WO2021227304A1 - 一种自动泊车路径规划的避障方法及泊车路径规划系统 - Google Patents

一种自动泊车路径规划的避障方法及泊车路径规划系统 Download PDF

Info

Publication number
WO2021227304A1
WO2021227304A1 PCT/CN2020/111804 CN2020111804W WO2021227304A1 WO 2021227304 A1 WO2021227304 A1 WO 2021227304A1 CN 2020111804 W CN2020111804 W CN 2020111804W WO 2021227304 A1 WO2021227304 A1 WO 2021227304A1
Authority
WO
WIPO (PCT)
Prior art keywords
parking path
vehicle
projection
obstacle
parking
Prior art date
Application number
PCT/CN2020/111804
Other languages
English (en)
French (fr)
Inventor
仝乐斌
吕兵兵
成航见
Original Assignee
惠州市德赛西威汽车电子股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 惠州市德赛西威汽车电子股份有限公司 filed Critical 惠州市德赛西威汽车电子股份有限公司
Publication of WO2021227304A1 publication Critical patent/WO2021227304A1/zh

Links

Images

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W30/00—Purposes of road vehicle drive control systems not related to the control of a particular sub-unit, e.g. of systems using conjoint control of vehicle sub-units
    • B60W30/06—Automatic manoeuvring for parking
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W30/00—Purposes of road vehicle drive control systems not related to the control of a particular sub-unit, e.g. of systems using conjoint control of vehicle sub-units
    • B60W30/08—Active safety systems predicting or avoiding probable or impending collision or attempting to minimise its consequences
    • B60W30/095—Predicting travel path or likelihood of collision
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W40/00—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models
    • B60W40/02—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models related to ambient conditions

Definitions

  • the invention relates to the technical field of parking path planning, in particular to an obstacle avoidance method for automatic parking path planning and a parking path planning system.
  • the automatic parking system has become a research hotspot of major domestic and foreign companies and institutions in recent years.
  • the automatic parking system has been gradually applied to some vehicles as a high-end configuration to assist the driver in completing the parking process. It mainly includes three parts: a recognition system, a path planning system, and a parking control system.
  • the recognition system In the entire parking process, the recognition system generally first recognizes the parking spaces in the space and the obstacle information around the parking spaces, and then transmits the recognized environmental information to the path planning system, which is based on the information collected by the sensors.
  • the parking control system converts the path information into control decisions according to the planned path, controls the vehicle's turning angle, speed, gear, etc., and feeds back the execution results to the central processing unit to facilitate further analysis and formulation by the central processing unit decision making. From the entire automatic parking process, it can be seen that the parking effect is closely related to path planning. The advantages of path planning directly lead to the success of automatic parking. Therefore, the path planning of the automatic parking system is an extremely important link.
  • the midpoint of the rear axle of the vehicle is generally selected as the vehicle reference point (that is, the center of mass point), and the planned parking path trajectory refers to the set of positions passed by the center of mass point, that is, the point set.
  • the actual vehicle volume is not considered. This leads to some centroid points that have no intersection with the centroid trajectories of surrounding vehicles or obstacles, but when considering the projected area of the vehicle and surrounding obstacles on the ground After that, it will cause the area to overlap, that is, there is a risk of collision.
  • the parking path planned by the collection of positions passed by the centroid when the volume of the vehicle or the volume of the surrounding obstacles is considered, it is very likely that the edge of the vehicle will interact with the obstacle during the parking process. There is a risk of friction or collision at the edges, as shown in Figure 1.
  • the dimensions of vehicles and obstacles are directly taken into account in the planning algorithm during path planning, it will greatly increase the difficulty of the solution, and the speed of the solution will decrease significantly, resulting in obvious delays and large errors. In this way, it is difficult to meet the requirements of automatic parking path planning.
  • the present invention provides an obstacle avoidance method for automatic parking path planning.
  • the method includes the following steps:
  • the vehicle contour is fitted to each node in turn;
  • the establishment of a coordinate system determines whether there is a vehicle profile that overlaps with the obstacle in both the X-axis direction and the Y-axis direction. If it is, the parking path trajectory is considered unqualified, and the parking path is re-planned; otherwise, The steps to consider the parking path trajectory to be qualified include:
  • the step of acquiring the first projection of the vehicle contour on the node on the X axis and the second projection on the Y axis includes:
  • the step of obtaining the third projection of the obstacle on the X axis and the fourth projection on the Y axis includes:
  • the step of sequentially fitting the vehicle contour to each node according to the pose corresponding to each node includes:
  • the rectangle is fitted to each node in turn as the contour of the vehicle.
  • the contour parameters of the vehicle are increased by preset expansion parameters, thereby simplifying the car body into a rectangle.
  • the obstacle information includes the position of the obstacle and the size of the obstacle; the obstacle information is obtained through a vehicle-mounted camera and a vehicle-mounted radar.
  • the optimal parking path trajectory is selected from the multiple parking path trajectories as the final parking path according to a preset index.
  • a parking path planning system based on the above-mentioned automatic parking path planning obstacle avoidance method, includes a vehicle positioning module, a parking space detection module, an obstacle detection module, a calculation module, and an interference detection module; the vehicle positioning module is used The vehicle position and pose are obtained; the parking space detection module is used to obtain the parking position; the obstacle detection module includes a vehicle-mounted camera and a vehicle-mounted radar for obtaining obstacle information; the calculation module is used to combine the vehicle position, Pose, obstacle information, and parking position are used to plan the parking path, obtain the parking path trajectory, and determine the position and posture of each node on the parking path trajectory; the interference detection module is used to fit the contour of the vehicle to each node, and Interference detection is performed on the vehicle contours and obstacles at each node to obtain a qualified parking path trajectory.
  • the present invention provides an obstacle avoidance method for automatic parking path planning.
  • the method performs obstacle collision detection on the parking path trajectory planned with the midpoint of the rear axle of the vehicle as the mass point, that is, by detecting each of the parking path trajectories
  • the node fits the vehicle contour, and considers the degree of coincidence between the vehicle contour and the obstacle projected on the X-axis and Y-axis of the coordinate system, so as to judge whether the parking path trajectory is qualified according to the projection coincidence degree, so as to avoid the risk of collision during parking.
  • This degree increases the parking success rate without increasing computing resources, which has important practical value.
  • FIG. 1 is a schematic diagram of a collision between a parking path trajectory planned by a mass point and an obstacle in the background art.
  • Fig. 2 is a schematic diagram of the positional relationship between the contour of the vehicle on the node and the obstacle in the coordinate system in the first embodiment.
  • FIG. 3 is a schematic diagram of awakening and expanding the vehicle and obstacles in Embodiment 1.
  • FIG. 3 is a schematic diagram of awakening and expanding the vehicle and obstacles in Embodiment 1.
  • Embodiment 4 is a schematic flowchart of the obstacle avoidance method for automatic parking path planning in Embodiment 1.
  • FIG. 5 is a schematic diagram of the internal connection relationship of the parking route planning system in Embodiment 1.
  • FIG. 5 is a schematic diagram of the internal connection relationship of the parking route planning system in Embodiment 1.
  • 1-Parking path planning system 2-interference detection module, 3-calculation module, 4-obstacle detection module, 5-parking detection module, 6-vehicle positioning module.
  • the embodiment of the present invention provides an obstacle avoidance method for automatic parking path planning, as shown in FIG. 4, including the following steps:
  • the parking path trajectory planning process takes the center point of the rear axle of the vehicle as the center of mass point (refer to the Ackerman steering principle), and the curve or straight line drawn on the path passed by the center of mass point is used as the parking path trajectory.
  • the obstacle information includes the position of the obstacle and the size of the obstacle, and the obstacle information is mainly obtained through a vehicle-mounted camera and a vehicle-mounted radar.
  • the reason why the vehicle-mounted camera and vehicle-mounted radar cooperate with each other to obtain obstacle information is because a single vehicle-mounted camera or a single vehicle-mounted radar cannot guarantee that the pose information of the obstacle is fully recognized, which will cause errors in the obstacle information collection. Affect the planning and judgment of the parking path trajectory. Of course, in order to further improve the accuracy of obstacle information collection, it is also possible to continue to integrate the parking space camera, or even the Beidou high-precision positioning system, which is not the only restriction here.
  • Selecting a number of nodes on the parking path trajectory according to the preset step length is essentially discretizing the parking path trajectory to obtain multiple discrete points, and using the discrete points as nodes for subsequent calculations.
  • the preset step length here is to ensure that several nodes are set at equal distances.
  • the preset step length can be set according to actual needs. Shorter to obtain a larger number of nodes, increase the number of obstacle interference detections, and improve the verification accuracy of the parking path trajectory; if there are fewer obstacles around the parking path trajectory, you can set the preset step length Set a longer length to reduce the number of nodes, thereby appropriately reducing the number of obstacle interference detections, and improving the verification efficiency of the parking path trajectory.
  • the contour parameters of the vehicle In order to fit the contour of the vehicle to each node, it is necessary to obtain the contour parameters of the vehicle first, and simplify the car body into a rectangle according to the contour parameters. Then according to the pose corresponding to each node, the rectangle is fitted to each node in turn as the contour of the vehicle.
  • the expansion parameters are preset to increase the contour parameters of the vehicle, thereby simplifying the vehicle body into a rectangle.
  • the car body can also be directly set as a rectangle, and the size of the rectangle can be assigned according to the length and width of the car.
  • the parking path trajectory and the obstacle are usually set within the same phase limit. Then obtain the first projection of the vehicle contour on the node on the X axis and the second projection on the Y axis, and at the same time obtain the third projection of the obstacle on the X axis and the fourth projection on the Y axis.
  • the overlap area mentioned here includes three cases: full overlap, partial overlap, and point overlap. That is, whether the first projection and the third projection completely overlap, or partially overlap, even if only one point is overlapped, they are all considered as overlaps.
  • the first projection and the third projection have overlapping areas; in the same way, whether the second projection and the fourth projection are completely overlapped, partially overlapped, or only one point overlaps, it is also regarded as overlapped between the second and fourth projections. area.
  • the first projection and the third projection have an overlapping area
  • the second projection and the fourth projection have an overlapping area
  • the performance is high. At this time, the trajectory of the parking path is unqualified and needs to be re-planned.
  • the parking path trajectory can be considered unqualified.
  • the parking path trajectory can be judged as unqualified.
  • the vehicle contour is projected on the X axis and the Y axis respectively, and the maximum projection X 1 and minimum projection X 2 of the vehicle contour on the X axis are obtained, and the line segment X 1 X 2 is regarded as In the first projection, the maximum projection Y 1 and the minimum projection Y 2 of the vehicle contour on the Y axis are obtained, and the line segment Y 1 Y 2 is regarded as the second projection.
  • the obstacle respectively projected to the X-axis and Y-axis a maximum value acquiring obstacle in the projection of the axis X 1 X 'and X 2 projected minimum', the segment X 1 'X 2' considered the third projection, the projection obstacle acquired maximum Y-axis Y 1 'and Y 2 minimum projection', a line segment Y 1 'Y 2' considered fourth projection, particularly as shown in FIG.
  • the blue square represents the contour of the vehicle fitted on the node
  • the red irregular curve represents the location and size of the obstacle.
  • the vehicle when planning the parking path, the vehicle is generally simplified into a mass point, and multiple parking path trajectories are planned. At this time, for the situation where there are multiple automatic parking path trajectories, it is necessary to follow the preset Index, select the optimal parking path trajectory from multiple parking path trajectories as the final parking path.
  • the projection distance between the vehicle contour on each node of the parking path and the obstacle on the X axis is larger, and the projection distance between the vehicle contour on each node on the parking path trajectory and the obstacle on the Y axis is also larger , It is considered that the parking path trajectory is safer.
  • the minimum distance between the first projection of the vehicle contour on each parking path trajectory and the third projection of the obstacle is calculated as the first A distance value; then, calculate the minimum distance between the second projection of the vehicle contour on each parking path trajectory and the fourth projection of the obstacle, as the second distance value; finally, the first distance value and the first distance
  • the two distance values are added together, and the parking path trajectory with the largest sum value is the optimal parking path trajectory.
  • the obstacle avoidance method for automatic parking path planning essentially awakens and expands the car and the obstacle in the coordinate system established at the beginning, as shown in FIG. 3.
  • the larger black frame is regarded as a car
  • the smaller black frame is regarded as an obstacle, so that the posture information of the car and the posture information of the obstacle can be ignored, which greatly simplifies the calculation difficulty. Therefore, only when the X-axis and Y-axis have intersections at the same time, the two black boxes will overlap, that is, the expanded car and the obstacle will collide.
  • the judgment method is simple in calculation and high in safety, and is suitable for popularization and application in a wide range.
  • the parking path planning system 1 includes a vehicle positioning module 6, a parking space detection module 5, an obstacle detection module 4, a calculation module 3 and an interference detection module 2.
  • the vehicle positioning module 6 is used to obtain the position and pose of the vehicle.
  • the parking space detection module 5 is used to obtain the parking position and posture.
  • the obstacle detection module 4 includes a vehicle-mounted camera and a vehicle-mounted radar for obtaining obstacle information.
  • the calculation module 3 is used to plan a parking path in combination with vehicle position, pose, obstacle information, and parking position, obtain the parking path trajectory, and determine the pose of each node on the parking path trajectory.
  • the interference detection module 2 is used to fit the contour of the vehicle to each node, and perform interference detection on the contour of the vehicle and the obstacle at each node to obtain a qualified parking path trajectory.

Landscapes

  • Engineering & Computer Science (AREA)
  • Automation & Control Theory (AREA)
  • Transportation (AREA)
  • Mechanical Engineering (AREA)
  • Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Control Of Driving Devices And Active Controlling Of Vehicle (AREA)
  • Steering Control In Accordance With Driving Conditions (AREA)
  • Traffic Control Systems (AREA)

Abstract

一种自动泊车路径规划的避障方法及泊车路径规划系统，包括：根据车辆初始位置、停车位以及障碍物信息规划泊车路径轨迹；在泊车路径轨迹上根据预设步长选取若干节点，确定各节点对应的位姿；根据各节点对应的位姿，将车辆轮廓依次拟合至各节点上；建立坐标系，判断是否存在与障碍物在X轴方向以及Y轴方向均发生投影重叠的车辆轮廓，若是，则认为泊车路径轨迹不合格，重新规划泊车路径；否则，认为泊车路径轨迹合格。该方法对以车辆后轴中点为质点规划的泊车路径轨迹进行了障碍物碰撞检测，避免了泊车过程中发生碰撞风险，一定程度增加了泊车成功率，同时也不会增加计算资源。

Description

一种自动泊车路径规划的避障方法及泊车路径规划系统 技术领域
本发明涉及泊车路径规划技术领域，尤其涉及一种自动泊车路径规划的避障方法及泊车路径规划系统。
背景技术
随着汽车保有量的增加，车位空间狭小，使得停车难逐渐成为了普遍现象，导致泊车过程中的安全事故频频发生。而自动泊车系统的出现，可以有效避免泊车过程中此类安全事故的发生。因此，自动泊车系统近年来成为了国内外各大企业和机构的研究热点。目前，自动泊车系统已经作为高端配置逐渐应用在了部分车辆上，以协助驾驶员完成泊车过程，其主要包括识别系统、路径规划系统以及泊车控制系统三个部分。在整个泊车过程中，一般先由识别系统对空间中的车位以及车位周围的障碍物信息进行识别，然后将所识别的环境信息传输给路径规划系统，由路径规划系统根据传感器采集到的信息建立对应的坐标系，通过约束条件判断车位是否符合要求，如果符合要求则将检测到的车位进行存储，并根据车辆的起始位置及姿态进行泊车路径规划，生成最终的泊车路径；若不符合自动泊车的要求，则控制车辆继续向前行驶，寻找新的车位，直至检测到可用的停车位为止。最后由泊车控制系统根据规划的路径将路径信息转化成控制决策，对车辆的转角、速度、档位等进行控制，同时将执行结果反馈给中央处理器，以方便中央处理器进一步分析和制定决策。从整个自动泊车过程可以看出，泊车效果的好坏与路径规划息息相关，路径规划的优势直接导致自动泊车的成功与否，所以自动泊车系统的路径规划是一个极其重要的环节。
在现有的自动泊车系统中，一般选取车辆的后轴中点作为车辆参考点(也就是质心点)，规划出来的泊车路径轨迹是指质心点经过的位置集合，即点集。整个路径规划过程中，是不考虑实际车辆体积大小的，这就导致有些质心点虽然与周围车辆或障碍物的质心点轨迹之间没有交集，但当考虑了车辆与周围障碍物在地面投影面积之后，就会造成面积存在重合的情况，也就是存在发生碰撞的风险。换句话说，以质心点经过的位置集合规划出的泊车路径，当考虑了车辆体积的大小或周围障碍物的体积大小之后，很有可能在泊车过程中，车辆边缘便会与障碍物边缘存在摩擦或碰撞的风险，如图1所示。然而，如果在路径规划时直接将车辆和障碍物尺寸考虑到规划算法中去，又会导致解算难度大大增加，解算速度会明显下降，从而出现明显的延迟现象，造成较大的误差，这样则难以满足自动泊车路径规划的要求。
发明内容
为了解决上述技术问题，本发明提供了一种自动泊车路径规划的避障方法，该方法包括以下步骤：
根据车辆初始位置、停车位以及障碍物信息规划泊车路径轨迹；
在泊车路径轨迹上根据预设步长选取若干节点，确定各节点对应的位姿；
根据各节点对应的位姿，将车辆轮廓依次拟合至各节点上；
建立坐标系，判断是否存在与障碍物在X轴方向以及Y轴方向均发生投影重叠的车辆轮廓，若是，则认为泊车路径轨迹不合格，重新规划泊车路径；否则，认为泊车路径轨迹合格。
进一步的，所述建立坐标系，判断是否存在与障碍物在X轴方向以及Y轴方向均发生投影重叠的车辆轮廓，若是，则认为泊车路径轨迹不合格，重新规划泊车路径；否则，认为泊车路径轨迹合格步骤，包括：
以泊车路径轨迹和障碍物所在平面建立坐标系；
获取节点上的车辆轮廓在X轴上的第一投影以及在Y轴上的第二投影；
获取障碍物在X轴上的第三投影以及在Y轴上的第四投影；
判断第一投影与第三投影、第二投影与第四投影是否均存在重叠区域，若是，则认为泊车路径轨迹不合格，重新规划泊车路径；否则，认为泊车路径轨迹合格。
进一步的，所述获取节点上的车辆轮廓在X轴上的第一投影以及在Y轴上的第二投影步骤，包括：
获取节点上的车辆轮廓在X轴上的投影最大值X 1和投影最小值X 2，将线段X 1X 2视为第一投影；
获取节点上的车辆轮廓在Y轴上的投影最大值Y 1和投影最小值Y 2，将线段Y 1Y 2视为第二投影。
进一步的，所述获取障碍物在X轴上的第三投影以及在Y轴上的第四投影步骤，包括：
获取障碍物在X轴上的投影最大值X 1’和投影最小值X 2’，将线段X 1’X 2’视为第三投影；
获取障碍物在Y轴上的投影最大值Y 1’和投影最小值Y 2’，将线段Y 1’Y 2’视为第四投影。
进一步的，所述根据各节点对应的位姿，将车辆轮廓依次拟合至各节点上步骤，包括：
获取车辆的轮廓参数，并根据轮廓参数将车体简化为长方形；
根据各节点对应的位姿，将长方形作为车辆轮廓依次拟合至各节点上。
进一步的，在所述获取车辆的轮廓参数，并根据轮廓参数将车体简化为长方形步骤中，通过预设膨胀参数，增大车辆的轮廓参数，进而将车体简化为长方形。
进一步的，所述障碍物信息包括障碍物位置以及障碍物大小；所述障碍物信息通过车载摄像头以及车载雷达获取。
进一步的，针对存在多个自动泊车路径轨迹的情况，根据预设指标，从多个泊车路径轨迹中筛选出最优泊车路径轨迹作为最终的停车路径。
一种泊车路径规划系统，基于上述的一种自动泊车路径规划的避障方法，包括车辆定位模块、车位检测模块、障碍物检测模块、计算模块以及干涉检测模块；所述车辆定位模块用于获取车辆位置以及位姿；所述车位检测模块用于获取车位位姿；所述障碍物检测模块包括车载摄像头和车载雷达，用于获取障碍物信息；所述计算模块用于结合车辆位置、位姿、障碍物信息以及车位位姿进行泊车路径规划，获取泊车路径轨迹以及确定泊车路径轨迹上各节点的位姿；所述干涉检测模块用于对各节点拟合车辆轮廓，并对各节点上的车辆轮廓与障碍物进行干涉检测，以获取合格的泊车路径轨迹。
本发明取得的技术效果如下：
本发明提供了一种自动泊车路径规划的避障方法，该方法对以车辆后轴中点为质点规划的泊车路径轨迹进行了障碍物碰撞检测，即通过对泊车路径轨迹上的各节点拟合车辆轮廓，考虑车辆轮廓与障碍物在坐标系X轴和Y轴上投影的重合度，从而根据投影重合度判断泊车路径轨迹是否合格，以避免泊车过程中发生碰撞风险，一定程度增加了泊车成功率，同时也不会增加计算资源，具有重要的实用价值。
附图说明
图1为背景技术中以质点规划的泊车路径轨迹与障碍物发生碰撞的示意图。
图2为实施例1中位于节点上的车辆轮廓与障碍物在坐标系中的位置关系示意图。
图3为实施例1中对车辆和障碍物进行惊醒膨胀的示意图。
图4为实施例1中自动泊车路径规划的避障方法流程示意图。
图5为实施例1中泊车路径规划系统的内部连接关系示意图。
附图标记：
1-泊车路径规划系统，2-干涉检测模块，3-计算模块，4-障碍物检测模块，5-车位检测模块，6-车辆定位模块。
具体实施方式
下面结合附图对本发明的较佳实施例进行详细阐述，以使本发明的优点和特征更易被本领域技术人员理解，从而对本发明的保护范围作出更为清楚的界定。
实施例1：
本发明实施例提供了一种自动泊车路径规划的避障方法，如图4所示，包括如下步骤：
101、根据车辆初始位置、停车位以及障碍物信息规划泊车路径轨迹。
泊车路径轨迹的规划过程是以车辆的后轴中点作为质心点的(参考阿克曼转向原理)，以质心点所经过的路径画出来的曲线或直线即作为泊车路径轨迹。
本实施例中，所说的障碍物信息包括障碍物位置以及障碍物大小，障碍物信息主要通过车载摄像头以及车载雷达来获取。这里之所以采用车载摄像头和车载雷达相互配合来获取障碍物信息，是因为单一的车载摄像头或单一的车载雷达均不能保证完全识别到障碍物的位姿信息，会造成障碍物信息采集出现误差，影响对泊车路径轨迹的规划和判断。当然，为了进一步提高障碍物信息采集的精准度，还可以继续融合停车位摄像头，甚至是北斗高精定位系统，在此不做唯一限制。
102、在泊车路径轨迹上根据预设步长选取若干节点，确定各节点对应的位姿。
在泊车路径轨迹上根据预设步长选取若干节点，其实质是对泊车路径轨迹进行离散化处理，以得到多个离散点，以离散点作为节点进行后续的计算。这里的预设步长是为了保证若干节点等距设置，预设步长可以根据实际需要进行设定，如泊车路径轨迹周围的障碍物数量较多，则可以将预设步长设置的较短一些，以获取更多数量的节点，增加障碍物干涉检测的次数，提高对泊车路径轨迹的验证准确性；若泊车路径轨迹周围的障碍物数量较少，则可以将预设步长设置的较长一些，以减少节点的数量，从而适当减少障碍物干涉检测的次数，提高对泊车路径轨迹的验证效率。
103、根据各节点对应的位姿，将车辆轮廓依次拟合至各节点上。
为了将车辆轮廓拟合至各节点上，需要先获取得到车辆的轮廓参数，并根据轮廓参数将车体简化为长方形。再根据各节点对应的位姿，将长方形作为车辆轮廓依次拟合至各节点上。本实施例中，通过预设膨胀参数，增大车辆轮廓参数的方法，进而将车体简化为了长方形。当然，为了进一步简化计算过程，也可以将车体直接设置成长方形，根据车长和车宽对长方形的大小进行赋值即可。
104、建立坐标系，判断是否存在与障碍物在X轴方向以及Y轴方向均发生投影重叠的车辆轮廓，若是，则认为泊车路径轨迹不合格，重新规划泊车路径；否则，认为泊车路径轨迹合格。
具体的，首先需要以泊车路径轨迹和障碍物所在平面建立坐标系(二维坐标系)，为了简 化分析过程，通常将泊车路径轨迹以及障碍物设置在同一相限内。然后获取节点上的车辆轮廓在X轴上的第一投影以及在Y轴上的第二投影，同时获取障碍物在X轴上的第三投影以及在Y轴上的第四投影。最后，将第一投影与第三投影相比较，将第二投影和第四投影相比较，判断第一投影与第三投影、第二投影与第四投影是否均存在重叠区域，若是，则认为泊车路径轨迹不合格，重新规划泊车路径；否则，认为泊车路径轨迹合格，并按照合格的泊车路径轨迹执行泊车过程。这里所说的重叠区域包括全部重合、部分重合以及点重合这三种情况，即无论是第一投影与第三投影完全重重合，还是部分重合，甚至是只有一个点是重合，都被视为第一投影和第三投影有重叠区域；同理，无论是第二投影和第四投影完全重合，还是部分重合，还是只有一个点重合，也都被视为第二投影和第四投影有重叠区域。当同时满足第一投影与第三投影有重叠区域，第二投影与第四投影有重叠区域，则认为车辆按照该泊车路径轨迹进行泊车时，车辆会与障碍物发生碰撞，泊车危险性是较高的，此时泊车路径轨迹是不合格的，是需要重新规划的。值得注意的是，只要泊车路径轨迹上有一个节点上的车辆轮廓与障碍物在X轴方向以及Y轴方向均发生投影重叠，就可认为泊车路径轨迹是不合格的。换句话说，只要泊车路径轨迹上有一个节点上的车辆轮廓满足条件，就可以判断泊车路径轨迹不合格，这里并不要求泊车路径轨迹上所有节点上的车辆轮廓均满足上述要求。
本实施例中，将车辆轮廓分别投影至X轴上和Y轴上，获取节点上的车辆轮廓在X轴上的投影最大值X 1和投影最小值X 2，将线段X 1X 2视为第一投影，获取节点上的车辆轮廓在Y轴上的投影最大值Y 1和投影最小值Y 2，将线段Y 1Y 2视为第二投影。同理，将障碍物也分别投影至X轴上和Y轴上，获取障碍物在X轴上的投影最大值X 1’和投影最小值X 2’，将线段X 1’X 2’视为第三投影，获取障碍物在Y轴上的投影最大值Y 1’和投影最小值Y 2’，将线段Y 1’Y 2’视为第四投影，具体如图2所示。在图2中，蓝色方块代表拟合在节点上的车辆轮廓，红色不规则曲线代表了障碍物的位置和大小。
作为优选的，在对泊车路径规划时，一般会将车辆简化为质点，规划出多条泊车路径轨迹，此时针对这种存在多个自动泊车路径轨迹的情况，就需要根据预设指标，从多个泊车路径轨迹中筛选出最优泊车路径轨迹作为最终的停车路径。通常泊车路径轨迹上各节点上的车辆轮廓与障碍物在X轴上的投影距离越大，且泊车路径轨迹上各节点上的车辆轮廓与障碍物在Y轴上的投影距离也越大，则认为该泊车路径轨迹就越安全。本实施例中，针对上述筛选出的多条合格泊车路径轨迹，先计算出每条泊车路径轨迹上的车辆轮廓的第一投影与障碍物的第三投影之间的最小距离，作为第一距离值；然后，再计算出每条泊车路径轨迹上的车辆轮廓的第二投影 与障碍物的第四投影之间的最小距离，作为第二距离值；最后将第一距离值与第二距离值相加，得到的和值最大的泊车路径轨迹即为最优泊车路径轨迹。
本实施例提供的自动泊车路径规划的避障方法本质上是把车和障碍物在一开始建立的坐标系下做惊醒膨胀，如图3所示。在图3中，将较大的黑框看成是车，将较小的黑框看成是障碍物，这样就可以不考虑车的姿态信息以及障碍物的姿态信息，大大简化了计算难度。所以只有当X轴和Y轴同时都有交点的时候，两个黑框才会重合，即膨胀后的车和障碍物才会发生碰撞。该判断方法计算简单，且安全性较高，适于大范围内的推广应用。
实施例2：
本实施例公开了一种泊车路径规划系统，用于实现实施例1中所说的一种自动泊车路径规划的避障方法。如图5所示，泊车路径规划系统1包括车辆定位模块6、车位检测模块5、障碍物检测模块4、计算模块3以及干涉检测模块2。所述车辆定位模块6用于获取车辆位置以及位姿。所述车位检测模块5用于获取车位位姿。所述障碍物检测模块4包括车载摄像头和车载雷达，用于获取障碍物信息。所述计算模块3用于结合车辆位置、位姿、障碍物信息以及车位位姿进行泊车路径规划，获取泊车路径轨迹以及确定泊车路径轨迹上各节点的位姿。所述干涉检测模块2用于对各节点拟合车辆轮廓，并对各节点上的车辆轮廓与障碍物进行干涉检测，以获取合格的泊车路径轨迹。
上面结合附图对本发明的实施方式作了详细说明，但是本发明并不限于上述实施方式，在本领域普通技术人员所具备的知识范围内，还可以在不脱离本发明宗旨的前提下作出各种变化。

Claims (9)

  1. 一种自动泊车路径规划的避障方法，其特征在于，该方法包括以下步骤：
    根据车辆初始位置、停车位以及障碍物信息规划泊车路径轨迹；
    在泊车路径轨迹上根据预设步长选取若干节点，确定各节点对应的位姿；
    根据各节点对应的位姿，将车辆轮廓依次拟合至各节点上；
    建立坐标系，判断是否存在与障碍物在X轴方向以及Y轴方向均发生投影重叠的车辆轮廓，若是，则认为泊车路径轨迹不合格，重新规划泊车路径；否则，认为泊车路径轨迹合格。
  2. 如权利要求1所述一种自动泊车路径规划的避障方法，其特征在于，所述建立坐标系，判断是否存在与障碍物在X轴方向以及Y轴方向均发生投影重叠的车辆轮廓，若是，则认为泊车路径轨迹不合格，重新规划泊车路径；否则，认为泊车路径轨迹合格步骤，包括：
    以泊车路径轨迹和障碍物所在平面建立坐标系；
    获取节点上的车辆轮廓在X轴上的第一投影以及在Y轴上的第二投影；
    获取障碍物在X轴上的第三投影以及在Y轴上的第四投影；
    判断第一投影与第三投影、第二投影与第四投影是否均存在重叠区域，若是，则认为泊车路径轨迹不合格，重新规划泊车路径；否则，认为泊车路径轨迹合格。
  3. 如权利要求2所述一种自动泊车路径规划的避障方法，其特征在于，所述获取节点上的车辆轮廓在X轴上的第一投影以及在Y轴上的第二投影步骤，包括：
    获取节点上的车辆轮廓在X轴上的投影最大值X 1和投影最小值X 2，将线段X 1X 2视为第一投影；
    获取节点上的车辆轮廓在Y轴上的投影最大值Y 1和投影最小值Y 2，将线段Y 1Y 2视为第二投影。
  4. 如权利要求2所述一种自动泊车路径规划的避障方法，其特征在于，所述获取障碍物在X轴上的第三投影以及在Y轴上的第四投影步骤，包括：
    获取障碍物在X轴上的投影最大值X 1’和投影最小值X 2’，将线段X 1’X 2’视为第三投影；
    获取障碍物在Y轴上的投影最大值Y 1’和投影最小值Y 2’，将线段Y 1’Y 2’视为第四投影。
  5. 如权利要求1所述一种自动泊车路径规划的避障方法，其特征在于，所述根据各节点对应的位姿，将车辆轮廓依次拟合至各节点上步骤，包括：
    获取车辆的轮廓参数，并根据轮廓参数将车体简化为长方形；
    根据各节点对应的位姿，将长方形作为车辆轮廓依次拟合至各节点上。
  6. 如权利要求5所述一种自动泊车路径规划的避障方法，其特征在于，在所述获取车辆的 轮廓参数，并根据轮廓参数将车体简化为长方形步骤中，通过预设膨胀参数，增大车辆的轮廓参数，进而将车体简化为长方形。
  7. 如权利要求1所述一种自动泊车路径规划的避障方法，其特征在于，所述障碍物信息包括障碍物位置以及障碍物大小；所述障碍物信息通过车载摄像头以及车载雷达获取。
  8. 如权利要求1所述一种自动泊车路径规划的避障方法，其特征在于，针对存在多个自动泊车路径轨迹的情况，根据预设指标，从多个泊车路径轨迹中筛选出最优泊车路径轨迹作为最终的停车路径。
  9. 一种泊车路径规划系统，基于权利要求1-8任一项所述的一种自动泊车路径规划的避障方法，其特征在于，包括车辆定位模块(6)、车位检测模块(5)、障碍物检测模块(4)、计算模块(3)以及干涉检测模块(2)；所述车辆定位模块(6)用于获取车辆位置以及位姿；所述车位检测模块(5)用于获取车位位姿；所述障碍物检测模块(4)包括车载摄像头和车载雷达，用于获取障碍物信息；所述计算模块(3)用于结合车辆位置、位姿、障碍物信息以及车位位姿进行泊车路径规划，获取泊车路径轨迹以及确定泊车路径轨迹上各节点的位姿；所述干涉检测模块(2)用于对各节点拟合车辆轮廓，并对各节点上的车辆轮廓与障碍物进行干涉检测，以获取合格的泊车路径轨迹。
PCT/CN2020/111804 2020-05-09 2020-08-27 一种自动泊车路径规划的避障方法及泊车路径规划系统 WO2021227304A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010386906.7 2020-05-09
CN202010386906.7A CN111674390B (zh) 2020-05-09 2020-05-09 一种自动泊车路径规划的避障方法及泊车路径规划系统

Publications (1)

Publication Number Publication Date
WO2021227304A1 true WO2021227304A1 (zh) 2021-11-18

Family

ID=72451806

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/111804 WO2021227304A1 (zh) 2020-05-09 2020-08-27 一种自动泊车路径规划的避障方法及泊车路径规划系统

Country Status (2)

Country Link
CN (1) CN111674390B (zh)
WO (1) WO2021227304A1 (zh)

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114030463A (zh) * 2021-11-23 2022-02-11 上海汽车集团股份有限公司 一种自动泊车系统的路径规划方法及装置
CN114179785A (zh) * 2021-11-22 2022-03-15 岚图汽车科技有限公司 一种基于面向服务的融合泊车控制系统、电子设备和车辆
CN114379543A (zh) * 2021-12-23 2022-04-22 浙江吉利控股集团有限公司 自动泊车方法、自动泊车控制装置和自动泊车系统
CN114435347A (zh) * 2022-02-24 2022-05-06 阿波罗智联(北京)科技有限公司 泊车轨迹的确定方法、装置、设备以及存储介质
CN114834447A (zh) * 2022-05-30 2022-08-02 远峰科技股份有限公司 基于自动泊车轨迹的碰撞风险预测方法及装置
CN114940164A (zh) * 2022-05-20 2022-08-26 重庆邮电大学 一种面向泊车场景的无人驾驶车辆行驶轨迹优化方法及系统
CN115201778A (zh) * 2022-09-09 2022-10-18 广州小鹏自动驾驶科技有限公司 不规则障碍物检测方法、车辆及计算机可读存储介质
CN115709484A (zh) * 2023-01-09 2023-02-24 常州检验检测标准认证研究院 一种移动机器人安全仿真检测方法及系统
CN116533989A (zh) * 2023-05-31 2023-08-04 南通大学 一种智能泊车辅助系统及其控制方法
CN116612458A (zh) * 2023-05-30 2023-08-18 易飒(广州)智能科技有限公司 基于深度学习的泊车路径确定方法与系统
EP4335710A4 (en) * 2022-06-30 2024-03-13 Momemta Suzhou Tech Co Ltd METHOD AND DEVICE FOR DETERMINING A LIMITED TRAVEL ROUTE, VEHICLE, STORAGE MEDIUM AND TERMINAL

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112558617B (zh) * 2021-02-19 2022-01-18 深圳佑驾创新科技有限公司 泊车控制方法、装置、计算机设备和存储介质
CN113119957B (zh) * 2021-05-26 2022-10-25 苏州挚途科技有限公司 泊车轨迹规划方法、装置及电子设备
CN113460040B (zh) * 2021-07-23 2023-02-03 广州小鹏自动驾驶科技有限公司 泊车路径确定方法、装置、车辆及存储介质
CN114973762A (zh) * 2022-06-30 2022-08-30 中汽创智科技有限公司 一种停车位管理方法、装置及系统
CN116050083A (zh) * 2022-12-16 2023-05-02 北京斯年智驾科技有限公司 自动驾驶死锁的仿真测试方法、装置、设备及介质

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101626921A (zh) * 2007-03-08 2010-01-13 丰田自动车株式会社 停车辅助装置
CN107672588A (zh) * 2017-08-29 2018-02-09 广州小鹏汽车科技有限公司 一种自动泊车路径障碍物碰撞检测方法、装置及系统
CN107672585A (zh) * 2017-08-29 2018-02-09 广州小鹏汽车科技有限公司 一种自动泊车路径规划方法及系统
US10303178B1 (en) * 2017-12-15 2019-05-28 Waymo Llc Collision mitigation static occupancy grid
US20200070823A1 (en) * 2018-08-30 2020-03-05 Baidu Online Network Technology (Beijing) Co., Ltd. Collision Detection Method and Apparatus Based on an Autonomous Vehicle, Device and Storage Medium
CN111089594A (zh) * 2019-12-30 2020-05-01 浙江大学 一种适用于多场景的自主泊车轨迹规划方法

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109895763A (zh) * 2018-05-17 2019-06-18 华为技术有限公司 基于超声波雷达的泊车车位检测方法和终端
CN110435638B (zh) * 2019-06-28 2021-01-01 惠州市德赛西威汽车电子股份有限公司 一种泊车位自动跟踪方法
CN111016886B (zh) * 2019-12-19 2021-07-30 合达信科技集团有限公司 一种基于b样条理论的自动泊车路径规划方法

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101626921A (zh) * 2007-03-08 2010-01-13 丰田自动车株式会社 停车辅助装置
CN107672588A (zh) * 2017-08-29 2018-02-09 广州小鹏汽车科技有限公司 一种自动泊车路径障碍物碰撞检测方法、装置及系统
CN107672585A (zh) * 2017-08-29 2018-02-09 广州小鹏汽车科技有限公司 一种自动泊车路径规划方法及系统
US10303178B1 (en) * 2017-12-15 2019-05-28 Waymo Llc Collision mitigation static occupancy grid
US20200070823A1 (en) * 2018-08-30 2020-03-05 Baidu Online Network Technology (Beijing) Co., Ltd. Collision Detection Method and Apparatus Based on an Autonomous Vehicle, Device and Storage Medium
CN111089594A (zh) * 2019-12-30 2020-05-01 浙江大学 一种适用于多场景的自主泊车轨迹规划方法

Cited By (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114179785B (zh) * 2021-11-22 2023-10-13 岚图汽车科技有限公司 一种基于面向服务的融合泊车控制系统、电子设备和车辆
CN114179785A (zh) * 2021-11-22 2022-03-15 岚图汽车科技有限公司 一种基于面向服务的融合泊车控制系统、电子设备和车辆
CN114030463A (zh) * 2021-11-23 2022-02-11 上海汽车集团股份有限公司 一种自动泊车系统的路径规划方法及装置
CN114030463B (zh) * 2021-11-23 2024-05-14 上海汽车集团股份有限公司 一种自动泊车系统的路径规划方法及装置
CN114379543A (zh) * 2021-12-23 2022-04-22 浙江吉利控股集团有限公司 自动泊车方法、自动泊车控制装置和自动泊车系统
CN114435347A (zh) * 2022-02-24 2022-05-06 阿波罗智联(北京)科技有限公司 泊车轨迹的确定方法、装置、设备以及存储介质
CN114435347B (zh) * 2022-02-24 2024-03-19 阿波罗智联(北京)科技有限公司 泊车轨迹的确定方法、装置、设备以及存储介质
CN114940164A (zh) * 2022-05-20 2022-08-26 重庆邮电大学 一种面向泊车场景的无人驾驶车辆行驶轨迹优化方法及系统
CN114834447A (zh) * 2022-05-30 2022-08-02 远峰科技股份有限公司 基于自动泊车轨迹的碰撞风险预测方法及装置
CN114834447B (zh) * 2022-05-30 2023-01-20 远峰科技股份有限公司 基于自动泊车轨迹的碰撞风险预测方法及装置
EP4335710A4 (en) * 2022-06-30 2024-03-13 Momemta Suzhou Tech Co Ltd METHOD AND DEVICE FOR DETERMINING A LIMITED TRAVEL ROUTE, VEHICLE, STORAGE MEDIUM AND TERMINAL
CN115201778A (zh) * 2022-09-09 2022-10-18 广州小鹏自动驾驶科技有限公司 不规则障碍物检测方法、车辆及计算机可读存储介质
CN115709484A (zh) * 2023-01-09 2023-02-24 常州检验检测标准认证研究院 一种移动机器人安全仿真检测方法及系统
CN116612458A (zh) * 2023-05-30 2023-08-18 易飒(广州)智能科技有限公司 基于深度学习的泊车路径确定方法与系统
CN116533989A (zh) * 2023-05-31 2023-08-04 南通大学 一种智能泊车辅助系统及其控制方法

Also Published As

Publication number Publication date
CN111674390B (zh) 2021-07-02
CN111674390A (zh) 2020-09-18

Similar Documents

Publication Publication Date Title
WO2021227304A1 (zh) 一种自动泊车路径规划的避障方法及泊车路径规划系统
CN110239535B (zh) 一种基于多传感器融合的弯道主动避撞控制方法
CN109987092B (zh) 一种车辆避障换道时机的确定方法及避障换道的控制方法
US20230135798A1 (en) Vehicle control device and vehicle control method
CN106926844B (zh) 一种基于实时环境信息的动态自动驾驶换道轨迹规划方法
CN109976329B (zh) 一种车辆避障换道路径的规划方法
US20190184982A1 (en) Method for automated parking of a vehicle
CN113916246B (zh) 一种无人驾驶避障路径规划方法和系统
JP7088135B2 (ja) 信号表示推定システム
US20200094829A1 (en) Driving support control device
US20200238980A1 (en) Vehicle control device
CN113479217A (zh) 一种基于自动驾驶的变道避障方法与系统
CN112277939B (zh) 一种对于避让前方压线车辆的偏移控制系统及方法
CN106184232A (zh) 一种基于驾驶员视角的车道偏离预警控制方法
CN113419534B (zh) 一种基于贝塞尔曲线的转向路段路径规划方法
US11934204B2 (en) Autonomous driving apparatus and method
CN114475664A (zh) 一种拥堵路段自动驾驶车辆变道协调控制方法
CN111397623A (zh) 一种基于最佳泊车起始点的路径规划方法
CN113335270B (zh) 一种泊车路径规划方法和装置
WO2021005392A1 (ja) 運転制御方法及び運転制御装置
CN114132305A (zh) 用于垂直与斜列泊车路径规划方法、系统、存储介质及车辆
JP6609292B2 (ja) 車外環境認識装置
CN113581181A (zh) 智能车辆超车轨迹的规划方法
CN116476840B (zh) 变道行驶方法、装置、设备及存储介质
CN115848359B (zh) 车位自适应泊入路径规划方法、车辆及存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20935779

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20935779

Country of ref document: EP

Kind code of ref document: A1