WO2021088528A1 - 一种无人车室外驾驶系统 - Google Patents

一种无人车室外驾驶系统 Download PDF

Info

Publication number
WO2021088528A1
WO2021088528A1 PCT/CN2020/116098 CN2020116098W WO2021088528A1 WO 2021088528 A1 WO2021088528 A1 WO 2021088528A1 CN 2020116098 W CN2020116098 W CN 2020116098W WO 2021088528 A1 WO2021088528 A1 WO 2021088528A1
Authority
WO
WIPO (PCT)
Prior art keywords
information
unmanned vehicle
vehicle
path
module
Prior art date
Application number
PCT/CN2020/116098
Other languages
English (en)
French (fr)
Inventor
鲁仁全
吴元清
刘奕杉
魏浩源
Original Assignee
广东工业大学
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 广东工业大学 filed Critical 广东工业大学
Priority to JP2020552241A priority Critical patent/JP2022502722A/ja
Publication of WO2021088528A1 publication Critical patent/WO2021088528A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/02—Control of position or course in two dimensions
    • G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
    • G05D1/0212—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
    • G05D1/0214—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory in accordance with safety or protection criteria, e.g. avoiding hazardous areas
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/02—Control of position or course in two dimensions
    • G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
    • G05D1/0212—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
    • G05D1/0223—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory involving speed control of the vehicle

Definitions

  • This application relates to the technical field of unmanned vehicle driving, in particular to an unmanned vehicle outdoor driving system.
  • Unmanned driving theory and technology have made considerable progress.
  • the number of cars has gradually increased, followed by congestion of urban routes and frequent traffic accidents.
  • the complexity of the industrial environment and tasks, as well as the diversification of war situations have prompted countries and enterprises to develop unmanned driving technology.
  • Unmanned driving can realize real-time and comprehensive perception of road conditions through on-board sensors, realize intelligent management and control of route traffic, avoid traffic accidents, and reduce travel pressure on urban routes; realize unmanned operations, intelligent production, and intelligent weapons.
  • unmanned driving technology has developed to a certain extent, it is still difficult to meet actual needs.
  • the purpose of this application is to propose a low- and medium-speed outdoor unmanned driving solution to solve the problems of low positioning accuracy of existing unmanned driving solutions and low recognition rate of dynamic obstacles.
  • An unmanned vehicle outdoor driving system includes a vehicle path generation module, a navigation information collection module, a vehicle path tracking module, a driving assistance module and a GUI visual interface, wherein:
  • the vehicle path generation module is used to store multiple path information from the start point to the end point.
  • the path information is collected in advance by the collection vehicle.
  • the location information collection process is: the collection vehicle plans a driving route from the start point to the end point, and then keeps it in the collection vehicle During the process of driving at a constant speed from the starting point to the ending point, GPS data is continuously collected to form a GPS trajectory sequence from the starting point to the ending point; the GPS trajectory sequence and path attributes are used to generate a csv format map as the path information from the starting point to the ending point;
  • the navigation information acquisition module includes a base station system, a vehicle integrated navigator, a vehicle GPS antenna, and a vehicle base station receiving station located on an unmanned vehicle.
  • the base station system includes a base station GPS antenna, a base station transmitting station, and a base station positioning receiver; a vehicle GPS antenna receiving
  • the base station system receives the second GPS signal through the base station GPS antenna, and then transmits the second GPS signal to the base station positioning receiver.
  • the calculated information is transmitted to the base station transmitting station.
  • the transmitting station uses the solution information to issue the RTK message through the radio frequency antenna.
  • the vehicle base station receives the RTK message, it inputs the RTK message and the first GPS signal into the vehicle integrated navigator at the same time. Accurate GPS positioning information is used as the current GPS positioning information of the unmanned vehicle;
  • the vehicle path tracking module is used to track the selected path information by the unmanned vehicle, including: traversing the path information stored in the vehicle path generation module according to the starting point and end point of the path to be driven by the unmanned vehicle, and selecting the corresponding path information;
  • the navigation information acquisition module obtains the current GPS positioning information of the unmanned vehicle. For example, when the GPS positioning information of the current position matches the GPS positioning information of the starting point in the selected route information, the path tracking starts, and the GPS positioning information of the unmanned vehicle is continuously obtained And compare with the track points in the GPS track sequence in the selected path information, and use the lateral control method to control the steering wheel angle of the unmanned vehicle to realize the unmanned vehicle's tracking of the selected path information;
  • the driving assistance module includes an image acquisition module, a target detection module, and a display decision-making module.
  • the image acquisition module collects road information in front of the unmanned vehicle through a vehicle-mounted camera to obtain driving road conditions images; after the target detection module acquires the driving road conditions images, first It is preprocessed, and the preprocessed driving image is recognized by the target detection algorithm.
  • the decision-making module combines the location information and distance information of the obstacle, the safe driving area of the unmanned vehicle, and the control information of the lateral control method to determine whether the obstacle will cause a safety hazard to the driving of the unmanned vehicle, and makes a decision based on the judgment result. Deceleration, acceleration, and braking decisions of unmanned vehicles;
  • the GUI visual interface is used to display the path information of the unmanned vehicle, the current GPS positioning information of the unmanned vehicle, and the location information and distance information of obstacles detected by the target detection algorithm.
  • the traversing the path information stored in the vehicle path generation module and selecting the corresponding path information includes:
  • the display decision module After the display decision module makes the unmanned vehicle deceleration, acceleration, and braking decisions, it converts the decision information into a control signal, and sends the decision information to the unmanned vehicle's underlying electronic control system through the unmanned vehicle's communication system.
  • the control system controls the unmanned vehicle's actuators to perform corresponding control actions.
  • the lateral control method adopts the Stanley control algorithm.
  • the target detection algorithm adopts the YOLO V3 algorithm.
  • the preprocessing process includes image filtering and normalization to improve the detection accuracy.
  • This application uses the integrated navigation instrument to output the differential position signal, obtains centimeter-level high-precision positioning, and realizes the precise positioning of the unmanned vehicle; at the same time, the target detection algorithm is adopted to realize the identification of dynamic obstacles, which greatly improves the unmanned vehicle's The adaptability of dynamic road environment to meet the demand for unmanned driving in medium and low speed road conditions.
  • Figure 1 is a schematic diagram of the structure of an unmanned vehicle outdoor driving system
  • Figure 2 is a schematic diagram of the principle of the application
  • Figure 3 is a schematic diagram of the process of unmanned vehicle path tracking
  • Figure 4 is a schematic diagram of the target detection process.
  • This application provides an unmanned vehicle outdoor driving system, as shown in FIG. 1, including a vehicle path generation module, a navigation information collection module, a vehicle path tracking module, a driving assistance module, and a GUI visualization interface; as shown in FIG. 2, this Apply to realize outdoor unmanned driving through trajectory tracking combined with dynamic target detection; that is, first collect the corresponding GPS sequence information for different paths to generate path information corresponding to different paths, and the unmanned vehicle selects the corresponding path information according to the path that needs to be driven. Combining the lateral control method to track the selected path information, and at the same time help visual target detection and other functions to realize the adaptation of the unmanned vehicle to the dynamic scene.
  • the modules of this application will be described in detail below.
  • the vehicle path generation module is used to store multiple path information from the start point to the end point.
  • the path information is collected in advance by the collection vehicle.
  • the location information collection process is: the collection vehicle plans a driving route from the start point to the end point, and then keeps it in the collection vehicle In the process of driving at a constant speed from the starting point to the ending point, GPS data is continuously collected to form a GPS track sequence from the starting point to the ending point; the GPS track sequence and path attributes are used to generate a csv format map as the path information from the starting point to the ending point.
  • the vehicle path generation module serves as a path information storage module and stores pre-established path information.
  • the collection vehicle can use other manned vehicles to collect route information to form a route information database.
  • Each route information stores a GPS trajectory sequence from a start point to an end point and other route attributes.
  • the GPS trajectory sequence is composed of a series of GPS data, and each GPS data corresponds to a waypoint on the path; then for a waypoint on the path, its GPS data and path attributes include: waypoint number, latitude , Longitude, heading angle, distance from the previous point, curvature, speed attribute, GPS status, etc.
  • the GPS data of each waypoint is collected to form a GPS trajectory sequence.
  • a csv format map can be generated and saved to obtain Path information corresponding to different start and end points.
  • the navigation information acquisition module includes a base station system, a vehicle integrated navigator, a vehicle GPS antenna, and a vehicle base station receiving station located on the unmanned vehicle.
  • the base station system includes base station GPS antenna, base station transmitting station, and base station positioning receiver.
  • the current GPS positioning information of the unmanned vehicle is continuously obtained through the navigation information acquisition module; the process of acquiring GPS positioning information each time includes:
  • the vehicle-mounted GPS antenna receives the first GPS signal
  • the base station system receives the second GPS signal through the base station GPS antenna, and then transmits the second GPS signal to the base station positioning receiver.
  • the base station positioning receiver completes the calculation, the calculated information is transmitted to the base station
  • the base station transmitting station uses the solution information to issue RTK messages through the radio frequency antenna.
  • the vehicle base station receives the RTK message, it inputs the RTK message and the first GPS signal into the vehicle integrated navigator at the same time. After comparison, more accurate, centimeter-level GPS positioning information is obtained as the current GPS positioning information of the unmanned vehicle.
  • the vehicle integrated navigator parses the received data frame into longitude, latitude, heading angle, tilt angle, roll angle, GPS status and other information.
  • the GPS status is divided into 4 levels. When the GPS status is 4 or 3, the GPS signal is available; when the GPS status is 1 or 2, it means that the surrounding buildings are blocked and the GPS signal is poor and unavailable.
  • the vehicle path tracking module is used to track the selected path information by the unmanned vehicle, as shown in Figure 3, including: traverse the path information stored in the vehicle path generation module according to the starting point and end point of the path to be driven by the unmanned vehicle, and select Corresponding path information. That is, after the unmanned vehicle is started, the current position of the unmanned vehicle is taken as the starting point A, and the position that the unmanned vehicle is expected to reach is taken as the end point B. First, you need to query the stored path information to find the path information whose starting point and ending point are A and B in the path information. The specific process is as follows:
  • the current position of the unmanned vehicle as A as the starting point of the path that the unmanned vehicle needs to travel, and the end point of the path that the unmanned vehicle needs to travel is B; traverse the stored path information, first find multiple paths with the end point of B in all the path information Information, and then select the path information from which the starting point of the path information is closest to the position A as the selected path information.
  • This selection method is adopted because the current position of the unmanned vehicle may not strictly match the starting position of any path information in the saved path information, so the path information with the closest starting position is selected as the selected path information. It is necessary to control the unmanned vehicle to drive to the starting point in the selected route information, and then track according to the selected route information, including:
  • the lateral control method adopts the Stanley control algorithm; in the tracking process, the driving assistance module can be activated at the same time to assist decision-making for target detection.
  • the driving assistance module includes an image acquisition module, a target detection module, and a display decision module, as shown in Figure 4, where the image acquisition module collects road surface information in front of the unmanned vehicle through a vehicle-mounted camera to obtain driving road conditions images; optionally, the vehicle-mounted camera Set in front of the cab, in order to ensure real-time and complete image information, the camera adopts a 720P/30FPS acquisition mode.
  • the target detection module acquires the driving road condition image, it first preprocesses it, including image filtering, normalization, etc., to improve the detection accuracy and reduce the influence of noise and light factors in the collection and transmission process.
  • the preprocessed driving image is identified by the target detection algorithm.
  • the obstacle is marked by the display decision module.
  • the position of the obstacle can be framed by the target frame, and the obstacle's position can be obtained.
  • the display decision module combines the location information of obstacles, distance information and the safe driving area of the unmanned vehicle, and the control information of the Stanley control algorithm to determine whether the obstacle will cause safety hazards to the driving of the unmanned vehicle , And make decisions on deceleration, acceleration, and braking of unmanned vehicles based on the judgment results.
  • the decision information is converted into control signals and sent to the unmanned vehicle's underlying electronic control system through the unmanned vehicle's communication system.
  • the unmanned vehicle's actuators (such as brake pedals, accelerator pedals, etc.) are controlled by the underlying electronic control system to perform the corresponding Control actions such as deceleration, acceleration, and braking.
  • the GUI visualization interface is used to display the path information selected by the unmanned vehicle, the current GPS positioning information of the unmanned vehicle, the location information and distance information of the obstacles detected by the target detection algorithm, that is, the image information can be displayed through the GUI visualization interface. To intuitively understand the current driving situation of unmanned vehicles.

Landscapes

  • Engineering & Computer Science (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Traffic Control Systems (AREA)
  • Navigation (AREA)
  • Control Of Driving Devices And Active Controlling Of Vehicle (AREA)

Abstract

一种无人车室外驾驶系统，包括车辆路径生成模块、导航信息采集模块、车辆路径跟踪模块、辅助驾驶模块以及GUI可视化界面，该无人车室外驾驶系统通过轨迹跟踪结合动态目标检测实现室外无人驾驶；即首先对不同的路径采集对应的GPS序列信息生成对应于不同路径的路径信息，无人车根据需要行驶的路径，选择对应的路径信息，结合横向控制方法跟踪所选的路径信息，同时辅助视觉目标检测等功能实现无人车对动态场景的适应；该无人车室外驾驶系统提高了无人车定位精度以及对动态路面环境的适应能力，从而满足对中低速路况的无人驾驶需求。

Description

一种无人车室外驾驶系统
本申请要求2019年11月7日提交中国专利局、申请号为201911079884.3、发明名称为“一种无人车室外驾驶系统”的发明专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本申请涉及无人车驾驶技术领域，具体涉及一种无人车室外驾驶系统。
背景技术
随着人工智能理论、传感器技术、计算机技术、车辆技术的蓬勃发展，无人驾驶理论和技术有了相当的进展。当前，伴随人们生活水平的提高，汽车数量逐渐增加，随之而来的是城市路径的拥挤，交通事故的频发。此外工业环境和任务的复杂化、战争态势的多样化都促使各国和企业开发无人驾驶技术。无人驾驶可以通过车载传感器实时、全面感知路面状况，实现路径车流的智能管控，避免交通事故的发生，减轻城市路径的出行压力；实现无人化作业、智能化生产、智能化武器等。目前，无人驾驶技术虽然有一定的发展，但是依然难以满足现实需求。
发明内容
本申请的目的是提出一种中低速室外无人驾驶方案，解决现有无人驾驶方案定位精度低，对动态障碍物识别率不高等问题。
为了实现上述任务，本申请采用以下技术方案：
一种无人车室外驾驶系统，包括车辆路径生成模块、导航信息采集模块、车辆路径跟踪模块、辅助驾驶模块以及GUI可视化界面，其中：
车辆路径生成模块用于存储多个起点到终点的路径信息，所述路径信息由采集车辆提前采集，位置信息的采集过程为：采集车辆规划好自起点至终点的行车路线，然后在采集车辆保持恒定车速从起点行驶至终点的过程中，不断采集GPS数据，从而形成自起点至终点的GPS轨迹序列；利用GPS轨迹序列及路径属性生成csv格式地图，作为从起点到终点的路径信息；
导航信息采集模块包括基站系统，以及位于无人车上的车载组合导航仪、车载GPS天线以及车载基站接收电台，其中基站系统包括基站GPS天线、基站发射电台、基站定位接收机；车载GPS天线接收第一GPS信号，基站系统通过基 站GPS天线接收第二GPS信号，然后将第二GPS信号传输到基站定位接收机上，在基站定位接收机上解算完成后将解算信息传输到基站发射电台，基站发射电台利用解算信息通过射频天线发布RTK报文，车载基站接收电台接收所述RTK报文后，将RTK报文、第一GPS信号同时输入到车载组合导航仪中，通过差分对比后得到更加准确的GPS定位信息作为无人车当前的GPS定位信息；
车辆路径跟踪模块用于实现无人车对所选路径信息的跟踪，包括：根据无人车所要行驶路径的起点、终点，遍历车辆路径生成模块中存储的路径信息，选择对应的路径信息；通过导航信息采集模块获取无人车当前的GPS定位信息，如当前位置的GPS定位信息与所选路径信息中的起点的GPS定位信息相匹配时开始路径跟踪，通过不断获取无人车的GPS定位信息并与所选路径信息中的GPS轨迹序列中的轨迹点进行比对，并采用横向控制方法操控无人车的方向盘打角，以实现无人车对所选路径信息的跟踪；
辅助驾驶模块包括图像采集模块、目标检测模块以及显示决策模块，其中图像采集模块通过车载摄像头采集无人车前方路面信息，获取行驶路况图像；所述目标检测模块获取所述行驶路况图像后，首先对其进行预处理，将预处理后的行驶路况图像利用目标检测算法进行识别，当识别出障碍物时，通过显示决策模块对障碍物进行标注，并获取障碍物的位置信息和距离信息；显示决策模块结合障碍物的位置信息、距离信息和无人车的安全行区域、所述横向控制方法的操控信息，判断障碍物是否会对无人车的行驶造成安全隐患，并根据判断结果做出无人车减速、加速、刹车决策；
GUI可视化界面用于显示无人车所选路径信息、无人车当前的GPS定位信息、目标检测算法检测出的障碍物的位置信息和距离信息。
进一步地，所述遍历车辆路径生成模块中存储的路径信息，选择对应的路径信息，包括：
记无人车当前位置为A，作为无人车所需要行驶路径的起点，无人车需要行驶的路径终点为B；遍历存储的路径信息，首先找到所有路径信息中终点为B的路径信息，再从其中选择路径信息的起点与位置A最接近的路径信息，作为所选路径信息；
控制无人车行驶至所选的路径信息中起点。
进一步地，所述显示决策模块做出无人车减速、加速、刹车决策后，将决策信息转换成控制信号，并通过无人车的通信系统发送给无人车底层电控系统，通过底层电控系统控制无人车的执行机构执行对应的控制动作。
进一步地，所述横向控制方法采用Stanley控制算法。
进一步地，所述目标检测算法采用YOLO V3算法。
进一步地，所述目标检测模块获取所述行驶路况图像后，进行预处理的过程包括图像滤波、归一化，以提高检测准确率。
本申请具有以下技术特点：
本申请采用组合导航仪器输出差分位置信号，得到厘米级高精度定位，实现了对无人车的精准定位；同时采用了目标检测算法实现了对动态障碍物的识别，大大提高了无人车对动态路面环境的适应能力，从而满足对中低速路况的无人驾驶需求。
附图说明
图1为一种无人车室外驾驶系统的结构示意图；
图2为本申请的原理示意图；
图3为无人车路径跟踪的流程示意图；
图4为目标检测过程的示意图。
具体实施方式
本申请提供了一种无人车室外驾驶系统，如图1所示，包括车辆路径生成模块、导航信息采集模块、车辆路径跟踪模块、辅助驾驶模块以及GUI可视化界面；如图2所示，本申请通过轨迹跟踪结合动态目标检测实现室外无人驾驶；即首先对不同的路径采集对应的GPS序列信息生成对应于不同路径的路径信息，无人车根据需要行驶的路径，选择对应的路径信息，结合横向控制方法跟踪所选的路径信息，同时助视觉目标检测等功能实现无人车对动态场景的适应。下面具体对本申请的各个模块进行详细说明。
1.车辆路径生成模块
车辆路径生成模块用于存储多个起点到终点的路径信息，所述路径信息由采集车辆提前采集，位置信息的采集过程为：采集车辆规划好自起点至终点的行车路线，然后在采集车辆保持恒定车速从起点行驶至终点的过程中，不断采集GPS 数据，从而形成自起点至终点的GPS轨迹序列；利用GPS轨迹序列及路径属性生成csv格式地图，作为从起点到终点的路径信息。
车辆路径生成模块作为路径信息的存储模块，存储有预先制定好的路径信息。所述采集车辆可以利用其他有人驾驶车辆，用于采集路径信息，构成路径信息数据库，每一条路径信息中存储从一个起点至一个终点的GPS轨迹序列及其他路径属性。所述的GPS轨迹序列是由一系列的GPS数据组成的，每个GPS数据对应路径上的一个路点；那么对于路径上的一个路点，其GPS数据和路径属性包括：路点编号，纬度，经度，航向角，距离上一点距离，曲率，速度属性，GPS状态等。当采集车辆从起点行驶至终点时，即采集到了每个路点的GPS数据，形成GPS轨迹序列，同时利用路点的路径属性以及GPS轨迹序列，可生成csv格式地图，对其进行保存，得到对应于不同起点、终点的路径信息。
2.导航信息采集模块
导航信息采集模块包括基站系统，以及位于无人车上的车载组合导航仪、车载GPS天线以及车载基站接收电台。其中基站系统包括基站GPS天线、基站发射电台、基站定位接收机。
无人车启动之后，即通过导航信息采集模块不间断地获取无人车当前的GPS定位信息；其中每一次获取GPS定位信息的过程包括：
车载GPS天线接收第一GPS信号，基站系统通过基站GPS天线接收第二GPS信号，然后将第二GPS信号传输到基站定位接收机上，在基站定位接收机上解算完成后将解算信息传输到基站发射电台，基站发射电台利用解算信息通过射频天线发布RTK报文，车载基站接收电台接收所述RTK报文后，将RTK报文、第一GPS信号同时输入到车载组合导航仪中，通过差分对比后得到更加准确的、厘米级的GPS定位信息作为无人车当前的GPS定位信息。
车载组合导航仪将接收到的数据帧解析成经度、纬度、航向角、倾斜角、翻滚角、GPS状态等信息。GPS状态划分为4个级别，当GPS状态为4或者3的时候GPS信号可用；当GPS状态为1或者2的时候，表示周围有建筑物遮挡造成GPS信号不佳，不可用。
3.车辆路径跟踪模块
车辆路径跟踪模块用于实现无人车对所选路径信息的跟踪，如图3所示，包括：根据无人车所要行驶路径的起点、终点，遍历车辆路径生成模块中存储的路 径信息，选择对应的路径信息。即，无人车启动之后，无人车当前的位置作为起点A，无人车期望到达的位置作为终点B。首先需要查询存储的路径信息，找到路径信息中的起点、终点为A和B的路径信息，具体过程如下：
记无人车当前位置为A，作为无人车所需要行驶路径的起点，无人车需要行驶的路径终点为B；遍历存储的路径信息，首先找到所有路径信息中终点为B的多条路径信息，再从其中选择路径信息的起点与位置A最接近的路径信息，作为所选路径信息。采用这样的选择方法是因为无人车的当前位置可能不与已保存的路径信息中任意一条路径信息的起点位置严格匹配，所以选择起点位置最接近的路径信息作为所选路径信息，此时只需要控制无人车行驶至所选的路径信息中起点，即可按照所选路径信息进行跟踪，包括：
通过导航信息采集模块获取无人车当前的GPS定位信息，如当前位置的GPS定位信息与所选路径信息中的起点的GPS定位信息相匹配时(即当前无人车已行驶到所选路径信息中的起点位置)开始路径跟踪，通过不断获取无人车的GPS定位信息并与所选路径信息中的GPS轨迹序列中的轨迹点进行比对，并采用横向控制方法操控无人车的方向盘打角，以实现无人车对所选路径信息的跟踪。可选地，本申请的一个实施例中，所述横向控制方法采用Stanley控制算法；在跟踪过程中，可同时启动辅助驾驶模块进行目标检测的辅助决策。
4.辅助驾驶模块
辅助驾驶模块包括图像采集模块、目标检测模块以及显示决策模块，如图4所示，其中图像采集模块通过车载摄像头采集无人车前方路面信息，获取行驶路况图像；可选地，所述车载摄像头设置在驾驶室前方，为了保证实时性和图像信息的完整，摄像头采用720P/30FPS的采集模式。所述目标检测模块获取所述行驶路况图像后，首先对其进行预处理，包括图像滤波、归一化等，以提高检测准确率，减少采集和传输过程中的噪声和光照因素影响。
将预处理后的行驶路况图像利用目标检测算法进行识别，当识别出障碍物时，通过显示决策模块对障碍物进行标注，具体可以用目标框将障碍物的位置框出，并获取障碍物的位置信息和距离信息；显示决策模块结合障碍物的位置信息、距离信息和无人车的安全行区域、所述Stanley控制算法的操控信息，判断障碍物是否会对无人车的行驶造成安全隐患，并根据判断结果做出无人车减速、加速、刹车决策。例如检测到某障碍物距离无人车20m，位置在无人车正前方偏右30°， 而如果当前道路平直，根据当前车速，无人车行驶的安全区域在无人车正前方左右±5°以内，而通过操控信息判断无人车无需转向，因此该障碍物不会对无人车行驶造成安全隐患，无需减速；而如果可能对无人车行驶造成影响，则判断结果为减速，然后通过Stanley控制算法控制方向盘打角，等避开障碍物之后，再进行加速。
将决策信息转换成控制信号，并通过无人车的通信系统发送给无人车底层电控系统，通过底层电控系统控制无人车的执行机构(例如刹车踏板、油门踏板等)执行对应的减速、加速、刹车等控制动作。
5.GUI可视化界面
GUI可视化界面用于显示无人车所选路径信息、无人车当前的GPS定位信息、目标检测算法检测出的障碍物的位置信息和距离信息，即图像信息均可以通过GUI可视化界面进行显示，以直观地了解无人车当前的行驶情况。
以上所述的实施例只为阐明本系统的所作的举例，并不能理解为对本专利的限制。相关技术人员可以在上述说明基础上做出更符合项目需求的不同形式改造。凡在本发明的精神和原则之内所作的任何修改、等同替换和改进等，均应包含在本发明权利要求的保护范围之内。

Claims (6)

  1. 一种无人车室外驾驶系统，其特征在于，包括车辆路径生成模块、导航信息采集模块、车辆路径跟踪模块、辅助驾驶模块以及GUI可视化界面，其中：
    车辆路径生成模块用于存储多个起点到终点的路径信息，所述路径信息由采集车辆提前采集，位置信息的采集过程为：采集车辆规划好自起点至终点的行车路线，然后在采集车辆保持恒定车速从起点行驶至终点的过程中，不断采集GPS数据，从而形成自起点至终点的GPS轨迹序列；利用GPS轨迹序列及路径属性生成csv格式地图，作为从起点到终点的路径信息；
    导航信息采集模块包括基站系统，以及位于无人车上的车载组合导航仪、车载GPS天线以及车载基站接收电台，其中基站系统包括基站GPS天线、基站发射电台、基站定位接收机；车载GPS天线接收第一GPS信号，基站系统通过基站GPS天线接收第二GPS信号，然后将第二GPS信号传输到基站定位接收机上，在基站定位接收机上解算完成后将解算信息传输到基站发射电台，基站发射电台利用解算信息通过射频天线发布RTK报文，车载基站接收电台接收所述RTK报文后，将RTK报文、第一GPS信号同时输入到车载组合导航仪中，通过差分对比后得到更加准确的GPS定位信息作为无人车当前的GPS定位信息；
    车辆路径跟踪模块用于实现无人车对所选路径信息的跟踪，包括：根据无人车所要行驶路径的起点、终点，遍历车辆路径生成模块中存储的路径信息，选择对应的路径信息；通过导航信息采集模块获取无人车当前的GPS定位信息，如当前位置的GPS定位信息与所选路径信息中的起点的GPS定位信息相匹配时开始路径跟踪，通过不断获取无人车的GPS定位信息并与所选路径信息中的GPS轨迹序列中的轨迹点进行比对，并采用横向控制方法操控无人车的方向盘打角，以实现无人车对所选路径信息的跟踪；
    辅助驾驶模块包括图像采集模块、目标检测模块以及显示决策模块，其中图像采集模块通过车载摄像头采集无人车前方路面信息，获取行驶路况图像；所述目标检测模块获取所述行驶路况图像后，首先对其进行预处理，将预处理后的行驶路况图像利用目标检测算法进行识别，当识别出障碍物时，通过显示决策模块对障碍物进行标注，并获取障碍物的位置信息和距离信息；显示决策模块结合障碍物的位置信息、距离信息和无人车的安全行区域、所述横向控制方法的操控信 息，判断障碍物是否会对无人车的行驶造成安全隐患，并根据判断结果做出无人车减速、加速、刹车决策；
    GUI可视化界面用于显示无人车所选路径信息、无人车当前的GPS定位信息、目标检测算法检测出的障碍物的位置信息和距离信息。
  2. 如权利要求1所述的无人车室外驾驶系统，其特征在于，所述遍历车辆路径生成模块中存储的路径信息，选择对应的路径信息，包括：
    记无人车当前位置为A，作为无人车所需要行驶路径的起点，无人车需要行驶的路径终点为B；遍历存储的路径信息，首先找到所有路径信息中终点为B的路径信息，再从其中选择路径信息的起点与位置A最接近的路径信息，作为所选路径信息；
    控制无人车行驶至所选的路径信息中起点。
  3. 如权利要求1所述的无人车室外驾驶系统，其特征在于，所述显示决策模块做出无人车减速、加速、刹车决策后，将决策信息转换成控制信号，并通过无人车的通信系统发送给无人车底层电控系统，通过底层电控系统控制无人车的执行机构执行对应的控制动作。
  4. 如权利要求1所述的无人车室外驾驶系统，其特征在于，所述横向控制方法采用Stanley控制算法。
  5. 如权利要求1所述的无人车室外驾驶系统，其特征在于，所述目标检测算法采用YOLO V3算法。
  6. 如权利要求1所述的无人车室外驾驶系统，其特征在于，所述目标检测模块获取所述行驶路况图像后，进行预处理的过程包括图像滤波、归一化，以提高检测准确率。
PCT/CN2020/116098 2019-11-07 2020-09-18 一种无人车室外驾驶系统 WO2021088528A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP2020552241A JP2022502722A (ja) 2019-11-07 2020-09-18 自律運転車両の屋外運転システム

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201911079884.3A CN111026106A (zh) 2019-11-07 2019-11-07 一种无人车室外驾驶系统
CN201911079884.3 2019-11-07

Publications (1)

Publication Number Publication Date
WO2021088528A1 true WO2021088528A1 (zh) 2021-05-14

Family

ID=70201016

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/116098 WO2021088528A1 (zh) 2019-11-07 2020-09-18 一种无人车室外驾驶系统

Country Status (3)

Country Link
JP (1) JP2022502722A (zh)
CN (1) CN111026106A (zh)
WO (1) WO2021088528A1 (zh)

Cited By (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113320543A (zh) * 2021-06-29 2021-08-31 东软睿驰汽车技术（沈阳）有限公司 行车方法、装置、车辆及存储介质
CN113534793A (zh) * 2021-06-23 2021-10-22 中煤科工开采研究院有限公司 一种井工煤矿自动驾驶磁寻迹定位方法
CN113778103A (zh) * 2021-09-19 2021-12-10 南京农业大学 一种基于北斗导航和视觉技术自动驾驶系统
CN114112421A (zh) * 2021-09-30 2022-03-01 东风汽车集团股份有限公司 自动驾驶的测试方法、装置、电子设备及存储介质
CN114205398A (zh) * 2021-12-10 2022-03-18 奇瑞汽车股份有限公司 车辆通信方法及系统
CN114234992A (zh) * 2021-12-06 2022-03-25 山东省国土测绘院 一种基于gis的导航方法
CN114401524A (zh) * 2021-11-26 2022-04-26 江苏徐工工程机械研究院有限公司 一种矿山无人化运输的rtk基站监测系统及方法
CN114415666A (zh) * 2021-12-21 2022-04-29 东风悦享科技有限公司 一种交互式多驾驶模式自动驾驶小车控制方法
CN114415689A (zh) * 2022-01-25 2022-04-29 无锡市金沙田科技有限公司 一种跟随车的决策方法及系统
CN115376322A (zh) * 2022-08-25 2022-11-22 安徽仓擎机器人有限公司 基于电子地图的港口作业车辆调度系统及方法
CN115571117A (zh) * 2022-11-21 2023-01-06 安徽蔚来智驾科技有限公司 车辆纵向控制方法、计算机设备、存储介质及车辆
CN115775062A (zh) * 2023-02-10 2023-03-10 深圳市永泰光电有限公司 基于物联网的汽车摄像头与雷达数据采集系统及方法
CN115166634B (zh) * 2022-05-18 2023-04-11 北京锐士装备科技有限公司 一种多手段结合的无人机飞手定位方法及系统
CN117356546A (zh) * 2023-12-01 2024-01-09 南京禄口国际机场空港科技有限公司 一种机场草坪用的喷雾车的控制方法、系统及存储介质

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111026106A (zh) * 2019-11-07 2020-04-17 广东工业大学 一种无人车室外驾驶系统
CN111459172B (zh) * 2020-05-20 2023-10-31 中国北方车辆研究所 围界安防无人巡逻车自主导航系统
CN114379584B (zh) * 2021-12-17 2023-12-22 阿波罗智联(北京)科技有限公司 车辆控制方法、装置、车辆、设备和存储介质
CN114489056A (zh) * 2021-12-31 2022-05-13 深圳华侨城卡乐技术有限公司 一种无轨车的路线控制方法、设备、系统以及储存介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105741595A (zh) * 2016-04-27 2016-07-06 常州加美科技有限公司 一种基于云端数据库的无人驾驶车辆导航行车方法
US20180357901A1 (en) * 2017-05-11 2018-12-13 Here Global B.V. Vehicle boarding area
CN109765887A (zh) * 2018-12-21 2019-05-17 费子偕 一种自动驾驶控制方法
CN109945882A (zh) * 2019-03-27 2019-06-28 上海交通大学 一种无人驾驶车辆路径规划与控制系统及方法
CN111026106A (zh) * 2019-11-07 2020-04-17 广东工业大学 一种无人车室外驾驶系统

Family Cites Families (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN200979501Y (zh) * 2006-04-19 2007-11-21 吉林大学 完备汽车运动状态测量系统
CN102621566A (zh) * 2012-02-20 2012-08-01 刘瑜 室外移动机器人的精确定位系统
WO2018008627A1 (ja) * 2016-07-05 2018-01-11 国立研究開発法人宇宙航空研究開発機構 衛星測位システム及び衛星測位方法
US20180194344A1 (en) * 2016-07-29 2018-07-12 Faraday&Future Inc. System and method for autonomous vehicle navigation
CN106548135A (zh) * 2016-10-17 2017-03-29 北海益生源农贸有限责任公司 一种路面障碍检测方法
CN107894771A (zh) * 2017-12-01 2018-04-10 浙江力邦合信智能制动系统股份有限公司 一种小车全向移动控制系统及方法
CN108596058A (zh) * 2018-04-11 2018-09-28 西安电子科技大学 基于计算机视觉的行车障碍物测距方法
CN108594821A (zh) * 2018-05-08 2018-09-28 陕西同力重工股份有限公司 无人驾驶矿车定位和寻迹方法
CN109101017A (zh) * 2018-07-27 2018-12-28 江苏盛海智能科技有限公司 一种无人车寻迹路线规划方法及终端
CN108711344A (zh) * 2018-08-01 2018-10-26 南京精智新信息科技有限公司 驾培驾考考试评判系统
CN109814550B (zh) * 2018-12-29 2021-06-22 北京中云智车科技有限公司 一种用于封闭园区的无人运输车
CN109895100B (zh) * 2019-03-29 2020-10-16 深兰科技（上海）有限公司 一种导航地图的生成方法、装置及机器人

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105741595A (zh) * 2016-04-27 2016-07-06 常州加美科技有限公司 一种基于云端数据库的无人驾驶车辆导航行车方法
US20180357901A1 (en) * 2017-05-11 2018-12-13 Here Global B.V. Vehicle boarding area
CN109765887A (zh) * 2018-12-21 2019-05-17 费子偕 一种自动驾驶控制方法
CN109945882A (zh) * 2019-03-27 2019-06-28 上海交通大学 一种无人驾驶车辆路径规划与控制系统及方法
CN111026106A (zh) * 2019-11-07 2020-04-17 广东工业大学 一种无人车室外驾驶系统

Cited By (22)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113534793A (zh) * 2021-06-23 2021-10-22 中煤科工开采研究院有限公司 一种井工煤矿自动驾驶磁寻迹定位方法
CN113320543B (zh) * 2021-06-29 2024-03-22 东软睿驰汽车技术（沈阳）有限公司 行车方法、装置、车辆及存储介质
CN113320543A (zh) * 2021-06-29 2021-08-31 东软睿驰汽车技术（沈阳）有限公司 行车方法、装置、车辆及存储介质
CN113778103A (zh) * 2021-09-19 2021-12-10 南京农业大学 一种基于北斗导航和视觉技术自动驾驶系统
CN114112421B (zh) * 2021-09-30 2023-06-20 东风汽车集团股份有限公司 自动驾驶的测试方法、装置、电子设备及存储介质
CN114112421A (zh) * 2021-09-30 2022-03-01 东风汽车集团股份有限公司 自动驾驶的测试方法、装置、电子设备及存储介质
CN114401524A (zh) * 2021-11-26 2022-04-26 江苏徐工工程机械研究院有限公司 一种矿山无人化运输的rtk基站监测系统及方法
CN114234992B (zh) * 2021-12-06 2024-06-04 山东省国土测绘院 一种基于gis的导航方法
CN114234992A (zh) * 2021-12-06 2022-03-25 山东省国土测绘院 一种基于gis的导航方法
CN114205398B (zh) * 2021-12-10 2023-08-22 奇瑞汽车股份有限公司 车辆通信方法及系统
CN114205398A (zh) * 2021-12-10 2022-03-18 奇瑞汽车股份有限公司 车辆通信方法及系统
CN114415666A (zh) * 2021-12-21 2022-04-29 东风悦享科技有限公司 一种交互式多驾驶模式自动驾驶小车控制方法
CN114415689A (zh) * 2022-01-25 2022-04-29 无锡市金沙田科技有限公司 一种跟随车的决策方法及系统
CN114415689B (zh) * 2022-01-25 2024-05-28 无锡市金沙田科技有限公司 一种跟随车的决策方法及系统
CN115166634B (zh) * 2022-05-18 2023-04-11 北京锐士装备科技有限公司 一种多手段结合的无人机飞手定位方法及系统
CN115376322A (zh) * 2022-08-25 2022-11-22 安徽仓擎机器人有限公司 基于电子地图的港口作业车辆调度系统及方法
CN115376322B (zh) * 2022-08-25 2024-02-13 安徽仓擎机器人有限公司 基于电子地图的港口作业车辆调度系统及方法
CN115571117A (zh) * 2022-11-21 2023-01-06 安徽蔚来智驾科技有限公司 车辆纵向控制方法、计算机设备、存储介质及车辆
CN115775062A (zh) * 2023-02-10 2023-03-10 深圳市永泰光电有限公司 基于物联网的汽车摄像头与雷达数据采集系统及方法
CN115775062B (zh) * 2023-02-10 2023-05-02 深圳市永泰光电有限公司 基于物联网的汽车摄像头与雷达数据采集系统及方法
CN117356546A (zh) * 2023-12-01 2024-01-09 南京禄口国际机场空港科技有限公司 一种机场草坪用的喷雾车的控制方法、系统及存储介质
CN117356546B (zh) * 2023-12-01 2024-02-13 南京禄口国际机场空港科技有限公司 一种机场草坪用的喷雾车的控制方法、系统及存储介质

Also Published As

Publication number Publication date
JP2022502722A (ja) 2022-01-11
CN111026106A (zh) 2020-04-17

Similar Documents

Publication Publication Date Title
WO2021088528A1 (zh) 一种无人车室外驾驶系统
US11970160B2 (en) Traffic signal response for autonomous vehicles
US11908328B2 (en) Determining changes in a driving environment based on vehicle behavior
USRE49650E1 (en) System and method for automatically detecting key behaviors by vehicles
EP3703028A1 (en) Operation of a vehicle in the event of an emergency
EP3724606B1 (en) Using prediction models for scene difficulty in vehicle routing
EP4071661A1 (en) Automatic driving method, related device and computer-readable storage medium
US20190079526A1 (en) Orientation Determination in Object Detection and Tracking for Autonomous Vehicles
CN110422177A (zh) 一种车辆控制方法、装置及系统
GB2594900A (en) Data driven rule books
US11774259B2 (en) Mapping off-road entries for autonomous vehicles
US11885639B2 (en) Generating scouting objectives
KR102548079B1 (ko) 내비게이션 정보의 이용 가능성에 기반한 자율 주행 차량의 동작
JP7462837B2 (ja) 低信頼度の物体検出条件における車両動作のための注釈及びマッピング
CN115963785A (zh) 用于运载工具的方法、系统和设备以及存储介质
CN116312007A (zh) 一种通感一体化设备引导车辆自动驾驶系统
US20240208481A1 (en) Map-Based Braking

Legal Events

Date Code Title Description
ENP Entry into the national phase

Ref document number: 2020552241

Country of ref document: JP

Kind code of ref document: A

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

Ref document number: 20883826

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: 20883826

Country of ref document: EP

Kind code of ref document: A1