Parking control method and system and vehicle

Abstract

The invention provides a parking control method and system and a vehicle, and the method comprises the steps: obtaining the parameter information and parking information of the vehicle, the parameter information comprises the width of the vehicle and the transverse opening distance of a vehicle door, and the parking information comprises the position of the vehicle, the position of a parking space, and the information of an obstacle in a parking environment; according to the parameter information and the obstacle information, getting-off space evaluation is carried out, and the transverse safety distance between the vehicle and the obstacle is determined; constructing a safe driving corridor according to the parking information, and searching a target parking path in the safe driving corridor by taking the transverse safe distance as a constraint condition of path search; and controlling the vehicle to complete parking according to the target parking path. In the automatic parking process, the getting-off space is evaluated, the transverse safety distance between the vehicle and the obstacle obtained through evaluation serves as the parking path constraint, the vehicle is controlled to complete parking, it can be effectively ensured that a driver and passengers obtain sufficient vehicle door opening space, getting-off is safely and conveniently completed, and the user experience is improved.

Description

Technical Field

[0001] This application relates to the technical field of vehicle driving assistance, and particularly to a parking control method, system and vehicle. Background Art

[0002] With the rapid development of autonomous driving technology, automatic parking, as one of the core functions of intelligent vehicles, has become a key technology to enhance the driving experience and convenience. The automatic parking technology aims to integrate various technologies such as environmental perception, path planning and safety guarantee to enable the vehicle to autonomously search for and park in a parking space without manual intervention, bringing a more convenient and comfortable driving experience to users.

[0003] In the related art, automatic parking mainly focuses on the vehicle being able to park safely and accurately in a parking space that meets geometric constraint conditions, while ignoring the convenience and safety of the driver and passengers getting out of the vehicle. That is, after the actual parking is completed, the vehicle may be parked too close to surrounding obstacles, making it impossible for the driver and passengers to open the door to get out of the vehicle or making it unsafe to open the door to get out of the vehicle, thus causing difficulties or dangers for the driver and passengers to open the door. In order to ensure that the driver and passengers can get out of the vehicle conveniently and safely, some set a fixed distance threshold as a standard, and when parking, the distance between the vehicle and the obstacle is restricted. However, this method is difficult to adapt to the dynamically changing parking environment and the diverse requirements of parking space layouts. Especially in the case of angled parking spaces or the presence of moving vehicles around, as well as in urban crowded areas where parking spaces are compact and there are many surrounding obstacles, there are still obvious shortcomings in ensuring the safety and convenience of the driver and passengers getting out of the vehicle. Summary of the Invention

[0004] In view of the above disadvantages, this application discloses a parking control method, system and vehicle, which are used to solve the technical problem that the safety and convenience of the driver and passengers getting out of the vehicle cannot be guaranteed in parking technology.

[0005] In a first aspect, this application provides a parking control method, the method includes: obtaining parameter information and parking information of the vehicle, the parameter information includes the vehicle width and the lateral opening distance of the door, and the parking information includes the vehicle position, the parking space position and the obstacle information in the parking environment; performing an off - vehicle space evaluation according to the parameter information and the obstacle information to determine the lateral safety distance between the vehicle and the obstacle; constructing a safe driving corridor according to the parking information, and using the lateral safety distance as a constraint condition for path search, and searching for a target parking path in the safe driving corridor; controlling the vehicle to complete parking according to the target parking path.

[0006] In an embodiment of the present application, the determination of the lateral safety distance between the vehicle and the obstacle includes: calculating the width of the obstacle, the width of the vehicle, and the lateral opening distance of the vehicle door to obtain the theoretical safety distance between the vehicle and the obstacle, where the obstacle information includes the width of the obstacle; calculating the lateral safety distance based on the theoretical safety distance and a preset safety distance margin, and the safety distance margin is a compensation distance set to prevent the vehicle from colliding with the obstacle.

[0007] In an embodiment of the present application, the construction of the safe driving corridor according to the parking information includes: determining a plurality of candidate positions of the vehicle according to the vehicle position, the parking space position, and at least one obstacle position, where the obstacle information includes the obstacle positions of at least one obstacle, and the candidate positions are positions where the vehicle will not collide with all obstacles; determining the total potential energy value of each candidate position according to the distance between each candidate position and each obstacle; if the total potential energy value of the target candidate position is less than or equal to a preset safety potential energy threshold, then determining the target candidate position as the target position; generating the safe driving corridor according to the target positions.

[0008] In an embodiment of the present application, the determination of the total potential energy value of each candidate position according to the distance between each candidate position and each obstacle includes: calculating the potential energy values of each obstacle by respectively calculating the obstacle positions, the candidate positions, the lateral safety distance, and a preset sensitivity coefficient, where the sensitivity coefficient represents the sensitivity of the potential energy to the change in the obstacle distance; calculating the sum of the potential energy values of each obstacle to obtain the total potential energy value.

[0009] In an embodiment of the present application, taking the lateral safety distance as a constraint condition for path search and searching for a target parking path in the safe driving corridor includes: monitoring the potential field gradient in the safe driving corridor, and determining an initial parking path according to the direction of the potential field gradient descent, where the distance between each node in the initial parking path and each obstacle is greater than or equal to the lateral safety distance; locally adjusting the initial parking path according to the real-time change of the safe driving corridor to obtain the target parking path, where the distance between each node in the target parking path and each obstacle is greater than or equal to the lateral safety distance.

[0010] In an embodiment of the present application, the method for obtaining the obstacle information includes: acquiring environmental perception data of the parking environment, where the environmental perception data includes lidar data, ultrasonic data, and image data; fusing the lidar data, the ultrasonic data, and the image data to obtain a multi-source environment model; identifying obstacles in the multi-source environment model to obtain the obstacle information of at least one obstacle, where the obstacle information includes obstacle position, obstacle width, obstacle height, and obstacle type.

[0011] In an embodiment of the present application, the step of fusing the lidar data, the ultrasonic data, and the image data to obtain a multi-source environment model includes: constructing a data fusion model according to the characteristics of the lidar, ultrasonic sensor, and camera, where the data fusion model includes a prediction sub-model and an update sub-model, the prediction sub-model includes a system model matrix, a control input influence matrix, and a process noise covariance matrix, and the update sub-model includes a measurement model matrix; inputting the real-time control input vector of the obstacle into the prediction sub-model to obtain a predicted state vector and a predicted state covariance matrix, where the predicted state vector and the predicted state covariance matrix are the state vector and state covariance matrix of the obstacle at the current moment; determining the measurement value vector and measurement noise covariance matrix of the obstacle at the current moment according to the lidar data, the ultrasonic data, and the image data; and correcting the predicted state vector and the predicted state covariance matrix according to the measurement value vector, the measurement noise covariance matrix, and the update sub-model to obtain a corrected state vector and state covariance matrix, thereby forming the multi-source environment model.

[0012] In an embodiment of the present application, the step of controlling the vehicle to complete parking according to the target parking path includes: if the distance between the target node in the target parking path and each obstacle is less than the lateral safety distance, determining the target parking mode of the vehicle from multiple parking modes, where the multiple parking modes are, from high to low priority, the path lateral adjustment mode, the delay mode, and the remote control mode; if the target parking mode is the path lateral adjustment mode, laterally adjusting the target parking path and controlling the vehicle to complete parking according to the adjusted target parking path; if the target parking mode is the delay mode, starting the first-level prompt and the delay mode, and controlling the vehicle to complete parking according to the target parking path, where the first-level prompt includes suggesting that the driver and passengers get off the vehicle before parking and switching the parking mode to the delay mode; if the target parking mode is the remote control mode, starting the second-level prompt and the remote control mode, and after receiving the remote parking instruction, controlling the vehicle to complete parking according to the target parking path, where the second-level prompt includes asking the driver and passengers to get off the vehicle before parking and switching the parking mode to the remote control mode.

[0013] In a second aspect, the present application provides a parking control system, which includes: a data acquisition module for acquiring parameter information and parking information of a vehicle, where the parameter information includes the vehicle width and the lateral opening distance of the vehicle door, and the parking information includes the vehicle position, the parking space position, and obstacle information in the parking environment; a space evaluation module for performing a getting-off space evaluation based on the parameter information and the obstacle information to determine the lateral safety distance between the vehicle and the obstacle; a path planning module for constructing a safe driving corridor according to the parking information and using the lateral safety distance as a constraint condition for path search to search for a target parking path in the safe driving corridor; and a parking control module for controlling the vehicle to complete parking according to the target parking path.

[0014] In a third aspect, the present application provides a vehicle that uses the parking control method described in the first aspect or includes the parking control system described in the second aspect.

[0015] As described above, a parking control method, system, and vehicle provided by an embodiment of the present application have the following beneficial effects: First, parameter information and parking information of the vehicle are acquired. Among them, the parameter information includes the vehicle width and the lateral opening distance of the vehicle door, and the parking information includes the vehicle position, the parking space position, and obstacle information in the parking environment. Then, a getting-off space evaluation is performed based on the parameter information and the obstacle information to determine the lateral safety distance between the vehicle and the obstacle. Next, a safe driving corridor is constructed according to the parking information, and the lateral safety distance is used as a constraint condition for path search to search for a target parking path in the safe driving corridor. Finally, the vehicle is controlled to complete parking according to the target parking path. During the automatic parking process, through a safe getting-off space evaluation mechanism, the getting-off space is evaluated in advance, that is, by comprehensively considering the vehicle width, the lateral opening distance of the vehicle door, and the obstacle information, the lateral safety distance between the vehicle and the obstacle is determined, and this lateral safety distance is used as a parking path constraint to ensure that there is enough opening space for the vehicle door after parking as the goal, and the vehicle is controlled to complete parking, thereby effectively ensuring that the driver and passengers can obtain sufficient vehicle door opening space, enabling the driver and passengers to get off safely and conveniently, and improving the driving and riding experience of users.

[0016] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The accompanying drawings here are incorporated into and form a part of this specification, showing embodiments consistent with this application, and are used together with the specification to explain the principles of this application. Obviously, the accompanying drawings in the following description are only some embodiments of this application, and for those of ordinary skill in the art, other accompanying drawings can be obtained based on these drawings without creative efforts. In the accompanying drawings: Figure 1 is a schematic diagram of the implementation environment of a parking control system shown in an exemplary embodiment of this application; Figure 2 is a flowchart of a parking control method shown in an exemplary embodiment of this application; Figure 3 is a flowchart of a specific parking control method shown in an exemplary embodiment of this application; Figure 4 is a block diagram of a parking control system shown in an exemplary embodiment of this application; Figure 5 is a block diagram of another parking control system shown in an exemplary embodiment of this application. Detailed implementation manners

[0018] The following will describe the implementation manners of this application with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of this application from the content disclosed in this specification. This application can also be implemented or applied through other different specific implementation manners. Various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this application. It should be understood that the preferred embodiments are only for explaining this application and not for limiting the protection scope of this application.

[0019] It should be noted that the diagrams provided in the following embodiments only schematically illustrate the basic concept of this application. Therefore, only the components related to this application are shown in the diagrams, rather than being drawn according to the number, shape, and size of the components in actual implementation. The type, quantity, and ratio of each component in actual implementation can be arbitrarily changed, and the component layout type may also be more complex.

[0020] In the following description, a large number of details are explored to provide a more thorough explanation of the embodiments of this application. However, it is obvious to those skilled in the art that the embodiments of this application can be implemented without these specific details. In other embodiments, well-known structures and devices are shown in the form of block diagrams rather than in detail to avoid making the embodiments of this application difficult to understand.

[0021] Automatic parking technology has corresponding technical applications in aspects such as environmental perception, path planning, and safety assurance. However, through research by the inventors of this application, it is found that automatic parking mainly focuses on the vehicle being able to park safely and accurately into a parking space that meets geometric constraint conditions, while ignoring the convenience and safety of the driver and passengers getting out of the vehicle. That is, after actual parking is completed, the vehicle may be parked too close to surrounding obstacles, making it impossible for the driver and passengers to open the door to get out of the vehicle or unable to safely open the door to get out of the vehicle, thus causing difficulties or dangers for the driver and passengers to open the door. In order to ensure that the driver and passengers can get out of the vehicle conveniently and safely, some solutions set a fixed distance threshold as a standard. When parking, the distance between the vehicle and the obstacle is restricted. However, this method is difficult to adapt to the dynamically changing parking environment and the diverse requirements of parking space layouts. Especially in the face of diagonal parking spaces or the presence of moving vehicles around, as well as in urban crowded areas where parking spaces are compact and there are many surrounding obstacles, the fixed threshold cannot adjust the safety margin in a timely manner, and there are still obvious shortcomings in ensuring the safety and convenience of the driver and passengers getting out of the vehicle. Therefore, how to ensure the safety and convenience of the driver and passengers getting out of the vehicle during automatic parking has become an urgent problem to be solved.

[0022] Therefore, please refer to Figure 1 , Figure 1 which is a schematic diagram of the implementation environment of a parking control system shown in an exemplary embodiment of this application. As Figure 1 shown, this implementation environment includes a vehicle 110 and a parking control system 120. Among them, the parking control system 120 is embedded in the vehicle 110 and is used to implement the parking control of the vehicle 110. The parking control system 120 includes but is not limited to a car machine system, an in-vehicle computer, etc. During automatic parking, through a safe getting-out space evaluation mechanism, the getting-out space is evaluated in advance, that is, by comprehensively considering the vehicle width, the lateral opening distance of the door, and the obstacle information, the lateral safety distance between the vehicle and the obstacle is determined, and this lateral safety distance is used as a parking path constraint. With the goal of ensuring that there is enough opening space for the door after parking, the vehicle is controlled to complete parking, thereby effectively ensuring that the driver and passengers can obtain sufficient door opening space, enabling the driver and passengers to get out of the vehicle safely and conveniently, and improving the driving experience of users.

[0023] Please refer to Figure 2 , Figure 2 which is a flowchart of a parking control method shown in an exemplary embodiment of this application. This method can be applied to the Figure 1 shown implementation environment. It should be understood that this method can also be applicable to other exemplary implementation environments, and this embodiment does not limit the implementation environment to which this method applies.

[0024] As Figure 2As shown, in an exemplary embodiment, the parking control method at least includes steps S210 to S240, which are introduced in detail as follows: Step S210, obtain the parameter information and parking information of the vehicle. The parameter information includes the vehicle width and the lateral opening distance of the door, and the parking information includes the vehicle position, the parking space position, and the obstacle information in the parking environment.

[0025] Step S220, based on the parameter information and the obstacle information, conduct an evaluation of the getting-off space to determine the lateral safety distance between the vehicle and the obstacle.

[0026] Step S230, construct a safe driving corridor according to the parking information, and use the lateral safety distance as a constraint condition for path search to search for a target parking path in the safe driving corridor.

[0027] Step S240, control the vehicle to complete parking according to the target parking path.

[0028] Among them, the lateral opening distance of the door refers to the maximum horizontal distance from the outermost edge of the door to the side of the vehicle body when the door is opened, that is, the space required by the door in the lateral direction; the lateral safety distance refers to the minimum lateral distance that must be maintained between the vehicle and the obstacle to avoid collision between the door and the obstacle when the door is opened based on the lateral opening distance of the door; the safe driving corridor refers to the drivable area of the vehicle, within which it indicates that the vehicle is always moving within a safe range.

[0029] In addition, the vehicle width can be obtained from the cloud database, and the lateral opening distance of the door can be obtained from the door control system. Among them, the lateral opening distance of the door can be the distance corresponding to the fully opened door, or the distance corresponding to the user's personalized opening angle; the vehicle position can be obtained based on the in-vehicle positioning system, the parking space position can be obtained by calling the digital map of the parking lot, or based on the camera, and the obstacle information in the parking environment can be obtained based on lidar, ultrasonic sensors, cameras, and their combinations.

[0030] In step S210, obtaining the parameter information and parking information of the vehicle provides input for subsequent evaluation of the getting-off space and parking path planning.

[0031] In step S220, by comprehensively considering the vehicle width, the door opening requirement, and the obstacle information, an evaluation of the getting-off space is conducted to determine the lateral safety distance between the vehicle and the obstacle, that is, the minimum distance to be reserved between the vehicle and the obstacle after parking. Different door opening requirements or different obstacle information result in different lateral safety distances, thus formulating a dynamic evaluation mechanism for the safe getting-off space.

[0032] In step S230, according to the vehicle position, parking space position, and obstacle information, a safe driving corridor is constructed, which can ensure that the vehicle parks within a safe range. The lateral safety distance is used as a constraint for path search, and the target parking path is searched in the safe driving corridor, meeting the requirements of the lateral safety distance throughout the process, and ensuring that there is enough space for getting off the vehicle after parking.

[0033] In step S240, according to the target parking path, the vehicle is controlled to accurately track the target parking path, so as to safely complete parking and leave enough space for getting off the vehicle for the driver and passengers.

[0034] In this embodiment, during the automatic parking process, through the safe getting-off space evaluation mechanism, the getting-off space is evaluated in advance, that is, by comprehensively considering the vehicle width, the lateral opening distance of the door, and the obstacle information, the lateral safety distance between the vehicle and the obstacle is determined, and this lateral safety distance is used as a parking path constraint. With the goal of ensuring that there is enough opening space for the door after parking, the vehicle is controlled to complete parking, thus effectively ensuring that the driver and passengers can obtain sufficient door opening space, enabling them to get off the vehicle safely and conveniently, and improving the user's driving and riding experience.

[0035] In addition, in this embodiment, based on the changes in parking progress and the dynamic changes in the environment, the safe driving corridor and the target parking path will also be dynamically updated.

[0036] In one embodiment, the method for obtaining obstacle information includes: acquiring environmental perception data of the parking environment, where the environmental perception data includes lidar data, ultrasonic data, and image data; fusing the lidar data, ultrasonic data, and image data to obtain a multi-source environmental model; identifying obstacles in the multi-source environmental model to obtain obstacle information of at least one obstacle, and the obstacle information includes obstacle position, obstacle width, obstacle height, and obstacle type.

[0037] Among them, the lidar data is captured by a lidar and is three-dimensional point cloud data. Each point cloud data point can be represented as , where is the index of the point cloud data point, represents the spatial coordinate value of the rd point cloud data point in the X-axis direction, represents the spatial coordinate value of the th point cloud data point in the Y-axis direction, represents the spatial coordinate value of the th point cloud data point in the Z-axis direction; the ultrasonic data is captured by an ultrasonic sensor to supplement the detection of close-range obstacles, especially soft obstacles (such as pedestrians), and provides a series of distance measurement values , where represent different ultrasonic sensors; the image data is captured by a camera, which is a high-resolution image for visual recognition and scene understanding, generating image frames, and each pixel can be represented as where is the index of the pixel point, represents the pixel point 's color component value in the red channel, represents the pixel point 's color component value in the green channel, represents the pixel point 's color component value in the blue channel. The lidar provides high-precision three-dimensional point cloud data, which can clearly present the spatial structure of the surrounding environment; the ultrasonic sensor focuses on detecting close-range obstacles, especially showing excellent performance for soft obstacles; the camera captures rich visual images to assist in scene understanding and target recognition.

[0038] In this embodiment, considering that most parking control schemes rely on a single sensor to perceive environmental data, but a single sensor cannot guarantee the accuracy of environmental perception. For example, although the ultrasonic sensor can detect close-range obstacles, in a complex environment, it is extremely vulnerable to environmental noise interference, resulting in an increased false detection rate. In a noisy urban street or parking lot, the surrounding noise signals may be misjudged as obstacle signals. The camera is seriously affected by lighting conditions. In harsh environments such as low light, rain, and snow, the image quality drops significantly, making the accuracy of obstacle detection based on image recognition decrease significantly. Moreover, a single sensor has weak real-time tracking ability for moving obstacles such as pedestrians and vehicles, and has a low recognition rate for unstructured obstacles such as temporarily placed roadblocks and irregular sundries. Therefore, in this embodiment, a multi-sensor fusion technology is adopted to organically combine the lidar, camera, and ultrasonic sensor to form a complementary perception system, making up for the limitations of a single sensor and achieving a full-range perception of the vehicle's surrounding environment. In this way, based on the multi-source environment model obtained from the full-range perception for obstacle recognition, the vehicle's surrounding environment can be stably perceived regardless of light changes, obstacle occlusion, or bad weather conditions, ensuring the accuracy of obstacle recognition in complex environments.

[0039] In a possible embodiment, a deep learning algorithm is used to identify obstacles in the multi-source environment model and determine the obstacle boundary and obstacle type. The obstacle boundary is represented as where, , are the center point coordinates, representing the obstacle position, , are the obstacle width and obstacle height respectively, is the index of the obstacle. For the definition of the obstacle boundary, semantic segmentation technology can be used, and the bounding box of the obstacle can be predicted through the segmentation network, significantly enhancing the ability to identify obstacles. Obstacle types include vehicles, pedestrians, and stone piers, etc.

[0040] Exemplarily, the deep learning algorithm is CNN (Convolutional Neural Network), and the embodiments of this application do not limit the deep learning algorithm. Feature extraction is performed from the multi-source environment model through the CNN algorithm, and then the obstacles are classified through the fully connected layer, and the probability that each pixel belongs to the obstacle is output. , define a threshold to determine whether the pixel belongs to the obstacle: Formula (1) where represents the pixel whether it is recognized as an obstacle (1 means yes, 0 means no).

[0041] In a possible embodiment, the obstacle types also include ground obstacles and off-ground obstacles. If it is a ground obstacle, the obstacle height refers to the vertical distance from the ground to the top of the obstacle. If it is an off-ground obstacle, the obstacle height is the vertical distance from the lowest point of the obstacle to the ground.

[0042] Exemplarily, ground obstacles such as stone piers and pedestrians on the ground, and off-ground obstacles such as hanging warning lights and branches extended by big trees.

[0043] In a possible embodiment, the obstacle height is used to filter out obstacles that do not affect the opening of the car door.

[0044] In this embodiment, if the obstacle is a ground obstacle, the obstacle height is compared with the height of the lower edge of the car door from the ground. If the obstacle height is less than the height of the lower edge of the car door from the ground, it is determined that the car door will not collide with the obstacle when opening; if the obstacle is an off-ground obstacle, the obstacle height is compared with the height of the upper edge of the car door from the ground. If the obstacle height is greater than the height of the upper edge of the car door from the ground, it is determined that the car door will not collide with the obstacle when opening. In this way, the height dimension is introduced to intelligently filter the obstacles. In parking control, the calculation of the lateral safety distance and the planning of the target parking path do not need to consider these obstacles, which significantly improves the parking efficiency while ensuring parking safety.

[0045] In one embodiment, lidar data, ultrasonic data, and image data are fused to obtain a multi-source environment model, including: constructing a data fusion model according to the characteristics of lidar, ultrasonic sensors, and cameras. The data fusion model includes a prediction sub-model and an update sub-model. The prediction sub-model includes a system model matrix, a control input influence matrix, and a process noise covariance matrix, and the update sub-model includes a measurement model matrix; inputting the real-time control input vector of the obstacle into the prediction sub-model to obtain a predicted state vector and a predicted state covariance matrix, where the predicted state vector and the predicted state covariance matrix are the state vector and the state covariance matrix of the obstacle at the current moment; determining the measurement value vector and the measurement noise covariance matrix of the obstacle at the current moment according to the lidar data, ultrasonic data, and image data; and correcting the predicted state vector and the predicted state covariance matrix according to the measurement value vector, the measurement noise covariance matrix, and the update sub-model to obtain a corrected state vector and a state covariance matrix, thereby forming a multi-source environment model.

[0046] Among them, the state vector is used to describe the position, speed, acceleration, etc. of the obstacle (i.e., the objects around the vehicle).

[0047] In this embodiment, considering the traditional multi-source data fusion method, when fusing the three-dimensional point cloud data generated by lidar and the visual image data captured by a camera, it is difficult to effectively handle the non-linear dynamic environment. In the actual parking scenario, the modeling accuracy of the environment is insufficient, and it cannot accurately reflect the real situation around the vehicle, affecting subsequent path planning and safety assessment. Therefore, multi-source data fusion based on the Kalman filter algorithm can better handle the non-linear dynamic environment, not only improving the fusion efficiency of multi-source heterogeneous data, but also improving the accuracy of environment modeling.

[0048] In addition, in this embodiment, according to the characteristics of lidar, ultrasonic sensors, and cameras, a system model matrix, a control input influence matrix, a process noise covariance matrix, and a measurement model matrix are designed. After constructing the data fusion model, inputting the real-time control input vector of the obstacle into the prediction sub-model can predict the state vector and the state covariance matrix of the obstacle at the current moment. Then, according to the lidar data, ultrasonic data, and image data, the measurement value vector and the measurement noise covariance matrix of the obstacle are determined. Combining with the update sub-model, the predicted state vector and the predicted state covariance matrix can be corrected to obtain the final state vector and the state covariance matrix of the obstacle, thereby forming a multi-source environment model. Subsequently, the prediction and correction of the state vector and the state covariance matrix of the obstacle at the next moment are continuously performed to obtain a real-time multi-source environment model, providing reliable environmental data support for subsequent obstacle recognition, getting-off safety space assessment, and parking path planning.

[0049] Exemplarily, the data fusion model is a Kalman filter model, that is, the fusion of multi-source data is realized based on the Kalman filter, and its core lies in two steps: the prediction stage and the update stage. The expression of the prediction sub-model is: Formula (2) Formula (3) Where represents the predicted state vector at the current time, represents the predicted state vector at the previous time, represents the system model matrix, represents the control input influence matrix, represents the control input vector, represents the predicted state covariance matrix at the current time, represents the predicted state covariance matrix at the previous time, represents the process noise covariance matrix; The expression of the update sub-model is: Formula (4) Formula (5) Formula (6) Where represents the Kalman gain, represents the predicted state covariance matrix at the current time, represents the measurement model matrix, represents the measurement noise covariance matrix, represents the corrected state vector at the current time, represents the predicted state vector at the current time, represents the measurement value vector, represents the corrected state covariance matrix at the current time, represents the identity matrix.

[0050] In the prediction sub-model, integrates the object space position information obtained by the lidar (reflected in ), the distance information measured by the ultrasonic sensor and the object feature information recognized by the camera (such as shape, color, etc. indirectly reflected in the state vector); is the system model matrix, which is used to describe the transfer of the system state over time. Its association with the sensor data lies in that the setting of its parameters needs to consider the environmental change laws reflected by different sensor data; is the control input influence matrix, is the control input vector, and these two are mainly associated with the position and speed of the obstacle; is the predicted state covariance matrix, which is used to measure the uncertainty of the predicted state of the obstacle. Its calculation depends on the accuracy and reliability of different sensor data. For example, the accuracy of lidar data, the measurement error of ultrasonic sensors, and the recognition accuracy of cameras, etc., will all affect the calculation of the covariance matrix; is the process noise covariance matrix, which is used to represent the noise interference in the system process. Sensors will inevitably be affected by various noises during data acquisition, such as the measurement noise of lidar, the environmental noise interference of ultrasonic sensors, and the image noise of cameras, etc. These noise factors will all be reflected in .

[0051] In the update sub-model, is the Kalman gain, which is used to balance the weights of the predicted value and the measured value. Its calculation is closely related to the accuracy of sensor data; is the measurement model matrix, which is used to map the obstacle state to the measurement space, that is, to map the three-dimensional point cloud data measured by lidar, the distance measurement values of ultrasonic sensors, and the visual feature information of cameras, etc. to a unified measurement space for data fusion; is the measured value vector, which directly corresponds to the data collected by each sensor, that is, the three-dimensional point cloud data of lidar , the distance measurement values of ultrasonic sensors and the pixel information of the camera The feature vector obtained after processing; is the measurement noise covariance matrix, which is used to represent the noise interference in the measurement process. The measurement noise characteristics of different sensors are different. For example, the measurement noise distribution of lidar, the measurement error range of ultrasonic sensors, and the image noise characteristics of cameras, etc. will all affect the value of.

[0052] In one embodiment, determining the lateral safety distance between the vehicle and the obstacle includes: calculating the obstacle width, the vehicle width and the lateral opening distance of the door to obtain the theoretical safety distance between the vehicle and the obstacle. The obstacle information includes the obstacle width; calculating the lateral safety distance based on the theoretical safety distance and a preset safety distance margin. The safety distance margin is a compensation distance set to prevent the vehicle from colliding with the obstacle.

[0053] In this embodiment, considering that in automatic parking, the door needs to have enough space to safely open the door and enable the driver and passengers to get off safely and conveniently. However, the setting method of the fixed distance threshold cannot dynamically adapt to various parking scenarios and various environmental changes, and cannot effectively ensure that the driver and passengers can get off safely and conveniently. Therefore, a dynamic lateral safety distance calculation method is proposed, that is, by combining the obstacle width, vehicle width and door lateral opening distance, the theoretical safety distance is first determined, and this theoretical safety distance varies according to the different widths of obstacles in the environment. At the same time, to avoid accidental collisions, a safety margin, that is, a safety distance margin, is added on the basis of the theoretical safety distance. In this way, the deficiency of the static safety distance model is effectively solved, and the reliability of the obtained lateral safety distance is ensured.

[0054] Exemplarily, the safety distance margin is a compensation distance set to prevent the vehicle from colliding with obstacles due to obstacle recognition errors, changes in the parking environment and parking position deviations.

[0055] In this way, the calculated lateral safety distance is more in line with the actual situation. No matter how complex the parking space layout is and how dynamically the surrounding obstacles change, there will be enough space on both sides of the vehicle for the driver and passengers to get off safely and conveniently after the vehicle stops.

[0056] Exemplarily, the calculation formula for the lateral safety distance is: Formula (7) Wherein, represents the lateral safety distance, represents the vehicle width, represents the obstacle width, represents the door lateral opening distance, represents the safety distance margin.

[0057] In a possible embodiment, if multiple obstacles are identified, the lateral safety distance corresponding to the vehicle and each obstacle needs to be calculated.

[0058] In one embodiment, a safe driving corridor is constructed according to the parking information, including: determining multiple candidate positions of the vehicle according to the vehicle position, parking space position and at least one obstacle position, the obstacle information includes the obstacle positions of at least one obstacle, and the candidate positions are positions where the vehicle will not collide with all obstacles; determining the total potential energy value of each candidate position according to the distance between each candidate position and each obstacle; if the total potential energy value of the target candidate position is less than or equal to a preset safe potential energy threshold, then determine the target candidate position as the target position; generating a safe driving corridor according to each target position.

[0059] In this embodiment, to ensure that the vehicle is always in a safe state in a complex environment during automatic parking, a safe driving corridor construction algorithm is proposed to dynamically generate a passage through which the vehicle can drive safely by real-time analysis and calculation of the environmental information around the vehicle.

[0060] In this embodiment, the total potential energy value is used to measure the safe distance between the vehicle and the surrounding obstacles at a certain candidate position, that is, to evaluate the possibility of the vehicle colliding at a certain candidate position. The smaller the total potential energy value, the safer the candidate position.

[0061] In this embodiment, all positions where the vehicle will not collide with any obstacles are first selected from the parking environment as candidate positions for forming the safe driving corridor. Then, according to whether the total potential energy value of each candidate position meets the condition of the safe potential energy threshold, the target positions that finally form the safe driving corridor are screened out from the candidate positions. In this way, by constructing the safe driving corridor through the potential energy field, the safest vehicle driving area can be selected, optimizing the robustness of the environmental modeling, adapting to various complex environments, and effectively reducing the collision risk of the vehicle.

[0062] Exemplarily, the safe driving corridor is defined by the following formula: Formula (8) Wherein, represents the safe driving corridor of the vehicle at moment, represents a candidate position of the vehicle at moment, represents the free configuration space of the vehicle at moment, represents the total potential energy value of the vehicle at the candidate position at moment, represents the safe potential energy threshold.

[0063] In this exemplary embodiment, The set composed of all candidate positions of the vehicle that meet specific conditions The specific condition is from the free configuration space of the vehicle at , which means that the candidate position is a position where the vehicle can theoretically reach and will not collide with any known obstacles. At the same time, the candidate position also needs to meet the condition that , that is, the total potential energy value of the vehicle at the candidate position at moment cannot exceed the potential energy safety threshold , so as to select the safest vehicle driving area and form a safe driving corridor.

[0064] In addition, as the vehicle moves and the states of surrounding obstacles change (such as the movement of other vehicles, the walking of pedestrians, etc.), the total potential energy value of the vehicle at different positions at each moment will also change accordingly. When the total potential energy value of the vehicle at a certain candidate position satisfies , this position is in a safe state. Connecting all these positions that meet the safety conditions forms a dynamically changing area where the vehicle can drive, that is, a safe driving corridor. The shape and position of this safe driving corridor will be adjusted in real time as the vehicle drives and the surrounding environment dynamically changes. For example, when a stationary obstacle starts to move closer to the vehicle, the new total potential energy value of this position will be calculated in real time. If it does not meet the safety conditions, the boundary of the safe driving corridor will also contract in the direction away from the obstacle accordingly to ensure that the vehicle is always in a safe area.

[0065] In one embodiment, determining the total potential energy value of each candidate position according to the distance between each candidate position and each obstacle includes: calculating the potential energy values of each obstacle by respectively calculating the positions of each obstacle, the candidate position, the lateral safety distance, and a preset sensitivity coefficient, where the sensitivity coefficient represents the sensitivity of the potential energy to the change in the distance of the obstacle; calculating the sum of the potential energy values of each obstacle to obtain the total potential energy value.

[0066] Among them, the potential energy value of each obstacle is used to measure the safe distance between the vehicle and a certain obstacle at a certain candidate position, that is, to evaluate the possibility of a collision between the vehicle and a certain obstacle at a certain candidate position. The smaller the potential energy value, the smaller the possibility of a collision between the vehicle and that certain obstacle at that certain candidate position, that is, that certain obstacle is farther away from the vehicle.

[0067] In addition, the sensitivity coefficient represents the sensitivity of the potential energy function to the change in the distance of the obstacle. That is to say, it reflects the change rate or amplitude of the potential energy value when the distance between the vehicle and the obstacle changes. If the sensitivity coefficient is high, it means that even a small change in the distance between the vehicle and the obstacle will cause a significant change in the potential energy value, indicating that the potential energy is more sensitive to the change in distance. If the sensitivity coefficient is low, it means that the response of the potential energy value to the change in distance is relatively sluggish. The sensitivity coefficient can adjust the steepness of the potential energy function. By reasonably setting the sensitivity coefficient, the sensitivity of the potential energy function to the change in the distance between the vehicle and the obstacle can be controlled, thereby optimizing the generation effect of the safe driving corridor.

[0068] In this embodiment, to ensure that the vehicle is always in a safe state in a complex environment during automatic parking and at the same time ensure that the driver and passengers have enough space to get out of the vehicle, all obstacles are comprehensively considered, and the total potential energy value of the vehicle at a certain candidate position is calculated comprehensively. Moreover, for the calculation of the total potential energy value of each candidate position, not only the distance between the vehicle and the obstacles is considered, but also the lateral safety distance and the sensitivity coefficient are introduced, so that the potential energy value is closely related to the safety space required for safe getting out of the vehicle and is closely related to the sensitivity to the change in the distance of the obstacles. In this way, the generation effect of the safe driving corridor can be effectively optimized.

[0069] In this embodiment, when calculating the potential energy value of each obstacle, the corresponding lateral safety distance is used.

[0070] Exemplarily, the calculation formula for the total potential energy value is: Formula (9) Wherein, represents the total potential energy value of the vehicle at the candidate position at time , represents the currently perceived number of obstacles, represents a candidate position of the vehicle at time, represents the th obstacle position at time, represents the lateral safety distance, represents the sensitivity coefficient.

[0071] In this exemplary embodiment, can be the coordinate information of the vehicle or a set of parameters that can describe the attitude of the vehicle in space; is updated in real time to ensure that the potential energy function can accurately reflect the real-time distance between the vehicle and each obstacle.

[0072] In a possible embodiment, if the parking space is in an environment with dense obstacles, a larger sensitivity coefficient is set to ensure higher safety; if the parking space is in an environment with sparse obstacles, a smaller sensitivity coefficient is set to avoid overreaction.

[0073] As a possible embodiment, the judgment criteria for dense obstacles and sparse obstacles can be jointly determined according to the number and distance of the surrounding obstacles. For example, if the number of obstacles detected within a preset distance range is greater than or equal to a preset threshold, it is considered that the obstacles are dense; conversely, if the number of obstacles is less than the preset threshold, it is considered that the obstacles are sparse. In addition, the sensitivity coefficient can vary gradiently according to the density of the obstacles.

[0074] In one embodiment, the lateral safety distance is used as a constraint condition for path search, and a target parking path is searched in the safe driving corridor, including: monitoring the potential field gradient in the safe driving corridor, and determining an initial parking path according to the potential field gradient descent direction, where the distance between each node in the initial parking path and each obstacle is greater than or equal to the lateral safety distance; according to the real-time change of the safe driving corridor, locally adjusting the initial parking path to obtain a target parking path, where the distance between each node in the target parking path and each obstacle is greater than or equal to the lateral safety distance.

[0075] In this embodiment, considering conventional path planning algorithms, such as the A-star algorithm (a search algorithm) and the RRT (Rapidly-exploring Random Trees) algorithm, after the initial planning is completed, they lack the ability of dynamic adjustment. When the environment suddenly changes during the parking process, such as the sudden appearance of obstacles or the sudden entry of other vehicles, they cannot quickly respond and re-plan the path, thus increasing the collision risk. Therefore, in order to ensure that the vehicle always travels in a safe area and there is enough space for the driver and passengers to get off after parking, a parking path is searched in the constructed safe driving corridor, and the lateral safety distance is used as a constraint condition for path search, and local path adjustment is performed in real time to ensure the completion of the search for the target parking path and obtain a better parking path. In this way, applying the graph search algorithm combined with the constraint condition to plan the parking path in the safe driving corridor can not only avoid obstacles but also ensure that the driver and passengers can get off safely and conveniently after parking.

[0076] In this embodiment, the real-time potential field gradient in the safe driving corridor is used to guide the search for the vehicle's parking path. The potential field gradient reflects the change trend of the potential value in space. The initial parking path is determined according to the potential field gradient descent direction, that is, driving towards the area with a lower potential value, so as to stay away from obstacles to the greatest extent and remain within the safe corridor. In addition, the distance between each node in the determined initial parking path and each obstacle should also be greater than or equal to the lateral safety distance. That is to say, if the initial parking path selected according to a certain potential field gradient descent direction does not meet this condition, that is, there is a node that does not meet the condition, then this path will be regarded as infeasible, and a path in the direction of another potential field gradient descent will be selected as the initial parking path. Among them, each node in the initial parking path represents a position of the vehicle. At any node, the respective lateral safety distances of each obstacle are used as a standard to judge whether the distance between each obstacle and the node meets the condition.

[0077] Exemplarily, the initial parking path is determined according to the direction in which the potential field gradient drops fastest. However, if the condition of the lateral safety distance is not met, the direction in which the potential field gradient drops second fastest is selected to determine the initial parking path, and so on, until an initial parking path that meets the lateral safety distance condition is determined.

[0078] In this embodiment, during the entire parking process, considering the changes in the parking progress and the dynamic changes in the surrounding environment, which will affect the safety of the initial parking path and the getting-off space after parking, therefore, the initial parking path will be locally adjusted in real time. That is, when it is monitored that the initially planned parking path does not meet the safe getting-off condition based on the changes in the parking progress and the dynamic changes in the surrounding environment, the dynamic trajectory replanning mechanism will be immediately triggered, and according to the real-time obstacle information, the initial parking path will be locally adjusted dynamically and flexibly to obtain the target parking path. During the dynamic replanning process, not only must the direction of the path be the direction in which the potential field gradient drops, but also the lateral safety distance condition must be met between each node and each obstacle, so as to ensure parking safety and sufficient getting-off space for the driver and passengers. In the face of the adjustment of the parking path, the driving direction and speed of the vehicle will be correspondingly adjusted. For example, when it is detected that the potential energy value on one side of the vehicle is relatively high, which means there is a nearby obstacle in that direction, the vehicle will automatically adjust its steering to make the vehicle drive towards the other side with a lower potential energy value, thus avoiding collision with the obstacle. This path optimization mechanism based on the potential field will continue to run, being able to flexibly handle various emergencies during the parking process, ensuring the driving safety of the vehicle in real time, creating conditions for the driver and passengers to get off safely eventually, and greatly improving the success rate and safety of parking.

[0079] In addition, as a possible embodiment, during the parking process, the lateral safety distance is referred to twice. One is that when constructing the safe driving corridor, the lateral safety distance is introduced into the calculation of the total potential energy value of each candidate position of the vehicle. The other is that in the path search, the lateral safety distance is introduced as a constraint condition. Considering the lateral safety distance twice in this way can effectively ensure that there is sufficient getting-off space for the driver and passengers after parking.

[0080] In a possible embodiment, the path planning also needs to consider the turning radius. Among them, the turning radius is directly related to the maneuverability and safety of parking, and it is necessary to meet the actual kinematic characteristics of the vehicle, and a curvature constraint is introduced to limit the degree of bending of the path to ensure the smoothness of the parking path.

[0081] In one embodiment, controlling a vehicle to complete parking according to a target parking path includes: if the distance between a target node in the target parking path and each obstacle is less than the lateral safety distance, determining a target parking mode of the vehicle from multiple parking modes, where the multiple parking modes are, from highest to lowest priority, the path lateral adjustment mode, the delay mode, and the remote control mode; if the target parking mode is the path lateral adjustment mode, laterally adjusting the target parking path and controlling the vehicle to complete parking according to the adjusted target parking path; if the target parking mode is the delay mode, starting the first-level prompt and the delay mode, and controlling the vehicle to complete parking according to the target parking path, where the first-level prompt includes suggesting that the driver and passengers get out of the vehicle before parking and switching the parking mode to the delay mode; if the target parking mode is the remote control mode, starting the second-level prompt and the remote control mode, and after receiving a remote parking instruction, controlling the vehicle to complete parking according to the target parking path, where the second-level prompt includes asking the driver and passengers to get out of the vehicle before parking and switching the parking mode to the remote control mode.

[0082] In this embodiment, considering that there is still a situation where the driver and passengers do not have enough space to get out of the vehicle after parking is completed, the requirements of human-machine collaboration are not fully considered. Especially in extreme scenarios such as extremely narrow parking spaces or complex obstacle distributions, the driver and passengers can only be forced to take risks getting out of the vehicle or abandon parking, greatly reducing the user experience. Therefore, a hierarchical decision-making mechanism is set up, that is, according to the real-time situation of the obstacles and the result of path replanning, different-priority parking modes are used to complete parking. In this way, it can ensure that a safe and convenient parking solution can be provided for the driver and passengers in various complex situations.

[0083] In this embodiment, a three-level decision-making mechanism is designed, corresponding to three parking modes, which are, from highest to lowest priority, the path lateral adjustment mode, the delay mode, and the remote control mode. Among them, the path lateral adjustment mode means controlling the vehicle to complete parking by laterally adjusting the path, the delay mode means controlling the vehicle to complete parking by delaying the parking, and the remote control mode means controlling the vehicle to complete parking by remote control, fully considering the safety requirements and human-machine collaboration in different parking scenarios.

[0084] Exemplarily, determining the target parking mode of the vehicle from multiple parking modes includes: if the end point in the target parking path does not meet the condition of the lateral safety distance and all the driver and passengers in the vehicle can get out of the vehicle from the driver's side, determining the target parking mode as the path lateral adjustment mode; if the driver and passengers in the vehicle cannot all get out of the vehicle from the driver's side or the lateral safety distance is still not met after lateral adjustment of the path, determining the target parking mode as the delay mode; if the lateral safety distance is still not met after delaying the parking, determining the target parking mode as the remote control mode.

[0085] In this exemplary embodiment, when the target parking path does not meet the safe getting-off condition, an attempt will be made to solve the problem by slightly deviating from the original parking path first, aiming to ensure that there is enough space for getting off on at least one side of the vehicle during parking, so as to maximize the safety and convenience of the driver and passengers getting off without affecting the overall parking efficiency; if the safe getting-off condition still cannot be met after the path deviation adjustment, the first-level prompt and delayed parking strategy will be activated, that is, a prompt will be sent to the driver and passengers in time, suggesting them to get off before parking starts. At the same time, the vehicle will adopt the delayed parking strategy, that is, wait until the safe getting-off condition is met and then continue to complete the parking operation, providing sufficient time for the driver and passengers to get off safely and effectively reducing the safety risk during the getting-off process; in an extremely narrow parking space, if the first two strategies cannot effectively ensure safe getting-off, the second-level prompt and remote control parking strategy will be activated, which requires the driver and passengers to get off in advance, and then remotely control the vehicle to complete the parking operation through a specially developed in-vehicle APP (Application) or intelligent device, completely avoiding the safety risk of getting off during parking and providing a safer and more flexible parking solution for users.

[0086] In addition, for the first-level prompt and the second-level prompt, the node for prompting the driver and passengers to get off is the node in the target parking path that meets the lateral safety distance condition.

[0087] Please refer to Figure 3 , Figure 3 which is a flowchart of a specific parking control method shown in an exemplary embodiment of the present application. As Figure 3 shown, the steps of this specific parking control method at least include steps S310 to S390, which are described in detail as follows: Step S310, configure sensors and use the sensors to sense environmental data; Step S320, perform multi-source data fusion on the environmental data sensed by multiple sensors; Step S330, perform obstacle recognition according to the fused multi-source environmental model; Step S340, determine the lateral safety distance according to the obstacle recognition result and the parameter information of the vehicle; Step S350, construct a safe driving corridor according to the obstacle recognition result, the vehicle position and the parking space position; Step S360, use the lateral safety distance as a constraint condition for path search and search for an initial parking path in the safe driving corridor; Step S370, dynamically re-plan the initial parking path in real time to obtain the target parking path; Step S380, determine the target parking mode based on the target parking path; Step S390: Control the vehicle to complete parking according to the target parking mode and the target parking path or the laterally adjusted target parking path.

[0088] In this way, by integrating multi-source environment perception technology, parking path planning and replanning technology, and multi-level intelligent decision-making technology, the obstacle recognition accuracy and real-time performance in complex scenarios are effectively improved, and the contradiction between automatic parking and the safe getting-off of the driver and passengers is effectively solved, thus providing users with a safer and more convenient parking experience.

[0089] For the above parking control method, first obtain the parameter information and parking information of the vehicle. Among them, the parameter information includes the vehicle width and the lateral opening distance of the door, and the parking information includes the vehicle position, the parking space position, and the obstacle information in the parking environment. Then, according to the parameter information and the obstacle information, conduct an evaluation of the getting-off space to determine the lateral safety distance between the vehicle and the obstacle. Next, construct a safe driving corridor based on the parking information, and use the lateral safety distance as a constraint condition for path search to search for the target parking path in the safe driving corridor. Finally, control the vehicle to complete parking according to the target parking path. During the automatic parking process, through the safe getting-off space evaluation mechanism, the getting-off space is evaluated in advance, that is, by comprehensively considering the vehicle width, the lateral opening distance of the door, and the obstacle information, determine the lateral safety distance between the vehicle and the obstacle, and use this lateral safety distance as a parking path constraint to ensure that there is enough opening space for the door after parking as the goal, control the vehicle to complete parking, so as to effectively ensure that the driver and passengers can obtain sufficient door opening space, enabling the driver and passengers to get off safely and conveniently, and improving the driving experience of users.

[0090] Please refer to Figure 4 , Figure 4 which is a block diagram of a parking control system shown in an exemplary embodiment of the present application. This system can be applied to Figure 1 the implementation environment shown. It should be understood that this system can also be applicable to other exemplary implementation environments, and this embodiment does not limit the implementation environment applicable to this system.

[0091] As Figure 4 shown, in an exemplary embodiment, the parking control system 400 at least includes a data acquisition module 410, a space evaluation module 420, a path planning module 430, and a parking control module 440, which are introduced in detail as follows: The data acquisition module 410 is used to acquire the parameter information and parking information of the vehicle. The parameter information includes the vehicle width and the lateral opening distance of the door, and the parking information includes the vehicle position, the parking space position, and the obstacle information in the parking environment; A space evaluation module 420, configured to perform alighting space evaluation according to parameter information and obstacle information, and determine a lateral safety distance between the vehicle and the obstacle; A path planning module 430, configured to construct a safe driving corridor according to parking information, and use the lateral safety distance as a constraint condition for path search, and search for a target parking path in the safe driving corridor; A parking control module 440, configured to control the vehicle to complete parking according to the target parking path.

[0092] It should be noted that the parking control system provided in the above embodiment and the parking control method provided in the above embodiment belong to the same concept. The content of the operations performed by each module has been described in detail in the method embodiment, and will not be repeated here.

[0093] Please refer to Figure 5 , Figure 5 , which is a block diagram of another parking control system shown in an exemplary embodiment of the present application. This system can be applied to Figure 1 the implementation environment shown. It should be understood that this system can also be applied to other exemplary implementation environments. This embodiment does not limit the implementation environment applicable to this system.

[0094] As Figure 5 shown, in an exemplary embodiment, this another parking control system 500 at least includes an environment perception and processing module 510, a safe alighting space evaluation module 520, and an intelligent decision-making module 530, which are introduced in detail as follows: The environment perception and processing module 510 includes a data acquisition device 511, a data fusion device 512, and an obstacle recognition device 513. Among them, the data acquisition device 511 is configured to acquire environmental data sensed by various sensors, the data fusion device 512 is configured to perform multi-source data fusion on the environmental data sensed by multiple sensors, and the obstacle recognition device 513 is configured to perform obstacle recognition according to the fused multi-source environment model; The safe alighting space evaluation module 520 includes a safety distance calculation device 521, a corridor construction device 522, a path planning device 523, and a path replanning device 524. Among them, the safety distance calculation device 521 is configured to determine a lateral safety distance according to the obstacle recognition result and the parameter information of the vehicle, the corridor construction device 522 is configured to construct a safe driving corridor according to the obstacle recognition result, the vehicle position, and the parking space position, the path planning device 523 is configured to use the lateral safety distance as a constraint condition for path search, and search for an initial parking path in the safe driving corridor, and the path replanning device 524 is configured to perform dynamic replanning on the initial parking path in real time to obtain a target parking path; The intelligent decision-making module 530 includes a parking mode determination device 531, which is used to determine the target parking mode based on the target parking path; The control module 540 includes a parking control device 541, which is used to control the vehicle to complete parking according to the target parking mode and the target parking path or the laterally adjusted target parking path.

[0095] It should be noted that the parking control system provided in the above embodiments and the parking control method provided in the above embodiments belong to the same concept. The content of the operations performed by each module has been described in detail in the method embodiments and will not be repeated here.

[0096] In addition, the parking control system 500 includes all the functions of the parking control system 400. The data acquisition device 511, data fusion device 512, and obstacle recognition device 513 in the parking control system 500 are equivalent to the data acquisition module 410 in the parking control system 400. The safety distance calculation device 521 in the parking control system 500 is equivalent to the space evaluation module 420 in the parking control system 400. The corridor construction device 522, path planning device 523, and path replanning device 524 in the parking control system 500 are equivalent to the path planning module 430 in the parking control system 400. The parking control device 541 in the parking control system 500 is equivalent to the parking control module 440 in the parking control system 400.

[0097] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present application. Among them, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code, and the above-mentioned module, program segment, or part of code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram or flowchart, and the combination of blocks in the block diagram or flowchart, can be implemented by a dedicated hardware-based system for performing the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0098] The units involved in the embodiments described in the present application can be implemented in software or in hardware, and the described units can also be provided in a processor. Among them, the names of these units do not constitute a limitation to the unit itself in some cases.

[0099] The above embodiments are only used to exemplarily illustrate the principles and effects of the present application, rather than to limit the present application. Any person familiar with this technology can modify or change the above embodiments without departing from the spirit and scope of the present application. Therefore, all equivalent modifications or changes completed by those with ordinary knowledge in the technical field without departing from the spirit and technical ideas disclosed by the present application should still be covered by the claims of the present application.

Claims

1. A parking control method, characterized in that The method includes: Obtaining parameter information and parking information of the vehicle, where the parameter information includes the vehicle width and the lateral opening distance of the vehicle door, and the parking information includes the vehicle position, the parking space position, and obstacle information in the parking environment; Performing an evaluation of the getting-off space based on the parameter information and the obstacle information to determine the lateral safety distance between the vehicle and the obstacle; Constructing a safe driving corridor according to the parking information, and using the lateral safety distance as a constraint condition for path search to search for a target parking path in the safe driving corridor; Controlling the vehicle to complete parking according to the target parking path.

2. The parking control method according to claim 1, wherein The determining the lateral safety distance between the vehicle and the obstacle includes: Calculating the obstacle width, the vehicle width, and the lateral opening distance of the vehicle door to obtain the theoretical safety distance between the vehicle and the obstacle, where the obstacle information includes the obstacle width; Calculating the lateral safety distance based on the theoretical safety distance and a preset safety distance margin, where the safety distance margin is a compensation distance set to prevent the vehicle from colliding with the obstacle.

3. The parking control method according to claim 1, characterized in that The constructing a safe driving corridor according to the parking information includes: Determining a plurality of candidate positions of the vehicle according to the vehicle position, the parking space position, and at least one obstacle position, where the obstacle information includes the obstacle positions of at least one obstacle, and the candidate positions are positions where the vehicle will not collide with all obstacles; Determining the total potential energy value of each candidate position according to the distance between each candidate position and each obstacle; If the total potential energy value of the target candidate position is less than or equal to a preset safe potential energy threshold, then determining the target candidate position as the target position; Generating the safe driving corridor according to the target positions.

4. The parking control method according to claim 3, wherein The determining the total potential energy value of each candidate position according to the distance between each candidate position and each obstacle includes: Calculating the potential energy values of each obstacle by respectively calculating the obstacle positions, the candidate position, the lateral safety distance, and a preset sensitivity coefficient, where the sensitivity coefficient represents the sensitivity of the potential energy to the change in the obstacle distance; Calculating the sum of the potential energy values of each obstacle to obtain the total potential energy value.

5. The parking control method according to claim 3, wherein The using the lateral safety distance as a constraint condition for path search to search for a target parking path in the safe driving corridor includes: Monitoring the potential field gradient in the safe driving corridor, and determining an initial parking path according to the direction of the potential field gradient descent, where the distance between each node in the initial parking path and each obstacle is greater than or equal to the lateral safety distance; Performing a local adjustment on the initial parking path according to the real-time change of the safe driving corridor to obtain a target parking path, where the distance between each node in the target parking path and each obstacle is greater than or equal to the lateral safety distance.

6. The parking control method according to claim 1, characterized in that The obtaining manner of the obstacle information includes: Obtaining the environmental perception data of the parking environment, where the environmental perception data includes lidar data, ultrasonic data, and image data; Fusing the lidar data, the ultrasonic data, and the image data to obtain a multi-source environmental model; Identify obstacles in the multi-source environment model to obtain the obstacle information of at least one obstacle, where the obstacle information includes obstacle position, obstacle width, obstacle height, and obstacle type.

7. The parking control method according to claim 6, characterized in that, The fusion of the lidar data, the ultrasonic data, and the image data to obtain a multi-source environment model includes: Construct a data fusion model according to the characteristics of the lidar, ultrasonic sensor, and camera. The data fusion model includes a prediction sub-model and an update sub-model. The prediction sub-model includes a system model matrix, a control input influence matrix, and a process noise covariance matrix. The update sub-model includes a measurement model matrix; Input the real-time control input vector of the obstacle into the prediction sub-model to obtain a predicted state vector and a predicted state covariance matrix, where the predicted state vector and the predicted state covariance matrix are the state vector and state covariance matrix of the obstacle at the current moment; Determine the measurement value vector and measurement noise covariance matrix of the obstacle at the current moment according to the lidar data, the ultrasonic data, and the image data; According to the measurement value vector, the measurement noise covariance matrix, and the update sub-model, correct the predicted state vector and the predicted state covariance matrix to obtain a corrected state vector and state covariance matrix, forming the multi-source environment model.

8. The parking control method according to any one of claims 1 to 7, characterized in that, The control of the vehicle to complete parking according to the target parking path includes: If the distance between the target node in the target parking path and each obstacle is less than the lateral safety distance, determine the target parking mode of the vehicle from multiple parking modes. The multiple parking modes are, from high to low priority, the path lateral adjustment mode, the delay mode, and the remote control mode; If the target parking mode is the path lateral adjustment mode, perform a lateral adjustment on the target parking path and control the vehicle to complete parking according to the adjusted target parking path; If the target parking mode is the delay mode, activate the first-level prompt and the delay mode, and control the vehicle to complete parking according to the target parking path. The first-level prompt includes suggesting that the driver and passengers get out of the vehicle before parking and switching the parking mode to the delay mode; If the target parking mode is the remote control mode, activate the second-level prompt and the remote control mode, and after receiving the remote parking instruction, control the vehicle to complete parking according to the target parking path. The second-level prompt includes asking the driver and passengers to get out of the vehicle before parking and switching the parking mode to the remote control mode.

9. A parking control system, characterized in that, The system includes: A data acquisition module for acquiring the parameter information and parking information of the vehicle. The parameter information includes the vehicle width and the lateral opening distance of the car door. The parking information includes the vehicle position, the parking space position, and the obstacle information in the parking environment; A space evaluation module for performing a getting-out space evaluation according to the parameter information and the obstacle information to determine the lateral safety distance between the vehicle and the obstacle; A path planning module, where a user constructs a safe driving corridor according to the parking information, and uses the lateral safety distance as a constraint condition for path search to search for a target parking path in the safe driving corridor; A parking control module, configured to control a vehicle to complete parking according to the target parking path.

10. A vehicle, characterized in that, Use the parking control method according to any one of claims 1 to 8, or include the parking control system according to claim 9.

Images