WO2019047640A1 - 一种在自动驾驶系统中进行雨刷控制的方法和装置 - Google Patents
一种在自动驾驶系统中进行雨刷控制的方法和装置 Download PDFInfo
- Publication number
- WO2019047640A1 WO2019047640A1 PCT/CN2018/098619 CN2018098619W WO2019047640A1 WO 2019047640 A1 WO2019047640 A1 WO 2019047640A1 CN 2018098619 W CN2018098619 W CN 2018098619W WO 2019047640 A1 WO2019047640 A1 WO 2019047640A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- information
- control
- automatic driving
- road image
- wiper
- Prior art date
- Legal status (The legal status 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 status listed.)
- Ceased
Links
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60S—SERVICING, CLEANING, REPAIRING, SUPPORTING, LIFTING, OR MANOEUVRING OF VEHICLES, NOT OTHERWISE PROVIDED FOR
- B60S1/00—Cleaning of vehicles
- B60S1/02—Cleaning windscreens, windows or optical devices
- B60S1/04—Wipers or the like, e.g. scrapers
- B60S1/06—Wipers or the like, e.g. scrapers characterised by the drive
- B60S1/08—Wipers or the like, e.g. scrapers characterised by the drive electrically driven
- B60S1/0818—Wipers or the like, e.g. scrapers characterised by the drive electrically driven including control systems responsive to external conditions, e.g. by detection of moisture, dirt or the like
Definitions
- the present invention relates to the field of automatic driving, and more particularly to a method and apparatus for performing wiper control in an automatic driving system.
- a control such as a rain sensor is used to implement the control of the wiper.
- this method requires additional installation of a rain sensor or the like.
- the rain sensor has poor accuracy and cannot be applied to a complex happensing. For example, when the glass is covered by dust mist, the rain sensor may not be able to sense this change, so that the wiper operation cannot be effectively driven in this case, which affects the image acquisition quality of the automatic driving.
- a method of performing wiper control in an automatic driving system includes the following steps:
- control prediction information including the wiper Forecast information.
- a control device for performing wiper control in an automatic driving system wherein the control device includes:
- the road image information and the corresponding real control information are used as training data to train the automatic driving model to obtain a device for controlling the prediction information corresponding to the road image information during the automatic driving process.
- a computer device in accordance with the present invention includes a memory, a processor, and a computer program stored on the memory and operative on the processor, wherein the method is implemented when the processor executes the program.
- a computer readable storage medium on which is stored a computer program, characterized in that the method is implemented when executed by a processor.
- the present invention has the following advantages: in the present solution, training is performed by acquiring real control information corresponding to road image information, especially based on real control information of the user in rain and snow weather, which can effectively improve the automatic driving. The accuracy of the prediction for wiper control. Moreover, in this solution, it is not necessary to specifically install other equipment in the vehicle, but only based on the input of road image information of the camera to which the automatic driving system is connected. It avoids erroneous operations due to inaccuracy of external devices such as rain sensors, and can adapt to more complex environmental conditions.
- FIG. 1 is a flow chart showing a method of a wiper control method in an automatic driving system in accordance with a preferred embodiment of the present invention
- FIG. 2 is a block diagram showing the structure of a wiper control device in an automatic driving system in accordance with a preferred embodiment of the present invention.
- control device performing the method of the invention is implemented by a computer device.
- the computer has a memory, a processor, and a computer program stored on the memory and operative on the processor, the processor executing the method as the access device performs when executing the corresponding computer program.
- a method corresponding to the control device can be implemented by storing a corresponding computer program on a computer readable storage medium such that the processor executes the program.
- the computer device comprises an electronic device capable of automatically performing numerical calculation and/or information processing according to an instruction set or stored in advance, the hardware of which includes but is not limited to a microprocessor, an application specific integrated circuit (ASIC), Programming gate arrays (FPGAs), digital processors (DSPs), embedded devices, and more.
- the computer device can include a network device and/or a user device.
- the computer device comprises a user device and/or a network device that can control the vehicle for automatic driving.
- the user equipment includes, but is not limited to, any electronic product that can be embedded in the vehicle and can interact with the user in a touch manner, for example, an embedded smart navigation device, a tablet computer, or the like.
- the network device includes, but is not limited to, a single network server, a server group composed of multiple network servers, or a cloud computing-based cloud composed of a large number of hosts or network servers, where the cloud computing is distributed computing.
- a super virtual computer consisting of a group of loosely coupled computers.
- the network device according to the present invention can control the vehicle to perform automatic driving by communicating with the vehicle.
- the network where the network device is located includes, but is not limited to, the Internet, a wide area network, a metropolitan area network, a local area network, a VPN network, and the like. It should be noted that the user equipment, the network equipment, and the network are only examples, and other existing or future network devices and networks may be applicable to the present invention, and are also included in the scope of the present invention. The way is included here.
- FIG. 1 there is illustrated a flow chart of a method of a wiper control method in an automated driving system in accordance with a preferred embodiment of the present invention.
- the method according to the invention comprises a step S1 and a step S2.
- step S1 the control device acquires real control information corresponding to a road image information.
- the road image information includes image information of the traveled road photographed by a camera corresponding to the automatic control system.
- the real control information includes, but is not limited to, wiper gear position information selected by the user.
- the real control information further includes but is not limited to any one of the following:
- control device can obtain the real control information based on the sensor in the vehicle or the interface with the vehicle numerical control system, and details are not described herein again.
- step S2 the control device uses the road image information and its corresponding real control information as training data to train the automatic driving model to obtain control corresponding to the road image information during the automatic driving process.
- Prediction information the control prediction information including wiper prediction information.
- the wiper prediction information includes gear position probability information corresponding to each gear position of the wiper.
- the current wiper has four gear positions of 0, 1, 2, and 3, the 0 gear is off, and the 1-3 gears are sequentially incremented, and the gear prediction information may include a percentage of the probability of each of the four gears. Since only one gear can be selected at a time, the sum of the probabilities of the four gears is one.
- control prediction information includes probability prediction information for each driving selection corresponding to the road image information
- control prediction information further includes but is not limited to any one of the following:
- Speed prediction information for example, current probability prediction for speed adjustment operation; for example, prediction of target speed, and the like.
- Direction prediction information for example, the following steering wheel angle prediction, or the adjustment operation prediction of the steering wheel angle.
- control means determines the automatic driving strategy based on the obtained control prediction information.
- control device determines an automatic driving strategy corresponding to the corresponding road image information based on the predetermined control mechanism based on the predetermined decision mechanism.
- the prediction information having the highest probability is selected as the operation at the time of automatic driving or the like.
- control device may also determine the wiper control information in the automatic driving measurement in conjunction with real-time related information related to the wiper.
- the control device obtains road image information showing dripping rain, and determines that the wiper prediction information of the automatic driving model is: 0 gear 0%, 1 gear 0%, 2 gear 45%, and 3 gear 55%. Then, the control device acquires current weather information, and selects 3 files to operate based on the current real-time weather report: rainstorm yellow warning.
- the control device filters the road image information and its corresponding real control information based on the predetermined screening conditions, based on the filtered road image information and
- the automatic driving model is trained corresponding to the real control information.
- control device judges that the difference between the obtained real control information and the control information normally corresponding to the road image information is larger than a predetermined threshold, it is screened out.
- control means determines the overall loss information of the automatic driving model based on the predicted control information and the real control information.
- the control device determines the overall loss information of the current automatic driving model by using, for example, a cross entropy function, based on the control prediction information corresponding to the road image information and the real control information corresponding to the road image information.
- the control prediction information that the loss information satisfies the predetermined requirement is obtained.
- the automatic driving model itself according to the present invention can be constructed in various ways, for example, using a neural network model or the like, and the manner in which the model is constructed does not affect the implementation of the present invention.
- the training by acquiring the real control information corresponding to the road image information, especially based on the real control information of the user in the rain and snow weather, can effectively improve the accuracy of the prediction of the wiper control in the automatic driving.
- FIG. 2 there is illustrated a block diagram of a control device for a wiper in an automated driving system in accordance with a preferred embodiment of the present invention.
- the control device comprises an acquisition device 101 and a training device 102.
- the acquisition device 101 acquires real control information corresponding to a road image information.
- the road image information includes image information of the traveled road photographed by a camera corresponding to the automatic control system.
- the real control information includes, but is not limited to, wiper gear position information selected by the user.
- the real control information further includes but is not limited to any one of the following:
- the acquisition device 101 can obtain the real control information based on sensors in the vehicle or an interface with the vehicle numerical control system, and details are not described herein again.
- the training device 102 uses the road image information and its corresponding real control information as training data to train the automatic driving model to obtain control prediction information corresponding to the road image information during the automatic driving process, and the control
- the prediction information includes wiper prediction information.
- the wiper prediction information includes gear position probability information corresponding to each gear position of the wiper.
- the current wiper has four gear positions of 0, 1, 2, and 3, the 0 gear is off, and the 1-3 gears are sequentially incremented, and the gear prediction information may include a percentage of the probability of each of the four gears. Since only one gear can be selected at a time, the sum of the probabilities of the four gears is one.
- control prediction information includes probability prediction information for each driving selection corresponding to the road image information
- control prediction information further includes but is not limited to any one of the following:
- Speed prediction information for example, current probability prediction for speed adjustment operation; for example, prediction of target speed, and the like.
- Direction prediction information for example, the following steering wheel angle prediction, or the adjustment operation prediction of the steering wheel angle.
- the control device determines the automatic driving strategy based on the obtained control prediction information.
- control device determines an automatic driving strategy corresponding to the corresponding road image information based on the predetermined control mechanism based on the predetermined decision mechanism.
- the prediction information having the highest probability is selected as the operation at the time of automatic driving or the like.
- control device may also determine the wiper control information in the automatic driving measurement in conjunction with real-time related information related to the wiper.
- the control device obtains road image information showing dripping rain, and determines that the wiper prediction information of the automatic driving model is: 0 gear 0%, 1 gear 0%, 2 gear 45%, and 3 gear 55%. Then, the control device acquires current weather information, and selects 3 files to operate based on the current real-time weather report: rainstorm yellow warning.
- the training device 102 filters the road image information and its corresponding real control information based on the predetermined screening conditions, based on the filtered road image information and The corresponding automatic control information is used to train the automatic driving model.
- the training device 102 determines that the difference between the obtained real control information and the control information normally corresponding to the road image information is greater than a predetermined threshold, it is screened out.
- the training device 102 determines the overall loss information of the automatic driving model based on the predicted control information and the real control information.
- the training device 102 determines the overall loss information of the current automatic driving model by using, for example, a cross entropy function, based on the control prediction information corresponding to a road image information and the real control information corresponding to the road image information.
- control prediction information in which the loss information satisfies the predetermined requirement is obtained.
- the automatic driving model itself according to the present invention can be constructed in various ways, for example, using a neural network model or the like, and the manner in which the model is constructed does not affect the implementation of the present invention.
- the training by acquiring the real control information corresponding to the road image information, especially based on the real control information of the user in the rain and snow weather, can effectively improve the accuracy of the prediction of the wiper control in the automatic driving.
Landscapes
- Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Mechanical Engineering (AREA)
- Traffic Control Systems (AREA)
- Control Of Driving Devices And Active Controlling Of Vehicle (AREA)
Abstract
本发明的目的是提供一种自动驾驶系统中的雨刷控制的方法和装置。根据本发明的方案包括以下步骤:获取一道路图像信息对应的用户对雨刷的真实控制信息;将所述道路图像信息及其对应的真实控制信息作为训练数据,来对自动驾驶模型进行训练,以获得自动驾驶过程中,该道路图像信息对应的控制预测信息,所述控制预测信息包括雨刷预测信息。本方案优点在于:通过获取与道路图像信息对应的真实控制信息进行训练,尤其基于用户在雨雪天气的真实控制信息,能够有效地提高自动驾驶中对于雨刷控制的预测的准确性。避免了由于雨量传感器等外部设备不精确等原因造成的错误操作,并能够适应更多更复杂的环境情况。
Description
相关申请的交叉引用
本专利申请要求于2017年9月5日提交的、申请号为201710792882.3、申请人为百度在线网络技术(北京)有限公司、发明名称为“一种在自动驾驶系统中进行雨刷控制的方法和装置”的中国专利申请的优先权,该申请的全文以引用的方式并入本申请中。
本发明涉及自动驾驶领域,尤其涉及一种在自动驾驶系统中进行雨刷控制的方法和装置。
伴随着深度学习的发展,通过端到端的深度学习实现自动驾驶便是自动驾驶领域的主要研究方向。在端到端的自动驾驶中,自动驾驶系统通常以单目摄像头进行输入,摄像头安装在汽车内侧,挡风玻璃上。由于摄像头采集到的图像质量,对端到端的安全驾驶非常重要,因此,在雨雪天气,不开启雨刷,就会影响采集的图像质量,影响驾驶安全。
现有技术中采用诸如雨量传感器等方式来实现对雨刷的控制,然而,一方面,该种方式需要额外安装雨量传感器等设备,另一方面,雨量传感器的准确性较差,并且无法适用于复杂情况。例如,在玻璃被尘雾遮盖时,雨量传感器可能无法感应到这一变化,从而无法有效地在此情况下驱动雨刷操作,影响自动驾驶的图像获取质量。
发明内容
本发明的目的是提供一种自动驾驶系统中的雨刷控制的方法和 装置。
根据本发明的方案,提供了一种在自动驾驶系统中进行雨刷控制的方法,其中,所述方法包括以下步骤:
-获取一道路图像信息对应的用户对雨刷的真实控制信息;
-将所述道路图像信息及其对应的真实控制信息作为训练数据,来对自动驾驶模型进行训练,以获得自动驾驶过程中,该道路图像信息对应的控制预测信息,所述控制预测信息包括雨刷预测信息。
根据本发明的方案,提供了一种在自动驾驶系统中进行雨刷控制的控制装置,其中,所述控制装置包括:
用于获取一道路图像信息对应的用户对雨刷的真实控制信息的装置;
用于将所述道路图像信息及其对应的真实控制信息作为训练数据,来对自动驾驶模型进行训练,以获得自动驾驶过程中,该道路图像信息对应的控制预测信息的装置。
根据本发明的一种计算机设备,包括存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序,其特征在于,所述处理器执行所述程序时实现所述的方法。
根据本发明的一种计算机可读的存储介质,其上存储有计算机程序,其特征在于,该程序被处理器执行时实现所述的方法。
与现有技术相比,本发明具有以下优点:本方案中,通过获取与道路图像信息对应的真实控制信息进行训练,尤其基于用户在雨雪天气的真实控制信息,能够有效地提高自动驾驶中对于雨刷控制的预测的准确性。并且,本方案中无需专门在车辆中安装其他设备,而仅基于自动驾驶系统所连接的摄像头的道路图像信息的输入即可。避免了由于雨量传感器等外部设备不精确等原因造成的错误操作,并能够适应更多更复杂的环境情况。
通过阅读参照以下附图所作的对非限制性实施例所作的详细描述,本发明的其它特征、目的和优点将会变得更明显:
图1示意出了根据本发明的一种优选实施例自动驾驶系统中的雨刷控制方法的方法流程图;
图2示意出了根据本发明的一种优选实施例自动驾驶系统中的雨刷控制装置的结构示意图。
附图中相同或相似的附图标记代表相同或相似的部件。
下面结合附图对本发明作进一步详细描述。
其中,执行本发明方法的控制装置通过计算机设备来实现。所述计算机具有存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序,该处理器执行程序相应的计算机程序时实现如存取装置所执行的方法。此外,通过在计算机可读存储介质上存储相应的计算机程序,以使得处理器执行该程序时可实现如控制装置对应的方法。
其中,所述计算机设备包括一种能够按照事先设定或存储的指令,自动进行数值计算和/或信息处理的电子设备,其硬件包括但不限于微处理器、专用集成电路(ASIC)、可编程门阵列(FPGA)、数字处理器(DSP)、嵌入式设备等。所述计算机设备可包括网络设备和/或用户设备。
优选地,所述计算机设备包括可控制车辆进行自动驾驶的用户设备和/或网络设备。
其中,所述用户设备包括但不限于任何一种可内嵌于所述车辆中并可与用户触控方式进行人机交互的电子产品,例如,内嵌智能导航设备、平板电脑等。
其中,所述网络设备包括但不限于单个网络服务器、多个网络服务器组成的服务器组或基于云计算(Cloud Computing)的由大量主机或网络服务器构成的云,其中,云计算是分布式计算的一种,由一群松散耦合的计算机集组成的一个超级虚拟计算机。其中,根据本发明的网络设备可通过与车辆的通信来控制车辆进行自动驾驶。
其中,所述网络设备所处的网络包括但不限于互联网、广域网、 城域网、局域网、VPN网络等。需要说明的是,所述用户设备、网络设备以及网络仅为举例,其他现有的或今后可能出现网络设备以及网络如可适用于本发明,也应包含在本发明保护范围以内,并以引用方式包含于此。
参照图1,图1示意出了根据本发明的一种优选实施例自动驾驶系统中的雨刷控制方法的方法流程图。
根据本发明的方法包括步骤S1和步骤S2。
在步骤S1中,控制装置获取一道路图像信息对应的真实控制信息。
其中,所述道路图像信息包括由自动控制系统对应的摄像头所拍到的所行驶道路的图像信息。
其中,所述真实控制信息包括但不限于用户所选择的雨刷档位信息。
优选地,所述真实控制信息还包括但不限于以下任一项:
1)车辆的行驶方向信息;
2)车辆的行驶速度信息。
其中,本领域技术人员应可理解,控制装置可基于车辆内的传感器或者与车辆数控系统的接口,来获得所述真实控制信息,此处不再赘述。
接着,在步骤S2中,所述控制装置将所述道路图像信息及其对应的真实控制信息作为训练数据,来对自动驾驶模型进行训练,以获得自动驾驶过程中,该道路图像信息对应的控制预测信息,所述控制预测信息包括雨刷预测信息。
其中,所述雨刷预测信息包括与该雨刷的各个档位对应的档位概率信息。
例如,当前雨刷有0,1,2,3四个档位,0档为关,1-3档的摇动速冻依次递增,档位预测信息可包括该四个档位各自所占概率的百分比,由于每次只能选择一个档位,因此该四个档位的概率之和为1。
其中,所述控制预测信息包括针对与该道路图像信息对应的各项 驾驶选择的概率预测信息
优选地,所述控制预测信息还包括但不限于以下任一项:
1)速度预测信息;例如,当前对于速度调整操作的概率预测;又例如,目标速度的预测等。
2)方向预测信息,例如,接下去的方向盘角度预测,或者,方向盘角度的调整操作预测等。
根据本发明的一个优选实施例,根据本发明的方法,控制装置基于所获得的控制预测信息,来确定自动驾驶策略。
具体地,控制装置基于预定的决策机制,来基于所述控制预测信息中确定相应的道路图像信息对应的自动驾驶策略。
例如,选择概率最大的预测信息来作为自动驾驶时的操作等。
优选地,控制装置还可结合与所述雨刷相关的实时相关信息,来确定自动驾驶测量中的雨刷控制信息。
例如,控制装置获得呈现点滴小雨的道路图像信息,并确定自动驾驶模型的雨刷预测信息为:0档0%,1档0%,2档45%,以及3档55%。则控制装置获取当前的天气信息,并基于当前的实时天气报:暴雨黄色警告,选择3档进行操作。
根据本发明的一个优选实施例,根据本发明的方法,控制装置基于预定筛选条件,对所述道路图像信息及其对应的真实控制信息进行筛选,以基于筛选后的所述道路图像信息及其对应的真实控制信息,对所述自动驾驶模型进行训练。
例如,当控制装置判断所获得的真实控制信息与道路图像信息通常对应的控制信息的差距大于预定阈值时,将其筛除。
根据本发明的又一个优选实施例,控制装置基于所述预测控制信息与所述真实控制信息来确定所述自动驾驶模型的整体损失信息。
其中,控制装置基于与一道路图像信息对应的控制预测信息,以及该道路图像信息对应的真实控制信息,利用诸如交叉熵函数等方式,来确定当前的自动驾驶模型的整体损失信息。
以获得损失信息满足预定要求的控制预测信息。
需要说明的是,本领域技术人员应可理解,根据本发明的自动驾驶模型本身可采用多种方式来构建,例如,采用神经网络模型等,其模型的构建方式并不影响本发明的实施。
本方案中,通过获取与道路图像信息对应的真实控制信息进行训练,尤其基于用户在雨雪天气的真实控制信息,能够有效地提高自动驾驶中对于雨刷控制的预测的准确性。并且,本方案中无需专门在车辆中安装其他设备,而仅基于自动驾驶系统所连接的摄像头的道路图像信息的输入即可。避免了由于雨量传感器等外部设备不精确等原因造成的错误操作,并能够适应更多更复杂的环境情况。
参照图2,图2示意出了根据本发明的一种优选实施例自动驾驶系统中的雨刷的控制装置的结构示意图。
根据本发明的控制装置包括获取装置101和训练装置102。
获取装置101获取一道路图像信息对应的真实控制信息。
其中,所述道路图像信息包括由自动控制系统对应的摄像头所拍到的所行驶道路的图像信息。
其中,所述真实控制信息包括但不限于用户所选择的雨刷档位信息。
优选地,所述真实控制信息还包括但不限于以下任一项:
1)车辆的行驶方向信息;
2)车辆的行驶速度信息。
其中,本领域技术人员应可理解,获取装置101可基于车辆内的传感器或者与车辆数控系统的接口,来获得所述真实控制信息,此处不再赘述。
接着,训练装置102将所述道路图像信息及其对应的真实控制信息作为训练数据,来对自动驾驶模型进行训练,以获得自动驾驶过程中,该道路图像信息对应的控制预测信息,所述控制预测信息包括雨刷预测信息。
其中,所述雨刷预测信息包括与该雨刷的各个档位对应的档位概率信息。
例如,当前雨刷有0,1,2,3四个档位,0档为关,1-3档的摇动速冻依次递增,档位预测信息可包括该四个档位各自所占概率的百分比,由于每次只能选择一个档位,因此该四个档位的概率之和为1。
其中,所述控制预测信息包括针对与该道路图像信息对应的各项驾驶选择的概率预测信息
优选地,所述控制预测信息还包括但不限于以下任一项:
1)速度预测信息;例如,当前对于速度调整操作的概率预测;又例如,目标速度的预测等。
2)方向预测信息,例如,接下去的方向盘角度预测,或者,方向盘角度的调整操作预测等。
根据本发明的一个优选实施例,根据本发明的方案,控制装置基于所获得的控制预测信息,来确定自动驾驶策略。
具体地,控制装置基于预定的决策机制,来基于所述控制预测信息中确定相应的道路图像信息对应的自动驾驶策略。
例如,选择概率最大的预测信息来作为自动驾驶时的操作等。
优选地,控制装置还可结合与所述雨刷相关的实时相关信息,来确定自动驾驶测量中的雨刷控制信息。
例如,控制装置获得呈现点滴小雨的道路图像信息,并确定自动驾驶模型的雨刷预测信息为:0档0%,1档0%,2档45%,以及3档55%。则控制装置获取当前的天气信息,并基于当前的实时天气报:暴雨黄色警告,选择3档进行操作。
根据本发明的一个优选实施例,根据本发明的方案,训练装置102基于预定筛选条件,对所述道路图像信息及其对应的真实控制信息进行筛选,以基于筛选后的所述道路图像信息及其对应的真实控制信息,对所述自动驾驶模型进行训练。
例如,当训练装置102判断所获得的真实控制信息与道路图像信息通常对应的控制信息的差距大于预定阈值时,将其筛除。
根据本发明的又一个优选实施例,训练装置102基于所述预测控制信息与所述真实控制信息来确定所述自动驾驶模型的整体损失信息。
其中,训练装置102基于与一道路图像信息对应的控制预测信息,以及该道路图像信息对应的真实控制信息,利用诸如交叉熵函数等方式,来确定当前的自动驾驶模型的整体损失信息。
从而获得损失信息满足预定要求的控制预测信息。
需要说明的是,本领域技术人员应可理解,根据本发明的自动驾驶模型本身可采用多种方式来构建,例如,采用神经网络模型等,其模型的构建方式并不影响本发明的实施。
本方案中,通过获取与道路图像信息对应的真实控制信息进行训练,尤其基于用户在雨雪天气的真实控制信息,能够有效地提高自动驾驶中对于雨刷控制的预测的准确性。并且,本方案中无需专门在车辆中安装其他设备,而仅基于自动驾驶系统所连接的摄像头的道路图像信息的输入即可。避免了由于雨量传感器等外部设备不精确等原因造成的错误操作,并能够适应更多更复杂的环境情况。
对于本领域技术人员而言,显然本发明不限于上述示范性实施例的细节,而且在不背离本发明的精神或基本特征的情况下,能够以其他的具体形式实现本发明。因此,无论从哪一点来看,均应将实施例看作是示范性的,而且是非限制性的,本发明的范围由所附权利要求而不是上述说明限定,因此旨在将落在权利要求的等同要件的含义和范围内的所有变化涵括在本发明内。不应将权利要求中的任何附图标记视为限制所涉及的权利要求。此外,显然“包括”一词不排除其他单元或步骤,单数不排除复数。系统权利要求中陈述的多个单元或装置也可以由一个单元或装置通过软件或者硬件来实现。第一,第二等词语用来表示名称,而并不表示任何特定的顺序。
Claims (12)
- 一种在自动驾驶系统中进行雨刷控制的方法,其中,所述方法包括以下步骤:-获取一道路图像信息对应的用户对雨刷的真实控制信息;-将所述道路图像信息及其对应的真实控制信息作为训练数据,来对自动驾驶模型进行训练,以获得自动驾驶过程中,该道路图像信息对应的控制预测信息,所述控制预测信息包括雨刷预测信息。
- 根据权利要求1所述的方法,其中,所述雨刷预测信息包括雨刷的各个档位分别对应的档位概率信息。
- 根据权利要求1或2所述的方法,其中,所述方法包括以下步骤:-基于预定筛选条件,对所述道路图像信息及其对应的真实控制信息进行筛选,以基于筛选后的所述道路图像信息及其对应的真实控制信息,对所述自动驾驶模型进行训练。
- 根据权利要求1至3中任一项所述的方法,其中,所述方法还包括以下步骤:-基于所述预测控制信息与所述真实控制信息来确定所述自动驾驶模型的整体损失信息。
- 根据权利要求1至4中任一项所述的方法,其中,所述方法还包括以下步骤:-基于所述控制预测信息,确定用于自动驾驶的驾驶策略信息。
- 一种在自动驾驶系统中进行雨刷控制的控制装置,其中,所述控制装置包括:用于获取一道路图像信息对应的用户对雨刷的真实控制信息的装置;用于将所述道路图像信息及其对应的真实控制信息作为训练数据,来对自动驾驶模型进行训练,以获得自动驾驶过程中,该道路图像信息对应的控制预测信息的装置。
- 根据权利要求6所述的控制装置,其中,所述控制预测信息包括雨刷预测信息,所述雨刷预测信息包括雨刷的各个档位分别对应的档位 概率信息。
- 根据权利要求6或7所述的控制装置,其中,所述控制装置进一步包括:用于基于预定筛选条件,对所述道路图像信息及其对应的真实控制信息进行筛选,以基于筛选后的所述道路图像信息及其对应的真实控制信息,对所述自动驾驶模型进行训练的装置。
- 根据权利要求6至8中任一项所述的控制装置,其中,所述控制装置还包括:用于基于所述预测控制信息与所述真实控制信息来确定所述自动驾驶模型的整体损失信息的装置。
- 根据权利要求6至9中任一项所述的控制装置,其中,所述控制装置还用于:-基于所述控制预测信息,确定用于自动驾驶的驾驶策略信息。
- 一种计算机设备,包括存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序,其特征在于,所述处理器执行所述程序时实现如权利要求1-5中任一所述的方法。
- 一种计算机可读的存储介质,其上存储有计算机程序,其特征在于,该程序被处理器执行时实现如权利要求1-5中任一所述的方法。
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710792882.3A CN107738627A (zh) | 2017-09-05 | 2017-09-05 | 一种在自动驾驶系统中进行雨刷控制的方法和装置 |
| CN201710792882.3 | 2017-09-05 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2019047640A1 true WO2019047640A1 (zh) | 2019-03-14 |
Family
ID=61235239
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2018/098619 Ceased WO2019047640A1 (zh) | 2017-09-05 | 2018-08-03 | 一种在自动驾驶系统中进行雨刷控制的方法和装置 |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN107738627A (zh) |
| WO (1) | WO2019047640A1 (zh) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107738627A (zh) * | 2017-09-05 | 2018-02-27 | 百度在线网络技术(北京)有限公司 | 一种在自动驾驶系统中进行雨刷控制的方法和装置 |
| CN116442953B (zh) * | 2023-03-29 | 2025-12-16 | 上汽通用五菱汽车股份有限公司 | 环境参数自学习系统、识别方法与智能辅助驾驶方法 |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH06171469A (ja) * | 1992-03-31 | 1994-06-21 | Mazda Motor Corp | 車両用ワイパ制御装置 |
| US5453676A (en) * | 1994-09-30 | 1995-09-26 | Itt Automotive Electrical Systems, Inc. | Trainable drive system for a windshield wiper |
| JP2014024354A (ja) * | 2012-07-24 | 2014-02-06 | Tokai Rika Co Ltd | ワイパ制御装置 |
| GB2536683A (en) * | 2015-03-26 | 2016-09-28 | Denso Corp | Vehicle equipment control |
| CN106945637A (zh) * | 2016-12-23 | 2017-07-14 | 惠州市德赛西威汽车电子股份有限公司 | 一种基于行车记录仪的雨刮控制系统及其方法 |
| CN106945638A (zh) * | 2017-03-21 | 2017-07-14 | 重庆长安汽车股份有限公司 | 一种雨刮控制系统 |
| CN107738627A (zh) * | 2017-09-05 | 2018-02-27 | 百度在线网络技术(北京)有限公司 | 一种在自动驾驶系统中进行雨刷控制的方法和装置 |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN202518228U (zh) * | 2012-04-18 | 2012-11-07 | 长安大学 | 一种汽车雨刷控制装置 |
| CN102815281B (zh) * | 2012-07-25 | 2015-03-11 | 江西好帮手电子科技有限公司 | 一种汽车雨刷控制装置及其控制方法 |
| CN104704541B (zh) * | 2012-10-04 | 2017-09-26 | 三菱电机株式会社 | 车载信息处理装置 |
| US9335178B2 (en) * | 2014-01-28 | 2016-05-10 | GM Global Technology Operations LLC | Method for using street level images to enhance automated driving mode for vehicle |
| JP2017001597A (ja) * | 2015-06-15 | 2017-01-05 | トヨタ自動車株式会社 | 自動運転装置 |
| CN105699095B (zh) * | 2016-01-27 | 2018-11-13 | 常州加美科技有限公司 | 一种无人驾驶车辆的测试方法 |
| CN106772687A (zh) * | 2016-12-06 | 2017-05-31 | 上海博泰悦臻电子设备制造有限公司 | 基于行驶车辆的天气信息推送系统及方法 |
| CN106873596B (zh) * | 2017-03-22 | 2018-12-18 | 北京图森未来科技有限公司 | 一种车辆控制方法及装置 |
-
2017
- 2017-09-05 CN CN201710792882.3A patent/CN107738627A/zh active Pending
-
2018
- 2018-08-03 WO PCT/CN2018/098619 patent/WO2019047640A1/zh not_active Ceased
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH06171469A (ja) * | 1992-03-31 | 1994-06-21 | Mazda Motor Corp | 車両用ワイパ制御装置 |
| US5453676A (en) * | 1994-09-30 | 1995-09-26 | Itt Automotive Electrical Systems, Inc. | Trainable drive system for a windshield wiper |
| JP2014024354A (ja) * | 2012-07-24 | 2014-02-06 | Tokai Rika Co Ltd | ワイパ制御装置 |
| GB2536683A (en) * | 2015-03-26 | 2016-09-28 | Denso Corp | Vehicle equipment control |
| CN106945637A (zh) * | 2016-12-23 | 2017-07-14 | 惠州市德赛西威汽车电子股份有限公司 | 一种基于行车记录仪的雨刮控制系统及其方法 |
| CN106945638A (zh) * | 2017-03-21 | 2017-07-14 | 重庆长安汽车股份有限公司 | 一种雨刮控制系统 |
| CN107738627A (zh) * | 2017-09-05 | 2018-02-27 | 百度在线网络技术(北京)有限公司 | 一种在自动驾驶系统中进行雨刷控制的方法和装置 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107738627A (zh) | 2018-02-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP7301077B2 (ja) | 自律走行システムのテスト方法、装置、電子機器、プログラム及びコンピュータ読み取り可能な記録媒体 | |
| JP7222868B2 (ja) | オブジェクト挙動のリアルタイム予測 | |
| US11962664B1 (en) | Context-based data valuation and transmission | |
| CN110692025B (zh) | 轨迹生成与执行架构 | |
| JP7271467B2 (ja) | 自動運転の制御方法、装置、電子機器及び記憶媒体 | |
| US20230102802A1 (en) | Map change detection | |
| CN114676844A (zh) | 用于自动驾驶的方法、装置、设备、计算机可读存储介质 | |
| US11481913B2 (en) | LiDAR point selection using image segmentation | |
| US20210237717A1 (en) | Method for controlling vehicle, and vehicle | |
| CN109255341B (zh) | 障碍物感知错误数据的提取方法、装置、设备及介质 | |
| JP6808775B2 (ja) | 複数のキューを利用したオブジェクト追跡 | |
| JP2022020764A (ja) | サンプル偏差データの取得方法、装置及び電子機器 | |
| US11629961B2 (en) | Crosswalk detection | |
| CN111652153B (zh) | 场景自动识别方法、装置、无人车及存储介质 | |
| DE102017101466A1 (de) | Verfolgen von objekten in einer dynamischen umgebung zur verbesserten lokalisierung | |
| CN108602509A (zh) | 基于运动规划来操作自动驾驶车辆的方法和系统 | |
| US12296846B2 (en) | Methods and apparatus for natural language based scenario discovery to train a machine learning model for a driving system | |
| JP7143269B2 (ja) | 車両挙動予測のための圧縮環境特徴の表示 | |
| EP3989031B1 (en) | Systems and methods for fusing road friction data to enhance vehicle maneuvering | |
| JP6908674B2 (ja) | 自動運転車両を動作させるための所定のキャリブレーションテーブルに基づく車両制御システム | |
| US12024100B2 (en) | Device-level fault detection | |
| CN113743356A (zh) | 数据的采集方法、装置和电子设备 | |
| CN112419722A (zh) | 交通异常事件检测方法、交通管控方法、设备和介质 | |
| CN113688760A (zh) | 自动驾驶的数据识别方法、装置、计算机设备和存储介质 | |
| US11887409B2 (en) | Device health code broadcasting on mixed vehicle communication networks |
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: 18853428 Country of ref document: EP Kind code of ref document: A1 |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| 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 18/09/2020) |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 18853428 Country of ref document: EP Kind code of ref document: A1 |