An unmanned aerial vehicle based inter-parking lot induction system

Abstract

The present application relates to parking management technical field, especially to a kind of parking lot between based on unmanned aerial vehicle's induction system, including parking platform, induction algorithm module and multiple unmanned aerial vehicles, wherein parking platform is configured to access the data of all parking lots in area, monitors parking stall occupancy, when parking lot approaches saturation, issue queuing event confirmation instruction to unmanned aerial vehicle at its entrance;Unmanned aerial vehicle responds to instruction, configured to take off and shoot the road around entrance, judge whether there is vehicle queuing and slow traffic, if it occurs, report to parking platform;Parking platform is configured to receive the event reported by unmanned aerial vehicle and send it to induction algorithm module;Induction algorithm module is configured to receive event information, output induction strategy and path planning to corresponding unmanned aerial vehicle in combination with data;Unmanned aerial vehicle is further configured to receive induction strategy and path planning, and execute induction task.It can monitor parking lot and surrounding road condition in real time through unmanned aerial vehicle, quickly generate induction strategy and path in combination with algorithm, effectively relieve parking congestion, improve regional parking efficiency.

Description

Technical Field

[0001] This invention relates to the technical field of parking management, and more particularly to a parking guidance system based on unmanned aerial vehicles (UAVs). Background Technology

[0002] With the continuous improvement of urbanization and the sustained growth of motor vehicle ownership, parking difficulties have arisen, especially during commuting hours and holidays, causing inconvenience to citizens' travel and daily life.

[0003] In the prior art, patent CN110364016B discloses a method for guiding parking using drones. License plate recognition devices are installed at both the entrance and exit of the parking lot. A drone ground communication base station and drone docking points are set up inside the parking lot. Multiple drones are installed at the docking points. The drones can scan and recognize vehicle license plate numbers and vehicle images. By establishing a parking location information model, parking space sensing devices collect parking space information. When the license plate recognition devices at the entrance and exit of the parking lot detect a vehicle entering, the drone ground communication base station uses a linear regression algorithm to calculate the matching degree S(p) for different parking spaces. The drones are then controlled to guide parking according to the matched target parking space and route, thus solving the problems of parking difficulties for car owners and parking lot management difficulties.

[0004] However, the aforementioned technical solutions can only provide parking guidance within a single parking lot. For parking lot entrance queues caused by sudden surges in vehicle traffic, the parking guidance system and drone equipment cannot handle the situation since the event occurs outside the parking lot. Furthermore, the prolonged queues at the entrance exacerbate the chaos in the entry and parking lot order, impairing the traffic efficiency of nearby roads. Summary of the Invention

[0005] This invention provides a parking guidance system based on drones that can achieve efficient traffic diversion and optimized allocation of parking space resources across parking lots, effectively solving the problems in the background art.

[0006] The parking guidance system based on drones of the present invention includes a parking platform, a guidance algorithm module, and multiple drones, wherein...

[0007] The parking platform is configured to access all parking lot data in a specific area. The data includes at least the parking lot name, number of parking spaces, parking lot coordinates, parking lot lanes, latitude and longitude coordinates of each lane, real-time number of occupied parking spaces, and parking records.

[0008] The parking platform is further configured to monitor the occupancy of parking spaces in each parking lot based on data. When the parking lot is detected to be close to saturation, a queuing event confirmation instruction is sent to the drones at each entrance of the parking lot.

[0009] In response to the queuing event confirmation command, the drone is configured to take off and photograph the road conditions around the parking lot entrance. It uses image recognition algorithms to determine whether there are vehicle queues and slow traffic. If a queuing event is determined to have occurred, the event is reported to the parking platform.

[0010] The parking platform is further configured to receive events reported by drones and send the event information to the guidance algorithm module;

[0011] The guidance algorithm module is configured to receive queuing event information sent by the parking platform, combine it with the data parking lot guidance algorithm to output guidance strategies and path planning to the entrance of the destination parking lot, and send them to the corresponding drones;

[0012] The drone is further configured to receive guidance strategies and path plans sent by the guidance algorithm module, broadcast guidance information above the queuing vehicles based on the guidance strategy, prompt the vehicles to follow it to the destination parking lot, guide the vehicles to the destination parking lot according to the path plan, and end the guidance mission and return to the starting point when it flies to the entrance of the destination parking lot.

[0013] In one possible design, the parking platform determines whether the parking lot is nearing saturation based on either: the real-time saturation of parking spaces is greater than a preset threshold; or, the real-time remaining number of parking spaces is less than a preset threshold.

[0014] In one possible design, the queuing event information reported by the drone includes at least the event number, parking lot name, entrance lane number, queue length, number of vehicles in the queue, speed of surrounding vehicles, and reporting time.

[0015] In one possible design, the parking guidance algorithm includes:

[0016] Traverse all parking lots within 0.5km of the current parking lot that have a real-time saturation of less than or equal to 60%;

[0017] Each eligible parking lot is evaluated using an incentive assessment index, which is:

[0018] =50* +50* ;

[0019] in, As an inductive evaluation index, This is the saturation recommendation index. The distance recommendation index;

[0020] Parking lots are ranked from highest to lowest based on their guidance assessment index, and the number of guided vehicles each parking lot can accommodate is calculated. The number of guided vehicles each parking lot can accommodate is:

[0021] = (80% - )* ;

[0022] in, Let x be the number of guided vehicles that the x-th parking lot can accommodate. This represents the real-time parking saturation of the x-th parking lot. Let x be the number of parking spaces in the x-th parking lot;

[0023] Compare the number of vehicles in the queue with the number of guided vehicles that the first-priority parking lot can accommodate. If the number of vehicles in the queue is less than or equal to the number of guided vehicles that the first-priority parking lot can accommodate, then all vehicles in the queue will be guided to the first-priority parking lot.

[0024] If the number of vehicles in the queue exceeds the number of guided vehicles that the first-priority parking lot can accommodate, then the number of guided vehicles that the first-priority parking lot can accommodate will be allocated to that parking lot. The remaining number of vehicles in the queue will then be compared with the number of guided vehicles that the second-priority parking lot can accommodate, until all vehicles in the queue are allocated or the maximum capacity of all eligible parking lots is reached.

[0025] In one possible design, the calculation rule for the saturation recommendation index is as follows:

[0026] When the real-time parking saturation of the xth parking lot When it is greater than 60%, the saturation recommendation index is... =0;

[0027] When the real-time parking saturation of the xth parking lot When the saturation recommendation index is less than or equal to 60% and greater than or equal to 20%, the saturation recommendation index is... =2.5*(0.6- );

[0028] When the real-time parking saturation of the xth parking lot When it is less than 20%, the saturation recommendation index is... The value is 1.

[0029] In one possible design, the rule for calculating the distance recommendation index is as follows:

[0030] When the x-th parking lot is a straight-line distance When the distance is greater than 0.5km, the recommended distance index is... =0;

[0031] When the x-th parking lot is a straight-line distance When the distance is less than or equal to 0.5km and greater than or equal to 0.1km, the recommended distance index is... =2.5*(0.5- );

[0032] When the x-th parking lot is a straight-line distance When the distance is less than 0.1km, the recommended distance index is... The value is 1.

[0033] In one possible design, the specific area is set as a city, district, street, or business district.

[0034] In one possible design, for parking lots where entry queues frequently occur, the entrance to the passage is set as the initial point for drones, and several drones are deployed to each parking lot based on the number of vehicles in the queue.

[0035] In one possible design, after flying a certain distance, the drone uses an image recognition algorithm to identify whether there are queuing vehicles following it. If not, it flies back to the starting point and continues to broadcast guidance information.

[0036] In one possible design, the drone guides vehicles to their destination parking lot according to a planned route, with its flight speed matching the current road traffic speed.

[0037] The technical solution of this invention can achieve the following technical effects:

[0038] This invention constructs a parking platform covering all parking lots in a specific area, combined with the real-time monitoring and guidance capabilities of drones. This solves the problem of queuing at the entrance of a single parking lot due to excessive instantaneous traffic. When a parking lot is nearing saturation, the drone can quickly confirm and report the queuing event. The guidance algorithm module generates the optimal guidance strategy based on factors such as the saturation and distance of surrounding parking lots. The drone then broadcasts the information and guides vehicles to suitable parking lots, alleviating the entry pressure on the original parking lot, avoiding chaos in entry and parking order, and improving the traffic efficiency of surrounding roads. At the same time, by rationally allocating guided vehicles, the parking space resources of surrounding parking lots can be fully utilized, optimizing the parking layout in the area and providing citizens with a smoother and more convenient parking experience. Attached Figure Description

[0039] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, 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 recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0040] Figure 1 This is a schematic diagram of the parking guidance system based on unmanned aerial vehicles (UAVs) in this invention. Detailed Implementation

[0041] This application will now be described with reference to the accompanying drawings.

[0042] like Figure 1 As shown, this invention discloses a parking guidance system based on drones, comprising a parking platform, a guidance algorithm module, and multiple drones. The parking platform is configured to access data from all parking lots in a specific area. This data includes at least the parking lot name, number of parking spaces, parking lot coordinates, parking lot lanes, latitude and longitude coordinates of each lane, real-time occupied parking spaces, and parking records. The parking platform is further configured to monitor the parking space occupancy of each parking lot based on the data. When a parking lot is detected to be nearing saturation, a queuing event confirmation command is sent to the drones at each entrance of that parking lot. In response to the queuing event confirmation command, the drones are configured to take off and photograph the surrounding roads at the parking lot entrance. Using image recognition algorithms, they determine whether vehicle queuing or slow traffic has occurred. If a queuing event is detected, the drones report the event to the parking platform. The parking platform is further configured to receive the events reported by the drones and send the event information to the guidance algorithm module. The guidance algorithm module is configured to receive the queuing event information sent by the parking platform, combine it with the data-driven parking guidance algorithm to output a guidance strategy and a path plan to the destination parking lot entrance lane, and send this to the corresponding drone. The drone is further configured to receive guidance strategies and path plans sent by the guidance algorithm module, broadcast guidance information above the queuing vehicles based on the guidance strategy, prompt the vehicles to follow it to the destination parking lot, guide the vehicles to the destination parking lot according to the path plan, and end the guidance mission and return to the starting point when it flies to the entrance of the destination parking lot.

[0043] In this embodiment, by accessing all parking lot data in a specific area and monitoring parking space occupancy through a parking platform, the system can promptly detect when parking lots are nearing saturation. Drones respond to queuing event confirmation commands, take off to photograph and determine if vehicle queues or slow traffic are occurring, enabling accurate event reporting and ensuring timely attention to issues. The guidance algorithm module outputs guidance strategies and path planning, with drones broadcasting information and guiding vehicles to their destination parking lot, effectively alleviating queuing problems at single parking lot entrances, optimizing entry and in-lot order, and improving traffic efficiency on nearby roads. Furthermore, the parking platform's integration and real-time monitoring of parking lot data across the entire area provides a more comprehensive view of vehicle dispatching, preventing concentrated pressure on local parking lots. The drones' proactive patrols and event confirmations enhance the sensitivity to parking-related issues, reducing oversights. The guidance algorithm module, combined with strategies generated from multi-dimensional data, makes vehicle diversion more scientific and reasonable, reducing the time cost for drivers searching for parking spaces, improving the travel experience, and reducing road congestion caused by parking difficulties.

[0044] Specific areas can be defined as a city, district, street, or business district. A city represents a larger administrative region, suitable for coordinating parking conflicts around large POIs (Points of Interest) throughout the city, such as transportation hubs, universities, and large hospitals, and can solve cross-regional parking resource allocation problems. A district can coordinate parking conflicts in core district-level areas, such as district-level hospitals and district-owned commercial centers, achieving precise allocation of resources within the region. A street is a basic administrative unit, capable of quickly responding to localized queuing problems in small and medium-sized parking lots within the street, such as around communities and street-level shopping malls. A business district is a densely populated commercial activity area with a large number of parking lots and instantaneous parking demand, making it a high-frequency application scenario for parking guidance. This multi-dimensional regional division allows the system to flexibly adapt to different scales of management scope according to actual needs. Furthermore, this multi-dimensional regional division conforms to the administrative logic of urban management and the natural distribution patterns of commercial activities, facilitating integration with local traffic management departments and parking lot operators.

[0045] In some embodiments of the present invention, the criteria for determining whether a parking lot is nearing saturation by the parking platform are: the real-time saturation of parking spaces is greater than a preset threshold; or, the real-time remaining number of parking spaces is less than a preset threshold. In this embodiment, the real-time saturation of parking spaces refers to the proportion of currently used parking spaces to the total number of parking spaces, reflecting the current load level of the parking lot. A high occupancy rate indicates a shortage of remaining parking spaces, which may lead to queuing. The real-time remaining number of parking spaces refers to the number of available parking spaces, directly reflecting the number of vehicles that can be accommodated. A low number of remaining spaces directly indicates that the parking lot is about to become saturated. Both criteria determine whether a parking lot is nearing saturation from different perspectives, covering scenarios of parking lots of different sizes. For example, large parking lots may have many remaining parking spaces but a high occupancy rate, while small parking lots may have few remaining parking spaces but a occupancy rate that may not have reached the threshold, ensuring the comprehensiveness and accuracy of the determination and providing timely triggering conditions for subsequent drone confirmation of queuing events. The calculation formulas are as follows:

[0046] ; ;

[0047] Where R is the real-time saturation of berths, O is the number of berths occupied in real time, T is the total number of berths, and L is the number of berths remaining in real time.

[0048] In setting the judgment thresholds, the judgment threshold for real-time saturation of parking spaces is usually set to a higher value, such as 80% or above, while the judgment threshold for real-time remaining parking spaces is usually set to a lower value, such as 20% or below of the total number of parking spaces. In practical applications, the thresholds can be flexibly adjusted according to factors such as the size of the parking lot and the surrounding traffic flow to adapt to the needs of different scenarios.

[0049] The drone is equipped with an image acquisition module. Upon receiving a queuing event confirmation command, it immediately takes off and uses the image acquisition module to photograph the roads surrounding the parking lot entrance. The drone has built-in image recognition algorithms, such as the YOLOv5 algorithm, which can identify queue length, number of vehicles in the queue, and the speed of surrounding vehicles by analyzing the captured images. If the algorithm determines that there is a queue and slow traffic, the drone quickly reports the event information to the parking platform. To ensure that the parking platform and subsequent guidance algorithm modules have a comprehensive understanding of the queuing event, the queuing event information reported by the drone includes at least the event number, parking lot name, entrance lane number, queue length, number of vehicles in the queue, speed of surrounding vehicles, and reporting time. Among these features, the event number serves as a unique identifier for each queuing event, ensuring accurate tracking and management of each event within the system and preventing information confusion. The parking lot name and entrance lane number accurately pinpoint the specific location of the queuing event, allowing the parking platform and guidance algorithm module to clearly identify the parking lot and specific entrance where the problem occurs. The queue length reflects the spatial extent of vehicle congestion, while the number of vehicles in the queue quantifies the scale of vehicles requiring guidance. Both queue length and the number of vehicles in the queue directly reflect the severity of the queuing event. The surrounding vehicle traffic speed reflects the impact of the queuing event on surrounding roads; if the traffic speed is too low, it indicates that the queuing has severely affected road traffic efficiency. The reporting time records the specific moment the event occurred, providing data support for analyzing peak parking periods and optimizing the timeliness of guidance strategies, while also ensuring that the entire guidance process proceeds in an orderly manner according to time.

[0050] In some embodiments of the present invention, the parking lot guidance algorithm of the guidance algorithm module includes:

[0051] Traverse all parking lots within 0.5km of the current parking lot that have a real-time saturation of less than or equal to 60%;

[0052] Each eligible parking lot is evaluated using an incentive assessment index, which is:

[0053] =50* +50* ;

[0054] in, As an inductive evaluation index, This is the saturation recommendation index. The distance recommendation index;

[0055] Parking lots are ranked from highest to lowest based on their guidance assessment index, and the number of guided vehicles each parking lot can accommodate is calculated. The number of guided vehicles each parking lot can accommodate is:

[0056] = (80% - )* ;

[0057] in, Let x be the number of guided vehicles that the x-th parking lot can accommodate. This represents the real-time parking saturation of the x-th parking lot. Let x be the number of parking spaces in the x-th parking lot;

[0058] Compare the number of vehicles in the queue with the number of guided vehicles that the first-priority parking lot can accommodate. If the number of vehicles in the queue is less than or equal to the number of guided vehicles that the first-priority parking lot can accommodate, then all vehicles in the queue will be guided to the first-priority parking lot.

[0059] If the number of vehicles in the queue exceeds the number of guided vehicles that the first-priority parking lot can accommodate, then the number of guided vehicles that the first-priority parking lot can accommodate will be allocated to that parking lot. The remaining number of vehicles in the queue will then be compared with the number of guided vehicles that the second-priority parking lot can accommodate, until all vehicles in the queue are allocated or the maximum capacity of all eligible parking lots is reached.

[0060] In this embodiment, the 0.5km limit is because parking lots that are too far away are not attractive to drivers and are difficult to effectively guide vehicle diversion. The 60% limit is because parking lots with a real-time saturation exceeding 60% are already close to saturation and have limited capacity to accommodate new vehicles. If vehicles are guided to such parking lots, it may cause queuing again, thus losing the purpose of guidance. By simultaneously limiting both distance and real-time saturation, it can be ensured that the selected target parking lots are both within a reasonable distance range and have sufficient parking spaces to accommodate the guided vehicles. In the process of calculating the guidance evaluation index, the saturation recommendation index P x Reflects the availability of parking spaces in the target parking lot, and is located near the recommendation index D. xReflecting the convenience of both the target parking lot and the current queuing parking lot, a 50 / 50 weighting for each avoids decision-making biases caused by a single factor. This prevents prioritizing proximity at the expense of the parking lot's actual capacity, and avoids recommending excessively distant parking lots based solely on available spaces, thus balancing parking convenience and parking lot resource utilization. In calculating the number of guided vehicles each parking lot can accommodate, an 80% upper limit is used, with a 20% redundancy reserve to prevent queuing issues from recurring in the target parking lot in the short term, ensuring its operational order and meeting the parking needs of subsequent vehicles. This is also calculated proportionally based on the total number of parking spaces. This system enables parking lots of different sizes to receive an allocation that matches their capacity. When allocating vehicles according to the guidance assessment index, the parking lots ranked higher have better overall conditions in terms of distance and saturation. Allocating vehicles to these parking lots first can maximize the parking experience for car owners, such as closer proximity and shorter waiting times. When the number of vehicles in the queue exceeds the capacity of the first-ranked parking lot, the vehicles are allocated to subsequent parking lots in turn. This ensures that all vehicles that need to be guided can be reasonably diverted, avoiding resource waste or vehicle congestion, and ultimately achieving the goal of alleviating the queuing pressure of the original parking lot and improving the traffic efficiency of the surrounding roads.

[0061] Specifically, the calculation rules for the saturation recommendation index are as follows:

[0062] When the real-time parking saturation of the xth parking lot When it is greater than 60%, the saturation recommendation index is... The real-time parking saturation of the x-th parking lot is 0; When the saturation recommendation index is less than or equal to 60% and greater than or equal to 20%, the saturation recommendation index is... =2.5*(0.6- When the real-time parking saturation of the xth parking lot... When it is less than 20%, the saturation recommendation index is... The value is 1.

[0063] Among them, the real-time parking saturation of the xth parking lot This is an indicator reflecting the current occupancy rate of parking spaces in the parking lot. Its calculation method is consistent with the logic of real-time parking space saturation, and the calculation formula is as follows: ; This indicates the real-time number of occupied parking spaces in the x-th parking lot, i.e., the number of parking spaces currently in use. This represents the total number of parking spaces in the x-th parking lot.

[0064] In this embodiment, parking lots with a real-time saturation greater than 60% are already full and no longer suitable for attracting surrounding vehicles; in parking lots with a real-time saturation between 20% and 60%, S x The smaller P xThe higher the value, the more suitable the parking lot is as a destination for induced parking. The linear formula can accurately distinguish the carrying potential of parking lots with medium saturation, avoiding decision-making bias caused by overly coarse threshold division. Parking lots with a real-time saturation of less than 20% have sufficient vacant parking spaces and are fully capable of accommodating a large number of induced vehicles. They will not become saturated due to the addition of new vehicles in the short term. Setting this value to the highest value can prioritize recommending such parking lots.

[0065] The rules for calculating the distance recommendation index are as follows:

[0066] When the x-th parking lot is a straight-line distance When the distance is greater than 0.5km, the recommended distance index is... The straight-line distance is 0; when the x-th parking lot is... When the distance is less than or equal to 0.5km and greater than or equal to 0.1km, the recommended distance index is... =2.5*(0.5- When the x-th parking lot is a straight-line distance When the distance is less than 0.1km, the recommended distance index is... The value is 1.

[0067] Wherein, the straight-line distance of the xth parking lot It is calculated using a distance formula based on the geographical coordinates of the x-th parking lot and the currently queuing parking lot. The calculation formula is as follows: R is the Earth's radius, with an average value of 6371 km. , These are the latitudes of the current queuing parking lot and the xth parking lot, respectively, and need to be converted to radians; , These are the longitudes of the current parking lot and the xth parking lot, respectively, and need to be converted to radians; , ;

[0068] In this embodiment, parking lots with a straight-line distance exceeding 0.5km are completely unsuitable for guiding surrounding vehicles to park there; for parking lots with a straight-line distance between 0.1km and 0.5km, R x The smaller D is x The higher the value, the stronger the attraction of the parking lot to car owners. The linear formula can accurately distinguish the attractiveness of parking lots with medium distance, ensuring that the distance factor is reflected in the guidance assessment. Parking lots with a straight-line distance of less than 0.1km are extremely close and have the strongest attraction to car owners, with the highest probability of successful guidance. Setting them to the highest value can prioritize recommending such parking lots.

[0069] The drones are equipped with voice broadcasting modules. Once each drone receives the guidance strategy generated by the guidance algorithm module, it enters the guidance execution phase. The drones fly sequentially above the queuing vehicles and broadcast guidance information to drivers via the voice broadcasting system, such as "The parking lot ahead is full. Please follow me to parking lot P3, 300 meters away." During the guidance process, the drones strictly follow the planned path, and their flight speed matches the current traffic speed to ensure the safety and effectiveness of the guidance process. If the drones fly too fast, following vehicles may speed to catch up, increasing the risk of traffic accidents; if they fly too slowly, it may cause congestion behind them, exacerbating traffic pressure. The flight rhythm that matches the road traffic speed allows drivers to clearly identify and steadily follow the drones, avoiding interruptions due to speed differences, while maintaining consistency with the overall traffic flow speed, reducing interference with surrounding traffic, and ensuring that guided vehicles safely and orderly drive to the destination parking lot, ultimately achieving the goal of efficient traffic diversion.

[0070] During the drone guidance process, drivers may not follow in time due to reasons such as not receiving information in time or not knowing the destination. To ensure the diversion effect, the drone hovers at preset intervals. For example, after flying 300 meters or 1 minute, it needs to use the image acquisition module and image recognition algorithm to identify whether there are queued vehicles following. If not, it flies back above the queued vehicles to continue broadcasting guidance information and guides the vehicles to follow it to the nearby destination parking lot.

[0071] For parking lots where historical data shows frequent queues at the entrance, the entrance can be designated as the initial drone location, and the number of drones deployed can be adjusted based on the peak number of queuing vehicles. For example, if statistics show that a parking lot experiences severe congestion on average for every 10 queuing vehicles during peak hours, then a drone can be deployed for every 10 queuing vehicles to ensure timely response when congestion occurs.

[0072] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A UAV-based inter-parking lot induction system, characterized by, Includes a parking platform, a guidance algorithm module, and multiple drones, among which The parking platform is configured to access all parking lot data in a specific area. The data includes at least the parking lot name, total number of parking spaces, parking lot coordinates, parking lot lanes, latitude and longitude coordinates of each lane, real-time number of occupied parking spaces, and parking records. The parking platform is further configured to monitor the occupancy of parking spaces in each parking lot based on the data, and when the parking lot is detected to be close to saturation, issue a queuing event confirmation command to the drones at each entrance of the parking lot. In response to the queuing event confirmation command, the drone is configured to take off and photograph the road conditions around the parking lot entrance. It uses an image recognition algorithm to determine whether there are vehicle queues and slow traffic. If a queuing event is determined to have occurred, the drone will report the event to the parking platform. The parking platform is further configured to receive events reported by the drone and send the event information to the guidance algorithm module; The guidance algorithm module is configured to receive queuing event information sent by the parking platform, combine the data parking lot guidance algorithm to output guidance strategies and path planning to the entrance channel of the destination parking lot, and send them to the corresponding drone. The drone is further configured to receive the guidance strategy and path planning sent by the guidance algorithm module, broadcast guidance information above the queuing vehicles based on the guidance strategy, prompt the vehicles to follow it to the destination parking lot, guide the vehicles to the destination parking lot according to the path planning, and end the guidance task and return to the starting point when it flies to the entrance channel of the destination parking lot. The queuing event information reported by the drone includes at least the event number, parking lot name, entrance lane number, queue length, number of vehicles in the queue, speed of surrounding vehicles, and reporting time. The parking lot guidance algorithm includes: Traverse all parking lots within 0.5km of the current parking lot that have a real-time saturation of less than or equal to 60%; Each eligible parking lot is evaluated using an incentive assessment index, which is: =50* +50* ； wherein, is an induction assessment index, is a saturation recommendation index, is a distance recommendation index; Parking lots are ranked from highest to lowest based on their guidance assessment index, and the number of guided vehicles that each parking lot can accommodate is calculated. The number of guided vehicles that each parking lot can accommodate is: =（80%- ）* ； wherein, is the number of parking spaces available for the xth parking lot, is the real-time parking saturation for the xth parking lot, is the number of parking spaces for the xth parking lot; Compare the number of vehicles in the queue with the number of guided vehicles that the first-priority parking lot can accommodate. If the number of vehicles in the queue is less than or equal to the number of guided vehicles that the first-priority parking lot can accommodate, then all vehicles in the queue will be guided to the first-priority parking lot. If the number of vehicles in the queue exceeds the number of guided vehicles that the first-priority parking lot can accommodate, then the number of guided vehicles that the first-priority parking lot can accommodate will be allocated to that parking lot. The remaining number of vehicles in the queue will then be compared with the number of guided vehicles that the second-priority parking lot can accommodate, until all vehicles in the queue are allocated or the maximum capacity of all eligible parking lots is reached.

2. The parking guidance system based on unmanned aerial vehicles (UAVs) according to claim 1, characterized in that, The criteria for determining whether a parking lot is nearing saturation, as monitored by the parking platform, are as follows: The real-time saturation of berths is greater than the preset threshold; or the real-time remaining number of berths is less than the preset threshold.

3. The parking guidance system based on unmanned aerial vehicles (UAVs) according to claim 1, characterized in that, The calculation rule for the saturation recommendation index is as follows: When the real-time parking saturation of the xth parking lot When it is greater than 60%, the saturation recommendation index is... =0; When the real-time parking saturation of the xth parking lot When the saturation recommendation index is less than or equal to 60% and greater than or equal to 20%, the saturation recommendation index is... =2.5*(0.6- ); When the real-time parking saturation of the xth parking lot When it is less than 20%, the saturation recommendation index is... The value is 1.

4. The parking guidance system based on unmanned aerial vehicles (UAVs) according to claim 3, characterized in that, The calculation rule for the distance recommendation index is as follows: When the x-th parking lot is in a straight line distance When the distance is greater than 0.5km, the recommended distance index is... =0; When the x-th parking lot is in a straight line distance When the distance is less than or equal to 0.5km and greater than or equal to 0.1km, the recommended distance index is... ; When the x-th parking lot is in a straight line distance When the distance is less than 0.1km, the recommended distance index is... The value is 1.

5. The parking guidance system based on unmanned aerial vehicles (UAVs) according to claim 1, characterized in that, The specific area is defined as a city, district, street, or business district.

6. The parking guidance system based on unmanned aerial vehicles (UAVs) according to claim 1, characterized in that, For parking lots where queues frequently occur, the entrance to the passage is set as the initial point for drones, and several drones are deployed to each parking lot based on the number of vehicles in the queue.

7. The parking guidance system based on unmanned aerial vehicles (UAVs) according to claim 1, characterized in that, After flying a certain distance, the drone uses an image recognition algorithm to identify whether there are queuing vehicles following it. If not, it flies back to the starting point and continues to broadcast guidance information.

8. The parking guidance system based on unmanned aerial vehicles (UAVs) according to claim 1, characterized in that, When the drone guides the vehicle to the destination parking lot according to the route plan, its flight speed matches the current road traffic speed.

Images