Vehicle control method and device and vehicle

Abstract

The vehicle control method comprises the steps that demand information is obtained, the demand information comprises parking demand information and calling demand information, and the calling demand information comprises a preset calling point; vehicle scene information is obtained, a global scene model is generated based on the vehicle scene information, and the vehicle scene information comprises vehicle state information and vehicle environment information; generating an automatic parking strategy and first path information based on the parking demand information and a global scene model; controlling the vehicle to park autonomously according to the automatic parking strategy; generating a fixed-point calling strategy based on the calling demand information, the first path information and the global scene model; and controlling the vehicle to run to a preset calling point according to the fixed-point calling strategy. According to the invention, the process success rate and response efficiency of parking and calling are improved through the construction of the global scene model and the linkage decision of autonomous parking and fixed-point calling.

Description

Technical Field

[0001] This invention relates to the field of intelligent driving technology, and in particular to a vehicle control method, device, and vehicle. Background Technology

[0002] With the development of intelligent driving technology, autonomous parking and point-to-point summoning have become core intelligent functions of vehicles. The two constitute a two-way service system of autonomous parking and point-to-point summoning. Autonomous parking enables the vehicle to automatically park in the target parking space from the user's drop-off point, while point-to-point summoning enables the vehicle to automatically drive from the parking space to the user's designated summoning point, providing users with a convenient car use experience.

[0003] In existing technologies, autonomous parking and fixed-point summoning use independent decision-making logic. This separation of decision-making directly leads to a high interruption rate in the entire parking and summoning process, as well as significant command response delays, which cannot meet users' needs for smooth and efficient use of the functions.

[0004] To solve the above-mentioned technical problems, it is urgent to propose a decision-making and control scheme that integrates autonomous parking and fixed-point summoning. Summary of the Invention

[0005] This invention provides a vehicle control method that improves the success rate and response efficiency of parking and summoning processes by constructing a global scene model and linking it with autonomous parking and fixed-point summoning decisions.

[0006] According to one aspect of the present invention, a vehicle control method is provided, comprising: Obtain demand information, which includes parking demand information and summoning demand information, and the summoning demand information includes preset summoning points; Obtain vehicle scene information and generate a global scene model based on the vehicle scene information, wherein the vehicle scene information includes vehicle status information and vehicle environment information; Based on the parking demand information and the global scene model, an automatic parking strategy and first path information are generated. The vehicle is controlled to park autonomously according to the automatic parking strategy; A fixed-point summoning strategy is generated based on the summoning demand information, the first path information, and the global scene model. The vehicle is controlled to travel to the preset summoning point according to the fixed-point summoning strategy.

[0007] Optionally, the types of the global scene model include autonomous parking scenarios in open spaces, autonomous parking scenarios in narrow parking spaces, fixed-point summoning scenarios on open roads, fixed-point summoning scenarios in narrow passages, fixed-point summoning scenarios for emergency response, and parking summoning switching scenarios.

[0008] Optionally, after obtaining vehicle scene information and generating a global scene model based on the vehicle scene information, the process further includes: Determine the type of the global scene model; The weight parameters for security factors, efficiency factors, accuracy factors, and user preference factors are assigned based on the type of the global scene model.

[0009] Optionally, after obtaining vehicle scene information and generating a global scene model based on the vehicle scene information, the process further includes: Determine the type of the global scene model; Calculate the actual drivable distance between the vehicle's parking location and the preset call point; The remaining driving range of a vehicle is obtained based on vehicle status information; Compare the actual drivable distance with the vehicle's remaining range; When the actual drivable distance is less than the vehicle's remaining range, the weight parameters of safety factors, efficiency factors, accuracy factors, and user preference factors are assigned according to the type of the global scenario model. When the actual drivable distance is greater than the vehicle's remaining range, adjust the weights of safety and efficiency factors corresponding to the summoning scenario; The safety factors include the path feasibility sub-factor.

[0010] Optionally, generating an automatic parking strategy and first path information based on the parking demand information and the global scene model includes: using an improved Astar algorithm to generate an automatic parking strategy and first path information based on the global scene model, the weight parameters, and the parking demand information. The automatic parking strategy includes a parking driving path and a parking space location; the first path information includes parking path information and parking environment information.

[0011] Optionally, generating a fixed-point summoning strategy based on the summoning demand information, the first path information, and the global scene model includes: using the Dstar-Lite dynamic path algorithm to generate a fixed-point summoning strategy based on the global scene model, the first path information, the weight parameters, and the summoning demand information. The fixed-point summoning strategy includes the summoning driving path, driving speed, obstacle avoidance sequence, and docking information.

[0012] Optionally, the step of controlling the vehicle to park autonomously according to the automatic parking strategy includes: Based on the weight parameters corresponding to the parking summon switching scenario, the vehicle is controlled to enter the summon standby state; It can identify fixed-point summoning commands and exit the summoning standby state when it receives a fixed-point summoning command.

[0013] Optionally, the step of "controlling the vehicle to a preset summoning point according to the fixed-point summoning strategy" includes: The vehicle's movement is controlled based on the aforementioned fixed-point summoning strategy; Real-time acquisition of vehicle driving information and updated summoning request information, wherein the vehicle driving information includes obstacle information and sensor status information; The fixed-point summoning strategy is adjusted based on the vehicle driving information, the updated summoning demand information, and the vehicle scene information, and the vehicle is controlled to drive to the preset summoning point according to the adjusted fixed-point summoning strategy.

[0014] According to a second aspect of the present invention, a vehicle control system is provided, comprising: The information acquisition module is used to acquire vehicle scene information and demand information, including parking demand information and summoning demand information. The decision-making module, connected to the information acquisition module, is used to generate a global scene model based on the vehicle scene information. The parking planning module, connected to the information acquisition module and the decision-making module, is used to generate an automatic parking strategy and first path information based on the parking demand information and the global scene model. The execution control module is connected to the information acquisition module, the parking planning module, and the summoning planning module; the execution control module controls the vehicle to park autonomously according to the automatic parking strategy, and controls the vehicle to drive to the preset summoning point according to the fixed-point summoning strategy.

[0015] According to a third aspect of the present invention, a vehicle is provided, including a processor and a memory, the memory storing a computer program, the processor being configured to execute the computer program to implement the vehicle control method described above.

[0016] The vehicle control method in the technical solution provided in this embodiment of the invention generates a global scene model by acquiring vehicle status information and vehicle environment information, providing a unified information base for autonomous parking and fixed-point summoning. At the same time, when generating the automatic parking strategy, the first path information is generated simultaneously and used as the basis for generating the fixed-point summoning strategy. Furthermore, by synchronously acquiring parking and summoning demand information and establishing a sequential decision execution logic of parking first and then summoning, the collaborative linkage and dynamic decision-making of autonomous parking and fixed-point summoning are realized, thereby improving the success rate and response efficiency of the parking and summoning process.

[0017] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart of a vehicle control method provided in an embodiment of the present invention; Figure 2 This is a flowchart of step S2 of the vehicle control method provided in the embodiment of the present invention; Figure 3 This is a flowchart of step S6 of the vehicle control method provided in the embodiment of the present invention. Detailed Implementation

[0020] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0021] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, any variations of the terms "comprising" and "having" are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0022] To address the aforementioned technical problems, the embodiments of the present invention provide the following technical solutions: Figure 1 This is a flowchart of a vehicle control method provided in an embodiment of the present invention, see reference. Figure 1 This invention provides a vehicle control method, including: Step S1: Obtain demand information, which includes parking demand information and summoning demand information, and the summoning demand information includes preset summoning points; Step S2: Obtain vehicle scene information and generate a global scene model based on the vehicle scene information, wherein the vehicle scene information includes vehicle status information and vehicle environment information; Step S3: Generate an automatic parking strategy and first path information based on the parking demand information and the global scene model; Step S4: Control the vehicle to park autonomously according to the automatic parking strategy; Step S5: Generate a fixed-point summoning strategy based on the summoning demand information, the first path information, and the global scene model; Step S6: Control the vehicle to drive to the preset summoning point according to the fixed-point summoning strategy.

[0023] Specifically, the vehicle control method in this embodiment is based on the basic logic of perception-analysis-decision-execution, realizing the linkage between autonomous parking and fixed-point summoning. The specific workflow is as follows: First, the information acquisition module receives parking and summoning request information issued by the user and extracts the core parameters, such as the target parking space and the preset summoning point. Second, the information acquisition module, which is capable of multimodal perception, completes the synchronous collection of vehicle status information and vehicle environment information. After removing redundant and noisy data through a data fusion algorithm, the decision module constructs a global scene model to provide a unified environmental basis for parking and summoning decisions. Third, the parking planning module, based on the global scene model and combined with parking request information, performs path planning and strategy formulation to generate an automatically executable parking strategy and simultaneously generates the first path information as the basis for connecting autonomous parking and fixed-point summoning. Fourth, the execution control module... The system receives automatic parking strategy commands and controls the vehicle's steering, accelerator, brakes, gear shifting, and other mechanisms to complete autonomous parking. The fifth step involves the call planning module reusing the global scene model and first path information, combined with call request information, to plan the call driving path and formulate a strategy, generating a fixed-point call strategy. The sixth step involves the execution control module receiving the fixed-point call strategy command and controlling the vehicle to start from the parking space and travel along the planned call driving path to the user-defined preset call point, completing the fixed-point call. The automatic parking strategy refers to the complete execution plan for enabling the vehicle to autonomously park in the target parking space, including parking driving path planning, driving speed control, precise parking space matching, obstacle avoidance, and other execution rules. The fixed-point call strategy refers to the complete execution plan for enabling the vehicle to travel from the parking space to the preset call point, including call driving path planning, driving speed control, obstacle avoidance, precise parking, and other execution rules.

[0024] In this embodiment, to more intuitively reflect the overall situation of the vehicle's own state and the surrounding static and dynamic environment, and thus generate a global scene model, the vehicle scene information in this embodiment includes vehicle state information and vehicle environment information. Vehicle state information includes the vehicle's own operating status and hardware status data, specifically remaining battery power, power system status, braking system status, steering system status, sensor accuracy, vehicle range, and the vehicle's current position / attitude. Vehicle environment information includes external environment data surrounding the vehicle, specifically parking space information, obstacle information in the parking / summoning area, road structure parameters, traffic signals, traffic rules, slope and curve parameters, and road topology. Parking demand information refers to the personalized demands made by the user for autonomous parking, including parking priority (safety priority / precision priority / efficiency priority), parking mode, and parking speed limits. Summoning demand information refers to the personalized demands made by the user for fixed-point summoning, including preset summoning point coordinates, summoning point accuracy (±0.5m~±1m), path preference (safest path / shortest path / no-slope path), response priority (instant response / delayed response), and temporary adjustment permissions (allowing change of summoning point / emergency pause).

[0025] To better connect the autonomous parking and point-to-point summoning processes and reduce redundant data collection during point-to-point summoning, the driving path and surrounding environment data associated with the autonomous parking process are stored as the first path information after autonomous parking is completed. The first path information can be directly used as the starting path for point-to-point summoning and can be directly called in the point-to-point summoning strategy, which greatly improves the generation efficiency of the point-to-point summoning strategy.

[0026] This embodiment achieves basic integrated linkage between autonomous parking and fixed-point summoning, solving the problem of completely separated decision-making logic between the two in existing technologies. It realizes path reuse and data sharing between parking and summoning, improving the success rate and response efficiency of the parking and summoning process. By constructing a global scenario model, it achieves unified integration of vehicle status information and environmental information, avoiding repeated collection of environmental data during the parking and summoning stages, reducing data redundancy, and improving decision-making efficiency. It supports the collection of differentiated parking and summoning needs from users, realizing a direct correlation between needs and strategy formulation, and initially solving the problem of insufficient user need adaptation in existing technologies. Using the first path information as the connection node between parking and summoning, it realizes the initial coordination between the parking path and the summoning driving path, reducing the probability of process interruption.

[0027] In one possible implementation, the types of the global scene model include autonomous parking scenarios in open spaces, autonomous parking scenarios in narrow parking spaces, fixed-point summoning scenarios on open roads, fixed-point summoning scenarios in narrow passages, fixed-point summoning scenarios for emergency response, and parking summoning switching scenarios.

[0028] Specifically, this embodiment refines the entire process of autonomous parking, fixed-point summoning, and function switching into six categories. Each category corresponds to a unique weight parameter allocation rule, achieving precise matching between scenarios and weights, and realizing full coverage of the entire process of autonomous parking, fixed-point summoning, and parking-summoning switching. Special scenarios such as narrow scenarios and emergency scenarios are separately classified, making the strategy formulation more in line with the core needs of special scenarios. For example, safety and accuracy are prioritized in narrow scenarios, and efficiency is prioritized in emergency scenarios, thus improving the adaptability of the strategy.

[0029] In one implementation, Table 1 is a weight parameter allocation rule table according to this embodiment, as shown in Table 1: Table 1 Weight Parameter Allocation Rules

[0030] For example, in this embodiment, the autonomous parking scenario in an open space refers to a scenario where the vehicle performs autonomous parking in an open area with a parking space spacing of ≥50cm, no obstructions around it, and a passage width of ≥4m. Autonomous parking in narrow parking spaces: This refers to a scenario where a vehicle performs autonomous parking in a narrow area with a parking space spacing of <30cm and a passage width of ≤3m; Fixed-point summoning on open roads: This refers to a scenario where a vehicle performs a fixed-point summoning on an open road with no obstructions, a passage width of ≥4m, and no complex slopes / curves; Fixed-point summoning in narrow passages: This refers to a scenario where a vehicle performs a fixed-point summoning in a narrow area with a passage width of ≤3m and complex road conditions such as slopes / curves / intersections, typically in an underground parking garage; Emergency response fixed-point summoning: This refers to a scenario where a user initiates an immediate response request, requiring the vehicle to travel to a preset summoning point as quickly as possible; Parking summoning switching: This refers to a transitional scenario where the vehicle is switching between parking and summoning functions after completing autonomous parking but before performing a fixed-point summoning.

[0031] In one feasible approach Figure 2 This is a flowchart of step S2 of the vehicle control method provided in this embodiment of the invention, see reference. Figure 2 The step S2 is followed by: Step S21: Determine the type of the global scene model; Step S22: Assign weight parameters for security factors, efficiency factors, accuracy factors, and user preference factors based on the type of the global scene model.

[0032] Specifically, in this embodiment, the global scene model is identified based on the vehicle environment information, vehicle status information, and user demand information in the global scene model to determine the type of the global scene model. Then, according to the preset scene-weight matching rules, corresponding weight parameters are assigned to safety factors, efficiency factors, accuracy factors, and user preference factors for the determined global scene model type, and the sum of the weight parameters of each factor is 1.

[0033] After acquiring demand information and constructing a global scene model, the vehicle uses a scene recognition type algorithm, combined with information such as road structure, obstacle distribution, vehicle tasks (parking / summoning), and user needs from the global scene model, to determine the specific type of the current global scene model. The vehicle then retrieves the built-in scene-weight matching rule library and assigns corresponding weight parameters to safety factors, efficiency factors, accuracy factors, and user preference factors based on the determined global scene model type. These weight parameters are then synchronized in real-time to the parking planning module and the summoning planning module. When the parking planning module generates an automatic parking strategy, it uses each weight parameter as a decision coefficient, prioritizing factors with higher weight percentages. Similarly, when the summoning planning module generates a fixed-point summoning strategy, it uses each weight parameter as a decision coefficient, prioritizing factors with higher weight percentages. The scene recognition type algorithm can employ lightweight classification algorithms such as decision trees, Bayesian classification, or random forests, which can be selected based on actual scene requirements. Specific application details are not elaborated in this application.

[0034] In one possible implementation, before the fixed-point summoning strategy is generated, in order to achieve the matching between the vehicle's driving range and the fixed-point summoning driving route, and to ensure the feasibility of the summoning driving route, step S2 further includes: Step S21: Determine the type of the global scene model; Step S22: Calculate the actual drivable distance between the vehicle's parking location and the preset summoning point; Step S23: Obtain the vehicle's remaining driving range based on the vehicle status information; Step S24: Compare the actual drivable distance with the vehicle's remaining driving range; Step S25: When the actual drivable distance is less than the vehicle's remaining range, assign weight parameters for safety factors, efficiency factors, accuracy factors, and user preference factors according to the type of the global scenario model; when the actual drivable distance is greater than the vehicle's remaining range, adjust the weights of the safety factors and efficiency factors corresponding to the summoning scenario; wherein, the safety factors include path feasibility sub-factors.

[0035] To pre-determine the feasibility of the vehicle's remaining range and the summoning route before generating the fixed-point summoning strategy, the feasibility of the summoning route is ensured through distance calculation and range assessment. Specifically: First, the geographical coordinates of the parking location and the preset summoning point are extracted from the global scene model. The straight-line distance between the two is calculated using the geographical coordinates, and the actual drivable distance is calculated by combining the road structure in the global scene model. The vehicle's remaining range is extracted from the vehicle status information. Then, the actual drivable distance is compared with the vehicle's remaining range. When the actual drivable distance is less than the vehicle's remaining range, the summoning route is generated according to step S51. When the actual drivable distance is greater than the vehicle's remaining range, the weights of safety factors and efficiency factors corresponding to the summoning scenario are assigned according to the type of the global scene model. Safety factors include summoning route feasibility. In this case, the weight of route feasibility needs to be increased, and the weight of efficiency factors needs to be decreased. If the energy-optimal route still cannot meet the range requirement, the vehicle sends a low range reminder to the user through the vehicle terminal and mobile terminal, suspends the fixed-point summoning strategy formulation, and waits for further instructions from the user.

[0036] This embodiment implements a pre-judgment of the feasibility of the summoning route, which solves the problem in the prior art that does not consider the vehicle's range and is prone to process interruption due to insufficient power, thus improving the success rate of the fixed-point summoning process; in addition, this embodiment combines the actual drivable distance to judge the range, and the judgment result is more accurate, avoiding the error of judging based on straight-line distance.

[0037] In one possible implementation, step S3 includes: using an improved Astar algorithm to generate an automatic parking strategy and first path information based on the global scene model, the weight parameters, and the parking demand information; the automatic parking strategy includes a parking driving path and a parking space location; the first path information includes parking path information and parking environment information.

[0038] Specifically, in this embodiment, the improved Astar algorithm is used as the core algorithm for path planning. The global scene model, the assigned weight parameters, and parking demand information are input into the improved Astar algorithm to perform autonomous parking path planning and parking strategy formulation, generating an automatic parking strategy that includes the parking driving path and the parking space location. At the same time, the improved Astar algorithm generates first path information, which specifically includes parking path information and parking environment information. Then, autonomous parking is performed according to the generated automatic parking strategy, and the first path information is cached to provide data support for the subsequent fixed-point summoning strategy formulation.

[0039] This embodiment employs an improved Astar algorithm adapted to autonomous parking scenarios, integrating weight parameters and user requirements into the algorithm's cost function. This makes the parking strategy more aligned with scenario and user needs, improving parking accuracy and user experience. Furthermore, this embodiment caches parking path information during autonomous parking as the starting segment of the call path, which can be directly invoked during subsequent fixed-point call strategy generation. This achieves early connection between the parking and call paths, providing a foundation for path reuse in subsequent fixed-point calls and further reducing process interruption rates. On the other hand, caching the first path information in this embodiment avoids repeated collection of parking environment information during the fixed-point call stage, reducing data redundancy and improving the decision-making efficiency of fixed-point calls. The improved Astar algorithm also supports real-time correction of parking deviations, controlling parking space matching deviations to ≤20cm, significantly improving the accuracy of autonomous parking.

[0040] In one possible implementation, step S5 includes: step S51, using the Dstar-Lite dynamic path algorithm to generate a fixed-point summoning strategy based on the global scene model, the first path information, the weight parameters, and the summoning demand information, wherein the fixed-point summoning strategy includes the summoning driving path, driving speed, obstacle avoidance sequence, and docking information.

[0041] After autonomous parking is completed, this embodiment uses the Dstar-Lite dynamic path algorithm as the core algorithm for fixed-point summoning driving path planning. The global scene model, the first path information cached after autonomous parking, the allocated weight parameters, and the summoning request information are used as the algorithm input conditions. The Dstar-Lite dynamic path algorithm is used to plan the summoning driving path and formulate the strategy to generate a fixed-point summoning strategy. The fixed-point summoning strategy specifically includes the summoning driving path, driving speed, obstacle avoidance sequence, and parking information. The vehicle executes the fixed-point summoning according to the generated fixed-point summoning strategy and drives to the preset summoning point.

[0042] The summoning route refers to the specific driving trajectory of the vehicle from the target parking space to the preset summoning point, including the node coordinates and actions of turning, driving, obstacle avoidance, and parking; the driving speed refers to the preset driving speed of the vehicle in different road sections and scenarios during the fixed-point summoning process, including normal driving speed, slope / curve driving speed, deceleration speed near the summoning point, etc. Obstacle avoidance sequence refers to the priority of avoiding different types of obstacles during the vehicle's fixed-point summoning process, sorted from high to low according to the degree of danger of the obstacles; parking information refers to the precise parking rules of the vehicle at the preset summoning point, including deceleration nodes near the summoning point, parking position deviation requirements, parking execution actions, etc.

[0043] This embodiment leverages the incremental update advantage of the Dstar-Lite algorithm, combined with the dynamic scenario requirements, weight parameters, and user needs of fixed-point summoning, to achieve dynamic planning and strategy formulation of the summoning driving path. Specifically, the vehicle rasterizes the global scene model, transforming it into a raster map recognizable by the Dstar-Lite dynamic path algorithm. Simultaneously, the cached first path information, weight parameters, and summoning request information are integrated into the algorithm's cost function, prioritizing factors with high weighting during the algorithm's planning process. The Dstar-Lite dynamic path algorithm uses the parking space location after the vehicle is parked as the starting point and a preset summoning point... As the endpoint, the parking path and parking environment information from the first path information are reused to search for the summoning driving path and generate the summoning driving path. The Dstar-Lite algorithm combines the type of the global scene model and the road structure to assign corresponding driving speeds to different road segments, such as decelerating to 1~2km / h on slopes and setting the obstacle avoidance order according to the degree of obstacle danger. The stopping information is formulated in combination with the accuracy requirements in the summoning demand information. For example, the obstacle avoidance order is "pedestrians > moving vehicles > temporary obstacles > fixed obstacles". In order to meet the user's needs, the stopping information is to decelerate to 0.5km / h within 5m of the summoning point and the stopping deviation is ≤0.5m.

[0044] Specifically, this embodiment employs the Dstar-Lite dynamic path algorithm, which can quickly respond to dynamic environmental changes during the fixed-point summoning process, solving the problem of weak dynamic adjustment capability in existing technologies. Reusing parking data from the first path information avoids repeated collection of environmental information, reduces path planning time, and shortens the response latency of the summoning strategy to within 500ms. Clear driving speed and obstacle avoidance sequence make the vehicle's fixed-point summoning driving safer, increasing the accuracy of dynamic obstacle recognition to ≥99% and reducing collision risk by 95%. Refined parking information ensures that the vehicle's parking deviation at the preset summoning point is controlled within ≤0.5m, significantly improving the accuracy of fixed-point summoning. Integrating weight parameters and user needs into the algorithm cost function makes the summoning strategy more aligned with scenario and user needs, improving user adaptability and user experience.

[0045] In one possible implementation, step S4 is followed by: Based on the weight parameters corresponding to the parking summon switching scenario, the vehicle is controlled to enter the summon standby state; It can identify fixed-point summoning commands and exit the summoning standby state when it receives a fixed-point summoning command.

[0046] To ensure a rapid response to on-site summoning and enter a low-power waiting state, the vehicle is in Park (P) gear, the braking system remains active, the environmental perception sensors remain active with low power, and the decision-making module is in standby mode. This state allows for rapid wake-up by on-site summoning commands. Specifically, after autonomous parking is completed, the current global scene model type is modified to a parking summoning switching scene, and the corresponding weight parameters (safety factor 0.5, efficiency factor 0.2, accuracy factor 0.2, user preference factor 0.1) are retrieved. Based on these weight parameters, summoning standby control rules are formulated, prioritizing the safety and response efficiency of the standby state. Based on these rules, the vehicle is controlled to complete operations such as engaging Park, activating the braking system, and entering a low-power activation state, thus entering the summoning standby state. It can identify fixed-point summoning commands and exit the summoning standby state when it receives a fixed-point summoning command.

[0047] This embodiment implements smooth transition control between parking and summoning. Based on the weight parameters of the parking-summoning switching scenario, the vehicle enters a summoning standby state, achieving seamless connection between parking and summoning functions and improving the response speed of fixed-point summoning. In addition, in this embodiment, the vehicle is in a low-power activation / standby state in the summoning standby state, which not only ensures a fast response to fixed-point summoning but also reduces the vehicle's energy consumption and improves its range. The standby control rules formulated based on the weight parameters of the parking-summoning switching scenario can ensure the safety of the standby state and avoid safety accidents such as vehicle rollaway or collisions in the standby state.

[0048] In one feasible approach Figure 3 This is a flowchart of step S6 of the vehicle control method provided in this embodiment of the invention, see reference. Figure 3 Step S6 includes: Step S61: Control the vehicle's movement based on the fixed-point summoning strategy; Step S62: Obtain vehicle driving information and updated summoning request information in real time. The vehicle driving information includes obstacle information and sensor status information. Step S63: Adjust the fixed-point summoning strategy based on the vehicle driving information, the updated summoning demand information and the vehicle scene information, and control the vehicle to drive to the preset summoning point according to the adjusted fixed-point summoning strategy.

[0049] This embodiment utilizes control logic involving real-time data acquisition, data fusion, strategy adjustment, and execution feedback to enable the vehicle to respond dynamically to environmental changes and user needs during the point-to-point summoning process, achieving dynamic strategy adjustment. Specifically, the execution control module, based on the original point-to-point summoning strategy, controls the vehicle to begin driving along the summoning path to execute the point-to-point summoning; the information acquisition module, with an information acquisition cycle of 100ms, collects obstacle information and sensor status information during the vehicle's movement in real time, integrates them into vehicle driving information, and synchronizes it to the summoning planning module in real time; the vehicle's user need acquisition unit continuously receives user instructions, and if the user issues a need modification instruction, it immediately integrates it into updated summoning need information and synchronizes it to the summoning planning module; the summoning planning module fuses the real-time vehicle driving information, the updated summoning need information, and the original vehicle scene information to update the global scene model; the summoning planning module... Leveraging the incremental update advantage of the Dstar-Lite dynamic path algorithm, the original fixed-point summoning strategy is dynamically adjusted based on the updated global scene model, generating an adjusted fixed-point summoning strategy, which is then synchronized to the execution control module. The execution control module immediately controls the vehicle's movement according to the adjusted fixed-point summoning strategy, including adjusting the driving path, speed, and obstacle avoidance maneuvers. If the user issues an emergency stop command, the vehicle is immediately brought to a stop. These steps are repeated until the vehicle accurately stops at the preset summoning point according to the adjusted fixed-point summoning strategy, achieving dynamic closed-loop adjustment of the fixed-point summoning strategy. This solves the problems of weak dynamic adjustment capabilities and inability to cope with sudden obstacles and changes in user needs in existing technologies. The user need acquisition unit can be a mobile app, cloud platform, or in-vehicle terminal, etc.

[0050] This invention also provides a vehicle control system including: The information acquisition module is used to acquire vehicle scene information and demand information, including parking demand information and summoning demand information. The decision-making module, connected to the information acquisition module, is used to generate a global scene model based on the vehicle scene information. The parking planning module, connected to the information acquisition module and the decision-making module, is used to generate an automatic parking strategy and first path information based on the parking demand information and the global scene model. The summoning planning module, connected to the information acquisition module and the parking planning module, is used to generate a fixed-point summoning strategy based on the summoning demand information, the first path information and the global scene model. The execution control module is connected to the information acquisition module, the parking planning module, and the summoning planning module; the execution control module controls the vehicle to park autonomously according to the automatic parking strategy, and controls the vehicle to drive to the preset summoning point according to the fixed-point summoning strategy.

[0051] Specifically, the information acquisition module can realize multimodal environment perception, acquire user demand information, and acquire vehicle status information. The information acquisition module includes a multimodal environment perception unit, a user demand acquisition unit, and a vehicle status monitoring unit. The multimodal environment perception unit includes an on-board camera, LiDAR, ultrasonic radar, high-precision map, V2X communication components, and garage positioning components.

[0052] This invention also provides a vehicle including a processor and a memory, the memory storing a computer program, and the processor executing the computer program to implement the vehicle control method described above.

[0053] To help those skilled in the art better understand the technical solution of this application, this application also provides an application example, as follows: Hardware configuration: Multimodal environmental perception unit: front-facing 8MP binocular camera, blind spot LiDAR, 16 ultrasonic radars, centimeter-level high-precision map, 5G-V2X communication module, garage UWB positioning unit; User needs acquisition unit: mobile APP, in-vehicle central control screen (12.3 inches), voice interaction unit; Vehicle Condition Monitoring Unit: BMS Battery Management System, PCU Powertrain Control System, EPS Electronic Power Steering System, EHB Brake-by-Hand System, Sensor Fault Diagnosis Unit; Decision module: NVIDIA Orin-X chip, running Dstar-Lite dynamic path algorithm and scenario-based weight allocation algorithm; Execution control modules: Automatic Shift System (AGS), Chassis Domain Controller, Body Control System.

[0054] Application Example 1 demonstrates the integrated implementation of autonomous parking and fixed-point parking summoning in an underground parking garage. The specific implementation steps are as follows: Step 1: Information collection for the integrated execution of autonomous parking and fixed-point summoning in the underground parking garage; Step 2: The user gets out of the car at the entrance of the underground parking garage and initiates autonomous parking and fixed-point call-out standby commands through the APP; Step 3: The information acquisition module collects garage environment data (parking space size 5.0m×2.5m, ramp slope 15°, aisle width 3.0m) and dynamic obstacles (1 cleaning robot, speed 0.8km / h); acquires user requirements: parking precision priority, summoning point is garage entrance (accuracy ±0.5m), path preference no ramp (summoning via a gentle aisle); acquires vehicle status: battery 70%, sensors normal.

[0055] Step 4: The decision-making module generates a global scenario model and assigns weights: Autonomous parking stage (narrow passage, parking space): safety factor 0.4, efficiency factor 0.1, accuracy factor 0.4, user preference 0.1; Call and standby stage: safety factor 0.5, efficiency factor 0.2, accuracy factor 0.2, user preference 0.1.

[0056] Step 5: The decision module generates an autonomous parking strategy: plans the path "entrance → gentle passage → target parking space" at a speed of 1.5km / h and a precise parking deviation of 8cm. It also pre-plans the summoning path: caches the path from the parking space to the entrance without a ramp and marks the cleaning robot's activity area as a risk point.

[0057] Step 6: The execution control module controls the vehicle to park autonomously according to the autonomous parking strategy, and enters the standby state after parking is completed.

[0058] Step 7, Fixed-point summoning execution and adjustment: 10 minutes later, the user initiates a summoning. The multimodal environmental perception unit detects that the cleaning robot has left the risk area. The decision module reuses the pre-planned path and dynamically adjusts the speed (2km / h in the channel, reduced to 0.5km / h 5m before the entrance), accurately stopping at the designated position at the entrance (deviation 0.3m), with no obstacle collisions throughout the process.

[0059] Example 2: Dynamic obstacle avoidance and summoning point adjustment during fixed-point summoning. Application Example 2 illustrates dynamic obstacle avoidance and summoning point adjustment during fixed-point summoning. The specific implementation steps are as follows: Step 1, Information Collection: The vehicle has completed autonomous parking in the mall parking lot, and the user initiates a fixed-point summoning (the summoning point is the main entrance of the mall); the multimodal environmental perception unit detects dynamic obstacles in the summoning path (200m in length) (3 pedestrians crossing the passage and 1 temporarily parked vehicle); user requirements: safety first, allow temporary adjustment of the summoning point.

[0060] Step 2, weight allocation: The global scene model type is open road fixed-point summoning scene, with safety factor 0.4, efficiency factor 0.4, accuracy factor 0.1, and user preference 0.1.

[0061] Step 3, Fixed-point summoning strategy generation: Plan the path "parking space → side passage → mall main entrance" at a speed of 3km / h, and preset pedestrian avoidance strategy (stop and wait) and temporary vehicle detour strategy.

[0062] Step 4, Dynamic Adjustment: When the vehicle is 100m away, a pedestrian suddenly enters the main passage, and the decision module immediately triggers a stop (response time 80ms); at the same time, the user adjusts the call point to the side gate (50m away from the current location) through the APP, and the decision module replans the route in real time (switching to the side passage to reach the side gate directly), and finally stops accurately, without interruption throughout the process.

[0063] In addition, this embodiment also provides the following alternative methods: (a) Decision algorithm replacement 1. Path planning algorithm: The Dstar-Lite algorithm is adopted to replace the improved Astar algorithm, which is suitable for dynamic path adjustment in fixed-point summoning and improves obstacle avoidance response speed by 30%; 2. Weighting Allocation Algorithm: The entropy weighting method is used to replace the fixed weight allocation based on the scenario. The weights are automatically calculated dynamically according to the complexity of the environment and user needs, making it more adaptable.

[0064] (ii) Hardware configuration replacement 1. Environmental perception module: "LiDAR and vision fusion" is used to replace part of the ultrasonic radar, while ensuring the accuracy of obstacle recognition on the summoning path (deviation ≤3cm). 2. Positioning Unit: In scenarios without high-precision maps, "GPS and IMU inertial navigation" is used as a substitute to meet the requirement of accurate docking at the call point (deviation ≤ 0.8m).

[0065] (iii) Replacement of user interaction methods 1. Demand collection method: Replace the APP settings with a vehicle-to-everything (V2X) voice assistant. The user's voice commands "Quickly park" or "Summon to me" can be used to analyze the demand. 2. Call Point Positioning: Supports "User Location Sharing" to replace manual settings, such as automatic vehicle navigation to the user's real-time location with an accuracy of ±1m.

[0066] (iv) Replacement of execution control methods 1. Scene switching control: "Progressive power adjustment" is used instead of direct switching. After parking is completed, the sensor activation frequency is gradually reduced (standby state). 2. Emergency Pause Mechanism: Supports "vehicle networking and Bluetooth dual-mode triggering", allowing users to pause the call in an emergency via APP or Bluetooth remote control.

[0067] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0068] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A vehicle control method, characterized in that, include: Obtain demand information, which includes parking demand information and summoning demand information, and the summoning demand information includes preset summoning points; Obtain vehicle scene information and generate a global scene model based on the vehicle scene information, wherein the vehicle scene information includes vehicle status information and vehicle environment information; Based on the parking demand information and the global scene model, an automatic parking strategy and first path information are generated. The vehicle is controlled to park autonomously according to the automatic parking strategy; A fixed-point summoning strategy is generated based on the summoning demand information, the first path information, and the global scene model. The vehicle is controlled to travel to the preset summoning point according to the fixed-point summoning strategy.

2. The vehicle control method according to claim 1, characterized in that, The types of global scene models include autonomous parking scenarios in open spaces, autonomous parking scenarios in narrow parking spaces, fixed-point summoning scenarios on open roads, fixed-point summoning scenarios in narrow passages, fixed-point summoning scenarios for emergency response, and parking summoning switching scenarios.

3. The vehicle control method according to claim 2, characterized in that, After obtaining vehicle scene information and generating a global scene model based on the vehicle scene information, the process further includes: Determine the type of the global scene model; The weight parameters for security factors, efficiency factors, accuracy factors, and user preference factors are assigned based on the type of the global scene model.

4. The vehicle control method according to claim 3, characterized in that, After obtaining vehicle scene information and generating a global scene model based on the vehicle scene information, the process further includes: Determine the type of the global scene model; Calculate the actual drivable distance between the vehicle's parking location and the preset call point; The remaining driving range of a vehicle is obtained based on vehicle status information; Compare the actual drivable distance with the vehicle's remaining range; When the actual drivable distance is less than the vehicle's remaining range, the weight parameters of safety factors, efficiency factors, accuracy factors, and user preference factors are assigned according to the type of the global scenario model. When the actual drivable distance is greater than the vehicle's remaining range, adjust the weights of safety and efficiency factors corresponding to the summoning scenario; The safety factors include the path feasibility sub-factor.

5. The vehicle control method according to claim 3, characterized in that, The step of generating an automatic parking strategy and first path information based on the parking demand information and the global scene model includes: using an improved Astar algorithm to generate an automatic parking strategy and first path information based on the global scene model, the weight parameters, and the parking demand information. The automatic parking strategy includes a parking driving path and a parking space location; the first path information includes parking path information and parking environment information.

6. The vehicle control method according to claim 3, characterized in that, The step of generating a fixed-point summoning strategy based on the summoning demand information, the first path information, and the global scene model includes: using the Dstar-Lite dynamic path algorithm to generate a fixed-point summoning strategy based on the global scene model, the first path information, the weight parameters, and the summoning demand information. The fixed-point summoning strategy includes the summoning driving path, driving speed, obstacle avoidance sequence, and stopping information.

7. The vehicle control method according to claim 5, characterized in that, The step of controlling the vehicle to park autonomously according to the automatic parking strategy includes: Based on the weight parameters corresponding to the parking summon switching scenario, the vehicle is controlled to enter the summon standby state; It can identify fixed-point summoning commands and exit the summoning standby state when it receives a fixed-point summoning command.

8. The vehicle control method according to claim 5, characterized in that, The step of controlling the vehicle to drive to the preset summoning point according to the fixed-point summoning strategy includes: The vehicle's movement is controlled based on the aforementioned fixed-point summoning strategy; Real-time acquisition of vehicle driving information and updated summoning request information, wherein the vehicle driving information includes obstacle information and sensor status information; The fixed-point summoning strategy is adjusted based on the vehicle driving information, the updated summoning demand information, and the vehicle scene information, and the vehicle is controlled to drive to the preset summoning point according to the adjusted fixed-point summoning strategy.

9. A vehicle control system, characterized in that, include: The information acquisition module is used to acquire vehicle scene information and demand information, including parking demand information and summoning demand information. The decision-making module, connected to the information acquisition module, is used to generate a global scene model based on the vehicle scene information. The parking planning module, connected to the information acquisition module and the decision-making module, is used to generate an automatic parking strategy and first path information based on the parking demand information and the global scene model. The summoning planning module, connected to the information acquisition module and the parking planning module, is used to generate a fixed-point summoning strategy based on the summoning demand information, the first path information and the global scene model. The execution control module is connected to the information acquisition module, the parking planning module, and the summoning planning module; the execution control module controls the vehicle to park autonomously according to the automatic parking strategy, and controls the vehicle to drive to the preset summoning point according to the fixed-point summoning strategy.

10. A vehicle, characterized in that, It includes a processor and a memory, the memory storing a computer program, and the processor executing the computer program to implement the vehicle control method according to any one of claims 1-8.

Images