Apollo CyberRT调度器：实时并行计算的关键

自动驾驶
需积分: 5 15 下载量 157 浏览量 更新于2024-08-05 收藏 1.28MB PDF 举报
"Apollo Cyber调度器是百度Apollo自动驾驶平台中的关键组件，主要负责计算任务的实时并行处理。它作为一个独立的模块，是Apollo CyberRT框架的核心，将原本需要在内核空间进行的调度工作转移到了用户空间，简化了算法开发者的任务，使他们能够专注于算法本身的实现，而无需关注并发控制、同步以及资源调度等底层细节。 CyberRT Scheduler的设计特点包括： 1. 绑定线程与CPU：通过pthread_setschedparam和pthread_setaffinity_np等接口，Scheduler能够将Native Thread绑定到特定的CPU核心上，实现Per-CPU Per-Thread，优化性能。 2. Processor概念：在CyberRT中，封装了Native Thread的类被称为Processor，它类似操作系统的物理CPU。 3. 上下文切换：CyberRT Scheduler利用栈上下文切换，使得由上层空间创建的Coroutine（协程）能够在Processor上按序执行，类似于内核调度线程到CPU。 4. Run Queue设计：借鉴内核调度器的Run Queue Per-Processor模式，CyberRT Scheduler也采用了相似的设计，能够针对各个Processor进行调度。 新调度模型的优势在于可以根据自动驾驶算法任务的特性进行定制化资源分配和优先级设置。例如，算法任务可以构成有向无环图（DAG），调度器可以将有依赖关系的任务安排在同一个或相邻的CPU上，提高缓存局部性，减少通信延迟，从而提升整体计算效率和系统响应速度。 Apollo Cyber调度器的这种设计方法不仅提高了计算任务的并行执行效率，还降低了系统复杂性，对于实现高效、可靠的自动驾驶系统至关重要。通过这种方式，开发者能够更加专注于自动驾驶的核心算法，而无需担忧底层的系统优化问题，这极大地推动了自动驾驶技术的研发进程。"

CyberRT:Apollo Cyber​​RT自动驾驶框架

2695 浏览量
介绍 Apollo Cyber​​ RT 是一个开源、高性能的运行时框架，专为自动驾驶 (AD) 场景而设计。 基于中心化的计算模型，针对自动驾驶的高并发、低延迟、高吞吐进行了大幅优化。 在过去几年 AD 技术的发展过程中，我们从之前使用 Apollo 的经验中学到了很多东西。 行业在不断发展，阿波罗也在发展。 展望未来，Apollo 已经从开发转向生产。 随着现实世界中的批量部署，我们看到了对最高水平的稳健性和性能的需求。 这就是我们花了数年时间构建和完善 Apollo Cyber​​ RT 的原因，它满足了 AD 解决方案的这些要求。 使用 Apollo Cyber​​ RT 的主要好处： 加速发展 具有数据融合功能的定义明确的任务界面 一系列开发工具 大量传感器驱动程序 简化部署 高效、自适应的消息通信 具有资源意识的可配置用户级调度程序 便携，依赖更少 为您自己的自动驾驶汽车

技术文档 _ Apollo的坐标系.pdf

187 浏览量
Apollo坐标系是自动驾驶技术的核心组成部分，它涉及到多种坐标系的使用。本文将详细介绍 Apollo坐标系中的几个关键概念，包括全球地理坐标系、局部坐标系、UTM 坐标系和车辆坐标系。 一、全球地理坐标系 全球地理...

技术文档 _ Apollo传感器标定指南.pdf

190 浏览量
"Apollo传感器标定指南" 本指南详细介绍了Apollo传感器标定服务和三种标定工具：Camera-to-Camera Calibration（相机到相机的标定）、Camera-to-LiDAR Calibration（相机到激光雷达的标定）和Radar-to-Camera ...

for topic, msg, t in bag.read_messages(): File "/home/baidu/anaconda3/envs/python36/lib/python3.6/site-packages/rosbag/bag.py", line 2696, in read_messages yield self.seek_and_read_message_data_record((entry.chunk_pos, entry.offset), raw, return_connection_header) File "/home/baidu/anaconda3/envs/python36/lib/python3.6/site-packages/rosbag/bag.py", line 2859, in seek_and_read_message_data_record msg_type = _get_message_type(connection_info) File "/home/baidu/anaconda3/envs/python36/lib/python3.6/site-packages/rosbag/bag.py", line 1910, in _get_message_type message_type = genpy.dynamic.generate_dynamic(info.datatype, info.msg_def)[info.datatype] File "/home/baidu/anaconda3/envs/python36/lib/python3.6/site-packages/genpy/dynamic.py", line 180, in generate_dynamic mod = __import__(os.path.basename(tmp_file.name)[:-3]) File "/tmp/genpy_zyoi0xpr/tmpxiplcd9k.py", line 10 class ___apollo.test.CaseSample(genpy.Message): ^ SyntaxError: invalid syntax

138 浏览量
根据错误提示，有一个类名为"___apollo.test.CaseSample"的消息类型生成时出现了无效的语法。可能有几个原因导致这个错误： 1. 类名中包含了无效字符或特殊字符。 2. 类名中使用了保留字或关键字。 3. 类名中存在...

root@in_dev_docker:/apollo# bash scripts/msf_create_lossless_map.sh /apollo/hdmap/pcd_apollo/ 50 /apollo/hdmap/ /apollo/bazel-bin WARNING: Logging before InitGoogleLogging() is written to STDERR E0715 22:08:35.399576 6436 lossless_map_creator.cc:162] num_trials = 1 Pcd folders are as follows: /apollo/hdmap/pcd_apollo/ Resolution: 0.125 Dataset: /apollo/hdmap/pcd_apollo Dataset: /apollo/hdmap/pcd_apollo/ Loaded the map configuration from: /apollo/hdmap//lossless_map/config.xml. Saved the map configuration to: /apollo/hdmap//lossless_map/config.xml. Saved the map configuration to: /apollo/hdmap//lossless_map/config.xml. E0715 22:08:35.767315 6436 lossless_map_creator.cc:264] ieout_poses = 1706 Failed to find match for field 'intensity'. Failed to find match for field 'timestamp'. E0715 22:08:35.769896 6436 velodyne_utility.cc:46] Un-organized-point-cloud E0715 22:08:35.781770 6436 lossless_map_creator.cc:275] Loaded 245443D Points at Trial: 0 Frame: 0. F0715 22:08:35.781791 6436 base_map_node_index.cc:101] Check failed: false *** Check failure stack trace: *** scripts/msf_create_lossless_map.sh: line 11: 6436 Aborted (core dumped) $APOLLO_BIN_PREFIX/modules/localization/msf/local_tool/map_creation/lossless_map_creator --use_plane_inliers_only true --pcd_folders $1 --pose_files $2 --map_folder $IN_FOLDER --zone_id $ZONE_ID --coordinate_type UTM --map_resolution_type single root@in_dev_docker:/apollo# bash scripts/msf_create_lossless_map.sh /apollo/hdmap/pcd_apollo/ 50 /apollo/hdmap/

189 浏览量
/apollo/hdmap/pcd_apollo/ Resolution: 0.125 Dataset: /apollo/hdmap/pcd_apollo Dataset: /apollo/hdmap/pcd_apollo/ Loaded the map configuration from: /apollo/hdmap//lossless_map/config.xml. Saved the ...

/apollo/bazel-bin WARNING: Logging before InitGoogleLogging() is written to STDERR E0715 22:08:35.399576 6436 lossless_map_creator.cc:162] num_trials = 1 Pcd folders are as follows: /apollo/hdmap/pcd_apollo/ Resolution: 0.125 Dataset: /apollo/hdmap/pcd_apollo Dataset: /apollo/hdmap/pcd_apollo/ Loaded the map configuration from: /apollo/hdmap//lossless_map/config.xml. Saved the map configuration to: /apollo/hdmap//lossless_map/config.xml. Saved the map configuration to: /apollo/hdmap//lossless_map/config.xml. E0715 22:08:35.767315 6436 lossless_map_creator.cc:264] ieout_poses = 1706 Failed to find match for field 'intensity'. Failed to find match for field 'timestamp'. E0715 22:08:35.769896 6436 velodyne_utility.cc:46] Un-organized-point-cloud E0715 22:08:35.781770 6436 lossless_map_creator.cc:275] Loaded 245443D Points at Trial: 0 Frame: 0. F0715 22:08:35.781791 6436 base_map_node_index.cc:101] Check failed: false *** Check failure stack trace: *** scripts/msf_create_lossless_map.sh: line 11: 6436 Aborted (core dumped) $APOLLO_BIN_PREFIX/modules/localization/msf/local_tool/map_creation/lossless_map_creator --use_plane_inliers_only true --pcd_folders $1 --pose_files $2 --map_folder $IN_FOLDER --zone_id $ZONE_ID --coordinate_type UTM --map_resolution_type single

192 浏览量
根据您提供的信息，看起来您遇到了一些错误...4. 查看相关文档或论坛，了解其他用户是否遇到类似的问题，并尝试找到解决方案。 如果您需要更详细的帮助，建议您向Apollo开发团队或相关社区提问，以获取更专业的支持。

在ros项目中添加发送websocket wss消息的功能，修改如下代码并在CmakeLists.txt中添加依赖，实现将serialized_data发送到wss://autopilot-test.t3go.cn:443/api/v1/vehicle/push/message/LFB1FV696M2L43840。main.cpp：#include "ros/ros.h" #include "std_msgs/String.h" #include <boost/thread/locks.hpp> #include <boost/thread/shared_mutex.hpp> #include "third_party/apollo/proto/perception/perception_obstacle.pb.h" #include "t3_perception.pb.h" apollo::perception::PerceptionObstacles perception_obstacles_; void perceptionCallback(const std_msgs::String& msg) { ROS_WARN("t3 perceptionCallback parse"); if (perception_obstacles_.ParseFromString(msg.data)) { double timestamp = perception_obstacles_.header().timestamp_sec(); ROS_INFO("t3 perceptionCallback timestamp %f count:%d", timestamp, perception_obstacles_.perception_obstacle().size()); std::string data; perception_obstacles_.SerializeToString(&data); VehData veh_data; veh_data.set_messagetype(5); veh_data.set_messagedes("PerceptionObstacles"); veh_data.set_contents(data); std::string serialized_data; veh_data.SerializeToString(&serialized_data); } else { ROS_ERROR("t3 perceptionCallback parse fail!"); } } int main(int argc, char **argv) { ros::init(argc, argv, "listener"); ros::NodeHandle n; ros::Subscriber sub = n.subscribe("/perception_node/perception_objects", 1000, perceptionCallback); ros::spin(); return 0; }CMakeLists.txt：cmake_minimum_required(VERSION 3.0.2) project(t3) find_package(catkin REQUIRED COMPONENTS roscpp rospy pcl_ros std_msgs third_party ) find_package(Protobuf REQUIRED) include_directories(${Protobuf_INCLUDE_DIRS} ${CMAKE_CURRENT_BINARY_DIR}/..) find_package(Boost REQUIRED) include_directories(${Boost_INCLUDE_DIRS}) set(ixwebsocket_INCLUDE_DIR "/usr/local/include/ixwebsocket") set(ixwebsocket_LIBRARIES "/usr/local/lib/libixwebsocket.a") include_directories(${ixwebsocket_INCLUDE_DIR}) include_directories(${CATKIN_DEVEL_PREFIX}/${CATKIN_GLOBAL_INCLUDE_DESTINATION}/${PROJECT_NAME}) include_directories(${CATKIN_DEVEL_PREFIX}/${CATKIN_GLOBAL_INCLUDE_DESTINATION}/smartview) catkin_package(INCLUDE_DIRS ${PROJECT_INCLUDE_DIRS} DEPENDS ${GFLAGS_LIBRARIES} ) include_directories( ${catkin_INCLUDE_DIRS} ${PROTOBUF_INCLUDE_DIR} ${PROJECT_SOURCE_DIR}/.. ) add_executable(${PROJECT_NAME}_node src/main.cpp ) add_dependencies(${PROJECT_NAME}_node ${catkin_EXPORTED_TARGETS}) target_link_libraries(${PROJECT_NAME}_node ${catkin_LIBRARIES} ${PROTOBUF_LIBRARIES} smartview_proto ) install(TARGETS ${PROJECT_NAME}_node ARCHIVE DESTINATION ${CATKIN_PACKAGE_LIB_DESTINATION} LIBRARY DESTINATION ${CATKIN_PACKAGE_LIB_DESTINATION} RUNTIME DESTINATION ${CATKIN_GLOBAL_BIN_DESTINATION} )

116 浏览量
#include "third_party/apollo/proto/perception/perception_obstacle.pb.h" #include "t3_perception.pb.h" #include <ixwebsocket/IXWebSocket.h> using namespace ix; apollo::perception::...

def main(args, rest_args): cfg = Config(path=args.cfg) model = cfg.model model.eval() if args.quant_config: quant_config = get_qat_config(args.quant_config) cfg.model.build_slim_model(quant_config['quant_config']) if args.model is not None: load_pretrained_model(model, args.model) arg_dict = {} if not hasattr(model.export, 'arg_dict') else model.export.arg_dict args = parse_model_args(arg_dict) kwargs = {key[2:]: getattr(args, key[2:]) for key in arg_dict} model.export(args.save_dir, name=args.save_name, **kwargs) if args.export_for_apollo: if not isinstance(model, BaseDetectionModel): logger.error('Model {} does not support Apollo yet!'.format( model.class.name)) else: generate_apollo_deploy_file(cfg, args.save_dir) if name == 'main': args, rest_args = parse_normal_args() main(args, rest_args)这段代码中哪几句代码是def main(args, rest_args): cfg = Config(path=args.cfg) model = cfg.model model.eval() if args.quant_config: quant_config = get_qat_config(args.quant_config) cfg.model.build_slim_model(quant_config['quant_config']) if args.model is not None: load_pretrained_model(model, args.model) arg_dict = {} if not hasattr(model.export, 'arg_dict') else model.export.arg_dict args = parse_model_args(arg_dict) kwargs = {key[2:]: getattr(args, key[2:]) for key in arg_dict} model.export(args.save_dir, name=args.save_name, **kwargs) if args.export_for_apollo: if not isinstance(model, BaseDetectionModel): logger.error('Model {} does not support Apollo yet!'.format( model.class.name)) else: generate_apollo_deploy_file(cfg, args.save_dir) if name == 'main': args, rest_args = parse_normal_args() main(args, rest_args)这段代码中哪几句代码是def main(args, rest_args): cfg = Config(path=args.cfg) model = cfg.model model.eval() if args.quant_config: quant_config = get_qat_config(args.quant_config) cfg.model.build_slim_model(quant_config['quant_config']) if args.model is not None: load_pretrained_model(model, args.model) arg_dict = {} if not hasattr(model.export, 'arg_dict') else model.export.arg_dict args = parse_model_args(arg_dict) kwargs = {key[2:]: getattr(args, key[2:]) for key in arg_dict} model.export(args.save_dir, name=args.save_name, **kwargs) if args.export_for_apollo: if not isinstance(model, BaseDetectionModel): logger.error('Model {} does not support Apollo yet!'.format( model.class.name)) else: generate_apollo_deploy_file(cfg, args.save_dir) if name == 'main': args, rest_args = parse_normal_args() main(args, rest_args)这段代码中哪几句是将训练时保存的动态图模型文件导出成推理引擎能够加载的静态图模型文件

158 浏览量
这段代码中的以下几句是将训练时保存的动态图模型文件导出成推理引擎能够加载的静态图模型文件： ``` arg_dict = {} if not hasattr(model.export, 'arg_dict') else model.export.arg_dict args = parse_model_...

xmelab: *E,CUVMUR (/lsc/projects/apollo/users/jjiang1/verif/verif_apollo/verif_hardip/mppcie8l/src/project_benches/mppcie8l/tb/testbench/rtl_struct_define.sv,77|15): instance 'hdl_top.gfor_struct_eth_xgmii_cib_quad[0].i_gear_rx' of design unit 'eth_gmii_gear_rx_2x1' is unresolved in 'worklib.hdl_top:sv'.

393 浏览量
这个错误消息表明在 Verilog 设计中，位于 'hdl_top.gfor_struct_eth_xgmii_cib_quad[0].i_gear_rx' 实例的 'eth_gmii_gear_rx_2x1' 设计单元在 'worklib.hdl_top:sv' 中未解析。 要解决这个问题，您可以尝试以下几...

/apollo/bazel-bin WARNING: Logging before InitGoogleLogging() is written to STDERR E0607 13:23:56.281872 47757 lossless_map_creator.cc:162] num_trials = 1 Pcd folders are as follows: /apollo/hdmap/air/pcd_apollo/ Resolution: 0.125 Dataset: /apollo/hdmap/air/pcd_apollo/ Dataset: /apollo/hdmap/air/pcd_apollo/ Dataset: /apollo/hdmap/air/pcd_apollo/ Loaded the map configuration from: /apollo/hdmap/air//lossless_map/config.xml. Saved the map configuration to: /apollo/hdmap/air//lossless_map/config.xml. Saved the map configuration to: /apollo/hdmap/air//lossless_map/config.xml. E0607 13:23:56.706698 47757 lossless_map_creator.cc:264] ieout_poses = 2186 Failed to find match for field 'intensity'. Failed to find match for field 'timestamp'. E0607 13:23:56.708952 47757 velodyne_utility.cc:46] Un-organized-point-cloud E0607 13:23:56.719892 47757 lossless_map_creator.cc:275] Loaded 144353D Points at Trial: 0 Frame: 0. F0607 13:23:56.719911 47757 base_map_node_index.cc:101] Check failed: false *** Check failure stack trace: *** scripts/msf_create_lossless_map.sh: line 11: 47757 Aborted (core dumped) $APOLLO_BIN_PREFIX/modules/localization/msf/local_tool/map_creation/lossless_map_creator --use_plane_inliers_only true --pcd_folders $1 --pose_files $2 --map_folder $IN_FOLDER --zone_id $ZONE_ID --coordinate_type UTM --map_resolution_type single

162 浏览量
这段代码似乎是关于自动驾驶系统中的地图构建和定位相关的内容，其中使用了Velodyne激光雷达获取点云数据，并将其加载到地图中。但是在处理过程中出现了一些错误，如无法匹配点云数据中的“强度”和“时间戳”字段等...
疯狂的机器人
  • 粉丝: 9222
上传资源 快速赚钱

最新资源