CN-122009152-A - Method and system for generating parking path

CN122009152ACN 122009152 ACN122009152 ACN 122009152ACN-122009152-A

Abstract

The invention provides a method and a system for generating a parking path, and relates to the technical field of parking control, wherein the method comprises the steps of responding to a parking request from a target vehicle in the process that the target vehicle parks according to a designated parking path, and generating N alternative paths; according to each alternative path in the N alternative paths, scene information included in the alternative paths is deconstructed through a scene recognition model, scene scores corresponding to the alternative paths are obtained, the predicted passing duration of the target vehicle driving alternative paths is determined according to the scene information included in the alternative paths through a duration prediction model, the predicted energy consumption of the target vehicle driving alternative paths is determined according to the predicted passing duration and the vehicle state of the target vehicle, the total cost of the alternative paths is determined according to the scene scores corresponding to the alternative paths, the predicted passing duration and the predicted energy consumption, and the target parking path is determined from the N alternative paths according to the N total costs corresponding to the N alternative paths.

Inventors

ZHANG MENG
ZHU ZIFAN
LIU JIFENG
SHEN CHEN
FU BIN

Assignees

岚图汽车科技股份有限公司

Dates

Publication Date: 20260512
Application Date: 20260105

Claims (10)

1. A method for generating a parking path, which is applied to a cloud, the method comprising: In the process that a target vehicle parks according to a specified parking path, N alternative paths from the current position of the target vehicle to a target parking space are generated in response to a parking request from the target vehicle; For each of the N alternative paths, resolving scene information included in the alternative path through a scene recognition model to obtain a scene score corresponding to the alternative path, determining a predicted traffic duration of the target vehicle for driving the alternative path according to the scene information included in the alternative path through a duration prediction model, determining a predicted energy consumption of the target vehicle for driving the alternative path according to the predicted traffic duration and a vehicle state of the target vehicle, and determining a total cost of the alternative path according to the scene score corresponding to the alternative path, the predicted traffic duration and the predicted energy consumption; And determining a target parking path from the N alternative paths according to N total costs corresponding to the N alternative paths, and issuing the target parking path to the target vehicle so that the target vehicle updates the designated parking path according to the target parking path.
2. The method of claim 1, wherein the target parking space is located within a target parking lot indicated by the parking request; The method for generating N alternative paths from the current position of the target vehicle to the target parking space in response to the parking request from the target vehicle comprises the following steps: Responding to the parking request, and acquiring empty space data sent by field end equipment of the target parking lot from the target vehicle, wherein the empty space data comprises the number and/or the position of the parking spaces in an unoccupied state in the target parking lot; Determining a target parking space for parking the target vehicle from the target parking lot according to the current position and the empty parking space, and sending a scheduling instruction to the field terminal, wherein the scheduling instruction instructs the field terminal equipment to modify the state of the target parking space; Issuing an entrance gate cooperative instruction to M entrance gates of the target parking lot to acquire waiting traffic at the M entrance gates of the target parking lot from the M entrance gates; Determining a target entrance gate from the M entrance gates according to the waiting traffic at the M entrance gates; generating N alternative paths according to the current position, the target entrance gate and the target parking space, wherein each alternative path comprises the target entrance gate and the target parking space.
3. The method of claim 2, wherein the deconstructing scene information included in the alternative path by a scene recognition model to obtain a scene score corresponding to the alternative path, comprises: Identifying the alternative paths through the scene identification model, and determining Y target scene areas included on the alternative paths; Labeling Y target scene labels corresponding to the Y target scene areas on the alternative path, wherein the scene recognition model is obtained based on training of preset X scene labels, and the Y target scene labels belong to the X scene labels; issuing a field terminal cooperative instruction to the field terminal equipment so as to acquire a parking lot video of the target parking lot from the field terminal; Acquiring a road condition video acquired by a sensor of the target vehicle; For each target scene area in Y target scene areas, determining a target scene video corresponding to the target scene area from the parking lot video and/or the road condition video; Injecting Y target scene videos into the target scene labels corresponding to the target scene areas; Analyzing the Y target scene videos through the scene recognition model, determining vehicle congestion conditions and/or people flow densities in Y target scene areas, and obtaining risk scores corresponding to the Y target scene areas; and determining scene scores corresponding to the alternative paths according to the risk scores corresponding to the Y target scene areas.
4. The method of claim 3, wherein the determining, by a duration prediction model, a predicted passage duration for the target vehicle to travel the alternative path based on the scene information included in the alternative path comprises: Acquiring a vehicle state of the target vehicle, wherein the vehicle state comprises vehicle hardware parameters, and the vehicle hardware parameters at least comprise length, width, height, wheelbase and minimum turning radius of the vehicle; predicting the passing time length of the target vehicle in Y target scene areas according to the vehicle hardware parameters through the time length prediction model; And obtaining the predicted passing duration of the alternative path according to Y passing durations corresponding to the Y target scene areas.
5. The method of claim 4, wherein the vehicle state further comprises a type of energy consumption and an amount of energy consumption per unit time of the target vehicle; The determining the predicted energy consumption of the target vehicle to drive the alternative path according to the predicted traffic duration and the vehicle state of the target vehicle comprises the following steps: and determining the predicted energy consumption according to the energy consumption type of the target vehicle, the energy consumption per unit time and the predicted traffic duration.
6. The method of claim 5, wherein the determining the total cost of the alternative path based on the scene score, the predicted time of flight, and the predicted energy consumption for the alternative path comprises: Determining the driving style of the target vehicle according to the vehicle hardware parameters, the energy consumption type and the energy consumption in unit time of the target vehicle; Determining a weight ratio corresponding to the target vehicle according to the driving style of the target vehicle; And determining the total cost of the alternative path according to the weight ratio, the scene score corresponding to the alternative path, the predicted passing duration and the predicted energy consumption.
7. A system for generating a parking path, the system comprising: The target vehicle is used for parking according to the appointed parking path and sending a parking request to the cloud in the process of parking the target vehicle; The cloud end is used for responding to a parking request from a target vehicle, generating N alternative paths from the current position of the target vehicle to a target parking space, solving scene information included in the alternative paths through a scene recognition model for each of the N alternative paths to obtain scene scores corresponding to the alternative paths, determining a predicted passing time length of the target vehicle for driving the alternative paths according to the scene information included in the alternative paths through a time length prediction model, determining a predicted energy consumption of the target vehicle for driving the alternative paths according to the predicted passing time length and the vehicle state of the target vehicle, and determining total cost of the alternative paths according to the scene scores corresponding to the alternative paths, the predicted passing time length and the predicted energy consumption; and the target vehicle is further used for updating the appointed parking path according to the target parking path under the condition that the target parking path from the cloud is received, and parking is carried out according to the updated appointed parking path.
8. The system of claim 7, further comprising a head-end device disposed in the target parking area indicated by the parking request, wherein the head-end device is configured to: under the condition that the target vehicle invokes the empty space data of the target parking lot, the empty space data is sent to the target vehicle, and the empty space data comprises the number of parking spaces and/or the positions of the parking spaces in an unoccupied state in the target parking lot; The target vehicle is further used for responding to a parking request, calling the empty parking space data from the field terminal equipment, generating an initial parking path from the target vehicle to the target parking lot when the parking space number is not 0, and taking the initial parking path as a designated parking path.
9. The system of claim 8, wherein the field terminal device is further to: responding to a field end cooperative instruction from a cloud end, and sending a parking lot video of the target parking lot to the cloud end; The cloud end is further used for determining a target scene video corresponding to the target scene area from the parking lot video under the condition that the parking lot video is received.
10. The system of claim 8, further comprising M portal gates disposed in the target parking lot, wherein the M portal gates are configured to: responding to an entrance gate cooperative instruction from a cloud end, and sending waiting traffic at M entrance gates of the target parking lot to the cloud end; the cloud end is further used for determining a target entrance gate from the M entrance gates according to the waiting traffic flow of the M entrance gates.

Description

Method and system for generating parking path Technical Field The present invention relates to the field of parking control technologies, and in particular, to a method and a system for generating a parking path. Background With the rapid development and popularization of automatic driving technology, an automatic parking function has become a key feature for improving the intelligent level of vehicles and user experience. Currently, automated parking systems typically make path planning and decisions locally on the vehicle, relying on the perception of onboard sensors and limited computational resources. However, the target vehicle leading scheme has inherent limitations that in the face of complex and changeable parking lot environments, the global information is difficult to deeply fuse and predict for a long time by single vehicle computing force, and a pre-planned single path lacks enough flexibility and robustness when encountering sudden obstacles or scene changes, which may cause interruption, overlooking time or rapid increase of energy consumption in the parking process, and influence the parking efficiency and user experience. Disclosure of Invention The invention provides a method and a system for generating a parking path, which are used for solving the problem that the path cannot cope with environmental change due to the fact that a target vehicle confirms the parking path according to real-time perception of an on-board sensor in the related technology. In a first aspect, an embodiment of the present invention provides a method for generating a parking path, which is applied to a cloud, where the method includes: In the process that a target vehicle parks according to a specified parking path, N alternative paths from the current position of the target vehicle to a target parking space are generated in response to a parking request from the target vehicle; For each of the N alternative paths, resolving scene information included in the alternative path through a scene recognition model to obtain a scene score corresponding to the alternative path, determining a predicted traffic duration of the target vehicle for driving the alternative path according to the scene information included in the alternative path through a duration prediction model, determining a predicted energy consumption of the target vehicle for driving the alternative path according to the predicted traffic duration and a vehicle state of the target vehicle, and determining a total cost of the alternative path according to the scene score corresponding to the alternative path, the predicted traffic duration and the predicted energy consumption; And determining a target parking path from the N alternative paths according to N total costs corresponding to the N alternative paths, and issuing the target parking path to the target vehicle so that the target vehicle updates the designated parking path according to the target parking path. The method comprises the steps of receiving a parking request from a target vehicle, determining a target parking space for parking the target vehicle from the target parking lot according to the current position and the empty space data, sending a scheduling instruction to the field end, wherein the scheduling instruction indicates the field end device to modify the state of the target parking space, sending an entrance gate cooperative instruction to M entrance gates of the target parking lot, so as to obtain waiting traffic at M entrance gates of the target parking lot, according to the waiting traffic at the M entrance gates, determining the target parking space from the M entrance gates, and generating a scheduling instruction to the N entrance gates according to the current position and the target parking space. The method comprises the steps of determining Y target scene areas on an alternative path through scene recognition models, labeling Y target scene labels corresponding to the Y target scene areas on the alternative path, wherein the scene recognition models are obtained based on preset X scene labels, the Y target scene labels belong to the X scene labels, transmitting a scene end cooperative instruction to the scene end equipment to obtain parking lot videos of the target parking lot from the scene end, obtaining road condition videos acquired from sensors of target vehicles, determining target scene videos corresponding to the target scene areas in the parking lot videos and/or the road condition videos for each target scene area, injecting the Y target scene videos into the target scene labels corresponding to the target scene areas, determining the target traffic situation score according to the target traffic situation score in the target traffic situation score, and analyzing the target traffic situation score according to the target traffic situation score. The method comprises the steps of obtaining a vehicle state of a target vehicle, wherein the vehicle state comprises