.20×(12 / 12)+0.20×(5 / 6)+0.15×0.3000=0.601250.

[0114] c 35 (v2): 0.25×0.500+0.20×(1 / 11)+0.20×(11 / 12)+0.20×(5 / 6)+0.15×0.1500=0.515682.

[0115] c 32 (v3): 0.25×0.625+0.20×(2 / 11)+0.20×(4 / 12)+0.20×(2 / 6)+0.15×0.1875=0.354072.

[0116] c 62 (v4): 0.25×0.250+0.20×(2 / 7)+0.20×(2 / 12)+0.20×(1 / 6)+0.15×0.3000=0.231310.

[0117] c 52 (v5): 0.25×0.375+0.20×(3 / 10)+0.20×(2 / 12)+0.20×(1 / 6)+0.15×0.1250=0.239167.

[0118] c 63 (v6): 0.25×0.375+0.20×(2 / 8)+0.20×(1 / 12)+0.20×(1 / 6)+0.15×0.15625=0.217188.

[0119] c 21 (Idle): 0.25×0.625+0.20×(1 / 7)+0.20×(8 / 12)+0.20×(4 / 6)+0=0.451488.

[0120] c 41 (Idle): 0.25×0.375+0.20×(2 / 8)+0.20×(2 / 12)+0.20×(1 / 6)+0=0.210417.

[0121] c55 (Idle): 0.25×0.250+0.20×(1 / 10)+0.20×(1 / 12)+0.20×(0 / 6)+0=0.099167.

[0122] Risk level explanation: As shown in the calculation results, c 34 and c 35 The comprehensive congestion risk values ​​were 0.601250 and 0.515682, respectively, both belonging to high-risk cells; c 32 The overall congestion risk value is 0.354072, which falls into the medium-risk category. 34 and c 35 The high risk is mainly due to the high path criticality risk and obstruction risk, even though it is not adjacent to the entrance or exit.

[0123] (a) Calculating the total risk The total risk is only added to the risk of the cells containing the six parked vehicles; all 29 idle cells are not included in the total risk.

[0124]

[0125] Initial total risk: 0.601250 + 0.515682 + 0.354072 + 0.231310 + 0.239167 + 0.217188 = 2.158669.

[0126] 2.158669 is greater than the total risk threshold of 2.00, thus meeting the vehicle redistribution trigger condition.

[0127] (ii) Check the number of high-risk cells High-risk cells occupied by vehicles include c 34 and c 35 The number of high-risk cells occupied is 2.

[0128] Final assessment: The total risk has reached the trigger threshold, and there is currently no emergency vehicle retrieval task, the garage is in a task gap and there are available spaces, therefore, the safe execution window is entered.

[0129] III. Screening of Vehicles to be Moved Vehicles awaiting relocation are screened based on whether their current location risk reaches the first risk threshold of 0.50.

[0130] In the formula, Vmove(t) is the set of vehicles to be reassigned, and θmove is the first risk threshold, which is 0.50 in this embodiment.

[0131] Table 7 - Screening Results of Six Vehicles

[0132] The selection results were v1 and v2, both of which were added to the waiting set because their current location risk reached the first risk threshold of 0.50. The remaining four vehicles remained in their original positions.

[0133] IV. Selecting vacant spaces and calculating the risks after a vehicle is moved in. System traversal except for entry and exit cells c 13 and the already occupied cell c 32 c 34 c 35 c 52 c 62 and c 63 All 29 cells except those mentioned above. After checking the idle state, enterable state, and path reachability, all 29 cells meet the basic candidate conditions. Therefore, the basic candidate sets corresponding to vehicles v1 and v2 are the same.

[0134]

[0135] In the formula, Represents cell At time t, the system is in an idle state. Represents cell Vehicles are permitted to enter or pass through; Indicates vehicle Able to reach the cell from the current position In this embodiment, except for the entry / exit cells and the six occupied cells, the remaining 29 cells all simultaneously satisfy the above three conditions.

[0136] C base (v i ,t)={c 11 c 12 c 14 c 15 c 16 c 21 c 22 c 23 c 24 c 25 c 26 c 31 c 33 c 36 c 41 c 42 c 43 c 44 c 45 c 46 c 51 c 53 c 54 c 55 c 56 c 61c 64 c 65 c 66}

[0137] (a) First, conduct a pre-screening based on the current risk of vacant positions. The basic candidate vacancies are first pre-screened according to the second risk threshold of 0.30 in this embodiment:

[0138] In the formula, This forms the basic candidate set that meets the hard constraints of being free, accessible, and reachable. For vehicles The pre-candidate set composed of candidate target cells, The second risk threshold is set to 0.30 in this embodiment.

[0139]

[0140] As shown in Table 8, the cell with a current comprehensive congestion risk value not exceeding 0.30 is c. 41 c 51 c 55 c 61 c 64 c 65 and c 66 Therefore, the pre-candidate cell sets for vehicles v1 and v2 are both {c 41 c 51 c 55 c 61 c 64 c 65 c 66}

[0141] Table 9 - Results of Pre-screening of Candidate Cells

[0142] (ii) Calculate all pre-candidate vehicle-cell pairings For the 14 pairs formed by v1, v2 and 7 pre-candidate cells, the actual vehicle retrieval frequency is substituted into each pair to calculate the predicted retrieval frequency risk, the predicted comprehensive risk, and the risk improvement amount:

[0143]

[0144]

[0145]

[0146] In the formula, Indicates the assumption that the vehicle Move into empty space The risk of predicting subsequent car pick-up frequency This indicates a forecast of overall risk. This represents the risk improvement relative to the vehicle's original position. Since the candidate cell is currently idle, the original current vehicle retrieval frequency risk is 0, so a new weighted frequency risk is directly added to the current comprehensive risk of the idle space.

[0147] Table 10 - Predictive Risk Screening Results for All 14 Pairs

[0148] From Table 10, we can see that the final candidate target cell set for v1 is {c 41 c 55 c 65 c 66 The final candidate target cell set of v2 is {c 41 c 55 c 64 c 65 c 66}

[0149] V. Calculate matching costs and complete one-to-one allocation. For an M-row N-column grid-type parking garage, the maximum reference number of moves is Dmax = M + N. 2. In this embodiment, M=N=6, therefore Dmax=10. Distance term weight. Take 4 / 7, predict the weight of the risk item. Take 3 / 7.

[0150]

[0151]

[0152] In the formula, The minimum number of steps is given by Dmax = 10, and Dnorm is the normalized distance. For vehicles Assign to an empty space Matching costs.

[0153] Table 11 - Matching Cost Matrix of All Feasible Pairs

[0154] If both cars independently choose the lowest cost objective, they will both choose option c. 55 Therefore, a one-to-one global matching method is used to determine the allocation result.

[0155]

[0156]

[0157] In the formula, =1 indicates a vehicle Assigned to cells Otherwise, take 0; the first constraint ensures that each vehicle to be redistributed selects a target cell, and the second constraint ensures that each target cell receives at most one vehicle.

[0158] Optimal combination: v1→c 41 v2→c 55 The total cost is 0.350893 + 0.171250 = 0.522143.

[0159] Second-best combination: v1→c 55 v2→c 65 The total cost is 0.252500 + 0.271084 = 0.523584.

[0160] After comparing all allocation combinations that satisfy the one-to-one constraint, 0.522143 is the minimum total cost, therefore v1 is allocated to c. 41 Assign v2 to c 55 .

[0161] VI. Determine the movement paths of the two vehicles. Once the target space is determined, the path evaluation value is constructed using "number of steps already moved, estimated remaining steps to the target, and current cell risk":

[0162] In the formula, g(q) is the cumulative number of steps, h(q) is the Manhattan heuristic distance, ρ is the risk penalty coefficient, μ is the conflict penalty coefficient, and Iconflict represents the risk of occupancy conflict, reservation occupancy, reciprocal exchange, or short-term inaccessibility. In this embodiment, ρ=0.5 and μ=1 are used; the following candidate next cells have no conflict, so Iconflict=0, but the conflict term in the complete formula is retained.

[0163] v1 First step comparison: After c 33 At that time, f = 1 + 3 + 0.5 × 0.520833 = 4.260417; after c 44 At that time, f = 1 + 3 + 0.5 × 0.40833 = 4.204167. Therefore, it prioritizes entering c. 44 .

[0164] The executable path for v1 is: c 34 →c 44 →c 43 →c 42 →c 41 (Solid arrows in Table 12).

[0165] v2 First step comparison: After c 45 At that time, f = 1 + 1 + 0.5 × 0.330114 = 2.165057; in discussing c 34 At that time, v1 had not yet left c. 34 Therefore, c 34 Cells that are already occupied and do not meet the current passability conditions are not considered feasible cells for v2's current path expansion. Based on comparison, v2 prioritizes entering cell c. 45 .

[0166] The executable path for v2 is: c 35 →c 45 →c 55 (The dashed arrows in Table 12).

[0167] Since the two paths do not share the same mobile cell and there is no reciprocal exchange, execution time windows that do not conflict can be assigned to the two vehicles.

[0168] Table 12 - Vehicle Movement Paths and Final Layout (indicated by red arrows)

[0169] In Table 12, the dark red solid arrows represent the execution path c of v1. 34 →c 44 →c 43 →c 42 →c 41 The red dashed arrow indicates the execution path c of v2. 35 →c 45 →c 55 .

[0170] VII. Verification of the Predicted Effects of the Redistribution Scheme After the move is complete, c 34 and c 35 Become idle; c 41 Occupied by v1, the predicted risk is 0.285417; c 55 The vehicle is occupied by v2, with a predicted risk of 0.132917. The positions of the other four vehicles remain unchanged.

[0171]

[0172] The estimated total risk after redistribution is: 0.285417 + 0.132917 + 0.354072 + 0.231310 + 0.239167 + 0.217188 = 1.460071.

[0173] Expected risk reduction: 2.158669 1.460071 = 0.698598.

[0174] Before execution, there were 2 occupied high-risk cells; after execution, the risk of each occupied cell was less than 0.50, and the number of occupied high-risk cells decreased from 2 to 0.

[0175]

[0176] VIII. Implementation Results In this embodiment, the initial total risk of occupied cells in the garage is 2.158669, which meets the vehicle reallocation trigger condition. After vehicle screening, target cell pre-screening, predicted risk calculation, and one-to-one matching, v1 is moved from c... 34 Move to c 41 v2 will be changed from c 35 Move to c 55 The total risk of occupied cells after redistribution is 1.460071, a decrease of 0.698598 compared to before redistribution. The number of high-risk occupied cells decreased from 2 to 0, indicating that this round of vehicle redistribution released key passage positions in the middle and reduced the overall congestion risk of the garage layout.

[0177] Reference Figure 2 The following is a flowchart illustrating a smart garage vehicle reallocation method based on a cellular congestion risk field, provided in another embodiment of this application. Compared with the prior art, the smart garage vehicle reallocation method based on a cellular congestion risk field in this embodiment has the following differences: (1) Using the cell risk field as the basis for unified decision-making throughout the entire process: The garage space is modeled as a two-dimensional cell network, and the cell comprehensive congestion risk value is used to run through all stages of redistribution triggering, vehicle screening, target location selection, path generation, and effect evaluation, replacing the traditional scattered indicators, so that the decision-making logic is unified and coherent.

[0178] (2) Five-dimensional risk integration to construct a precise congestion risk model: The five major risk factors of entrance and exit distance, neighborhood occupancy, path criticality, obstruction impact and vehicle retrieval frequency are normalized and weighted to form a comprehensive risk value, which fully covers the key causes of garage congestion and can accurately identify high-risk parking locations.

[0179] (3) System-level triggering + vehicle-level screening redistribution mechanism: First, determine whether redistribution is needed based on the overall occupancy risk of the garage and the number of high-risk parking spaces; then, screen vehicles that need to be adjusted based on the current location risk of the vehicle, and prioritize the release of entrances and exits, key passages, and buffer areas to avoid ineffective scheduling and frequent disturbances.

[0180] (4) Benefit-cost equilibrium vehicle and parking space matching method: Instead of just selecting the nearest empty space, the original location risk is used as the release benefit and the target location risk is used as the occupation cost. The optimal matching is calculated by combining the movement cost, so as to realize the orderly migration of vehicles from high-risk locations to low-risk long-term parking locations.

[0181] (5) Incorporating risk and conflict into improvement path planning: In By incorporating cell risk and conflict penalties into the path evaluation function, the path priority shifts from "shortest steps" to low-risk, low-conflict, and executable priority, proactively avoiding congested areas and critical channels, and reducing congestion during the redistribution process.

[0182] (6) Quantifiable risk closed-loop evaluation and optimization: The optimization effect can be intuitively evaluated by the overall risk reduction before and after redistribution and the change in the number of high-risk parking spaces, and the risk weight can be corrected by feedback, forming a complete closed loop of "decision-execution-evaluation-optimization".

[0183] Corresponding to the above embodiment, a smart garage vehicle reallocation method based on a cellular congestion risk field, Figure 3 This diagram illustrates a structural block diagram of a smart garage vehicle reallocation system based on a cellular congestion risk field, as provided in one embodiment of this application. The system is applied to a smart garage, which comprises a two-dimensional discrete space consisting of M rows and N columns of minimum schedulable spatial units, each of which is considered a cell. For ease of explanation, only the parts relevant to the embodiment of this application are shown. References Figure 3 The intelligent garage vehicle redistribution system 20 based on cellular congestion risk field includes: a risk perception module 21, a risk calculation module 22, a first screening module 23, a second screening module 24, and a vehicle redistribution module 25.

[0184] Among them, the risk perception module 21 is used to determine the comprehensive congestion risk value of each cell based on the attribute information of multiple cells in the smart garage; Risk calculation module 22 determines the quantitative characteristics of congestion risk of the smart garage based on the comprehensive congestion risk value of each cell; The first screening module 23 is used to select vehicles whose comprehensive congestion risk value of the occupied cells is greater than a preset first risk threshold from multiple parked vehicles in the smart garage when the congestion risk quantification characteristics of the smart garage meet the preset triggering conditions, and to select them as vehicles to be reassigned. The second screening module 24 is used to, for each vehicle to be reassigned, calculate the predicted pick-up frequency risk of the vehicle moving into each candidate cell based on the vehicle's pick-up frequency and the distance from each candidate cell to the entrance / exit; determine the predicted comprehensive congestion risk value corresponding to each candidate cell based on the predicted pick-up frequency risk and the comprehensive congestion risk value corresponding to each candidate cell; calculate the risk difference between the comprehensive congestion risk value of the currently occupied cell of the vehicle to be reassigned and the predicted comprehensive congestion risk value corresponding to each candidate cell; from multiple candidate cells, select cells whose corresponding predicted comprehensive congestion risk value is less than a preset second risk threshold and whose corresponding risk difference is greater than a preset difference threshold, as candidate target cells of the vehicle to be reassigned, and construct multiple feasible pairs; each feasible pair includes the vehicle to be reassigned and one candidate target cell of the vehicle to be reassigned; the candidate cell is a cell whose corresponding operating status information is idle and allows vehicles to enter, and the second risk threshold is less than the first risk threshold; The vehicle reallocation module 25 is used to calculate, for each feasible pair, the cost required to move the vehicle to be reallocated in the feasible pair to the candidate target cell in the feasible pair, and obtain the matching cost corresponding to the feasible pair; based on the matching cost corresponding to each feasible pair, determine the target parking cell corresponding to each vehicle to be reallocated; based on the target parking cell corresponding to each vehicle to be reallocated, generate an executable movement path for each vehicle to be reallocated, and move each vehicle to be reallocated to the corresponding target parking cell along the executable movement path, so as to realize the reallocation of each vehicle to be reallocated.

[0185] In one embodiment of this application, the attribute information of each cell includes location information and operating status information; the operating status information includes idle / occupied identifier, passable / inaccessible identifier, and vehicle retrieval frequency; the risk perception module 21 is specifically used for: Based on the cell's location information, the congestion risk caused by the cell being close to the entrance / exit is determined, which is then used as the cell's entrance / exit distance risk. Based on the idle / occupied flag of each cell, the number of occupied cells in the k-level neighborhood of that cell is counted; based on the passable / impedable flag of each cell, the number of passable cells in the k-level neighborhood of that cell is counted; based on the number of occupied cells in the k-level neighborhood of that cell and the number of passable cells in the k-level neighborhood of that cell, the vehicle density around that cell is determined as the neighborhood occupancy risk of that cell. The number of candidate vehicle access paths passing through the cell is determined based on the cell's location information, which serves as the path criticality risk for that cell. The degree of impact on the movement of other vehicles after the cell is occupied is determined based on the cell's location information, which serves as the cell's obstruction risk. Calculate the matching deviation between the cell's location information and the vehicle retrieval frequency, and use it as the vehicle retrieval frequency risk for that cell; The overall congestion risk value of this cell is obtained by weighting and summing the risks of entrance / exit distance, neighborhood occupancy, path criticality, obstruction, and vehicle retrieval frequency.

[0186] In one embodiment of this application, the risk calculation module 22 is specifically used for: From the multiple cells of the smart garage, select the cells occupied by vehicles and treat them as multiple used cells; The total congestion risk of the smart garage is obtained by summing the combined congestion risk values ​​of multiple used cells. The total amount of congestion risk is used as a quantitative characteristic of congestion risk in smart parking garages; Among them, the total congestion risk of the smart parking garage meets the preset triggering conditions, including: The total congestion risk of the smart parking garage exceeds the preset total risk threshold.

[0187] In one embodiment of this application, the risk calculation module 22 is further used for: Based on the comprehensive congestion risk value, the multiple cells of the smart garage are divided into risk levels, resulting in high-risk cells, medium-risk cells, low-risk cells, and stable low-risk cells. Calculate the number of high-risk cells, and the proportion of cells that are occupied for a long time among high-risk and medium-risk cells; cells that are occupied for a long time refer to cells that have been continuously occupied for a duration greater than a preset time threshold. The total congestion risk, the number of high-risk cells, and the proportion of cells that are occupied for a long time among high-risk and medium-risk cells are used as quantitative characteristics of congestion risk in smart parking garages. Among them, the total congestion risk of the smart parking garage meets the preset triggering conditions, including: The total congestion risk in the smart parking garage exceeds the preset total risk threshold; or, The number of high-risk cells exceeds a preset threshold; or The proportion of cells that are occupied for a long time in high-risk and medium-risk cells is greater than the preset threshold.

[0188] In one embodiment of this application, for each feasible pairing, the vehicle reallocation module 25 is specifically used for: Calculate the number of movement steps required to move the vehicle to be reassigned in the feasible pair to the candidate target cell in the feasible pair; The cost required to move a vehicle to be reallocated from one feasible pair to a candidate target cell in that feasible pair is calculated based on the number of steps taken and the predicted comprehensive congestion risk value corresponding to the candidate target cell in that feasible pair.

[0189] In one embodiment of this application, the vehicle redistribution module 25 is specifically used for: Based on the matching cost corresponding to each feasible pair, the following objective function is constructed:

[0190] Solve the objective function to obtain the target parking cell corresponding to each vehicle to be redistributed; in, Represents decision variables, , Let m represent the matching cost for each feasible pair, m represent the number of vehicles to be redistributed, and n represent the number of candidate target cells, where m ≤ n.

[0191] In one embodiment of this application, for each vehicle to be redistributed, the vehicle redistribution module 25 is specifically used for: based on The path search algorithm generates an executable movement path for the vehicle to be reassigned; In each stage of generating an executable movement path, each candidate target cell for the next stage is evaluated based on a path evaluation function. For any candidate target cell q, the path evaluation function... for: ; in, This represents the cumulative number of steps the vehicle to be reassigned has taken from the starting point to the candidate target cell q. This represents the number of steps required for the vehicle to be reassigned to move from the candidate target cell q to the target parking cell. This represents the overall congestion risk value of candidate target cell q. This indicates whether the candidate cell q has occupancy conflicts, opposite swap conflicts, or short-term inaccessibility risks.

[0192] See Figure 4 , Figure 4 This is a schematic block diagram of an electronic device provided according to an embodiment of this application. Figure 4 The electronic device 300 in this embodiment may include one or more processors 301, one or more input devices 302, one or more output devices 303, and one or more memories 304. The processors 301, input devices 302, output devices 303, and memories 304 communicate with each other via a communication bus 305. The memories 304 store computer programs, including program instructions. The processors 301 execute the program instructions stored in the memories 304. Specifically, the processors 301 are configured to invoke the program instructions to perform the functions of each module / unit in the above-described device embodiments, for example... Figure 3 The functions of the risk perception module 21, risk calculation module 22, first screening module 23, second screening module 24, and vehicle reallocation module 25 are shown.

[0193] It should be understood that, in the embodiments of this application, the processor 301 may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0194] Input device 302 may include a touchpad, a fingerprint sensor (for collecting the user's fingerprint information and fingerprint orientation information), a microphone, etc., and output device 303 may include a display (LCD, etc.), a speaker, etc.

[0195] The memory 304 may include read-only memory and random access memory, and provides instructions and data to the processor 301. A portion of the memory 304 may also include non-volatile random access memory. For example, the memory 304 may also store preset constants such as a first risk threshold and a second risk threshold.

[0196] In specific implementations, the processor 301, input device 302, and output device 303 described in the embodiments of this application can execute the implementation method described in the intelligent garage vehicle reallocation method based on cellular congestion risk field provided in the embodiments of this application, or they can execute the implementation method of the electronic device described in the embodiments of this application, which will not be repeated here.

[0197] In another embodiment of this application, a computer-readable storage medium is provided. This computer-readable storage medium stores a computer program, which includes program instructions. When executed by a processor, the program instructions implement all or part of the processes in the methods described above. Alternatively, the computer program can instruct related hardware to complete the process. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include any entity or device capable of carrying computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0198] The computer-readable storage medium can be an internal storage unit of the electronic device in any of the foregoing embodiments, such as a hard disk or memory of the electronic device. The computer-readable storage medium can also be an external storage device of the electronic device, such as a plug-in hard disk, smart media card (SMC), secure digital card (SD), flash card, etc., equipped on the electronic device. Furthermore, the computer-readable storage medium can include both internal and external storage units of the electronic device. The computer-readable storage medium is used to store computer programs and other programs and data required by the electronic device. The computer-readable storage medium can also be used to temporarily store data that has been output or will be output.

[0199] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this application.

[0200] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the electronic devices and units described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0201] In the several embodiments provided in this application, it should be understood that the disclosed electronic devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the couplings or direct couplings or communication connections shown or discussed may be indirect couplings or communication connections through some interfaces or units, or they may be electrical, mechanical, or other forms of connection.

[0202] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of the embodiments of this application, depending on actual needs.

[0203] Furthermore, the modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module. The integrated modules described above can be implemented in hardware or as software functional modules.

[0204] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A vehicle reallocation method for intelligent parking garages based on cellular congestion risk fields, applied to intelligent parking garages, wherein the intelligent parking garage comprises a two-dimensional discrete space consisting of M rows and N columns of minimum schedulable spatial units, each minimum schedulable spatial unit being a cell, characterized in that... include: The comprehensive congestion risk value of each cell is determined based on the attribute information of multiple cells within the intelligent parking garage. The congestion risk quantification characteristics of the intelligent parking garage are determined based on the comprehensive congestion risk value of each cell; If the congestion risk quantification characteristics of the smart garage meet the preset triggering conditions, then from the multiple parked vehicles in the smart garage, vehicles whose comprehensive congestion risk value of the occupied cells is greater than the preset first risk threshold are selected as vehicles to be reassigned. For each vehicle to be reassigned, based on the vehicle retrieval frequency of the vehicle to be reassigned and the distance from each candidate cell to the entrance / exit, the predicted retrieval frequency risk of the vehicle to be reassigned moving into each candidate cell is calculated, and the predicted comprehensive congestion risk value corresponding to each candidate cell is determined based on the predicted retrieval frequency risk and the comprehensive congestion risk value corresponding to each candidate cell. Calculate the risk difference between the overall congestion risk value of the currently occupied cell of the vehicle to be reassigned and the predicted overall congestion risk value corresponding to each candidate cell; From multiple candidate cells, cells whose predicted comprehensive congestion risk value is less than a preset second risk threshold and whose corresponding risk difference is greater than a preset difference threshold are selected as candidate target cells for the vehicle to be reallocated, and multiple feasible pairs are constructed; each feasible pair includes the vehicle to be reallocated and one candidate target cell of the vehicle to be reallocated. The candidate cell is a cell whose corresponding operating status information is idle and allows vehicles to enter, and the second risk threshold is less than the first risk threshold. For each feasible pair, calculate the cost required to move the vehicle to be reassigned in the feasible pair to the candidate target cell in the feasible pair, and obtain the matching cost corresponding to the feasible pair. Based on the matching cost corresponding to each feasible pair, determine the target parking cell corresponding to each vehicle to be reassigned. Based on the target parking cell corresponding to each vehicle to be reassigned, an executable movement path is generated for each vehicle to be reassigned, and each vehicle to be reassigned is moved to the corresponding target parking cell along the executable movement path to achieve the reassignment of each vehicle to be reassigned.

2. The intelligent garage vehicle reallocation method based on cellular congestion risk field as described in claim 1, characterized in that, Each cell's attribute information includes location information and operating status information; the operating status information includes idle / occupied flags, passable / inaccessible flags, and vehicle retrieval frequency; Specifically, for each cell, a comprehensive congestion risk value is determined based on the attribute information of multiple cells within the intelligent parking garage, including: Based on the cell's location information, the congestion risk caused by the cell being close to the entrance / exit is determined, which is then used as the cell's entrance / exit distance risk. Based on the idle / occupied flag of each cell, the number of occupied cells in the k-level neighborhood of that cell is counted. Based on the passable / impeded flag of each cell, the number of passable cells in the k-level neighborhood of that cell is counted. Based on the number of occupied cells in the k-level neighborhood of that cell and the number of passable cells in the k-level neighborhood of that cell, the vehicle density around that cell is determined as the neighborhood occupancy risk of that cell. The number of candidate vehicle access paths passing through the cell is determined based on the cell's location information, which serves as the path criticality risk for that cell. The degree of impact on the movement of other vehicles after the cell is occupied is determined based on the cell's location information, which serves as the cell's obstruction risk. Calculate the matching deviation between the cell's location information and the vehicle retrieval frequency, and use it as the vehicle retrieval frequency risk for that cell; The overall congestion risk value of the cell is obtained by weighting and summing the risk of the entrance / exit distance, the risk of the neighborhood occupancy, the risk of the path criticality, the risk of obstruction, and the risk of vehicle retrieval frequency.

3. The intelligent garage vehicle reallocation method based on cellular congestion risk field as described in claim 1, characterized in that, The determination of the congestion risk quantification characteristics of the smart parking garage based on the comprehensive congestion risk value of each cell includes: From the multiple cells of the intelligent garage, cells occupied by vehicles are selected as multiple used cells; The total congestion risk of the smart garage is obtained by summing the combined congestion risk values ​​of multiple used cells. The total congestion risk is used as the quantitative characteristic of the congestion risk of the smart garage; The total congestion risk of the intelligent parking garage meets preset triggering conditions, including: The total congestion risk of the smart garage exceeds a preset total risk threshold.

4. The intelligent garage vehicle reallocation method based on cellular congestion risk field as described in claim 3, characterized in that, The method of determining the congestion risk quantification characteristics of the intelligent parking garage based on the comprehensive congestion risk value of each cell also includes: Based on the comprehensive congestion risk value, the multiple cells of the intelligent garage are divided into risk levels, resulting in high-risk cells, medium-risk cells, low-risk cells, and stable low-risk cells. Calculate the number of high-risk cells and the proportion of cells that are occupied for a long time among the high-risk cells and the medium-risk cells; the cells that are occupied for a long time refer to cells that have been continuously occupied for a duration greater than a preset time threshold. The total congestion risk, the number of high-risk cells, and the proportion of cells that are occupied for a long time among the high-risk cells and the medium-risk cells are used as the quantitative characteristics of the congestion risk of the smart garage. The total congestion risk of the intelligent parking garage meets preset triggering conditions, including: The total congestion risk of the intelligent parking garage exceeds a preset total risk threshold; or, The number of high-risk cells is greater than a preset threshold; or The proportion of cells that are occupied for a long time in the high-risk cells and medium-risk cells is greater than a preset threshold.

5. The intelligent garage vehicle reallocation method based on cellular congestion risk field as described in claim 1, characterized in that, For each feasible pair, calculate the cost required to move the vehicle to be reassigned in that feasible pair to a candidate target cell in that feasible pair, including: Calculate the number of movement steps required to move the vehicle to be reassigned in the feasible pair to the candidate target cell in the feasible pair; Based on the number of movement steps and the predicted comprehensive congestion risk value corresponding to the candidate target cell in the feasible pair, calculate the cost required to move the vehicle to be reallocated in the feasible pair to the candidate target cell in the feasible pair.

6. The intelligent garage vehicle reallocation method based on cellular congestion risk field as described in claim 1, characterized in that, Based on the matching cost corresponding to each feasible pair, the target parking cell corresponding to each vehicle to be reassigned is determined, including: Based on the matching cost corresponding to each feasible pair, the following objective function is constructed: Solve the objective function to obtain the target parking cell corresponding to each vehicle to be redistributed; in, Represents decision variables, , Let m represent the matching cost for each feasible pair, m represent the number of vehicles to be redistributed, and n represent the number of candidate target cells, where m ≤ n.

7. The intelligent garage vehicle reallocation method based on cellular congestion risk field as described in claim 1, characterized in that, For each vehicle to be reassigned, based on the target parking cell corresponding to that vehicle, an executable movement path is generated for that vehicle, including: based on The path search algorithm generates an executable movement path for the vehicle to be reassigned; In each stage of generating an executable movement path, each candidate target cell for the next stage is evaluated based on a path evaluation function. For any candidate target cell q, the path evaluation function... for: ； in, This represents the cumulative number of steps the vehicle to be reassigned has taken from the starting point to the candidate target cell q. This represents the number of steps required for the vehicle to be reassigned to move from the candidate target cell q to the target parking cell. This represents the overall congestion risk value of candidate target cell q. This indicates whether the candidate cell q has occupancy conflicts, reciprocal swapping conflicts, or short-term inaccessibility risks.

8. A vehicle reallocation system for intelligent parking garages based on cellular congestion risk fields, applied to intelligent parking garages, wherein the intelligent parking garage comprises a two-dimensional discrete space consisting of M rows and N columns of minimum schedulable spatial units, each minimum schedulable spatial unit being a cell, characterized in that... include: The risk perception module is used to determine the comprehensive congestion risk value of each cell based on the attribute information of multiple cells in the intelligent garage. The risk calculation module determines the quantification characteristics of the congestion risk of the intelligent garage based on the comprehensive congestion risk value of each cell; The first screening module is used to select vehicles whose comprehensive congestion risk value of the occupied cells is greater than a preset first risk threshold from multiple parked vehicles in the smart garage when the congestion risk quantification characteristics of the smart garage meet the preset triggering conditions, and then select them as vehicles to be reassigned. The second screening module calculates the predicted pick-up frequency risk of the vehicle to be reassigned to each candidate cell based on the vehicle pick-up frequency of the vehicle to be reassigned and the distance from each candidate cell to the entrance / exit for each candidate cell. Based on the predicted pick-up frequency risk and the comprehensive congestion risk value corresponding to each candidate cell, the module determines the predicted comprehensive congestion risk value corresponding to each candidate cell. Calculate the risk difference between the overall congestion risk value of the currently occupied cell of the vehicle to be reassigned and the predicted overall congestion risk value corresponding to each candidate cell; From multiple candidate cells, cells whose predicted comprehensive congestion risk value is less than a preset second risk threshold and whose corresponding risk difference is greater than a preset difference threshold are selected as candidate target cells for the vehicle to be reallocated, and multiple feasible pairs are constructed; each feasible pair includes the vehicle to be reallocated and one candidate target cell of the vehicle to be reallocated. The candidate cell is a cell whose corresponding operating status information is idle and allows vehicles to enter, and the second risk threshold is less than the first risk threshold. The vehicle reallocation module is used to calculate, for each feasible pair, the cost required to move the vehicle to be reallocated in the feasible pair to the candidate target cell in the feasible pair, and obtain the matching cost corresponding to the feasible pair. Based on the matching cost corresponding to each feasible pair, determine the target parking cell corresponding to each vehicle to be reassigned. Based on the target parking cell corresponding to each vehicle to be reassigned, an executable movement path is generated for each vehicle to be reassigned, and each vehicle to be reassigned is moved to the corresponding target parking cell along the executable movement path to achieve the reassignment of each vehicle to be reassigned.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 7.

Images