Data acquisition system and method

By introducing domain controllers, switches and industrial control machines into the intelligent driving data acquisition system, the problems of high equipment costs, complex networking and difficult problem positioning in the existing systems are solved, and the data acquisition function is simplified and cost reduction is achieved.

CN119472449AInactive Publication Date: 2025-02-18ZHEJIANG GEELY HLDG GRP CO LTD +1
View PDF 7 Cites 0 Cited by

Patent Information

Application Number
CN202510017915.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-01-06
Publication Date
2025-02-18
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The current intelligent driving data acquisition system has problems such as high equipment costs, complex networking and difficult problem positioning.

Method used

A data acquisition system is designed, including a domain controller, a switch and an industrial control machine. The domain controller interface is connected to multiple sensors on the vehicle. The lower computer node receives and processes sensor data, the switch transmits data to the industrial control machine, and the industrial control machine stores it. Processed data. The system reduces hardware costs and maintenance complexity through time synchronization and data compression processing.

Benefits of technology

It has achieved simplification of data acquisition functions, reduced equipment costs and maintenance complexity, and improved the efficiency of problem location and troubleshooting.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119472449A_ABST
    Figure CN119472449A_ABST
Patent Text Reader

Abstract

The invention provides a data acquisition system and method, the data acquisition system comprises a domain controller, a switch and an industrial personal computer, an interface of the domain controller is connected with a multi-path sensor on a vehicle, the domain controller comprises a lower computer node, and sensor data of the multi-path sensor is received and processed through the lower computer node. The sensor data is transmitted to an industrial personal computer through a switch, the industrial personal computer comprises an upper computer node which is in communication connection with the lower computer node, and the upper computer node receives and stores the processed sensor data transmitted by the domain controller and stores the processed sensor data. Furthermore, the data acquisition function of the acquisition end can be realized only by adding an industrial personal computer and a network switch, so that the equipment cost is greatly reduced, and the acquisition equipment is simple to maintain.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of data acquisition technology, and in particular to a data acquisition system and method. Background Art

[0002] The automotive industry is currently engaged in a race for intelligent driving technology, and companies are increasing their investment in research and development. The core challenge lies in improving the accuracy of perception algorithms, which requires the use of a large amount of labeled data to train and generalize the models to adapt to a wider range of applications.

[0003] However, obtaining high-quality data suitable for training is a key challenge. Current intelligent driving data collection mainly relies on vehicles equipped with dedicated acquisition boards, which are connected to on-board sensors through the data acquisition devices on board. These devices are responsible for collecting the original signals from the sensors and transmitting them to the server, which parses these signals to achieve synchronization and format conversion of sensor data. However, this solution has shortcomings. The complex networking of the acquisition system requires each data acquisition vehicle to be equipped with an acquisition device and a back-end server, which increases equipment costs and R&D costs. In addition, the acquisition system composed of many components has a relatively complex networking. Once a problem occurs, it is difficult to locate and troubleshoot the problem.

[0004] To this end, a new intelligent driving car data collection system was designed to address the technical difficulties in collecting these intelligent driving data. Summary of the invention

[0005] In order to overcome the problems existing in the related art, this specification provides a data acquisition system and method.

[0006] According to a first aspect of an embodiment of this specification, a data acquisition system is provided, the system comprising: A domain controller, wherein an interface of the domain controller is connected to a multi-channel sensor on a vehicle, and the domain controller includes a lower computer node for receiving and processing sensor data of the multi-channel sensor through the lower computer node; An industrial computer, the industrial computer comprising an upper computer node communicatively connected to the lower computer node, configured to receive and store the processed sensor data transmitted by the domain controller through the upper computer node, and store the processed sensor data; and The switch is used to establish a communication connection between the lower computer node and the upper computer node.

[0007] According to a data acquisition system provided by the present application, the system further includes an additional sensor installed on the vehicle; the additional sensor is connected to the switch, and the switch is used to send data of the additional sensor to the industrial computer; The industrial computer includes a data recording node, which is used to receive the data of the installed sensor sent by the switch and store the data of the installed sensor.

[0008] According to a data acquisition system provided by the present application, the domain controller includes a microcontroller unit for performing time synchronization between the additional sensor and the existing multi-channel sensors of the vehicle.

[0009] According to a data acquisition system provided by the present application, the microcontroller unit is used to send a time synchronization signal with a calibration timestamp to the additional sensor and the multi-channel sensor, so that the additional sensor and the multi-channel sensor automatically calibrate the internal clock according to the calibration timestamp to output data with a unified timestamp.

[0010] According to a data acquisition system provided by the present application, the lower computer node is also used to synchronize the sensor data in the time dimension to obtain synchronized data; and compress the synchronized data to obtain processed sensor data.

[0011] According to a data acquisition system provided by the present application, the lower computer node is used to store the sensor data in corresponding queues according to the data source of the sensor data; the data source and the queue have a one-to-one correspondence; the timestamps of the sensor data in the various queues are traversed; and the sensor data whose timestamp difference meets a threshold range is determined as synchronized data.

[0012] According to a data acquisition system provided by the present application, the host computer node is used to continue traversing the timestamps of the unsynchronized data in the queue until the amount of data in the queue is greater than the queue length, and then clear the data in the queue.

[0013] According to a data acquisition system provided by the present application, the system also includes a vehicle-mounted Ethernet converter, and the vehicle-mounted Ethernet converter establishes a two-way communication connection with the switch and the domain controller respectively.

[0014] According to a second aspect of an embodiment of this specification, a data acquisition method is provided, which is applied to any of the above data acquisition systems, and the method includes: Receiving sensor data collected by multiple sensors on the vehicle through the domain controller; Processing the sensor data by the domain controller; Transmitting the processed sensor data to the industrial computer through the switch; The processed sensor data is received and stored by the industrial computer.

[0015] According to a data collection method provided by the present application, the system further includes an additional sensor installed on the vehicle, wherein the additional sensor is connected to the switch. The method further comprises: Transmitting the data of the added sensor to the industrial computer through the switch; The data of the installed sensor sent by the switch is received through the industrial computer, and the data of the installed sensor is stored.

[0016] According to a data collection method provided by the present application, the domain controller includes a microcontroller unit, Before receiving the sensor data collected by the multi-channel sensors on the vehicle through the domain controller, the method further includes: The microcontroller unit is used to perform time synchronization between the added sensor and the existing multi-channel sensors of the vehicle.

[0017] According to a data acquisition method provided by the present application, the time synchronization of the additional sensor and the existing multi-channel sensors of the vehicle is performed by the microcontroller unit, including: The microcontroller unit sends a time synchronization signal with a calibration timestamp to the additional sensor and the multi-channel sensor, so that the additional sensor and the multi-channel sensor automatically calibrate the internal clock according to the calibration timestamp to output data with a unified timestamp.

[0018] According to a data collection method provided by the present application, the domain controller includes a lower node, The processing of the sensor data by the domain controller includes: In the time dimension, the sensor data is synchronized through the lower node to obtain synchronized data; The synchronized data is compressed to obtain processed sensor data.

[0019] According to a data collection method provided by the present application, in the time dimension, the sensor data is synchronized by the lower node to obtain synchronized data, including: According to the data source of the sensor data, the sensor data is stored in corresponding queues; the data source and the queue have a one-to-one correspondence; Traversing the timestamps of the sensor data in each queue; The sensor data whose timestamp difference meets the threshold range is determined as synchronized data.

[0020] According to a data collection method provided by the present application, the method further includes: The timestamps of the unsynchronized data in the queue are continuously traversed until the amount of data in the queue is greater than the queue length, and then the data in the queue is cleared.

[0021] The data acquisition system and method in the embodiments of this specification, compared with the current data acquisition networking which is relatively complex, has high hardware costs and is difficult to locate and troubleshoot once a problem occurs, the data acquisition system includes a domain controller, a switch and an industrial computer. The interface of the domain controller is connected to the multi-channel sensors on the vehicle. The domain controller includes a lower computer node, and the sensor data of the multi-channel sensors is received and processed through the lower computer node. The sensor data is transmitted to the industrial computer through the switch. The industrial computer includes an upper computer node that is communicatively connected to the lower computer node, and receives and stores the processed sensor data transmitted by the domain controller through the upper computer node, and stores the processed sensor data. Furthermore, the acquisition end only needs to add an industrial computer and a network switch to realize the data acquisition function, which greatly reduces the equipment cost, and the acquisition equipment is simple to maintain.

[0022] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present specification. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the specification and, together with the description, serve to explain the principles of the specification.

[0024] Figure 1 is an architecture diagram of a data acquisition system according to an exemplary embodiment of this specification; Figure 2 is a flowchart of a data collection method according to an exemplary embodiment of the present specification; Figure 3 It is a data acquisition link of a data acquisition system shown in this specification according to an exemplary embodiment. DETAILED DESCRIPTION

[0025] Here, the technical solutions in the embodiments (or "implementation methods") of the present application will be described clearly and completely in conjunction with the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements.

[0026] If there are terms involving directional indications or positional relationships in the embodiments of the present application (such as up, down, left, right, front, back, inside, outside, top, bottom, center, vertical, horizontal, longitudinal, transverse, length, width, counterclockwise, clockwise, axial, radial, circumferential, etc.), such terms are only used to explain the relative positional relationship, movement, etc. between the components under a certain specific posture (as shown in the accompanying drawings); if the specific posture changes, the directional indication or positional relationship will also change accordingly. In addition, the terms "first" and "second" in the embodiments of the present application are only used for the purpose of convenience of description and cannot be understood as indicating or implying relative importance.

[0027] The present application provides a data acquisition system and method. The present application is described in detail below in conjunction with the accompanying drawings. In the absence of conflict, the features of the following embodiments and implementations can be combined with each other.

[0028] At present, intelligent driving data collection mainly relies on vehicles equipped with dedicated acquisition boards, which are connected to on-board sensors through the data acquisition devices on board. In use, these devices are responsible for collecting the original signals of the sensors and transmitting them to the server, which analyzes these signals to achieve synchronization and format conversion of sensor data. However, this solution has several problems: (1) Expensive equipment costs: Each data collection vehicle needs to be equipped with a data collection device and a back-end server, which increases R&D costs.

[0029] (2) Network complexity: The acquisition system is composed of many components, including acquisition boxes, inertial navigation equipment, servers, time synchronizers, NASA storage boxes, etc. This makes the network more complex. Once a problem occurs, it is difficult to locate and troubleshoot the problem.

[0030] In order to solve the above technical problems, this specification provides a smart driving car data acquisition system.

[0031] The data acquisition system includes a domain controller, a switch and an industrial computer. Its system architecture is simple. The acquisition function can be realized by adding an industrial computer and a vehicle-mounted switch, which greatly reduces the equipment cost and makes the acquisition equipment easy to maintain.

[0032] As an example, see Figure 1 , Figure 1 is an architecture diagram of a data acquisition system provided in an embodiment of this specification.

[0033] A domain controller, wherein an interface of the domain controller is connected to a multi-channel sensor on a vehicle, and the domain controller comprises a lower computer node for receiving and processing sensor data of the multi-channel sensor through the lower computer node.

[0034] The multi-channel sensors on the vehicle include cameras, lidar, positioning sensors and other sensors related to intelligent driving. Usually, the multi-channel sensors that are already in the vehicle when it leaves the factory are connected to the domain controller, which receives the sensor data collected by the multi-channel sensors and processes the sensor data, such as compression processing and frame extraction processing, to reduce the pressure of data collection and storage.

[0035] In other embodiments, a lower computer node is established in the vehicle-mounted domain controller segment to receive sensor data from multiple sensors and process the sensor data, for example, compressing camera images in real time and collecting compressed image data to reduce storage costs.

[0036] An industrial computer, the industrial computer includes an upper computer node that is communicatively connected to the lower computer node, and is used to receive and store the processed sensor data transmitted by the domain controller through the upper computer node, and store the processed sensor data.

[0037] A host computer node is established on the industrial computer side, and a communication connection is established between the host computer node and the slave computer node. The sensor data transmitted by the slave computer node is received through the host computer node and saved in real time to ensure the reliability of the collected data.

[0038] The switch is used to establish a communication connection between the lower computer node and the upper computer node. It can be, but not limited to, TCP, Ethernet connection, etc. To a certain extent, a suitable communication connection standard can be selected according to actual needs, such as whether the communication bandwidth meets the number of sensors, the reliability of the communication connection, etc.

[0039] In this embodiment, the data acquisition system of the present application directly collects sensor data at the domain controller end. The acquisition end only needs to add an industrial computer and a network switch to realize the data acquisition function. Compared with the current board acquisition system equipment, the hardware cost is greatly reduced. In addition, the acquisition equipment is easy to maintain. If there is a problem with data transmission, you can first check the connection between the upper computer and the lower computer, instead of blindly searching in the entire complex system (including all devices such as domain controllers, switches and industrial computers). At the same time, this architecture also facilitates the maintenance and upgrade of the system. When the software or part of the hardware of the upper and lower nodes is updated, it will not cause too much interference to the entire system, because their functions are relatively independent and have a clear division of labor.

[0040] In some embodiments, the system further includes an additional sensor installed on the vehicle. The additional sensor described herein refers to a sensor added according to actual acquisition requirements, which may be, but is not limited to, a 360 laser radar, a blind laser radar, etc.

[0041] Continue to refer to Figure 1The installed sensor is connected to the switch, and the switch is used to send the data of the installed sensor to the industrial computer; the industrial computer includes a data recording node, which is used to receive the data of the installed sensor sent by the switch and store the data of the installed sensor.

[0042] In this embodiment, the switch can provide better network scalability and flexibility, can be equipped with multiple sensors, and manage data flow to avoid data conflicts.

[0043] In some embodiments, the microcontroller unit is used to send a time synchronization signal with a calibration timestamp to the added sensor and the multi-channel sensor, so that the added sensor and the multi-channel sensor automatically calibrate the internal clock according to the calibration timestamp to output data with a unified timestamp.

[0044] The data acquisition system includes data acquisition of multiple sensors. The data collected by each sensor will be fused. During the data fusion process, the timestamp of the sensor data shall be used as a benchmark. Therefore, it is very important to synchronize the time between the sensors.

[0045] Continue to refer to Figure 1 The domain controller MCU (Micro-Controller Unit) is used as the main clock source to achieve time synchronization between the original vehicle's multi-channel sensors and the added sensors. The time synchronization process will be described in detail in the following chapters.

[0046] In this embodiment, the output data of multiple sensors are provided with a unified and accurate timestamp, ensuring that the data collected by each sensor can be seamlessly integrated.

[0047] In some embodiments, the lower computer node is further used to synchronize the sensor data in the time dimension to obtain synchronized data; and to compress the synchronized data to obtain processed sensor data.

[0048] While the lower computer node is established on the domain controller to receive the sensor data, the lower computer node processes the sensor data, which may be, but is not limited to, compression of the sensor data, real-time frame extraction, data synchronization, etc. The synchronized data with a unified timestamp is processed and transmitted to the industrial computer for subsequent use.

[0049] As an example, the lower computer node is used to store the sensor data in corresponding queues according to the data source of the sensor data; the data source and the queue have a one-to-one correspondence; the timestamps of the sensor data in the various queues are traversed; and the sensor data whose timestamp difference meets the threshold range is determined as synchronized data.

[0050] As another example, the host computer node is used to continue traversing the timestamps of the unsynchronized data in the queue until the amount of data in the queue is greater than the queue length, and then clear the data in the queue.

[0051] In this embodiment, the domain controller only transmits data that meets the labeling requirements to the outside world to reduce the pressure on data transmission bandwidth and reduce data storage costs. There is no need to align the data of each camera in the post-processing stage of data acquisition, saving the time cost of data post-processing.

[0052] In some embodiments, continue to refer to Figure 1 The system also includes a vehicle-mounted Ethernet converter, which establishes a two-way communication connection with the switch and the domain controller respectively.

[0053] The Ethernet converter described in this article is mainly used for protocol conversion and interface matching to ensure that data of different protocols can be correctly transmitted between the two.

[0054] Based on the above data acquisition system, a data acquisition method is proposed.

[0055] Figure 2 This is a flow chart of a data collection method provided by an embodiment of this specification, comprising the following steps: Step 100: receiving sensor data collected by multiple sensors on the vehicle through the domain controller.

[0056] The data link of the data acquisition system is divided into the domain controller end data link and the industrial computer end data link. The data link and flow of the entire data acquisition system are as follows: Figure 3 As shown, Figure 3 It is the data acquisition link of the data acquisition system. The data link of this data acquisition system is divided into 4 categories: PTP time synchronization message, camera image data message, vehicle positioning data message, and lidar data message.

[0057] It should be noted that the multi-channel sensors on the vehicle include cameras, lidars, positioning sensors and other sensors related to intelligent driving, and the data of these sensors are received through the lower computer nodes of the domain controller.

[0058] As an example, see Figure 1 and Figure 3 The data collected by the data acquisition system includes the sensor data of the original vehicle (including the camera data and positioning data of the original vehicle) and the data of the added sensors (such as the added laser radar).

[0059] The original car's built-in camera data is generated by the on-board camera, which includes 10 cameras: front view, rear view, left front, left rear, right front, right rear, front surround, rear surround, left surround, and right surround. Figure 1 , the lower computer on the SOCA side collects data from the rear view, left front, left rear, right front, and right rear 5 cameras.

[0060] Therefore, the original vehicle's built-in camera data includes the front-view camera output data, the rear-view camera output data, the surround-view camera data (left front-view camera data, left rear-view camera data, right front-view camera data, right rear-view camera data), and the surround-view camera data (left surround-view camera data, right surround-view camera data, front surround-view camera data, and rear surround-view camera data). The original vehicle's camera data are all output by the domain controller.

[0061] The original vehicle comes with positioning data, including inertial navigation output data (imu) and satellite navigation positioning data (gnss), both of which are output by the original vehicle sensors.

[0062] After receiving the above data, the lower computer node compresses the data and forwards it to the upper computer node on the industrial computer side through the TCP protocol, so that after the upper computer receives the image message, it can be stored in the solid state drive in real time through the dump node.

[0063] As for the data of the installed sensors, the data of the installed sensors are received by the industrial computer.

[0064] As an example, the system further includes an additional sensor installed on the vehicle, and the additional sensor is connected to the switch.

[0065] The data collection method further comprises: Step 110, transmitting the data of the added sensor to the industrial computer through the switch; Step 120: receiving the data of the installed sensor sent by the switch through the industrial computer, and storing the data of the installed sensor.

[0066] Since the additional sensors are connected to the switch, the additional sensors (including 360° lidar data and blind lidar data) are connected to the industrial computer through the switch, and the data of the additional sensors are received by the industrial computer.

[0067] In some embodiments, the data acquisition system includes data acquisition of multiple vehicle-mounted sensors. The data collected by each sensor will be fused. During the data fusion process, the timestamp of the sensor data is used as a reference. Therefore, it is very important to synchronize the time between the sensors.

[0068] As an example, the domain controller includes a microcontroller unit, and before receiving the sensor data collected by the multi-channel sensors on the vehicle through the domain controller, the method further includes: Step a11, time synchronization of the additional sensor and the existing multi-channel sensors of the vehicle is performed through the microcontroller unit.

[0069] Continue to refer to Figure 3 , with the MCU of the ADCU (Automated Driving Control Unit) domain controller as the main clock source, it will send out PTP time synchronization messages, which are used for the time synchronization of various vehicle-mounted sensors to ensure the synchronization of the sensor's own clock source with the MCU clock source. After each sensor completes the time synchronization, the clock source of the entire system is exactly the same, so the timestamps carried by the data output by each sensor can be synchronized.

[0070] It should be noted that for the additional sensors, the MCU is the main clock source and will send a PTP time synchronization message to the additional sensors through the switch to synchronize the time of the additional sensors with the vehicle's existing multi-channel sensors.

[0071] As an example, the time synchronization of the additional sensor and the existing multi-channel sensors of the vehicle by the microcontroller unit includes: Step a12, sending a time synchronization signal with a calibration timestamp to the added sensor and the multi-channel sensor through the microcontroller unit, so that the added sensor and the multi-channel sensor automatically calibrate the internal clock according to the calibration timestamp to output data with a unified timestamp.

[0072] Continue to refer to Figure 3 The PTP time synchronization message is broadcasted by the MCU of ADCU, and forwarded to each camera end through SOCA and SOCB on the ADCU side; the positioning sensor can directly receive the PTP time synchronization message sent by MCU. At the same time, the PTP time synchronization message is forwarded to each sensor end through the switch.

[0073] After receiving the PTP synchronization signal from the master clock source, each sensor will automatically calibrate its internal clock so that its output data has a unified and accurate timestamp, ensuring that the sensor's own time source is synchronized with the MCU master time source, thereby ensuring data consistency in this acquisition system.

[0074] This time calibration process not only eliminates data inconsistencies caused by clock bias, but also provides a common time baseline for the integrated analysis of multi-source data, which is of great significance for the fusion of laser point cloud data with camera data. Ultimately, through this carefully designed time synchronization mechanism, the data acquisition system of this application can ensure that even in high-speed mobile environments, various sensors - from optical cameras to lidars to GPS locators - can provide time-consistent data, ensuring that the collected multi-sensor data is accurately fused.

[0075] Step 200: Processing the sensor data by the domain controller.

[0076] Since the raw sensor data collected by the current data collection system is not processed, a large amount of storage space is required, which increases the storage cost and makes the data storage cost high. In order to solve this problem, the present application pre-processes the sensor data on the domain controller side, collects the processed data, and reduces the storage cost.

[0077] In some embodiments, after receiving the sensor data, the lower computer node in the domain controller is difficult to transmit to the outside of the domain controller due to the large amount of data such as cameras and laser radars. Therefore, after the lower computer of the data acquisition system obtains the sensor data on the domain controller side, it compresses and extracts frames to reduce the data size, and then transmits it to the upper computer node via Ethernet for storage, thereby reducing storage pressure and cost.

[0078] As an example, compared with the data of positioning sensors and lidar data, the data of camera data is larger, taking camera images as an example. Camera image compression is implemented on the vehicle-mounted domain controller, and frame extraction is performed. After the compressed image is forwarded to the host computer node on the industrial computer side through the TCP protocol, after the host computer receives the image message, it is stored in the solid-state drive in real time through the dump node, which greatly reduces the pressure of data collection and storage. It should be noted that according to the data processing requirements of the domain controller and the industrial computer, a customized compression ratio is set. For example, when the compression ratio reaches 20:1, the data transmission efficiency is the highest.

[0079] In other embodiments, the data collected by the data acquisition system is used for subsequent assisted driving, and the data needs to be sorted before use to select valid data. In addition, in order to ensure the accuracy and effectiveness of the data in subsequent processing and application, it is necessary to synchronize and align the data from different sensors, synchronize and align the sensor data with the timestamp through the lower node, and only transmit sensor data that meets the labeling requirements to the outside, avoiding the transmission of a large amount of unsynchronized and possibly chaotic data, thereby reducing the amount of data transmission and effectively alleviating the pressure on the data transmission bandwidth.

[0080] As an example, the domain controller includes a lower node. The processing of the sensor data by the domain controller includes: Step 210, in the time dimension, synchronizing the sensor data through the lower node to obtain synchronized data; Step 220: compress the synchronized data to obtain processed sensor data.

[0081] The data synchronization described in this article refers to obtaining the data of each sensor with the same timestamp, ensuring that the data of all sensors are close enough in time to meet the needs of subsequent applications. It can be understood that all camera data are closely aligned in time, which can provide complete and synchronized multi-view data for subsequent complex applications. In addition, there is no need to align the sensor data in the post-processing stage of data acquisition, because the synchronization operation has been completed at the previous lower node end, so the time cost of aligning the data again in the subsequent processing process is saved, which improves the overall data processing efficiency.

[0082] The synchronized data is compressed and frame-extracted to obtain processed sensor data, which reduces the amount of data transmitted by the acquisition system. The processed sensor data is sent to the industrial computer and can be used directly to reduce the post-processing link of data acquisition.

[0083] As an example, step 210 may specifically include the following steps: Step b11, storing the sensor data in corresponding queues according to the data source of the sensor data; the data source and the queue have a one-to-one correspondence.

[0084] Create multiple independent queues for storing sensor data. The number of queues should be consistent with the number of sensors. For example, for camera data, if there are 10 cameras, set up 10 independent queues accordingly.

[0085] This design can make the data management of each sensor more orderly, facilitate the subsequent separate processing of each sensor data, and avoid confusion between different sensor data. For example, in a vehicle monitoring system, the data collected by cameras in different positions (such as front view, rear view, side view, etc.) have different uses and importance. Through separate queues, these data can be processed in a targeted manner.

[0086] Furthermore, considering the timeliness of data and memory management, the length of the queue is set. This length limits the amount of data stored in each queue, ensuring that the system does not occupy too much memory resources due to storing too much historical data. At the same time, during the data synchronization process, this length also provides a suitable range for timestamp comparison. A shorter queue length allows the system to process data faster and avoid processing long time series data, thereby improving synchronization efficiency. For example, 10 queues are created to store each camera data separately, and the queue length is 5.

[0087] There is a one-to-one correspondence between the data source and the queue. According to the source of the sensor data (such as the path of the sensor), the data is placed in the corresponding queue, and the sensor data of each channel is classified and stored in order. For example, if the data of camera 1 arrives, the system will recognize that it is the data from the first camera and put it into the first queue. This ensures that the data of each camera is stored in its respective queue in the order of receipt, providing an ordered data set for subsequent timestamp comparison.

[0088] Furthermore, by storing data in this way, the integrity of each camera's data can be easily checked. If the data from a camera does not arrive within a period of time, no new data will be added to the corresponding queue, which can help the system promptly detect problems in the data transmission process, such as camera failure, network interruption, etc.

[0089] Step b12, traversing the timestamps of the sensor data in each queue.

[0090] The timestamps of the sensor data in each queue are traversed at the same time. This simultaneous traversal method ensures that all camera data are compared at the same time, improving the accuracy of synchronization judgment. For example, the system starts from the head of each queue (the earliest stored data) and checks the timestamp of each data in turn to determine the time relationship between the camera data.

[0091] Step b13: Determine the sensor data whose timestamp difference satisfies the threshold range as synchronized data.

[0092] When the timestamp of each sensor data meets the threshold range, the corresponding sensor data is determined to be synchronized data. For example, the timestamp difference between 10 camera data is less than 5ms as the standard for data synchronization. The 5ms time difference can ensure that the images taken by each camera are close enough in time to meet the needs of subsequent applications. However, if the time difference is too large, the images taken by different cameras will be obviously inconsistent in time, affecting the accuracy of these applications. Therefore, it is crucial to set the threshold range reasonably.

[0093] When it is determined that the sensor data synchronization is completed, a synchronization mark will be set. After the lower computer node obtains this synchronization mark, it sends the synchronized data to the upper computer node.

[0094] When the timestamp difference of the sensor data of each channel does not meet the threshold range, the corresponding sensor data will not be uploaded to the host computer node, and these data will be temporarily stored in the queue for the next traversal.

[0095] As an example, the timestamps of the unsynchronized data in the queue are continuously traversed until the amount of data in the queue is greater than the queue length, and then the data in the queue is cleared.

[0096] When the amount of data in the queue reaches the set queue length, the queue needs to be emptied. This is to ensure that the queue can continuously store new data and avoid data loss due to the queue being full and unable to receive new data. After the queue is emptied, the system can start storing new data again and continue the data synchronization process.

[0097] This method of clearing the queue and continuing synchronization forms a cyclic data processing mechanism. By continuously receiving, storing, synchronizing, and sending camera data, the system can continuously provide the host computer with synchronized sensor data to ensure the normal operation of the entire system. For example, during the operation of the vehicle, the camera continuously collects data. According to this cyclic mechanism, the system can send the synchronized camera data to the host computer in real time for applications such as vehicle status monitoring.

[0098] Through the above embodiment, the lower computer node is used to synchronize the sensor data messages of each channel, so that the data acquisition system only acquires sensor data with aligned timestamps, reduces the amount of data transmitted by the acquisition system, and only transmits sensor data that meets the labeling requirements to the outside, reduces the pressure on data transmission bandwidth, and reduces data storage costs. At the same time, the collected sensor data can be directly used for labeling without post-processing operations, saving the time cost of data post-processing.

[0099] In other embodiments, for the data of the installed sensors, the data of the installed sensors are transmitted to the industrial computer through the switch, and the industrial computer performs data synchronization processing on the received data based on the timestamp of the data of the installed sensors and the timestamp of the synchronized data or the processed sensor data, and subsequent data applications provide valid data.

[0100] Step 300: transmitting the processed sensor data to the industrial computer through the switch.

[0101] Reference Figure 1The switch of the data acquisition system plays the role of a data transmission transfer station. The switch is connected to the 5-way laser radar, the vehicle Ethernet sensor and the industrial computer through the network cable. The vehicle Ethernet sensor is connected to the domain controller and the switch respectively. Therefore, the messages of the data acquisition system will be transferred through the switch and sent to the destination (such as the industrial computer).

[0102] In some embodiments, the system further includes a vehicle-mounted Ethernet converter, and the vehicle-mounted Ethernet converter establishes a two-way communication connection with the switch and the domain controller respectively.

[0103] The Ethernet converter is mainly used for protocol conversion and interface matching to ensure that data of different protocols can be correctly transmitted between the two. Therefore, the data of the sensor connected to the domain controller is sent from the domain controller, reaches the switch via the vehicle Ethernet converter, and is then forwarded by the switch to the industrial computer. The PTP time synchronization message is also forwarded to the lidar and industrial computer via the switch.

[0104] Step 400: receiving and storing the processed sensor data via the industrial computer.

[0105] After the industrial computer receives the processed sensor data and the data of the added sensor, it stores them in real time in the solid-state drive through the dump node to ensure the reliability of the collected data.

[0106] As an example, there are 5 laser radars in the data acquisition system, which are installed in the front (front laser radar), rear (rear laser radar), left (left laser radar), right (right laser radar) and roof (main laser radar). The laser radar data is sent to the industrial computer side through the switch via UDP messages. The laser radar data recording node on the industrial computer side receives the laser radar message, converts it and dumps it to the solid-state hard drive.

[0107] The present application provides a data acquisition system and method. Compared with the current data acquisition network which is relatively complex, has high hardware cost and is difficult to locate and troubleshoot once a problem occurs, the data acquisition system includes a domain controller, a switch and an industrial computer. The interface of the domain controller is connected to the multi-channel sensor on the vehicle. The domain controller includes a lower computer node, and the sensor data of the multi-channel sensor is received and processed by the lower computer node. The sensor data is transmitted to the industrial computer through the switch. The industrial computer includes an upper computer node that is communicatively connected to the lower computer node, and receives and stores the processed sensor data transmitted by the domain controller through the upper computer node, and stores the processed sensor data. Furthermore, the acquisition end only needs to add an industrial computer and a network switch to realize the data acquisition function, which greatly reduces the equipment cost, and the acquisition equipment is simple to maintain.

[0108] It should be noted that the technical solutions or technical features described in the above embodiments can be combined or supplemented with each other without causing conflicts. The scope of protection of this application is not limited to the precise structures described in the above embodiments and shown in the drawings; all modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of this application should be included in the scope of protection of this application.

Claims

1. A data acquisition system, characterized in that: The system comprises: A domain controller, wherein an interface of the domain controller is connected to a multi-channel sensor on a vehicle, and the domain controller includes a lower computer node for receiving and processing sensor data of the multi-channel sensor through the lower computer node; An industrial computer, the industrial computer comprising an upper computer node communicatively connected to the lower computer node, configured to receive and store the processed sensor data transmitted by the domain controller through the upper computer node, and store the processed sensor data; and The switch is used to establish a communication connection between the lower computer node and the upper computer node.

2. The data acquisition system according to claim 1, characterized in that: The system further comprises an additional sensor installed on the vehicle; the additional sensor is connected to the switch, and the switch is used to send data of the additional sensor to the industrial computer; The industrial computer includes a data recording node, which is used to receive the data of the installed sensor sent by the switch and store the data of the installed sensor.

3. The data acquisition system according to claim 2, characterized in that: The domain controller comprises a microcontroller unit for performing time synchronization between the additional sensor and the existing multi-channel sensors of the vehicle.

4. The data acquisition system according to claim 3, characterized in that: The microcontroller unit is used to send a time synchronization signal with a calibration timestamp to the added sensor and the multi-channel sensor, so that the added sensor and the multi-channel sensor automatically calibrate the internal clock according to the calibration timestamp to output data with a unified timestamp.

5. The data acquisition system according to claim 1, characterized in that: The lower computer node is also used to synchronize the sensor data in the time dimension to obtain synchronized data; and to compress the synchronized data to obtain processed sensor data.

6. The data acquisition system according to claim 5, characterized in that: The lower computer node is used to store the sensor data in corresponding queues according to the data source of the sensor data; the data source and the queue have a one-to-one correspondence; the timestamps of the sensor data in the various queues are traversed; and the sensor data whose timestamp difference meets the threshold range is determined as synchronized data.

7. The data acquisition system according to claim 6, characterized in that: The host computer node is used to continue traversing the timestamps of the unsynchronized data in the queue until the amount of data in the queue is greater than the queue length, and then clear the data in the queue.

8. The data acquisition system according to claim 1, characterized in that: The system also includes a vehicle-mounted Ethernet converter, which establishes a two-way communication connection with the switch and the domain controller respectively.

9. A data acquisition method, applied to any data acquisition system in claims 1-8, characterized in that: Receiving sensor data collected by multiple sensors on the vehicle through the domain controller; Processing the sensor data by the domain controller; Transmitting the processed sensor data to the industrial computer through the switch; The processed sensor data is received and stored by the industrial computer.

10. The data collection method according to claim 9, characterized in that: The system further comprises an additional sensor installed on the vehicle, wherein the additional sensor is connected to the switch. The method further comprises: Transmitting the data of the added sensor to the industrial computer through the switch; The data of the installed sensor sent by the switch is received through the industrial computer, and the data of the installed sensor is stored.

11. The data collection method according to claim 9, characterized in that: The domain controller comprises a microcontroller unit, Before receiving the sensor data collected by the multi-channel sensors on the vehicle through the domain controller, the method further includes: The microcontroller unit is used to perform time synchronization between the added sensor and the existing multi-channel sensors of the vehicle.

12. The data collection method according to claim 11, characterized in that: The time synchronization of the additional sensor and the existing multi-channel sensors of the vehicle by the microcontroller unit includes: The microcontroller unit sends a time synchronization signal with a calibration timestamp to the additional sensor and the multi-channel sensor, so that the additional sensor and the multi-channel sensor automatically calibrate the internal clock according to the calibration timestamp to output data with a unified timestamp.

13. The data collection method according to claim 9, characterized in that: The domain controller includes a lower node, The processing of the sensor data by the domain controller includes: In the time dimension, the sensor data is synchronized through the lower node to obtain synchronized data; The synchronized data is compressed to obtain processed sensor data.

14. The data collection method according to claim 13, characterized in that: In the time dimension, synchronizing the sensor data through the lower node to obtain synchronized data includes: According to the data source of the sensor data, the sensor data is stored in corresponding queues; the data source and the queue have a one-to-one correspondence; Traversing the timestamps of the sensor data in each queue; The sensor data whose timestamp difference meets the threshold range is determined as synchronized data.

15. The data collection method according to claim 14, characterized in that: The method further comprises: The timestamps of the unsynchronized data in the queue are continuously traversed until the amount of data in the queue is greater than the queue length, and then the data in the queue is cleared.

Citation Information

Patent Citations

  • Distributed domain controller system based on network architecture of automatic driving system

    CN109917765A

  • Multi-source data synchronous processing method, device and system

    CN114172610A

  • Multi-terminal forwarding equipment, automatic driving system and automatic driving vehicle

    CN116668472A

  • Vehicle sensor data processing method, application method, electronic equipment and medium

    CN117573737A

  • Time synchronization system for CAN-TSN heterogeneous network

    CN118101114A