Indoor and outdoor seamless parking navigation method and system for automatic driving vehicle

Abstract

The invention discloses an automatic driving vehicle indoor and outdoor seamless parking navigation method and system, and belongs to the technical field of automatic driving. The method is executed by a management cloud platform, and comprises the following steps: receiving a parking request which is sent by a vehicle terminal and contains a current position, a destination and vehicle kinetic parameters; based on the real-time parking space state and short-term demand prediction, dynamically distributing a target parking lot and a parking space for the vehicle through a decision model, and generating a reservation voucher; a multi-target optimization algorithm is adopted, real-time road conditions, a parking lot internal structure and vehicle kinetic parameters are fused, a whole-course seamless navigation path from a current position to a target parking space is generated, the path is of a unified data structure, and outdoor roads and indoor parking lot road sections are fused; and the path and the voucher are issued to the vehicle terminal through the V2X Internet of Vehicles, and the vehicle is controlled to realize seamless switching from outdoor to indoor and accurate parking. According to the invention, continuous, efficient and reliable navigation in the whole parking process is realized.

Description

Technical Field

[0001] This application relates to the field of autonomous driving technology, and in particular to a seamless indoor and outdoor parking navigation method and system for autonomous vehicles. Background Technology

[0002] With the rapid development of autonomous driving technology, automatic parking has become a research hotspot. However, most existing automatic parking solutions have limitations. First, most systems only focus on path planning and control within the parking lot, failing to organically integrate with urban road navigation before the vehicle arrives at the parking lot, creating two separate stages: "outdoor navigation" and "indoor parking." When a vehicle enters an indoor parking space (such as an underground garage) from the outdoors, signal switching (such as GPS signal loss) and sudden environmental changes often lead to navigation interruptions or the need for replanning, resulting in inconsistent user experience and safety risks.

[0003] Secondly, the existing system lacks dynamic and proactive management of parking resources. Vehicles may arrive at the parking lot only to find no spaces available, leading to ineffective parking and increasing traffic congestion and energy consumption. While some technologies enable parking information lookup, they are not deeply integrated with route planning for optimization.

[0004] Furthermore, traditional path planning algorithms typically only consider path length or time, rarely taking into account the complex road conditions of public roads, the structural constraints of narrow passages inside parking lots, and the vehicle's own dynamic characteristics (such as turning radius) in a synchronous and unified manner. This may result in a planned path that is theoretically "optimal" but is infeasible or uncomfortable for autonomous vehicles in actual execution. Summary of the Invention

[0005] This application provides a seamless indoor and outdoor parking navigation method and system for autonomous vehicles, aiming to solve the problems of the existing technology, such as the interruption of experience caused by the disconnection between indoor and outdoor navigation stages, the ineffective patrol caused by the lack of dynamic and forward-looking management of parking space resources, and the failure of path planning to synchronously and uniformly optimize public road conditions, parking lot internal structure and vehicle dynamics characteristics, resulting in infeasible or uncomfortable paths.

[0006] Firstly, a method for seamless indoor and outdoor parking navigation of autonomous vehicles is provided, the method being executed by a management cloud platform, the method comprising:

[0007] S1, Receive a parking request sent from the vehicle terminal, the parking request including at least the vehicle's current location, destination information and vehicle dynamic parameters;

[0008] S2, based on the parking request, the real-time parking space status of each connected parking lot, and the short-term parking space demand prediction information, the decision model dynamically allocates the target parking lot and target parking space to the vehicle and generates the corresponding reservation voucher.

[0009] S3. Using the path planning module, a multi-objective optimization algorithm is adopted to integrate real-time traffic information, high-precision map structure information inside the target parking lot, and vehicle dynamic parameters to generate a seamless navigation path from the vehicle's current position to the target parking space. The seamless navigation path is a unified data structure that seamlessly integrates outdoor public road sections and indoor parking lot internal road sections.

[0010] S4, via V2X vehicle-to-everything (V2X) network, the seamless navigation route and the reservation voucher are sent to the vehicle terminal, enabling the vehicle terminal to control the vehicle to seamlessly switch from outdoor public road navigation to indoor parking lot navigation and accurately park based on the received seamless navigation route.

[0011] Optionally, in the above scheme, the dynamic allocation in step S2 specifically includes:

[0012] The management cloud platform calculates the allocation scheme based on the vehicle's destination, estimated arrival time, real-time parking space status of each parking lot, and parking space shortage prediction results, through a decision model that weighs distance, cost, and parking space availability.

[0013] If the parking space shortage in the preferred parking lot is higher than a preset threshold, the vehicle will be automatically assigned to a backup parking lot and parking space, and the assignment plan and comparison information will be pushed to the vehicle terminal for confirmation.

[0014] In the above scheme, optionally, in step S3, the optimization objectives of the multi-objective optimization algorithm include at least two of the following: total path travel time, energy consumption, and path smoothness;

[0015] The constraints of the multi-objective optimization algorithm include: traffic rules and real-time road conditions on public roads, the geometry of passages and obstacles inside parking lots, and the minimum turning radius and dynamic limits of vehicles.

[0016] In the above scheme, optionally, in step S3, the route planning module takes travel time and traffic rule compliance as the main optimization objectives when planning public road segments;

[0017] When planning the internal road sections of a parking lot, the main optimization objectives are path smoothness, safety, and compliance with the vehicle's mechanical steering limits.

[0018] In the above scheme, optionally, the seamless navigation path generated in step S3 is a unified data structure that seamlessly integrates and encodes outdoor public road segments and indoor parking lot internal road segments;

[0019] This allows the vehicle terminal to control the vehicle to travel from outdoors to indoors without having to replan the route due to a change in the positioning signal source, and the received control commands are continuous and uninterrupted.

[0020] Optionally, in the above scheme, the seamless handover in step S4 is specifically implemented as follows:

[0021] When vehicles are driving outdoors, they mainly rely on GNSS signals for positioning and navigation;

[0022] When a vehicle enters a parking lot and causes GNSS signal attenuation, positioning dominance is smoothly transferred to a UWB or Bluetooth beacon positioning system provided by the field facility.

[0023] During this process, the seamless navigation path used by the vehicle terminal remains unchanged, and the vehicle control system continues to execute coherent vehicle control commands.

[0024] Optionally, after step S4, the above scheme may further include:

[0025] S5. After the vehicle successfully parks in the target parking space, the vehicle terminal or site facility sends parking confirmation information to the management cloud platform.

[0026] The management cloud platform updates the parking space status to "occupied" and starts the billing process;

[0027] If the vehicle does not arrive within the reservation validity period, the management cloud platform will release the reservation voucher and update the parking space status according to the preset strategy.

[0028] In a second aspect, an indoor / outdoor seamless parking navigation system for autonomous vehicles, used to perform the method described above, the system comprising:

[0029] The vehicle terminal, configured on autonomous vehicles, is used to initiate parking requests, receive navigation routes and reservation credentials, and execute autonomous driving control.

[0030] The management cloud platform communicates with the vehicle terminal via V2X vehicle networking for dynamically allocating parking spaces, generating reservation vouchers, and planning seamless navigation routes throughout the entire process;

[0031] The site-end facilities are deployed inside the parking lot to provide high-precision indoor positioning signals and real-time parking space status data.

[0032] Optionally, in the above scheme, the management cloud platform integrates at least:

[0033] The parking management module is used to maintain high-precision maps and real-time parking status of each parking lot, and to manage the lifecycle of parking space reservation vouchers;

[0034] The demand forecasting module is used to predict the short-term parking space demand of each parking lot based on historical data and real-time requests.

[0035] The path planning module is used to generate the seamless navigation path throughout the entire process using a multi-objective optimization algorithm.

[0036] Optionally, in the above scheme, the vehicle terminal includes a V2X communication unit, a high-precision positioning unit, and an environmental perception sensor group;

[0037] The site-end facilities include indoor high-precision positioning beacons, parking space status detection sensors, and site-end communication gateways.

[0038] Compared with the prior art, this application has at least the following beneficial effects:

[0039] Based on further analysis and research into existing technical problems, this application recognizes that the aforementioned issues of fragmented navigation, static resource management, and singular planning constraints are not isolated phenomena. Their root cause lies in the lack of a central hub capable of overseeing the entire system and conducting cross-environmental collaborative scheduling and planning. This application constructs an execution architecture with a management cloud platform as the single decision-making core, deeply integrating dynamic parking space reservation (based on real-time and predictive data) with integrated indoor and outdoor path planning (based on multi-objective optimization and a unified data structure). This achieves a seamless, dynamic, feasible, and comfortable autonomous driving parking effect at the system level, from public roads to indoor parking spaces, through a "reservation-driven planning, planning-guaranteed reservation" approach.

[0040] This application also has at least the following effects:

[0041] 1. Truly seamless experience: By integrating end-to-end route planning and indoor and outdoor positioning technologies, the system eliminates navigation transitions and provides users with a smooth and seamless parking experience.

[0042] 2. Optimal global efficiency: The dynamic reservation mechanism avoids ineffective parking and optimizes the utilization rate of parking resources and regional traffic flow at the system level; multi-objective path planning improves parking efficiency and economy at the individual vehicle level.

[0043] 3. Enhanced safety and comfort: The path planning process fully considers vehicle dynamics and complex environmental constraints, resulting in a path that better matches the handling characteristics of autonomous vehicles. This avoids aggressive or impractical driving commands, thereby improving safety and passenger comfort. Attached Figure Description

[0044] Figure 1A flowchart illustrating a seamless indoor and outdoor parking navigation method for an autonomous vehicle, provided as an embodiment of this application;

[0045] Figure 2 A flowchart illustrating a seamless indoor and outdoor parking navigation method for an autonomous vehicle, provided as another embodiment of this application;

[0046] Figure 3 This application provides an overall architecture diagram of an indoor and outdoor seamless parking navigation system for autonomous vehicles, as one embodiment of the present application.

[0047] Figure 4 This application provides a data interaction diagram of an indoor and outdoor seamless parking navigation system for autonomous vehicles, as one embodiment of the present application.

[0048] Figure 5 A schematic diagram illustrating the workflow of a path planning module performing multi-objective optimization according to an embodiment of this application;

[0049] Figure 6 This is a schematic diagram illustrating a scenario where a vehicle achieves seamless navigation from an outdoor road to an indoor parking space, as provided in one embodiment of this application. Detailed Implementation

[0050] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0051] In the description of this application, unless otherwise stated, the terms "including", "comprising", "having", etc., also mean "not limited to" (certain units, components, materials, steps, etc.).

[0052] In the technical solution of this application, the user information (including but not limited to user personal information, user image information, user device information, such as location information) and data (including but not limited to data used for analysis, stored data, and displayed data) involved are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of the relevant data all comply with the relevant laws, regulations, and standards of the relevant countries and regions, and necessary confidentiality measures have been taken. They do not violate public order and good morals, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0053] In some embodiments, reference Figures 1-6 A seamless indoor and outdoor parking navigation method for autonomous vehicles is provided, the method being executed by a management cloud platform, the method comprising:

[0054] S1, Receive a parking request sent from the vehicle terminal, the parking request including at least the vehicle's current location, destination information and vehicle dynamic parameters;

[0055] S2, based on the parking request, the real-time parking space status of each connected parking lot, and the short-term parking space demand prediction information, the decision model dynamically allocates the target parking lot and target parking space to the vehicle and generates the corresponding reservation voucher.

[0056] S3. Using the path planning module, a multi-objective optimization algorithm is adopted to integrate real-time traffic information, high-precision map structure information inside the target parking lot, and vehicle dynamic parameters to generate a seamless navigation path from the vehicle's current position to the target parking space. The seamless navigation path is a unified data structure that seamlessly integrates outdoor public road sections and indoor parking lot internal road sections.

[0057] S4, via V2X vehicle-to-everything (V2X) network, the seamless navigation route and the reservation voucher are sent to the vehicle terminal, enabling the vehicle terminal to control the vehicle to seamlessly switch from outdoor public road navigation to indoor parking lot navigation and accurately park based on the received seamless navigation route.

[0058] In some embodiments, the seamless indoor and outdoor parking navigation method for autonomous vehicles clearly demonstrates how the core functions of the system are implemented through the specific execution process of the management cloud platform. This method uses the cloud platform as a single control hub, tightly coupling dynamic resource scheduling with cross-environment path planning to form a complete technological closed loop.

[0059] Specifically, in step S1, the management cloud platform acts as an information aggregation point, receiving structured requests from vehicle terminals and providing precise input for subsequent decision-making. Step S2 is the core manifestation of the "dynamic reservation" function in this embodiment: the cloud platform does not simply query based on a static parking space list, but integrates real-time status, short-term prediction, and decision-making models to perform intelligent resource allocation and locking, generating a reservation voucher with timeliness and exclusivity, thereby avoiding invalid parking from the source. Step S3 then takes over the output of step S2 (i.e., the determined target parking space) and generates a "seamless" path. The key to this step is that the path planning module uses a multi-objective optimization algorithm, which, within a unified optimization framework, simultaneously processes real-time traffic information of outdoor roads, the precise geometric structure of indoor parking lots, and the physical passability constraints of vehicles (such as dynamic parameters), and outputs a navigation path with a unified data structure and spatial continuity. This ensures that the indoor and outdoor sections are organically integrated during the path planning stage. Finally, in step S4, the cloud platform sends out the complete navigation instructions and reservation voucher together through V2X vehicle networking, enabling the vehicle terminal to control the vehicle to autonomously complete the entire process from driving on public roads to precise parking indoors according to this pre-planned, uninterrupted route, truly realizing a seamless "end-to-end" autonomous driving experience.

[0060] This embodiment, through the above four logically rigorous steps, concretizes how to solve the problems of fragmented navigation phase, lack of foresight in resource management, and single planning constraints mentioned in the background art.

[0061] In some embodiments, in step S2, the dynamic allocation specifically includes:

[0062] The management cloud platform calculates the allocation scheme based on the vehicle's destination, estimated arrival time, real-time parking space status of each parking lot, and parking space shortage prediction results, through a decision model that weighs distance, cost, and parking space availability.

[0063] If the parking space shortage in the preferred parking lot is higher than a preset threshold, the vehicle will be automatically assigned to a backup parking lot and parking space, and the assignment plan and comparison information will be pushed to the vehicle terminal for confirmation.

[0064] In some embodiments, the input information used by the management cloud platform for allocation is multi-dimensional: it includes both the immediate intent of vehicle requests (destination, estimated arrival time) and system data reflecting the current state and future trends of resources (real-time parking space status, parking space availability prediction results). Through comprehensive weighing and optimization calculations of multiple objectives such as distance, cost, and parking space availability, the decision model can output a more reasonable allocation scheme from a global system perspective, rather than just the optimal choice for the user alone. When the system predicts that the preferred parking lot is about to become saturated (susceptibility exceeds the threshold), it does not passively wait for reservation conflicts to occur, but proactively and preventively guides vehicles to alternative parking lots with more capacity. This mechanism effectively smooths the load between parking lots at the system level, avoiding localized congestion.

[0065] In some embodiments, in step S3, the optimization objectives of the multi-objective optimization algorithm include at least two of the following: total path travel time, energy consumption, and path smoothness.

[0066] The constraints of the multi-objective optimization algorithm include: traffic rules and real-time road conditions on public roads, the geometry of passages and obstacles inside parking lots, and the minimum turning radius and dynamic limits of vehicles.

[0067] In some embodiments, step S3 uses the total travel time, energy consumption, and path smoothness as optimization objectives, either individually or in combination. This means that the planning module no longer pursues the theoretical optimum of a single dimension (such as the shortest distance), but rather strives to find a feasible solution that achieves the best balance among multiple aspects such as efficiency, economy, and travel experience. For example, a slightly detour route with fewer traffic lights and smoother starts and stops may have better overall performance in terms of total time and energy consumption.

[0068] More importantly, the constraints are, for the first time, uniformly and explicitly incorporated into three key categories at the algorithm level: public road rules and real-time traffic conditions: ensuring the social compliance and real-time feasibility of the path; parking lot internal geometry: using static obstacles such as passage width, column positions, and curve angles as hard constraints to avoid planning impassable paths; and vehicle dynamics limits: explicitly using parameters such as the minimum turning radius of vehicles as constraints, fundamentally ensuring that every planned path—especially narrow curves inside the parking lot—is physically feasible and safe for the specific vehicle performing the task.

[0069] Therefore, in this embodiment, step S3 transforms the requirement of "generating a seamless navigation path from the vehicle's current location to the target parking space" from an abstract functional description into a technical task with a clear optimization objective, strict constraints, and a specific algorithmic process. This ensures that the path ultimately issued to the vehicle is a pre-verified, fully validated, and truly safe and comfortable "end-to-end" optimal path that takes into account legal rules, environmental structure, and vehicle physical characteristics, thereby solving the fundamental problem in the background technology where "the planned path is infeasible or uncomfortable in actual execution."

[0070] In some embodiments, in step S3, the route planning module prioritizes travel time and traffic rule compliance as the main optimization objectives when planning public road segments.

[0071] When planning the internal road sections of a parking lot, the main optimization objectives are path smoothness, safety, and compliance with the vehicle's mechanical steering limits.

[0072] In some embodiments, when planning public road segments in step S3, the system prioritizes travel time and traffic rule compliance. This is because efficiency and legal compliance are core requirements in open public road traffic, and the optimization algorithm needs to prioritize dynamic and static rules such as real-time traffic flow, traffic lights, and lane regulations in order to plan efficient and legal driving routes.

[0073] When planning the internal road sections of a parking lot, the optimization focus shifts to path smoothness, safety, and compliance with the vehicle's mechanical steering limits. This is because the parking lot environment is characterized by narrow passages, static and complex obstacles (such as pillars and walls), low driving speeds, and extremely high requirements for handling precision. At this stage, "smoothness" is directly related to ride comfort and control stability, "safety" is reflected in maintaining sufficient distance from static obstacles, and "compliance with the vehicle's mechanical steering limits" (such as the minimum turning radius) is a fundamental prerequisite for the path to be physically executed by the vehicle. This shift in optimization objectives ensures that the final "last hundred meters" of the path generated for the vehicle is not only passable but also easily executed smoothly, accurately, and safely by the autonomous driving system.

[0074] The differentiated optimization strategy disclosed in this embodiment enables the generated "seamless navigation path" to achieve a balance between macro-efficiency and micro-feasibility in its internal logic. Specifically, the public road sections ensure the efficiency and compliance of the overall journey, while the parking lot internal sections guarantee the comfort, safety, and absolute feasibility of the parking maneuver, ultimately fulfilling the promise of a "seamless" and "reliable" user experience. This effectively solves the inherent problem in the background technology of path planning being difficult to synchronize and uniformly consider different environmental constraints.

[0075] In some embodiments, the seamless navigation path generated in step S3 is a unified data structure that seamlessly integrates and encodes outdoor public road segments and indoor parking lot internal road segments.

[0076] This allows the vehicle terminal to control the vehicle to travel from outdoors to indoors without having to replan the route due to a change in the positioning signal source, and the received control commands are continuous and uninterrupted.

[0077] In some embodiments, this unified data structure serves as the foundation for seamless integration at the physical layer. By encoding and representing outdoor public road segments and indoor parking lot segments in an integrated manner (e.g., using the same coordinate system, continuous segment node IDs, and unified geometric and attribute description specifications), the path blurs the technical boundaries between indoor and outdoor environments at the data level, forming a logically continuous and complete digital twin path.

[0078] It is precisely this unified data structure that enables the vehicle terminal to achieve a continuous and uninterrupted command flow during control execution. Upon receiving this path, the vehicle terminal can directly interpret it into a series of sequential control commands (such as steering, acceleration, and braking). When the vehicle moves from a GNSS signal coverage area into an indoor area relying on UWB / Bluetooth beacons, the switching of the positioning signal source is only used to match the vehicle's specific position on the predetermined continuous path in real time, without triggering any path recalculation, splicing, or command sequence reset. This not only solves the problem of navigation interruption but also simplifies the vehicle terminal's processing logic and improves the reliability and real-time performance of the entire system.

[0079] In some embodiments, the seamless switching in step S4 is specifically implemented as follows:

[0080] When vehicles are driving outdoors, they mainly rely on GNSS signals for positioning and navigation;

[0081] When a vehicle enters a parking lot and causes GNSS signal attenuation, positioning dominance is smoothly transferred to a UWB or Bluetooth beacon positioning system provided by the field facility.

[0082] During this process, the seamless navigation path used by the vehicle terminal remains unchanged, and the vehicle control system continues to execute coherent vehicle control commands.

[0083] In some embodiments, the system acknowledges and anticipates the inevitable switching of positioning signal sources. Instead of forcing the use of a single signal, it designs a leadership handover process. When the vehicle is outdoors primarily using GNSS signals, and enters an indoor area where signals attenuate, the system smoothly transitions positioning leadership to a UWB or Bluetooth beacon positioning system deployed by the field facility. This "smoothness" means that the handover process is achieved through sensor fusion algorithms, ensuring that the vehicle's position calculation does not experience jumps or interruptions, providing continuous and reliable state input to the control layer.

[0084] Regardless of how the underlying positioning signal source switches, the "highest command" upon which the vehicle control system relies—the seamless navigation path—remains constant and unique. Therefore, the control system consistently executes a coherent sequence of vehicle control commands derived from the same global path. Switching positioning sources is only used to more accurately determine the vehicle's real-time progress along this fixed path, without triggering path replanning, command sequence resets, or interruptions to control logic.

[0085] In some embodiments, after step S4, the method further includes:

[0086] S5. After the vehicle successfully parks in the target parking space, the vehicle terminal or site facility sends parking confirmation information to the management cloud platform.

[0087] The management cloud platform updates the parking space status to "occupied" and starts the billing process;

[0088] If the vehicle does not arrive within the reservation validity period, the management cloud platform will release the reservation voucher and update the parking space status according to the preset strategy.

[0089] In some embodiments, the system receives a confirmation signal that the operation was successful by automatically sending a "parking confirmation message" through the vehicle terminal or site facilities. Based on this, the management cloud platform atomically updates the parking space status from "reserved" to "occupied," achieving precise synchronization between the digital world resource status and the actual physical world situation, and providing a reliable trigger point for initiating billing.

[0090] "Initiating the billing process" signifies that the service has moved from a simple navigation aid to a commercial operation phase, giving the entire technical solution a complete commercial implementation logic.

[0091] The rule that "if the vehicle does not arrive within the reservation validity period, then... release the reservation voucher and update the parking space status" is the ultimate guarantee of the dynamic reservation concept in this embodiment. It prevents "zombie reservations" from occupying resources for a long time due to plan changes or accidents, and ensures that parking space resources can be automatically reclaimed and re-entered into the scheduling pool according to preset strategies, thereby maintaining the overall resource utilization efficiency and service robustness of the system.

[0092] In some embodiments, an indoor and outdoor seamless parking navigation system for autonomous vehicles is used to perform the method provided in the embodiments described above, the system comprising:

[0093] The vehicle terminal, configured on autonomous vehicles, is used to initiate parking requests, receive navigation routes and reservation credentials, and execute autonomous driving control.

[0094] The management cloud platform communicates with the vehicle terminal via V2X vehicle networking for dynamically allocating parking spaces, generating reservation vouchers, and planning seamless navigation routes throughout the entire process;

[0095] The site-end facilities are deployed inside the parking lot to provide high-precision indoor positioning signals and real-time parking space status data.

[0096] In some embodiments, the management cloud platform integrates at least:

[0097] The parking management module is used to maintain high-precision maps and real-time parking status of each parking lot, and to manage the lifecycle of parking space reservation vouchers;

[0098] The demand forecasting module is used to predict the short-term parking space demand of each parking lot based on historical data and real-time requests.

[0099] The path planning module is used to generate the seamless navigation path throughout the entire process using a multi-objective optimization algorithm.

[0100] In some embodiments, the vehicle terminal includes a V2X communication unit, a high-precision positioning unit, and an environmental perception sensor group;

[0101] The site-end facilities include indoor high-precision positioning beacons, parking space status detection sensors, and site-end communication gateways.

[0102] In this embodiment, a seamless indoor and outdoor parking navigation system for autonomous vehicles based on dynamic reservation is provided. The system includes:

[0103] Vehicle terminal: As the vehicle's "intelligent driver," it perceives the surrounding environment through high-precision positioning units (such as GNSS / IMU fusion positioning) and environmental perception sensors (such as cameras and LiDAR). Its core is the V2X communication unit, which is responsible for real-time data exchange with the cloud platform.

[0104] The management cloud platform is the "command center" of the entire system. It comprises three core modules:

[0105] Parking management module: Maintains a dynamic map covering multiple parking lots, updates the status of each parking space (vacant / occupied / reserved) in real time, and manages the lifecycle of parking space reservations.

[0106] Demand forecasting module: Using time series analysis or machine learning algorithms, it predicts the parking space demand trend of each parking lot in the next 15-60 minutes, providing data support for dynamic scheduling.

[0107] Path planning module: This is the core technology of this invention. It employs a multi-objective optimization algorithm (such as a constrained genetic algorithm or an A* variant) to simultaneously plan paths for vehicles on both public road segments and within parking lots, within a unified framework. Optimization objectives include not only minimizing time but also minimizing energy consumption and ensuring the smoothest path. Constraints encompass traffic rules, lane markings, static obstacles such as pillars and walls within the parking lot, and dynamic limitations such as the vehicle's minimum turning radius.

[0108] Parking facility: These are the system's "tentacles" in the parking lot. They include UWB ultra-wideband positioning base stations or Bluetooth beacons deployed within the parking lot to provide centimeter-level indoor positioning; parking space status sensors (such as video stakes or geomagnetic sensors) to confirm the actual status of parking spaces; and parking communication gateways to ensure smooth indoor and outdoor network connectivity.

[0109] Example 1

[0110] refer to Figures 2-4 A user drives an autonomous vehicle equipped with the vehicle terminal of this invention to a downtown business district. When approximately 2 kilometers from the destination, the vehicle terminal automatically sends a parking request to the management cloud platform. Upon receiving the request, the platform's parking management module checks real-time data from two adjacent smart parking garages, A and B. Based on historical data, the demand prediction module determines that parking garage A will be nearly full in 15 minutes, while parking garage B will still have ample parking spaces.

[0111] Based on the strategy of "closest proximity and ample parking spaces," the platform assigned the vehicle parking space B305 on the 3rd floor of Parking Garage B and generated a reservation voucher valid for 20 minutes. Subsequently, the route planning module was activated. It obtained the current city road congestion situation (showing an accident on one main road), a high-precision map of Parking Garage B (including information such as lane width and turning angles), and the vehicle's minimum turning radius. Through multi-objective optimization calculations, it generated a route: guiding the vehicle to bypass the congested main road, reach the entrance of Parking Garage B via a secondary road, and then planning a smooth, non-sharp in-parking route directly to parking space B305.

[0112] The complete route and reservation voucher are sent to the vehicle's terminal. The vehicle drives automatically along the route, and when it enters the parking garage with weak GPS signal, it seamlessly switches to the UWB positioning system inside the building, and finally smoothly and accurately parks in parking space B305, sending a parking success confirmation to the platform.

[0113] Example 2

[0114] In a high-end office park setting, an autonomous vehicle equipped with an energy-saving mode sends a parking request. During route planning, the management cloud platform, based on the vehicle user's settings, prioritizes "optimal energy efficiency" as the primary optimization objective. The route planning module generates a path that minimizes sudden acceleration, braking, and frequent turns, and prioritizes lanes with gentler gradients to enter the underground parking garage. While this path may not be the absolute shortest in terms of time, it significantly reduces overall energy consumption.

[0115] This application provides a safe, efficient, and comfortable complete parking solution for autonomous vehicles through the deep integration of dynamic reservation and seamless indoor and outdoor path planning, which strongly promotes the practical application of smart parking and autonomous driving technologies.

[0116] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

Claims

1. A seamless indoor and outdoor parking navigation method for autonomous vehicles, characterized in that, The method is executed by a management cloud platform, and the method includes: S1, Receive a parking request sent from the vehicle terminal, the parking request including at least the vehicle's current location, destination information and vehicle dynamic parameters; S2, based on the parking request, the real-time parking space status of each connected parking lot, and the short-term parking space demand prediction information, the decision model dynamically allocates the target parking lot and target parking space to the vehicle and generates the corresponding reservation voucher. S3. Using the path planning module, a multi-objective optimization algorithm is adopted to integrate real-time traffic information, high-precision map structure information inside the target parking lot, and vehicle dynamic parameters to generate a seamless navigation path from the vehicle's current position to the target parking space. The seamless navigation path is a unified data structure that seamlessly integrates outdoor public road sections and indoor parking lot internal road sections. S4, via V2X vehicle-to-everything (V2X) network, the seamless navigation route and the reservation voucher are sent to the vehicle terminal, enabling the vehicle terminal to control the vehicle to seamlessly switch from outdoor public road navigation to indoor parking lot navigation and accurately park based on the received seamless navigation route.

2. The method according to claim 1, characterized in that, In step S2, the dynamic allocation specifically includes: The management cloud platform calculates the allocation scheme based on the vehicle's destination, estimated arrival time, real-time parking space status of each parking lot, and parking space shortage prediction results, through a decision model that weighs distance, cost, and parking space availability. If the parking space shortage in the preferred parking lot is higher than a preset threshold, the vehicle will be automatically assigned to a backup parking lot and parking space, and the assignment plan and comparison information will be pushed to the vehicle terminal for confirmation.

3. The method according to claim 1, characterized in that, In step S3, the optimization objectives of the multi-objective optimization algorithm include at least two of the following: total path travel time, energy consumption, and path smoothness. The constraints of the multi-objective optimization algorithm include: traffic rules and real-time road conditions on public roads, the geometry of passages and obstacles inside parking lots, and the minimum turning radius and dynamic limits of vehicles.

4. The method according to claim 3, characterized in that, In step S3, the route planning module prioritizes travel time and compliance with traffic rules as the main optimization objectives when planning public road segments. When planning the internal road sections of a parking lot, the main optimization objectives are path smoothness, safety, and compliance with the vehicle's mechanical steering limits.

5. The method according to claim 1, characterized in that, The seamless navigation path generated in step S3 is a unified data structure that seamlessly integrates and encodes outdoor public road segments and indoor parking lot internal road segments; This allows the vehicle terminal to control the vehicle to travel from outdoors to indoors without having to replan the route due to a change in the positioning signal source, and the received control commands are continuous and uninterrupted.

6. The method according to claim 1, characterized in that, In step S4, the seamless handover is specifically implemented as follows: When vehicles are driving outdoors, they mainly rely on GNSS signals for positioning and navigation; When a vehicle enters a parking lot and causes GNSS signal attenuation, positioning dominance is smoothly transferred to a UWB or Bluetooth beacon positioning system provided by the field facility. During this process, the seamless navigation path used by the vehicle terminal remains unchanged, and the vehicle control system continues to execute coherent vehicle control commands.

7. The method according to claim 1, characterized in that, Following step S4, the following is also included: S5. After the vehicle successfully parks in the target parking space, the vehicle terminal or site facility sends parking confirmation information to the management cloud platform. The management cloud platform updates the parking space status to "occupied" and starts the billing process; If the vehicle does not arrive within the reservation validity period, the management cloud platform will release the reservation voucher and update the parking space status according to the preset strategy.

8. An indoor and outdoor seamless parking navigation system for autonomous vehicles, used to perform the method as described in any one of claims 1-7, characterized in that, The system includes: The vehicle terminal, configured on autonomous vehicles, is used to initiate parking requests, receive navigation routes and reservation credentials, and execute autonomous driving control. The management cloud platform communicates with the vehicle terminal via V2X vehicle networking for dynamically allocating parking spaces, generating reservation vouchers, and planning seamless navigation routes throughout the entire process; The site-end facilities are deployed inside the parking lot to provide high-precision indoor positioning signals and real-time parking space status data.

9. The system according to claim 8, characterized in that, The management cloud platform integrates at least the following: The parking management module is used to maintain high-precision maps and real-time parking status of each parking lot, and to manage the lifecycle of parking space reservation vouchers; The demand forecasting module is used to predict the short-term parking space demand of each parking lot based on historical data and real-time requests. The path planning module is used to generate the seamless navigation path throughout the entire process using a multi-objective optimization algorithm.

10. The system according to claim 8, characterized in that, The vehicle terminal includes a V2X communication unit, a high-precision positioning unit, and an environmental perception sensor group. The site-end facilities include indoor high-precision positioning beacons, parking space status detection sensors, and site-end communication gateways.

Images