WO2020103532A1 - 一种多轴电客车自导向方法 - Google Patents

一种多轴电客车自导向方法

Info

Publication number
WO2020103532A1
WO2020103532A1 PCT/CN2019/104833 CN2019104833W WO2020103532A1 WO 2020103532 A1 WO2020103532 A1 WO 2020103532A1 CN 2019104833 W CN2019104833 W CN 2019104833W WO 2020103532 A1 WO2020103532 A1 WO 2020103532A1
Authority
WO
WIPO (PCT)
Prior art keywords
vehicle
control method
pid control
axis electric
self
Prior art date
Application number
PCT/CN2019/104833
Other languages
English (en)
French (fr)
Inventor
王睿
杨颖�
杜求茂
赵青选
陈勇
陈平安
王虎高
Original Assignee
中车株洲电力机车有限公司
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 中车株洲电力机车有限公司 filed Critical 中车株洲电力机车有限公司
Publication of WO2020103532A1 publication Critical patent/WO2020103532A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/50—Context or environment of the image
    • G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
    • G06V20/588—Recognition of the road, e.g. of lane markings; Recognition of the vehicle driving pattern in relation to the road
    • 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
    • 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
    • G05D1/0268—Control of position or course in two dimensions specially adapted to land vehicles using internal positioning means
    • G05D1/027—Control of position or course in two dimensions specially adapted to land vehicles using internal positioning means comprising intertial navigation means, e.g. azimuth detector
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/35—Categorising the entire scene, e.g. birthday party or wedding scene
    • G06V20/38—Outdoor scenes
    • G06V20/39—Urban scenes

Definitions

  • the invention belongs to the field of urban transportation, and particularly relates to a multi-axis electric bus self-directing method.
  • the multi-axis electric bus self-guiding system is divided into two types: contact type and non-contact type.
  • Contact type guidance has curbs and guide rails.
  • Non-contact type guidance has optical and electromagnetic methods.
  • the physical guide systems that use guide rails include the Raul system and the Bombardier rail system.
  • the Rail system in the Tianjin Development Zone Rail Line 1 and Shanghai Zhangjiang Tram are both used.
  • Optical guidance requires the use of image processing technology, using the camera in front of the vehicle to scan the ground guidance marking line, and transmit the collected image data to the on-board computer in real time.
  • the on-board computer transmits these data to dynamic parameters such as vehicle speed, deviation, wheel angle, etc. And analysis and processing, and then send instructions to the steering system to control the vehicle's driving direction, so that the actual running trajectory of the vehicle and the ground guide marking line basically match.
  • the optical method mainly uses the lane line recognition technology, that is, the special track line on the road surface is recognized by the on-board camera, and the virtual track is intelligently controlled , So as to transmit the running information to the train "brain” (central control unit), according to the "brain” instructions, while ensuring that the train achieves normal actions such as traction, braking, steering, etc., it can accurately control the train to travel on the established "virtual track ", To achieve intelligent operation.
  • the train "brain” central control unit
  • the high stability optical method requires the camera to have a high shooting frequency and high resolution. Therefore, it also requires the storage capacity of the controller to be large enough, the arithmetic processing capability of the main controller is strong enough, and the data transmission capability of the ground transmission device is stable enough. At this stage, it is difficult to be commercialized on a large scale and has higher requirements on the controller.
  • An object of the present invention is to provide a self-guided method for a multi-axis electric bus in view of the above-mentioned shortcomings in the prior art.
  • the technical solutions adopted by the present invention are:
  • a multi-axis electric bus self-guided method which is characterized by the following steps:
  • Step A Input the original images captured by the cameras on the front and rear sides of the vehicle to the on-board equipment and preprocess the original images to obtain the preprocessed images;
  • Step B For the pre-processed image, the vehicle driving area is locally enhanced
  • Step C Extract the lane line
  • Step D Determine the position of the vehicle according to the position of the camera and the lane line;
  • Step E According to the vehicle travel constraints, the PID control method, fuzzy PID control method, predictive control method, and nonlinear control method are used to control the vehicle to run within the allowed trajectory.
  • the method further includes using a satellite positioning system and / or an inertial navigation system installed in the intermediate vehicle to determine the position of the vehicle.
  • the pre-processing process in step A includes image compression, gray-scale conversion, filtering processing, and equalization processing in this order.
  • the control system of the through passage of the steering mechanism is designed according to the PID control method: Where K d is the differential factor of the PID control system, K p is the proportional factor of the PID control system, and K i is the integral factor of the PID control system.
  • step E it also includes constructing a generation value function according to driving rules, comfort requirements, line length, noise limitation and other conditions to find the optimal vehicle running trajectory.
  • the present invention can improve the accuracy and redundancy of the vehicle's self-guided system, and comprehensively use the optical and satellite inertial navigation system to realize the vehicle's lane keeping function and multi-wheel synchronous steering function, using electrical signal transmission
  • the electronically controlled steering device is also easy to control the vehicle.
  • Figure 1 is the control principle diagram of the present invention.
  • Figure 2 is the layout of the camera and satellite antenna.
  • Figure 3 is a schematic diagram of satellite system positioning.
  • FIG. 4 is a schematic diagram of vehicle position recognition.
  • FIG. 5 is a schematic diagram of a lane keeping method for finely adjusting the turning angle of a straight line segment.
  • FIG. 6 is a schematic diagram of a vehicle crossing a curve.
  • the present invention is applied to a multi-axle heavy-duty passenger vehicle.
  • Cameras are provided on the front and rear sides of the vehicle to capture the front and rear pictures of the vehicle and pass them to the central processor.
  • Image compression, grayscale conversion, image preprocessing, partial image enhancement, lane line detection and other methods directly extract the lane line, so as to determine the position of the vehicle on the current route according to the lane line.
  • the image captured by the camera is input into the vehicle-mounted device and the image is compressed according to requirements, simplifying the calculation amount.
  • the RGB color model is converted into a binary model for analysis to further simplify the calculation.
  • Filtering methods such as mean filtering or histogram equalization are used to preprocess the image to remove possible optical noise.
  • An operator that sets the threshold of the lane line according to weather conditions records the area that meets the threshold, and extracts the lane line.
  • the lane line After extracting the lane line, lock the lane line, considering that there may be some clutter signals or factors such as obstacles, weather conditions, other vehicles covering the lane line, so the clutter signal is removed by the filter or the lane fitting method will be The lane line is complete.
  • Areas of interest can be selected for road conditions, that is, areas where vehicles are driving for local enhancement.
  • a 5 * 5 matrix can be used for straight lane lines.
  • this method can be used for comparison. The characteristics are shown in the following table:
  • the roads on which vehicles travel have lane lines, they are generally divided into special lane lines and ordinary lane lines.
  • the special lane lines designed for heavy-duty electric buses are relatively easy to identify. Therefore, the general lane lines are mainly analyzed here.
  • the position of the vehicle can be uniquely determined according to the position of the camera and the lane line in the image, thereby completing the lane line recognition.
  • a satellite positioning system and an inertial navigation system are installed in the middle of the vehicle at the same time, and the position of the vehicle is determined by the satellite system and the inertial navigation system to ensure that the vehicle is driving in the area where the lane line can be recognized.
  • a satellite can receive satellite signals and base station signals through an antenna
  • a system composed of several satellites and base stations can be used to determine the time ⁇ t of the GPS signal reaching the receiver based on the instantaneous position of the satellite as a known value, and then determine the distance according to the propagation speed
  • the geometric relationship is constructed according to the triangle rule. The more satellites, the more accurate the position.
  • the description image is shown in Figure 3.
  • the position of the vehicle can be uniquely determined according to the position of the lane line recognized by the camera, so that in the main controller, according to Vehicle constraints and PID control, fuzzy PID control, predictive control, nonlinear control and other methods to ensure that the vehicle runs within the allowed trajectory.
  • this method can infer the position of the vehicle by comparing the lane line information and provide a basis for the main controller to execute the control command.
  • the differential satellite signal system uses a differential satellite reference station with known accurate three-dimensional coordinates to obtain the pseudorange correction amount or position correction amount, and then sends this correction amount to the user in real time or afterwards to correct the user's measurement data to improve Satellite positioning accuracy.
  • This system can completely capture the position information of the vehicle through the multi-satellite system installed in the head car and the middle car. The more satellites, the higher the accuracy. The current can reach the centimeter level, and the worst can reach the sub-meter level.
  • the optical video system must be recognized under the condition of the lane line, and the recognition situation is also unstable. Therefore, the use of a differential satellite signal system can ensure that the vehicle runs within the lane line.
  • the positioning signal and map information given by the satellite directly provide the basis for the steering, while the optical system will only serve as a feedback system to give the system a feedback signal to improve the robustness and stability of the system.
  • An inertial navigation system (INS, hereinafter referred to as inertial navigation) is an autonomous navigation system that does not depend on external information or radiate energy to the outside.
  • the basic working principle of inertial navigation is based on Newton's law of mechanics.
  • you can get the Information such as speed, yaw angle and position, installed on the center axis of the vehicle can make up for the problem of vehicle satellite system errors, so as to ensure the accuracy of vehicle positioning, and consider the interference with the platform when installing the camera in the middle car .
  • base station satellite heavy-duty vehicles can be interconnected and interconnected, the transmission time is fixed, the position can be measured according to the speed, and then connected according to the triangle rule, so as to uniquely determine the specific position, considering the more satellites, The more complex its shape, the polygon method can be used to solve it, thereby improving accuracy.
  • the antenna position of each vehicle can be uniquely determined by the antenna position. If two antennas are used, when the satellite accuracy reaches the centimeter or even millimeter level, the plane attitude angle can also be determined according to the positions of the two antennas (Plane analysis with two points in line), if you use three antennas, you can determine the three-dimensional attitude angle (three-point plane analysis), the more antennas, the higher the recognition accuracy, and then according to the satellite system's own map system Accurately realize the positioning function. Therefore, the vehicle's position, attitude angle, etc. are determined according to the vehicle's position information and related systems.
  • the information of the received vehicle is fed back to the main controller through the interactive device, the terminal and the network cable.
  • the main controller comprehensively processes the information, the vehicle position is uniquely determined.
  • the vehicle trajectory circuit diagram in the vehicle controller, it can be According to the line conditions and the lane lines identified by the integrated camera, under severe weather conditions, the positioning is mainly achieved through the satellite system. Under the tunnel and bridge and other places where the satellite signal is lost, the camera will be used to identify the positioning, and the inertial navigation is in the middle.
  • the vehicle is positioned to provide a basis for controlling the vehicle's steering or fine-tuning the steering actuator to realize the lane keeping function.
  • the information on the front of the vehicle can be captured synchronously, displayed on the image, and the relevant image data is transferred to the processor, which is converted into an image with multiple pixels to obtain image information ,
  • the processor which is converted into an image with multiple pixels to obtain image information .
  • the vehicle is connected through the through channel, and the head and tail vehicles are determined by the camera.
  • the remaining steering mechanisms complete the wheel following function according to PID control, fuzzy PID control, predictive control, nonlinear control and other methods.
  • the satellite system and inertial navigation system installed on the vehicle are integrated to achieve positioning within the sub-meter level of the vehicle, thereby keeping the vehicle within the lane line.
  • the specific installation position is shown in Figure 2.
  • the satellite and inertial navigation system are integrated Identify the position of the vehicle in the satellite plane through the positioning device, and ensure that the vehicle is within the lane line, and the camera captures the lane line information and gives a feedback signal to the position determined by the satellite system, so as to more accurately determine the vehicle's steering and fine-tuning.
  • the optical recognition will be used as the basis to adjust Kp, Ki, Kd in PID in the algorithm, and Strengthen the proportion of non-linear algorithms, while using the inertial navigation system as a supplement, so that the positioning system can maintain the positioning at the submeter level or the centimeter level as much as possible, and when the satellite positioning is lost for too long or the satellite positioning system fails, the optical is directly switched
  • the system and the inertial navigation system replace the function of the satellite system within a certain length through the inertial navigation system.
  • the optical system continues to recognize the lane line to maintain the precise positioning of the vehicle.
  • the weather and other bad weather cover the lane line then the positioning function is directly realized through the satellite system and the inertial navigation system, and assisted by the inertial navigation to obtain the complete vehicle position. Therefore, this method can obtain the vehicle position in all weather and under any conditions To provide a basis for steering by wire.
  • the scenario analysis is as follows:
  • FIG. 5 The schematic diagram of the lane keeping method for fine-tuning the turning angle of the straight line segment is shown in FIG. 5, and the schematic diagram of the vehicle crossing curve method is shown in FIG. 6.
  • the main controller After calculating the running track, the main controller processes the running track as an electrical signal, and feeds back the calculated electrical signal to the actuator to complete the rotation requirements within the constraint line, so that the steering mechanism turns according to the predetermined track and performs according to the lane line.
  • the actuator controls the wheel pairs separately according to the electrical signals, so as to realize the simultaneous steering of multi-axis wheels, and complete the lane keeping and steering functions.
  • the actuator adopts an electronic control system, which directly transmits the electric signal to the electronically controlled steering mechanism.
  • the position of the wheel at the next moment can be known according to the running track of the vehicle.
  • it can still ensure that each wheel pair of the vehicle reaches the specified position according to the trajectory generated by the main controller at the next moment, so that Control multiple steering mechanisms, the steering mechanism adjusts according to the received electrical signals to complete the requirements of steering and lane keeping. Simplify or even eliminate the man-made steering operation, reduce the interference caused by human factors, thereby improving the efficiency of handling vehicle-related matters.
  • the vehicle obtains the vehicle's position through the navigation system and the video system, and provides a basis for the vehicle's driving according to the position. Because of the existence of the satellite system and the inertial navigation system, the vehicle can directly detect the vehicle's displacement and speed through the system. Attitude angle, etc., avoid the design of the wheel mechanism to monitor the wheel steering system. Directly passing the vehicle's body position provides a basis for the steering system's steering. At the same time, compared with the current vehicle first-wheel steering and rear-wheel following scheme, this scheme can provide the vehicle position more quickly and directly control each wheel, improving the redundancy of the system. Since this method can synchronize vehicle position data in real time, errors can be avoided. At the same time, it is also possible to fine-tune the rotation angle of the wheelset to avoid the occurrence of tail flicking.
  • the satellite system is mainly used to provide positioning and complete the path planning function in the main controller, and the camera is mainly used to achieve close-range lane line recognition and environment detection.
  • the integrated use of the positioning device of the satellite and the inertial navigation system can identify the position of the vehicle in the satellite plane and can ensure that the vehicle is within the lane line. To determine the vehicle's steering and fine-tuning.
  • the optical recognition will be mainly used as the basis to adjust the K p , K i , K in the PID of the algorithm d , and strengthen the proportion of the nonlinear algorithm, while using the inertial navigation system as a supplement, so that the positioning system can maintain the positioning accuracy at the centimeter level as much as possible, and when the satellite positioning is lost for too long or the satellite positioning system fails, the optical is directly switched
  • the system and the inertial navigation system replace the function of the satellite system within a certain length through the inertial navigation system.
  • the optical system continues to recognize the lane line to maintain the precise positioning of the vehicle.
  • the positioning function is directly realized through the satellite system and the inertial navigation system, and assisted by the inertial navigation to obtain the complete vehicle position.
  • K p , K i , K d can be adjusted , And adjust the algorithm of the non-linear system according to the climatic conditions, thus ensuring the requirement that both day and night conditions can be used.
  • this method can return to the original lane through the satellite system and continue to complete the operation after encountering the problem of sudden lane changes such as obstacles on the road, which has high redundancy, stability and robustness. Therefore, this method is less affected by the weather and the day and night environment, and can obtain the position of the vehicle in all weather and most extreme conditions, thereby providing a basis for steering by wire.
  • the invention is suitable for multi-axis heavy-duty electric passenger cars, which can improve the accuracy and redundancy of the vehicle's self-guided system, and comprehensively uses optical and satellite inertial navigation systems to realize the vehicle's lane keeping function and multi-wheel synchronous steering function.
  • the signal transmission to the electronically controlled steering device is also easy to control the vehicle; it provides a human-machine interaction interface to facilitate the driver to understand the vehicle information, and can also improve the driver's maneuverability during the assisted driving phase.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Automation & Control Theory (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Steering Control In Accordance With Driving Conditions (AREA)
  • Navigation (AREA)
  • Control Of Driving Devices And Active Controlling Of Vehicle (AREA)

Abstract

本发明公开了一种多轴电客车自导向方法，包括：步骤A.将车辆前后两侧摄像头捕获的原始图像输入到车载设备并对原始图像进行预处理，得到预处理后的图像；步骤B.针对预处理后的图像，对车辆行驶区域进行局部增强；步骤C.提取出车道线；步骤D.根据摄像头和车道线的位置确定车辆位置；步骤E.根据车辆行驶约束条件，运用PID控制方法、模糊PID控制方法、预测控制方法、非线性控制方法控制车辆在允许的轨迹内运行。本发明可以提升车辆自导向***的精确性和冗余性，并综合运用光学和卫星惯导导航***实现车辆的车道保持功能及多车轮同步转向功能，采用电信号传递给电控转向装置也易于控制车辆。

Description

一种多轴电客车自导向方法 技术领域
本发明属于城市交通领域，特别涉及一种多轴电客车自导向方法。
背景技术
因为人口数量增长、人口老龄化的严重以及交通拥挤现象严重，越来越多的人出行选择了载客量大，不需要自己驾驶的大型公共交通。因此，大型公共交通越来越成为人们出行的选择。考虑到BRT载客量小，重载大巴控制较难，有轨电车轨道铺设成本较高，因此解决这三个车辆目前面临的问题将是未来城市交通必然的发展方向。由于BRT载客量小，有轨电车铺设轨道是车辆本身的属性，较难实现，因此解决重载大巴控制较难的问题将成为了下一阶段交通行业最可行的解决方案。
目前，多轴电客车自导向***分为接触式和非接触式两种，接触式的导向有路缘石、导轨等方式，非接触式的导向有光学和电磁两种方式。
采用导轨的物理导向***有劳尔***和庞巴迪导轨***，天津开发区导轨电车1号线和上海张江有轨电车均采用的劳尔***。光学导向需要借助图像处理技术，利用车辆前方摄像头扫描地面导向标识线，将采集到的图像数据实时传输至车载计算机，车载计算机将这些数据与车辆的速度、偏离值、车轮角度等动态参数一并进行分析处理，然后给转向***传达指令，以控制车辆的行驶方向，使车辆实际运行轨迹与地面导向标识线基本相符。
然而，接触式自导向***易出现磨耗，稳定性较差，且需要重新设计道路，成本较高。非接触式的方法中，电磁方法则存在易受电磁干扰影响等问题，而现阶段光学方法主要采用车道线识别技术，即通过车载摄像头识别路面的专用轨道线，并对虚拟轨迹进行智能跟踪控制，从而将运行信息传送至列车“大脑”(中央控制单元)，根据“大脑”的指令，在保证列车实现牵引、制动、转向等正常动作的同时，能够精准控制列车行驶在既定“虚拟轨迹”上，实现智能运行。然而此方法受制于由于天气变化所带来的光线强弱、光线遮挡问题，光化学污染及地面车道线不够干净等因素的影响，而这些因素也直接影响到光学方法识别车道线，同时，精度高而且稳定性高光学方法要求相机拍摄频率较高，分辨率高，因此也要求控制器的存储量足够大，主控器的运算处理能力足够强，地面传输装置的数据传输能力足够稳定，因此在现阶段面临难以大规模商品化以及对控制器要求较高等问题。
此外，单一的光学方法在应对突发的换道以及其他偏离预定车道的情况下将难以再回到原车道完成车道线识别功能，因此，此方法的稳定性及鲁棒性不高。
发明内容
本发明的目的在于，针对上述现有技术的不足，提供一种多轴电客车自导向方法。
为解决上述技术问题，本发明所采用的技术方案是：
一种多轴电客车自导向方法，其特点是包括以下步骤：
步骤A.将车辆前后两侧摄像头捕获的原始图像输入到车载设备并对原始图像进行预处理，得到预处理后的图像；
步骤B.针对预处理后的图像，对车辆行驶区域进行局部增强；
步骤C.提取出车道线；
步骤D.根据摄像头和车道线的位置确定车辆位置；
步骤E.根据车辆行驶约束条件，运用PID控制方法、模糊PID控制方法、预测控制方法、非线性控制方法控制车辆在允许的轨迹内运行。
进一步地，所述步骤D中，还包括利用安装在中间车的卫星定位***和/或惯导***确定车辆位置。
作为一种优选方式，所述步骤A中的预处理过程依次包括：图像压缩、灰度转化、滤波处理、均衡处理。
作为一种优选方式，所述步骤E中，根据PID控制方法设计转向机构贯通道的控制***：
Figure PCTCN2019104833-appb-000001
其中K d为PID控制***的微分因子、K p为PID控制***的比例因子、K i为PID控制***的积分因子。
进一步地，所述步骤E中，还包括根据驾驶规则、舒适性要求、线路长度、噪音限制及其他条件构建代价值函数，找到最佳车辆运行轨迹。
与现有技术相比，本发明可以提升车辆自导向***的精确性和冗余性，并综合运用光学和卫星惯导导航***实现车辆的车道保持功能及多车轮同步转向功能，采用电信号传递给电控转向装置也易于控制车辆。
附图说明
图1为本发明控制原理图。
图2为摄像机及卫星天线布置图。
图3为卫星***定位原理图。
图4为车辆位置识别示意图。
图5为直线段微调转角的车道保持方法示意图。
图6为车辆过弯道示意图。
具体实施方式
如图1和图2所示，本发明应用于多轴重载客车，在车辆前后两侧均设置摄像头，用于捕获车辆前后的图片并传递给中央处理器，在中央处理器通过对图像输入，图像的压缩，灰度转化，图像预处理，局部图像增强，车道线检测等方法直接提取出车道线，从而根据车道线确定车辆在当前路线上的位置。
一、车道线识别与卫星惯导***
1.车道线识别
(1)图像输入与压缩
将摄像头捕获的图像输入到车载设备里并根据需求对图像进行压缩处理，简化计算量。
(2)灰度转化
将RGB色彩模型转化为2值模型进行分析，从而进一步简化计算。
(3)滤波处理及均衡处理
使用均值滤波等滤波方法或直方图均衡等方法对图像进行预处理，从而将可能存在的光学噪音去除。
(4)车道线识别
根据天气条件设置车道线阈值的算子，记录符合阈值的区域，从而提取车道线。
在提取车道线后，锁定车道线，考虑到可能存在一些杂波信号或诸如障碍物，天气条件，其它车辆等因素覆盖车道线，因此通过滤波器将杂波信号去除或者通过车道拟合方法将车道线补充完整。
针对道路条件可以选择感兴趣区域，即车辆行驶的区域进行局部增强，针 对直线车道线可以采用5*5矩阵，考虑到车道线具有特别的形状，因此可以运用此方法进行对比，一种可选择的特征如下表所示：
0 0 0 0 127
0 0 0 127 127
0 0 127 127 127
0 127 127 127 127
127 127 127 127 127
表1车道线特征矩阵
(5)车辆位置识别
考虑到车辆行驶的道路上具有车道线，一般分为特种车道线和普通车道线，相对而言针对重载电客车设计的特种车道线相对识别较为容易，因此这里主要分析一般车道线。
考虑到车辆在车道线两侧中间行驶，可以根据图像内摄像头和车道线的位置唯一确定车辆的位置，从而完成车道线识别。
同时，在车辆中间车同时安装卫星定位***和惯导***，通过卫星***和惯导***确定车辆位置，保证车辆在可以识别车道线的区域行驶。考虑到卫星可以通过天线接收卫星信号及基站的信号，根据数个卫星及基站构成的一个***，从而根据卫星瞬间位置作为已知测定GPS信号到达接收机的时间△t，然后根据传播速度确定距离，最终根据三角形法则构建几何关系，卫星数量越多，确定的位置也更准确。说明图像如图3所示。
车辆位置说明图如图4所示，当车辆在不同位置的时候，车道线的位置也有所不同，因此可以根据摄像头识别的车道线的位置唯一确定车辆的位置，从而在主控器内，根据车辆约束条件及PID控制，模糊PID控制，预测控制，非线性控制等多种方法保证车辆在允许的轨迹内运行。
因此，此方法能够通过比对车道线的信息反向推断出车辆的位置，并为主控器执行控制指令提供依据。
2.卫星***及惯导***
差分卫星信号***是利用已知精确三维坐标的差分卫星基准台，求得伪距修正量或位置修正量，再将这个修正量实时或事后发送给用户，对用户的测量数据进行修正，以提高卫星定位精度。此***通过安装在头车及中间车的多卫星***，可以完整的捕捉车辆的位置信息，卫星数量越多其精度越高，目前最高可达到厘米级别，最差也可以达到亚米级别，考虑到光学视频***必须在有车道线的条件下识别，而且识别情况也不稳定，因此，运用差分卫星信号***可以保证车辆在车道线内运行，同时在光学***不稳定的条件下，也可以直接按照卫星给出的定位信号及地图信息直接为转向提供依据，而光学***将只作为一个反馈***给***一个反馈信号，提高***的鲁棒性和稳定性。
惯性导航***(INS，以下简称惯导)是一种不依赖于外部信息、也不向外部辐射能量的自主式导航***。惯导的基本工作原理是以牛顿力学定律为基础，通过测量载体在惯性参考系的加速度，将它对时间进行积分，且把它变换到导航坐标系中，就能够得到在导航坐标系中的速度、偏航角和位置等信息，安装在车辆中轴线上，可以很好的弥补车辆卫星***出现误差的问题，从而保证车辆定位精度，而且考虑到中间车安装摄像头需要考虑和站台的干涉问题，一般不考虑使 用摄像头，所以在中间车往往考虑使用惯导***做辅助，在卫星***出现故障的情况下，根据惯导***提供车辆的位置信息，从而完成执行机构的转向功能。
如图2和图3所示，基站卫星重载车辆均可以互联互通，传输的时间一定，根据速度可以测算出位置，然后根据三角形法则连线，从而唯一确定具***置，考虑到卫星越多，其形状越复杂，可以利用多边形方法求解，从而提升精度。
考虑到天线位置已知，即可以通过天线位置唯一确定每一节车辆的天线位置，如果使用两根天线，当卫星精度达到厘米甚至毫米级别时，也可以根据两根天线的位置确定平面姿态角(两点成线的面分析)，如果使用三根天线则可以确定立体姿态角(三点成面的立体分析)，天线越多其识别精度越高，再根据卫星***自带的地图***即可以精确的实现定位功能。从而根据车辆位置信息及相关***确定车辆的位置，姿态角等。
天线接受信息后，通过交互器，终端及网线将接收车辆的信息反馈给主控器，主控器综合处理信息后，唯一确定车辆位置，考虑到车辆控制器内有车辆轨迹线路图，因此可以根据线路条件并综合摄像头识别出的车道线进行分析，在天气恶劣的条件下，主要通过卫星***实现定位，在隧道及桥下以及其它丢失卫星信号的地方将采用摄像头识别定位，惯导在中间车实现定位，从而为控制车辆的转向或者微调转向执行机构实现车道保持功能提供依据。
二.确定车辆位置信息
(1)通过安装在车辆上的摄像机可以同步捕捉到车辆前部的信息，显示在图像上，并将相关图像数据传到处理器中，转化为具有多个像素点的图像,从而得到图像信息，结合定位算法识别车道线及车辆位置，考虑到多轴重载电客车需要控制多个转向装置，也有向两个方向运行的需求。因此，需要设置前后两个摄像头共同识别车道线，从而完整的确定车辆和车道线的相对位置，并根据车道线判断车辆位置。
(2)车辆通过贯通道相连接，在头车和尾车通过摄像头确定，其余转向机构根据PID控制，模糊PID控制，预测控制，非线性控制等方法完成车轮跟随作用。而车上安装的卫星***及惯导***则通过综合定位，实现车辆的亚米级别以内的定位，从而将车辆保持在车道线内，具体安装位置如图2所示，卫星和惯导***综合通过定位装置识别出车辆在卫星平面内的位置，并保证车辆在车道线内，而摄像头捕捉车道线信息，给卫星***确定的位置一个反馈信号，从而更加精准的判断车辆的转向及微调。同时，当卫星***数据不够准确的情况下(诸如在隧道内，高楼效应，有障碍物在天空遮挡等)，将主要以光学识别作为基础，调整算法里PID里的Kp,Ki,Kd，并强化非线性算法的比重，同时使用惯导***作为补充，使定位***能够尽可能保持定位在亚米级别或厘米级别，而当卫星定位丢失的时间过长或卫星定位***故障，则直接切换光学***及惯导***，通过惯导***在一定长度内代替卫星***的功能，光学***继续识别车道线保持车辆的精准定位，完成这段运营并能够自导向回厂完成检修，而当阴雨，雷暴天气等恶劣天气覆盖车道线，则直接通过卫星***及惯导***实现定位功能，并通过惯导进行辅助，从而得到完整的车辆位置，因此，此方法可以在全天候，任何条件下得到车辆的位置，从而为线控转向提供依据。场景分析如下：
直线段微调转角的车道保持方法示意图为图5，车辆过弯道方法示意图为图6。
三.车辆控制以及执行
(1)如图6所示，在已知车辆位置的前提下，在控制器内部有车辆的轨迹约束线，每一个车轮不能超过车轮约束线，从而将其转为有约束条件的模型，再根据PID控制方法综合设计转向机构贯通道的控制***：
Figure PCTCN2019104833-appb-000002
主控器完成数据分析后，再根据传递函数矩阵保证***的稳定性和可控性，考虑到PID***只能控制线性***和较为简单的非线性***，因此还需要设计预测控制模型及增益调度等非线性控制，从而更加精确的确定车辆在约束线内运行的轨迹。主控器计算出运行轨迹后将运行轨迹处理为电信号，并将计算出的电信号反馈给执行机构完成在约束线内的转动要求，从而使转向机构按照预定轨迹转向，并根据车道线进行校核，并给出反馈信号，从而确保转向功能完成的准确和稳定；执行机构根据电信号分别对各轮对进行控制，从而实现多轴车轮同步转向，并完成车道保持及转向功能。
考虑在约束条件下可能有多值结果，而且考虑车辆动力学模型以及可能存在的误差，因此有必要生成多次样条曲线或通过仿真方法，即在主控器内根据驾驶规则，舒适性要求，每条线路的长度等需求进行评估，构建轨迹的代价值函数Y cost＝f(R，S，L，W，O)(R驾驶规则，S舒适性要求，L线路长度，W噪音，O其他条件)及Y opt＝Max(Y cost1,Y cost2,Y cost3……)来确定最终的行驶轨迹找到最佳的车辆轨迹，通过控制算法找到车辆的运行轨迹，从而构建其传递函数矩阵，并构建车辆动力学模型，将转向机构及贯通道构建起一个多维控制矩阵，保证车辆的鲁棒性和稳定性，同时，在约束条件下完成车辆轨迹与车轮的位置的对应，从而分析车辆是否需要微调各轮对，或者执行轮对转向指令，并以此为依据生成电信号，传递给执行机构。
(2)执行机构采用电控***，将电信号直接传递给电控转向机构，考虑到每个轮对均可以被控制，因此根据车辆运行轨迹，即可知道车轮下一时刻所处的位置，同时，由于贯通道柔性的存在，因此当出现微小误差及其他因为动力学条件产生误差的条件下，仍能够保证车辆每个轮对在下一时刻按照主控器生成的轨迹到达指定位置，从而分别控制多个转向机构，转向机构根据接收的电信号进行调整从而完成转向和车道保持的要求。简化甚至消除人为控制转向的操作，减轻因为人为因素产生的干扰，从而提高处理车辆相关各项事宜的效率。
(3)车辆通过导航***及视频***得到了车辆的位置，并根据位置为车辆的行驶提供依据，因为卫星***和惯导***的存在，车辆可以直接通过该***检测出车辆的位移，速度，姿态角等，避免了车轮机构监测车轮转向***的设计。直接通过车辆的车***置为转向***的转向提供依据。同时，相比于目前的车辆第一轮转向，后轮跟随方案，此方案能够更快捷的提供车辆位置，并直接控制各个车轮，提升了***的冗余性。由于此方法可以实时同步车辆位置数据，因而可以避免误差。同时也可以进行轮对转角的微调，避免甩尾现象的发生。
(4)针对全天候的运营要求，卫星***主要用于提供定位并在主控器内完成路径规划功能，摄像机主要用于实现近距离的车道线识别以及环境探测。综合运用卫星和惯导***的定位装置可以识别出车辆在卫星平面内的位置，并可以保证车辆在车道线内，而摄像头捕捉车道线信息，给卫星***确定的位置一个反馈信号，从而更加精准的判断车辆的转向及微调。同时，当卫星***数据不够准确的情况下(诸如在隧道内，高楼效应，有障碍物在天空遮挡等)，将主要以光学识别作为基础，调整算法里PID里的K p,K i,K d，并强化非线性算法的比重，同时使用惯导***作为补充，使定位***能够尽可能保持定位精度在厘米级别，而当 卫星定位丢失的时间过长或卫星定位***故障，则直接切换光学***及惯导***，通过惯导***在一定长度内代替卫星***的功能，光学***继续识别车道线保持车辆的精准定位，完成这段运营并能够自导向回厂完成检修，而当阴雨，雷暴天气等恶劣天气覆盖车道线，则直接通过卫星***及惯导***实现定位功能，并通过惯导进行辅助，从而得到完整的车辆位置，在日夜条件下，可以调整K p,K i,K d，并针对气候条件调整非线性***的算法，从而保证了日夜条件均可以使用的要求。同时，此方法当遇到道路出现障碍物等突发改变车道的问题后，可通过卫星***返回原有车道，继续完成运营，具有较高的冗余性，稳定性和鲁棒性。因此，此方法受天气以及日夜环境的影响较小，可以在全天候，大部分极端恶劣条件下得到车辆的位置，从而为线控转向提供依据。
本发明适用于多轴重载电客车，可以提升车辆自导向***的精确性和冗余性，并综合运用光学和卫星惯导导航***实现车辆的车道保持功能及多车轮同步转向功能，采用电信号传递给电控转向装置也易于控制车辆；提供了人机交互界面，方便司机了解车辆的信息，在辅助驾驶阶段也可以提升司机的操控性。
上面结合附图对本发明的实施例进行了描述，但是本发明并不局限于上述的具体实施方式，上述的具体实施方式仅仅是示意性的，而不是局限性的，本领域的普通技术人员在本发明的启示下，在不脱离本发明宗旨和权利要求所保护的范围情况下，还可做出很多形式，这些均属于本发明的保护范围之内。

Claims (5)

  1. 一种多轴电客车自导向方法，其特征在于，包括以下步骤：
    步骤A.将车辆前后两侧摄像头捕获的原始图像输入到车载设备并对原始图像进行预处理，得到预处理后的图像；
    步骤B.针对预处理后的图像，对车辆行驶区域进行局部增强；
    步骤C.提取出车道线；
    步骤D.根据摄像头和车道线的位置确定车辆位置；
    步骤E.根据车辆行驶约束条件，运用PID控制方法、模糊PID控制方法、预测控制方法、非线性控制方法控制车辆在允许的轨迹内运行。
  2. 如权利要求1所述的多轴电客车自导向方法，其特征在于，
    所述步骤D中，还包括利用安装在中间车的卫星定位***和/或惯导***确定车辆位置。
  3. 如权利要求1所述的多轴电客车自导向方法，其特征在于，所述步骤A中的预处理过程依次包括：图像压缩、灰度转化、滤波处理、均衡处理。
  4. 如权利要求1所述的多轴电客车自导向方法，其特征在于，所述步骤E中，根据PID控制方法设计转向机构贯通道的控制***：
    Figure PCTCN2019104833-appb-100001
    其中K d为PID控制***的微分因子、K p为PID控制***的比例因子、K i为PID控制***的积分因子。
  5. 如权利要求1所述的多轴电客车自导向方法，其特征在于，所述步骤E中，还包括根据驾驶规则、舒适性要求、线路长度、噪音限制构建代价值函数，找到最佳车辆运行轨迹。
PCT/CN2019/104833 2018-11-20 2019-09-09 一种多轴电客车自导向方法 WO2020103532A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201811382390.8 2018-11-20
CN201811382390.8A CN109492599A (zh) 2018-11-20 2018-11-20 一种多轴电客车自导向方法

Publications (1)

Publication Number Publication Date
WO2020103532A1 true WO2020103532A1 (zh) 2020-05-28

Family

ID=65696311

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/104833 WO2020103532A1 (zh) 2018-11-20 2019-09-09 一种多轴电客车自导向方法

Country Status (2)

Country Link
CN (1) CN109492599A (zh)
WO (1) WO2020103532A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113709344A (zh) * 2021-08-25 2021-11-26 佛吉亚歌乐电子(丰城)有限公司 一种图像数据处理方法、装置、设备及存储介质

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109492599A (zh) * 2018-11-20 2019-03-19 中车株洲电力机车有限公司 一种多轴电客车自导向方法
CN112224203A (zh) * 2019-06-28 2021-01-15 北汽福田汽车股份有限公司 车辆控制的方法、装置和汽车
CN110472578B (zh) * 2019-08-15 2020-09-18 宁波中车时代传感技术有限公司 基于车道弯曲度的车道线保持方法
CN110597252B (zh) * 2019-09-03 2021-01-05 安徽江淮汽车集团股份有限公司 自动驾驶汽车融合定位控制方法、装置、设备及存储介质
CN113064344B (zh) * 2021-03-19 2022-06-07 中山大学 一种多轴无人重载车辆的轨迹跟踪控制方法
CN113942522A (zh) * 2021-05-31 2022-01-18 重庆工程职业技术学院 一种智能驾驶行车安全保护***

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103226354A (zh) * 2013-02-27 2013-07-31 广东工业大学 基于光电导航的无人驾驶道路识别***
CN106627582A (zh) * 2016-12-09 2017-05-10 重庆长安汽车股份有限公司 单车道自动驾驶超越相邻车道车辆的路径规划***及方法
CN107200020A (zh) * 2017-05-11 2017-09-26 江苏大学 一种基于混杂理论的无人驾驶汽车自主转向控制***和方法
CN107972663A (zh) * 2018-01-03 2018-05-01 汽-大众汽车有限公司 一种基于智能驾驶技术的车辆控制***、装置及方法
CN108036794A (zh) * 2017-11-24 2018-05-15 华域汽车***股份有限公司 一种高精度地图生成***及生成方法
WO2018145602A1 (zh) * 2017-02-07 2018-08-16 腾讯科技（深圳）有限公司 车道的确定方法、装置及存储介质
CN109492599A (zh) * 2018-11-20 2019-03-19 中车株洲电力机车有限公司 一种多轴电客车自导向方法

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102167082B (zh) * 2011-03-21 2013-02-06 武汉理工大学 电动轮驱动汽车的差力主动转向***及其控制方法
CN105718865A (zh) * 2016-01-15 2016-06-29 武汉光庭科技有限公司 一种自动驾驶中基于双目摄像头的道路安全检测***及其方法
CN106441319B (zh) * 2016-09-23 2019-07-16 中国科学院合肥物质科学研究院 一种无人驾驶车辆车道级导航地图的生成***及方法
CN107065689B (zh) * 2017-05-27 2023-07-25 广东天物新材料科技有限公司 一种新型电动车供电装置

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103226354A (zh) * 2013-02-27 2013-07-31 广东工业大学 基于光电导航的无人驾驶道路识别***
CN106627582A (zh) * 2016-12-09 2017-05-10 重庆长安汽车股份有限公司 单车道自动驾驶超越相邻车道车辆的路径规划***及方法
WO2018145602A1 (zh) * 2017-02-07 2018-08-16 腾讯科技（深圳）有限公司 车道的确定方法、装置及存储介质
CN107200020A (zh) * 2017-05-11 2017-09-26 江苏大学 一种基于混杂理论的无人驾驶汽车自主转向控制***和方法
CN108036794A (zh) * 2017-11-24 2018-05-15 华域汽车***股份有限公司 一种高精度地图生成***及生成方法
CN107972663A (zh) * 2018-01-03 2018-05-01 汽-大众汽车有限公司 一种基于智能驾驶技术的车辆控制***、装置及方法
CN109492599A (zh) * 2018-11-20 2019-03-19 中车株洲电力机车有限公司 一种多轴电客车自导向方法

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113709344A (zh) * 2021-08-25 2021-11-26 佛吉亚歌乐电子(丰城)有限公司 一种图像数据处理方法、装置、设备及存储介质
CN113709344B (zh) * 2021-08-25 2023-08-01 佛吉亚歌乐电子(丰城)有限公司 一种图像数据处理方法、装置、设备及存储介质

Also Published As

Publication number Publication date
CN109492599A (zh) 2019-03-19

Similar Documents

Publication Publication Date Title
WO2020103532A1 (zh) 一种多轴电客车自导向方法
US11635764B2 (en) Motion prediction for autonomous devices
CN109520498B (zh) 一种用于虚拟轨道车辆的虚拟道岔***及方法
US10937178B1 (en) Image-based depth data and bounding boxes
US20200348672A1 (en) Safety and comfort constraints for navigation
US11748909B2 (en) Image-based depth data and localization
US10582137B1 (en) Multi-sensor data capture synchronizaiton
WO2021217420A1 (zh) 车道线跟踪方法和装置
CN109386155B (zh) 面向自动化停车场的无人泊车搬运机器人的对位方法
EP2948928B1 (en) Modifying behavior of autonomous vehicles based on sensor blind spots and limitations
CN103204162B (zh) 具有有效后转向的车道跟踪***
CN106950964B (zh) 无人电动大学生方程式赛车及其控制方法
CN109556615A (zh) 基于自动驾驶的多传感器融合认知的驾驶地图生成方法
CN112650220A (zh) 一种车辆自动驾驶方法、车载控制器及***
CN102700548A (zh) 具有前摄像机和后摄像机的稳健的车辆侧向控制
CN103400392A (zh) 基于变电站巡检机器人的双目视觉导航***及方法
DE112017007906T5 (de) Fahrzeugsteuervorrichtung, fahrzeugsteuerverfahren und programm
CN113791621B (zh) 一种自动驾驶牵引车与飞机对接方法及***
CN111137298B (zh) 一种车辆自动驾驶方法、装置、***和存储介质
US20190163201A1 (en) Autonomous Vehicle Sensor Compensation Using Displacement Sensor
WO2023125754A1 (zh) 一种自动成列行驶车辆***及控制方法
CN112837554A (zh) 基于双目相机的agv定位导航方法及***
CN112216142B (zh) 基于特定场景的车辆视觉定位***
Jun et al. Autonomous driving system design for formula student driverless racecar
EP3857865A1 (en) Image scan line timestamping

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: 19887844

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: 19887844

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 19887844

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 07.01.2022)

122 Ep: pct application non-entry in european phase

Ref document number: 19887844

Country of ref document: EP

Kind code of ref document: A1