Vehicle-road time sequence data acquisition method and system based on distributed architecture

Abstract

The invention relates to the technical field of vehicle-road cooperation, in particular to a vehicle-road time sequence data acquisition method and system based on a distributed architecture, and the method comprises the steps: respectively carrying out the time service of a roadside sensing subsystem and a data acquisition vehicle according to a received satellite time service signal, obtaining a track point sequence of a collection task, and carrying out the time service of the track point sequence; according to the task planning result, controlling the data acquisition vehicle to perform state switching, and when the data acquisition vehicle is in an uploading state, transmitting unuploaded data to a cloud server; the method comprises the following steps: acquiring and analyzing the state and driving track data of a data acquisition vehicle in real time, judging whether a current roadside sensing subsystem needs to perform data acquisition, controlling the roadside sensing subsystem to perform state switching based on a received judgment result, transmitting the acquired data which is not uploaded to a cloud server when the roadside sensing subsystem is in an uploading state, and transmitting the acquired data to the cloud server when the roadside sensing subsystem is in the uploading state. Data recording of roadside equipment is dynamically triggered, redundancy is reduced, a cross-intersection collaborative data acquisition mechanism is realized, and data acquisition cost is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle-road cooperation, and particularly to a vehicle-road simultaneous time-series data acquisition method and system based on a distributed architecture. Background Art

[0002] In a vehicle-road cooperation system, the synchronous acquisition of time-series data of vehicles and roadside devices (such as cameras, lidar, etc.) is a key basis for realizing vehicle-road cooperation perception and conducting end-to-end model training. In the prior art, the following problems mainly exist in data acquisition: 1. Redundant acquisition: Roadside devices continuously record data, resulting in a large amount of redundant data storage and transmission pressure. However, most of the recorded data is invalid data, wasting cloud storage and transmission network resources; 2. Low acquisition efficiency: When the vehicle end and the road end simultaneously acquire data, it is mainly through communication between testers via mobile phones or walkie-talkies. When the vehicle continuously passes through multiple intersections, it is necessary to manually start the data recording program and manually switch the data recording programs at different intersections. There is a lack of an associated cooperation mechanism between intersections, resulting in low efficiency of vehicle-road simultaneous time-series data acquisition and high labor costs; 3. Centralized storage bottleneck: Traditional solutions rely on a single data center to centrally store roadside data, resulting in high network bandwidth pressure and high backhaul latency. Moreover, the full-volume acquisition strategy requires a wired network connection to be built between roadside devices and the data center, with high cost.

[0003] Therefore, there is an urgent need for a high-efficiency vehicle-road simultaneous time-series data acquisition solution to dynamically trigger the data recording of roadside devices, reduce redundancy, implement a cross-intersection collaborative data acquisition mechanism, and reduce data acquisition costs. Summary of the Invention

[0004] The purpose of the present invention is to provide a vehicle-road simultaneous time-series data acquisition method and system based on a distributed architecture to dynamically trigger the data recording of roadside devices, reduce redundancy, implement a cross-intersection collaborative data acquisition mechanism, and reduce data acquisition costs.

[0005] To achieve the above purpose, in the first aspect, the present invention provides a vehicle-road simultaneous time-series data acquisition method based on a distributed architecture, including the following steps: Obtain the trajectory point sequence of the current acquisition task, control the data acquisition vehicle to switch states according to the task planning result, and when in the upload state, transmit the untransmitted data to the cloud server; Real-time obtain and analyze the state and driving trajectory data of the data acquisition vehicle, and determine whether the current roadside perception subsystem needs to perform data acquisition; Based on the received judgment result, control the roadside perception subsystem to switch its state, and when in the upload state, transmit the untransmitted data collected to the cloud server.

[0006] Wherein, the method further includes: Timing the roadside perception subsystem and the data collection vehicle respectively according to the received satellite timing signal.

[0007] Wherein, obtaining and parsing the state and driving trajectory data of the data collection vehicle in real time, and judging whether the current roadside perception subsystem needs to collect data includes: Parsing the sensor data of the camera and lidar in real time and transmitting the parsed perception data to the data collection sub-unit; Parsing the status data received by the roadside communication unit in real time, and calculating whether the current roadside perception subsystem needs to collect data according to the state of the data collection vehicle and the driving trajectory obtained by parsing.

[0008] Wherein, calculating whether the current roadside perception subsystem needs to collect data according to the state of the data collection vehicle and the driving trajectory obtained by parsing includes: Calculating whether the real-time vehicle state of the data collection vehicle is equal to 0 according to the state of the data collection vehicle and the driving trajectory obtained by parsing; If it is equal to 0, the data collection vehicle is not in the data collection state and does not need to collect data; If it is not equal to 0, the data collection vehicle is in the data collection state.

[0009] Wherein, after the data collection vehicle is in the data collection state, the method further includes: Traversing and calculating the distance between the perception center point of the current roadside perception subsystem and the vehicle driving trajectory point. If the distance is not less than the perception range radius of the current roadside perception subsystem, data collection is not required; If the distance is less than the perception range radius of the current roadside perception subsystem, judge whether the vehicle trajectory point has passed the perception range radius, and continue to calculate whether the current roadside perception subsystem needs to collect data based on the passed vehicle trajectory points.

[0010] Wherein, continuing to calculate whether the current roadside perception subsystem needs to collect data based on the passed vehicle trajectory points includes: Calculating the distance between the center point position of the data collection vehicle and the perception center point position of the current roadside perception subsystem. If the collection vehicle has currently left the perception range of the current roadside perception subsystem, data collection is not required; if the data collection vehicle is still within the perception range of the current roadside perception subsystem, the current roadside perception subsystem needs to collect data.

[0011] Among them, based on the received judgment result, the roadside perception subsystem is controlled to perform state switching, and when in the upload state, the un-uploaded data collected is transmitted to the cloud server, including: Based on the received judgment result, the roadside perception subsystem is controlled to switch states among the data collection state, the upload state, and the standby state; When the roadside perception subsystem needs to collect data, the system state switches from the standby state to the data collection state; When the roadside perception subsystem does not need to collect data, the system state switches from the data collection state to the standby state; When there is un-uploaded data in the edge storage unit of the roadside perception subsystem, the system state switches from the standby state to the upload state; when all data has been uploaded and there is no un-uploaded data in the edge storage unit, the system state switches from the upload state to the standby state.

[0012] Among them, the trajectory point sequence of the current collection task is obtained, the data collection vehicle is controlled to perform state switching according to the task planning result, and when in the upload state, the un-uploaded data is transmitted to the cloud server, including: The trajectory point sequence of the current collection task is obtained, and the data collection vehicle is controlled to switch states among the data collection state, the upload state, and the standby state according to the task planning result; When the system of the data collection vehicle is in the upload state, data is collected based on the set trajectory points of the data collection task, and the vehicle state and vehicle trajectory data are uploaded to the cloud server in real time.

[0013] In a second aspect, the present invention provides a vehicle-road synchronous data collection system based on a distributed architecture. The vehicle-road synchronous data collection system based on a distributed architecture includes a roadside perception subsystem, a data collection vehicle, and a cloud server. The roadside perception subsystem and the data collection vehicle are respectively connected to the cloud server; The roadside perception subsystem includes roadside sensors, a first PTP time server, a first aggregation switch, an edge storage unit, a roadside communication unit, and a first satellite receiving antenna; The data collection vehicle includes perception sensors, a second PTP time server, a second aggregation switch, an on-vehicle storage unit, an on-vehicle communication unit, and a second satellite receiving antenna; The cloud server includes a data transfer server and a cloud storage server; The first satellite receiving antenna receives the satellite time synchronization signal and transmits it to the first PTP time synchronizer. The first PTP time synchronizer simultaneously synchronizes the time for the roadside sensor, the first aggregation switch, and the edge storage unit. The roadside communication unit receives the status data from the cloud server through the network and sends the sensed data in the edge storage unit to the cloud server through the network. The edge storage unit receives the status data of the roadside communication unit and performs the corresponding data acquisition process; The second satellite receiving antenna receives the satellite time synchronization signal and transmits it to the second PTP time synchronizer. The second PTP time synchronizer simultaneously synchronizes the time for the vehicle-mounted sensor, the second aggregation switch, and the vehicle-mounted storage unit. The vehicle-mounted communication unit transmits the vehicle's real-time position data and acquisition status data to the cloud server through the network; The data transfer server receives the positioning data, sensed data, and status data uploaded by the roadside sensing subsystem and the data acquisition vehicle in real time, and forwards the data to the roadside communication unit and the vehicle-mounted communication unit. The large-volume original sensed data uploaded by the roadside sensing subsystem is directly stored in the cloud storage server.

[0014] A vehicle-road simultaneous time-series data acquisition method and system based on a distributed architecture according to the present invention includes a roadside sensing subsystem, a data acquisition vehicle, and a cloud server. The roadside sensing subsystem and the data acquisition vehicle are respectively synchronized according to the received satellite time synchronization signal to obtain the trajectory point sequence of the current acquisition task. According to the task planning result, the data acquisition vehicle is controlled to perform state switching, and when in the upload state, the unuploaded data is transmitted to the cloud server; the state and driving trajectory data of the data acquisition vehicle are obtained and parsed in real time, and it is judged whether the current roadside sensing subsystem needs to perform data acquisition. Based on the received judgment result, the roadside sensing subsystem is controlled to perform state switching, and when in the upload state, the uncollected data collected is transmitted to the cloud server to dynamically trigger the data recording of the roadside device, reduce redundancy, implement a cross-intersection collaborative data acquisition mechanism, and reduce the data acquisition cost. Description of the Drawings

[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art.

[0016] Figure 1 It is a schematic diagram of the steps of the vehicle-road simultaneous time-series data acquisition method provided by the present invention.

[0017] Figure 2 It is a hardware architecture diagram of the vehicle-road simultaneous time-series data acquisition system provided by the present invention.

[0018] Figure 3 This is the software architecture diagram of the vehicle-road simultaneous time-series data acquisition system provided by the present invention.

[0019] Figure 4 This is the flow chart for judging the roadside data acquisition requirements provided by the present invention.

[0020] Figure 5 This is the schematic diagram of the method for judging the roadside data acquisition requirements provided by the present invention.

[0021] Figure 6 This is the state machine switching logic diagram of the roadside perception subsystem provided by the present invention.

[0022] Figure 7 This is the state machine switching logic diagram of the data acquisition vehicle system provided by the present invention. Detailed implementation manners

[0023] The embodiments of the present invention will be described in detail below. The examples of the embodiments are shown in the accompanying drawings. The embodiments described below by referring to the drawings are exemplary and are intended to explain the present invention, but should not be construed as limiting the present invention.

[0024] The first embodiment of the present application is as follows: Please refer to Figures 1-7 , Figure 1 This is the step schematic diagram of the vehicle-road simultaneous time-series data acquisition method based on the distributed architecture provided by the present invention. Figure 2 This is the hardware architecture diagram of the vehicle-road simultaneous time-series data acquisition system provided by the present invention. Figure 3 This is the software architecture diagram of the vehicle-road simultaneous time-series data acquisition system provided by the present invention. Figure 4 This is the flow chart for judging the roadside data acquisition requirements provided by the present invention. Figure 5 This is the schematic diagram of the method for judging the roadside data acquisition requirements provided by the present invention. Figure 6 This is the state machine switching logic diagram of the roadside perception subsystem provided by the present invention. Figure 7 This is the state machine switching logic diagram of the data acquisition vehicle system provided by the present invention.

[0025] The present invention provides a vehicle-road simultaneous time-series data acquisition method based on a distributed architecture, including the following steps: S101. Obtain the trajectory point sequence of the current acquisition task, control the state switching of the data acquisition vehicle according to the task planning result, and when in the upload state, transmit the untransmitted data to the cloud server.

[0026] Specifically, the vehicle tester inputs the trajectory point sequence of the current acquisition task through the data acquisition task planning subunit. The state machine of the data acquisition vehicle system switches the state according to the task planning result, such as Figure 7As shown in the figure. When the system is in the data acquisition state, the system will start the data acquisition sub-unit to record the camera image data, lidar point cloud data, and integrated navigation and positioning data in real time; when the system is in the standby state, the system only runs the sensor driver sub-unit and the vehicle communication unit driver sub-unit; when the system is in the upload state, the system will start the data upload sub-unit to upload the un-uploaded data to the acquisition data receiving service of the cloud server through the wireless network.

[0027] Among them, the detailed process of the state switching of the data acquisition vehicle system according to the task planning result is as follows: The data acquisition vehicle system has three states, namely the data acquisition state, the upload state, and the standby state. The vehicle acquisition state of the data acquisition vehicle at this time represents the data acquisition state, represents the standby state, represents the upload state. The initial state of the system is the standby state, among which, indicates the vehicle state.

[0028] When the tester manually sets the data acquisition task track points through the data acquisition task planning sub-unit and clicks the start data acquisition button, the data acquisition task planning sub-unit outputs , and the state of the data acquisition vehicle system switches from the standby state to the data acquisition state; when the tester selects to stop the data acquisition action through the data acquisition task planning sub-unit, the data acquisition task planning sub-unit outputs , and the state of the data acquisition vehicle system switches from the data acquisition state to the standby state, among which, indicates whether the vehicle needs to perform data acquisition.

[0029] When and , that is, there is un-uploaded data in the vehicle storage unit, the system state switches from the standby state to the upload state; when all data has been uploaded and there is no un-uploaded data in the vehicle storage unit, that is, , the system state switches from the upload state to the standby state, among which, indicates whether the vehicle needs to perform data upload.

[0030] When the system is in the upload state, after the tester manually sets the data acquisition task track points through the data acquisition task planning sub-unit and clicks the start data acquisition button, the data acquisition task planning sub-unit outputs , then the system immediately switches from the upload state to the data acquisition state; the system state cannot directly switch from the data acquisition state to the upload state.

[0031] S102. Real-time obtain and parse the state and driving trajectory data of the data acquisition vehicle, and determine whether the current roadside perception subsystem needs to perform data acquisition.

[0032] Specifically, the roadside perception subsystem consists of a camera driving subunit, a lidar driving subunit, a data acquisition subunit, a data upload subunit, a roadside communication unit driving subunit, a roadside data acquisition requirement judgment subunit, and a roadside perception system state machine. The camera driving subunit and the lidar driving subunit parse the sensor data of the camera and the lidar in real time and transmit the parsed perception data to the data acquisition subunit; the roadside communication unit driving subunit parses the status data received by the roadside communication unit in real time and sends the status and driving trajectory data of the data acquisition vehicle to the roadside data acquisition requirement judgment subunit. The acquisition requirement judgment subunit calculates whether the current roadside system needs to perform data acquisition based on the status of the data acquisition vehicle and the driving trajectory of the vehicle as Figure 4 shown, and sends the judgment result to the roadside perception system state machine.

[0033] As Figure 4 and Figure 5 shown, the detailed steps of the judgment method include: Step 1: The real-time vehicle status of the data acquisition vehicle is , where and represent the real-time position of the vehicle, represents the real-time driving heading of the vehicle, represents the vehicle length, represents the vehicle width, represents the vehicle height, represents the real-time acquisition status of the vehicle, and judge the vehicle acquisition status of the data acquisition vehicle at this time. If , that is, the acquisition vehicle is not in the data acquisition state, then roadside data acquisition is not required, that is ; if , that is, the acquisition vehicle is in the data acquisition state, then go to Step 2; Step 2: The driving trajectory of the data acquisition vehicle is , where is the coordinate of the vehicle trajectory point, represents whether the vehicle has traveled through this trajectory point, represents that the vehicle has not traveled yet, represents that the vehicle has traveled through this trajectory point, where i 1,..., m; The state of this roadside perception subsystem is , where , is the coordinate of the perception center point of this roadside perception subsystem, is the perception range radius of this roadside perception subsystem, This is the system state of the roadside perception subsystem.

[0034] Traverse and calculate the distance between the perception center point of the roadside perception subsystem and the vehicle driving trajectory points , Among them

[0035] If , then the roadside perception subsystem does not need to perform data collection, that is , the judgment program ends, indicating non-existence; If , it is identified as judgment condition 1, then go to step 3, indicating existence; Step 3: If , that is, the th vehicle trajectory point has not been traveled yet, and this trajectory point is within the perception range of the roadside perception system, then the roadside perception subsystem needs to perform data collection, that is , the judgment program ends, i* represents the actual individual object, and i represents a general reference; If , that is, the th vehicle trajectory point has already been traveled, go to step 4; Step 4: Calculate the distance between the position of the center point of the collected vehicle and the position of the perception center point of the roadside perception subsystem ,

[0036] If , that is, the collected vehicle has currently left the perception range of the roadside perception subsystem, then data collection is not required, that is , the judgment program ends; If , that is, the collected vehicle is still within the perception range of the roadside perception subsystem, then the roadside perception subsystem needs to perform data collection, that is , the judgment program ends.

[0037] S103. Based on the received judgment result, control the roadside perception subsystem to perform state switching, and when in the upload state, transmit the untransmitted data collected to the cloud server.

[0038] Specifically, the roadside perception system state machine switches among the three states of data collection state, standby state, and upload state according to the information input, as Figure 6As shown, when the roadside perception subsystem is in the data collection state, the system will start the data collection subunit to record camera image data and lidar point cloud data in real time; when the system is in the standby state, the system only runs the camera driver subunit, the lidar driver subunit, the roadside data collection requirement judgment subunit, and the roadside communication unit driver subunit; when the system is in the upload state, the system will start the data upload subunit to transmit the unuploaded collected data to the collection data receiving service of the cloud server through a wired or wireless network.

[0039] The roadside perception subsystem has three states, namely the data collection state, the upload state, and the standby state. represents the data collection state. represents the standby state. represents the upload state. The initial state of the system is the standby state, where represents the roadside state.

[0040] When the roadside data collection requirement judgment subunit outputs , that is, the roadside perception subsystem needs to collect data, and the system state switches from the standby state to the data collection state; when the roadside data collection requirement judgment subunit outputs , that is, the roadside perception subsystem does not need to collect data, and the system state switches from the data collection state to the standby state, where indicates whether the roadside needs to collect data.

[0041] When and , that is, there is unuploaded data in the roadside edge storage unit, then the system state switches from the standby state to the upload state; when all data has been uploaded and there is no unuploaded data in the roadside edge storage unit, that is , the system state switches from the upload state to the standby state, where indicates whether the roadside needs to upload data.

[0042] When the system is in the upload state, the roadside data collection requirement judgment subunit outputs , then the system immediately switches from the upload state to the data collection state; the system state cannot directly switch from the data collection state to the upload state.

[0043] The method further includes: Timing the roadside perception subsystem and the data collection vehicle respectively according to the received satellite timing signal.

[0044] Please refer to Figure 2, the present invention provides a vehicle-road synchronous data acquisition system based on a distributed architecture. The vehicle-road synchronous data acquisition system based on the distributed architecture includes a roadside perception subsystem, a data acquisition vehicle, and a cloud server. The roadside perception subsystem and the data acquisition vehicle are respectively connected to the cloud server; The roadside perception subsystem includes roadside sensors, a first PTP time server, a first aggregation switch, an edge storage unit, a roadside communication unit, and a first satellite receiving antenna; The data acquisition vehicle includes perception sensors, a second PTP time server, a second aggregation switch, an in-vehicle storage unit, an in-vehicle communication unit, and a second satellite receiving antenna; The cloud server includes a data transfer server and a cloud storage server; The first satellite receiving antenna receives satellite timing signals and transmits them to the first PTP time server. The first PTP time server simultaneously times the roadside sensors, the first aggregation switch, and the edge storage unit. The roadside communication unit receives status data from the cloud server through the network and transmits the perception data in the edge storage unit to the cloud server through the network. The edge storage unit receives the status data of the roadside communication unit and performs corresponding data acquisition processes; The second satellite receiving antenna receives satellite timing signals and transmits them to the second PTP time server. The second PTP time server simultaneously times the in-vehicle sensors, the second aggregation switch, and the in-vehicle storage unit. The in-vehicle communication unit transmits vehicle real-time position data and acquisition status data to the cloud server through the network; The data transfer server receives positioning data, perception data, and status data uploaded by the roadside perception subsystem and the data acquisition vehicle in real time, and forwards the data to the roadside communication unit and the in-vehicle communication unit. The large-volume original perception data uploaded by the roadside perception subsystem is directly stored in the cloud storage server.

[0045] In this embodiment, the system mainly consists of a roadside perception subsystem, a data acquisition vehicle, and a cloud server.

[0046] The roadside perception subsystem includes roadside sensors, a first PTP time server, a first aggregation switch, an edge storage unit, a roadside communication unit, and a first satellite receiving antenna. Among them, the first satellite receiving antenna receives the satellite timing signal and transmits it to the first PTP time server. The first PTP time server simultaneously provides timing for the roadside sensors, the first aggregation switch, and the edge storage unit. The roadside communication unit receives status data from the cloud through the 4G / 5G wireless cellular network and sends the perception data in the edge storage unit to the cloud through the 4G / 5G wireless cellular network. The edge storage unit receives the status data of the roadside communication unit and executes the corresponding data acquisition sub-unit.

[0047] The data collection vehicle is equipped with perception sensors, a second PTP time server, a second aggregation switch, an on-vehicle storage unit, an on-vehicle communication unit, and a second satellite receiving antenna. Among them, the second satellite receiving antenna receives the satellite timing signal and transmits it to the second PTP time server. The second PTP time server simultaneously provides timing for the on-vehicle sensors, the second aggregation switch, and the on-vehicle storage unit. The on-vehicle communication unit transmits the vehicle's real-time position data and acquisition status data to the cloud server through the 4G or 5G mobile network.

[0048] The first aggregation switch and the second aggregation switch are normal aggregation switches used to transfer and exchange data.

[0049] The cloud server mainly includes a data transfer server and a cloud storage server. The data transfer server receives the positioning data, perception data, and status data uploaded from the roadside and the vehicle end in real time, and forwards the data to the roadside communication unit and the on-vehicle communication unit. The large-volume original perception data uploaded from the roadside is directly stored in the cloud storage server.

[0050] As shown in the appendix Figure 3 shown, among which, there are actually multiple edge storage units in the appendix Figure 3 For the sake of clarity of the attached drawings, the edge storage units from edge storage unit 1 to edge storage unit n - 1 are omitted, and only the structure shown in the appendix Figure 3 shown is retained. The roadside perception subsystem runs in the edge storage unit. The edge storage unit is composed of a camera drive sub-unit, a lidar drive sub-unit, a data acquisition sub-unit, a data upload sub-unit, a roadside communication unit drive sub-unit, a roadside data acquisition requirement judgment sub-unit, and a roadside perception system status mechanism. The camera drive sub-unit and the lidar drive sub-unit parse the sensor data of the camera and the lidar in real time and transmit the parsed perception data to the data acquisition sub-unit; the roadside communication unit drive sub-unit parses the status data received by the roadside communication unit in real time and sends the status and driving trajectory data of the data collection vehicle to the roadside data acquisition requirement judgment sub-unit. The acquisition requirement judgment sub-unit calculates whether the current roadside system needs to perform data acquisition based on the status of the data collection vehicle and the vehicle's driving trajectory. For example,Figure 4 as shown, and send the judgment result to the roadside perception system state machine.

[0051] According to the information input, the roadside perception system state machine switches among three states: data collection state, standby state, and upload state as Figure 6 shown. When the roadside perception subsystem is in the data collection state, the system starts the data collection sub-unit to record camera image data and lidar point cloud data in real time; when the system is in the standby state, the system only runs the camera driver sub-unit, lidar driver sub-unit, roadside data collection requirement judgment sub-unit, and roadside communication unit driver sub-unit; when the system is in the upload state, the system starts the data upload sub-unit to transmit the un-uploaded collected data to the collection data receiving service of the cloud server through wired or wireless networks.

[0052] The data collection vehicle runs in the on-vehicle storage unit. The on-vehicle storage unit includes a camera driver sub-unit, a lidar driver sub-unit, an integrated navigation driver sub-unit, a data collection sub-unit, a data upload sub-unit, an on-vehicle communication unit driver sub-unit, a data collection task planning sub-unit, and a data collection vehicle system state machine sub-unit. Among them, the driver sub-units always keep running, and the on-vehicle communication unit driver sub-unit will upload the vehicle state and vehicle trajectory data to the real-time data transmission cloud service of the cloud server in real time.

[0053] The vehicle tester inputs the trajectory point sequence of this collection task through the data collection task planning sub-unit. The data collection vehicle system state machine performs state switching according to the task planning result, as Figure 7 shown. When the system is in the data collection state, the system starts the data collection sub-unit to record camera image data, lidar point cloud data, and integrated navigation positioning data in real time; when the system is in the standby state, the system only runs the sensor driver sub-unit and the on-vehicle communication unit driver sub-unit; when the system is in the upload state, the system starts the data upload sub-unit to upload the un-uploaded data to the collection data receiving service of the cloud server through wireless networks.

[0054] In traditional data acquisition solutions, there is no real-time communication between the vehicle end and the roadside end during the acquisition process. The roadside end and the vehicle end do not know each other's acquisition status, or only confirm the acquisition status through phone calls between the testers at the roadside end and the vehicle end. This approach has low acquisition efficiency and high labor costs. The architecture of this solution addresses this problem. In this solution, the data acquisition vehicle and the roadside perception subsystem communicate in real time through the cloud data transfer server and the mobile cellular network. The vehicle end can obtain the status of the roadside perception system in real time, such as whether the sensors are working properly and whether they are in the acquisition state; the roadside perception system can also obtain the status of the vehicle end in real time, including the real-time positioning of the vehicle and whether the vehicle end is in the acquisition state. The vehicle end and the roadside can dynamically adjust the data acquisition strategy based on the status data of each other, eliminating the need for the testers at both ends to communicate by phone to confirm the acquisition task. This greatly improves the data acquisition efficiency and reduces the labor cost of data acquisition.

[0055] The beneficial effects of the present invention are as follows: 1. Precise data acquisition reduces the cost of post-data processing Traditional roadside data acquisition methods generally use the method of recording data continuously for 24 hours. This method causes great difficulties in selecting useful data from a large amount of data later.

[0056] This system can accurately identify and judge the roadside acquisition requirements through an intelligent roadside data acquisition requirement judgment algorithm, and realize data acquisition only when it is needed. As long as the data collected is useful data, it directly avoids the later data screening work and greatly reduces the cost of post-data processing.

[0057] 2. Distributed data storage reduces the system construction cost Traditional vehicle-road data acquisition systems build a fiber-optic wired data connection between roadside perception devices and the data center, and use multi-layer switches to upload raw data to the data center server for unified storage in real time. This method has the following disadvantages when the number of roadside devices is large: The cost of building a wired network is huge. Building fiber-optic cables, switches and other network devices from multiple intersections to the data center requires high construction and equipment procurement costs; The cost of building network storage in the data center is huge. The traditional method directly collects data from roadside sensors, and the data center needs to build a large enough storage capacity, which is costly.

[0058] This system deploys roadside edge storage devices on the roadside, enabling a large amount of raw data to be distributed and stored in the edge storage devices on each roadside. Instead of immediately transmitting the data back to the data center, it waits until the data collection task is completed and then transmits it to the data center through the wireless network in sequence. This not only significantly reduces the construction cost of the data center's storage capacity but also directly eliminates the construction cost of the wired network, further reducing the cost.

[0059] 3. Collaborative automated collection reduces the labor cost of data collection During the traditional vehicle-road data collection process, multiple testers are required to cooperate. For example, on the data collection vehicle, 1-2 testers are needed to drive the vehicle and start the vehicle-end data collection and communicate with other testers; in the data center, 1-2 testers are needed to communicate with the vehicle-end testers and dynamically activate the cloud data reception sub-unit according to the needs of the vehicle-end testers; at each roadside edge, 1 tester is needed to communicate with the vehicle-end testers and dynamically activate the data collection sub-unit of the roadside device. Thus, in the case of using the traditional method for vehicle-road data collection, the more intersections the data collection vehicle passes through during the collection task, the more testers are needed. This method has too high labor cost for data collection and low efficiency.

[0060] This system enables the vehicle-end, road-end, and cloud-end to work together. By automatically switching the working states of the finite state machines at the road-end and vehicle-end, it realizes the automated data collection work of the vehicle-end, road-end, and cloud-end. The road-end and vehicle-end can automatically switch states according to the states of each other, so there is no need to deploy testers at the cloud-end and road-end. Only one tester needs to be deployed at the vehicle-end throughout the testing process. The tester only needs to set the data collection task and activate the data collection sub-unit before driving the vehicle, and all subsequent collection tasks are fully automated, significantly reducing the labor cost of data collection.

[0061] The above-disclosed are only one or more preferred embodiments of the present application, and the scope of rights of the present application cannot be limited thereby. Those of ordinary skill in the art can understand all or part of the processes of implementing the above embodiments, and the equivalent changes made according to the claims of the present application still fall within the scope covered by the present application.

Claims

1. A method for collecting vehicle-road simultaneous data based on a distributed architecture, characterized in that: The following steps are involved: Obtain the trajectory point sequence of this collection task, control the data collection vehicle to switch states according to the task planning results, and transmit the unuploaded data to the cloud server when in the uploading state; Acquire and analyze the status and driving trajectory data of the data collection vehicle in real time, and determine whether the current roadside perception subsystem needs to collect data; Based on the received judgment results, the roadside perception subsystem is controlled to switch states, and when in the uploading state, the collected data that has not been uploaded is transmitted to the cloud server.

2. The method for collecting vehicle-road simultaneous data based on a distributed architecture as claimed in claim 1, characterized in that: The method further comprises: The roadside perception subsystem and data collection vehicle are timed according to the received satellite timing signals.

3. The method for collecting vehicle-road simultaneous data based on a distributed architecture as claimed in claim 1, characterized in that: Acquire and analyze the status and driving trajectory data of the data collection vehicle in real time, and determine whether the current roadside perception subsystem needs to collect data, including: Analyze the sensor data of the camera and lidar in real time and transmit the analyzed perception data to the data acquisition subunit; The status data received by the roadside communication unit is analyzed in real time, and the status and driving trajectory of the data acquisition vehicle obtained from the analysis are used to calculate whether the current roadside perception subsystem needs to collect data.

4. The method for collecting vehicle-road simultaneous data based on a distributed architecture as claimed in claim 3, characterized in that: The state and driving trajectory of the data collection vehicle obtained by analysis are used to calculate whether the current roadside perception subsystem needs to collect data, including: According to the state and driving trajectory of the data collection vehicle obtained by analysis, calculate whether the real-time vehicle state of the data collection vehicle is equal to 0; If it is equal to 0, the data collection vehicle is not in the data collection state and does not need to collect data; If it is not equal to 0, the data collection vehicle is in the data collection state.

5. The method for collecting vehicle-road simultaneous data based on a distributed architecture as claimed in claim 4, characterized in that: After the data collection vehicle is in a data collection state, the method further includes: The distance between the perception center point of the current roadside perception subsystem and the vehicle driving trajectory point is calculated by traversing. If the distance is not less than the perception range radius of the current roadside perception subsystem, data collection is not required. If the distance is less than the perception range radius of the current roadside perception subsystem, it is determined whether the vehicle trajectory point has passed the perception range radius, and based on the vehicle trajectory points that have passed, it is continued to calculate whether the current roadside perception subsystem needs to collect data.

6. The method for collecting vehicle-road simultaneous data based on a distributed architecture as claimed in claim 5, characterized in that: Based on the vehicle trajectory points that have passed, the current roadside perception subsystem continues to calculate whether data collection is required, including: Calculate the distance between the center point of the data collection vehicle and the perception center point of the current roadside perception subsystem. If the data collection vehicle has left the perception range of the current roadside perception subsystem, data collection is not required. If the data collection vehicle is still within the perception range of the current roadside perception subsystem, the current roadside perception subsystem needs to collect data.

7. The method for collecting vehicle-road simultaneous data based on a distributed architecture as claimed in claim 1, characterized in that: Based on the received judgment result, the roadside perception subsystem is controlled to switch state, and when in the uploading state, the collected but not uploaded data is transmitted to the cloud server, including: Based on the received judgment result, the roadside sensing subsystem is controlled to switch between the data acquisition state, the upload state and the standby state; When the roadside sensing subsystem needs to collect data, the system state switches from the standby state to the data collection state; When the roadside sensing subsystem does not need to collect data, the system state switches from the data collection state to the standby state; When there is unuploaded data in the edge storage unit of the roadside perception subsystem, the system state switches from the standby state to the upload state; when all data are uploaded and there is no unuploaded data in the edge storage unit, the system state switches from the upload state to the standby state.

8. The method for collecting vehicle-road simultaneous data based on a distributed architecture as claimed in claim 1, characterized in that: Obtain the trajectory point sequence of this collection task, control the data collection vehicle to switch states according to the task planning results, and transfer the unuploaded data to the cloud server when in the uploading state, including: Obtain the trajectory point sequence of this collection task, and control the data collection vehicle to switch between data collection state, upload state and standby state according to the task planning results; When the system of the data collection vehicle is in the uploading state, data collection is performed based on the set data collection task trajectory points, and the vehicle status and vehicle trajectory data are uploaded to the cloud server in real time.

9. A vehicle-road simultaneous sequential data collection system based on a distributed architecture, using the vehicle-road simultaneous sequential data collection method based on a distributed architecture as described in any one of claims 1 to 8, characterized in that: The vehicle-road simultaneous data collection system based on a distributed architecture includes a roadside sensing subsystem, a data collection vehicle and a cloud server, wherein the roadside sensing subsystem and the data collection vehicle are respectively connected to the cloud server; The roadside sensing subsystem includes a roadside sensor, a first PTP timer, a first convergence switch, an edge storage unit, a roadside communication unit, and a first satellite receiving antenna; The data collection vehicle includes a perception sensor, a second PTP timer, a second convergence switch, an on-board storage unit, an on-board communication unit, and a second satellite receiving antenna; The cloud server includes a data transfer server and a cloud storage server; The first satellite receiving antenna receives a satellite timing signal and transmits it to the first PTP timing device, the first PTP timing device simultaneously provides timing for the roadside sensor, the first convergence switch, and the edge storage unit, the roadside communication unit receives status data from the cloud server through the network, and sends the perception data in the edge storage unit to the cloud server through the network; The edge storage unit receives the status data of the roadside communication unit and performs a corresponding data collection process; The second satellite receiving antenna receives the satellite timing signal and transmits it to the second PTP timing device, and the second PTP timing device simultaneously provides timing for the vehicle-mounted sensor, the second convergence switch, and the vehicle-mounted storage unit, and the vehicle-mounted communication unit transmits the vehicle's real-time position data and collection status data to the cloud server through the network; The data transfer server receives the positioning data, perception data and status data uploaded by the roadside perception subsystem and the data collection vehicle in real time, and forwards the data to the roadside communication unit and the vehicle-mounted communication unit. The large volume of original perception data uploaded by the roadside perception subsystem is directly stored in the cloud storage server.

Images