Automatic driving vehicle test data acquisition system

Abstract

The embodiment of the invention relates to an automatic driving vehicle test data acquisition system. The system comprises a vehicle acquisition device, an evaluation client, a questionnaire page, a data processing service and a data storage system. The vehicle acquisition equipment is used for acquiring system and sensor data of a vehicle and a test track; the evaluation client is used for collecting personnel evaluation data; the questionnaire page is used for collecting response data of the subjective experience questionnaire; the data processing service is used for updating the basic edition-road test data set according to acquired data of the vehicle acquisition equipment, the evaluation client and the questionnaire page; and the server is also used for setting a label cluster according to the test track and setting a space-time version-label data set based on the label cluster. According to the method, human factor data dimensions are increased by adding evaluation and questionnaire data sets, and data support can be provided for positioning data blocks based on labels through a space-time version-label data set.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a test data acquisition system for autonomous vehicles. Background Technology

[0002] Before a vehicle's autonomous driving system is put into use, it needs to perform a large number of road tests (referred to as road tests). During the road tests, a series of vehicle-side data are continuously collected from the vehicle under test. If the vehicle under test has an anomaly during the road tests, anomaly analysis is performed based on the collected data.

[0003] Analysis of traditional road test data collection schemes revealed several shortcomings: 1) Lack of human factors dimension: Traditional schemes only collect machine-dimensional data objects such as system data items related to the autonomous driving system (e.g., speed, acceleration, steering wheel angle, throttle opening, brake opening, vehicle posture phase angle, etc.) and sensor data items (e.g., visual images, radar point clouds, etc.), without collecting evaluation data from the driver's and passengers' perspective or subjective experience data; 2) Inability to locate data blocks using spatiotemporal labels: Traditional schemes only perform time alignment on the collected data elements but do not label them, thus making it impossible to quickly locate the local data blocks required for anomaly analysis based on time period and / or region labels.

[0004] To address the aforementioned shortcomings, we propose the following improvement: 1) Increasing the collection of evaluation data from road testers during the road test process; 2) Increasing the collection of subjective experience questionnaire data for the road test task; 3) Adding associated spatiotemporal tags to the collected data elements. How to implement these improvements is the technical problem that this invention aims to solve. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of existing technologies by providing a data acquisition system for autonomous vehicle testing. This system includes: vehicle data acquisition equipment, an evaluation client, a questionnaire page, a data processing service, and a data storage system. The vehicle data acquisition equipment is used to collect system data items and sensor data items from the autonomous driving system, and to collect data on the test trajectory of the vehicle under test. The evaluation client is used to collect evaluation data provided by evaluators. The questionnaire page is used to provide subjective experience questionnaires to evaluators and to collect data on their responses. The data processing service is used to update the basic version of the road test dataset for the road test task based on the real-time collected data sent by the vehicle data acquisition equipment, the evaluation client, and the questionnaire page. A subset of this dataset contains the system dataset, sensor dataset, evaluation dataset, and questionnaire dataset. The data processing service is also used to set up label clusters based on the test trajectory and to label the system / sensor datasets based on the label clusters to obtain the corresponding spatiotemporal version of the label dataset. The data storage system is used to store the basic version of the road test dataset and the spatiotemporal version of the label dataset corresponding to each road test task. This invention aims to increase the dimensions of human factors data by adding assessment and questionnaire datasets; and by adding spatiotemporal label clusters and their corresponding spatiotemporal version-label datasets, it provides necessary data support for a retrieval mechanism that can locate data blocks based on labels.

[0006] To achieve the above objectives, embodiments of the present invention provide an autonomous vehicle test data acquisition system, the system comprising: vehicle acquisition equipment, evaluation client, questionnaire page, data processing service, and data storage system;

[0007] The data processing service is connected to the vehicle data collection device, the evaluation client, the questionnaire page, and the data storage system, respectively.

[0008] The vehicle data acquisition equipment is installed on the vehicle being tested;

[0009] The vehicle data acquisition equipment is used to store the system monitoring list and the sensor monitoring list;

[0010] The vehicle data acquisition device is also used to collect data values ​​of multiple types of system data specified in the system monitoring list in real time through the vehicle CAN bus during the road test of the vehicle under test, generate corresponding system data element sets, collect data files generated by multiple types of sensors specified in the sensor monitoring list in real time, generate corresponding sensor data element sets, and send the system data element sets and the sensor data element sets to the data processing service.

[0011] The vehicle data acquisition device is also used to collect the latitude and longitude information of the vehicle under test, as well as the information of the road, road segment, lane and intersection where the vehicle is located, in real time based on the built-in positioning device and the high-precision road map, and generate corresponding trajectory point data elements to send to the data processing service.

[0012] The evaluation client is used to collect data on the evaluation scores and comment texts made by the evaluators on each evaluation item they are responsible for during the road test of the vehicle under test, generate corresponding evaluation data elements, and send them to the data processing service.

[0013] The questionnaire page is used to provide the evaluator with a subjective experience questionnaire, and to collect data on the answer text of the current evaluator on each question item of the subjective experience questionnaire to generate corresponding question and answer data elements. The question and answer data elements corresponding to the subjective experience questionnaire are combined into a question and answer data element set and sent to the data processing service.

[0014] The data processing service is used to plan a corresponding basic version of the road test dataset for the current task on the data storage system before the road test task of the vehicle under test is executed; and to set the basic information of the basic version of the road test dataset.

[0015] The data processing service is also used to, during the road test of the vehicle under test, update the system dataset and sensor dataset of the basic version-road test dataset in real time based on the system data element set and sensor data element set received each time; update the evaluation dataset of the basic version-road test dataset in real time based on the evaluation data element received each time; update the test trajectory of the basic version-road test dataset in real time based on the trajectory point data element received each time; and update the questionnaire dataset of the basic version-road test dataset in real time based on the received question and answer data element set.

[0016] The data processing service is also used to: first, perform data element deduplication on the basic version-road test dataset after the road test task of the vehicle under test is completed; then, perform time standardization on the basic version-road test dataset; then, set spatiotemporal labels based on the test trajectory to obtain the corresponding label cluster; and then, label the data elements of the system dataset and the sensor dataset based on the label cluster to obtain the corresponding spatiotemporal version-label dataset and store it in the data storage system.

[0017] The data storage system is used to store the basic version-road test dataset and the spatiotemporal version-label dataset corresponding to each road test task.

[0018] Preferably, the system monitoring list consists of system data types from multiple autonomous driving systems;

[0019] When the sensor monitoring list is not empty, it consists of one or more sensor types; the sensor types include cameras, lidar, millimeter-wave radar, ultrasonic sensors, and infrared sensors;

[0020] The system data element set includes multiple system data elements; the system data elements in the system data element set correspond one-to-one with the system data types in the system monitoring list;

[0021] When the sensor data element set is not empty, it consists of one or more sensor data elements; the sensor data elements in the sensor data element set correspond one-to-one with the sensor types in the sensor monitoring list;

[0022] Each road test task of each of the aforementioned vehicles corresponds to a task identifier, and each road test task includes multiple evaluation subjects, and each road test task will be assigned a corresponding evaluation personnel group.

[0023] The assessment subjects include at least the following subject types: illegal / irregular behavior assessment subject, lateral driving behavior assessment subject, longitudinal driving behavior assessment subject, other vehicle interaction behavior assessment subject, driving comfort assessment subject, excellent handling behavior assessment subject, and vehicle signal light control behavior assessment subject; each assessment subject consists of one or more assessment items, and each assessment item corresponds to an assessment item identifier and a set of assessment requirements;

[0024] The evaluation team consists of one or more of the evaluation personnel; each evaluation personnel corresponds to a personnel identifier and is responsible for one or more of the evaluation subjects;

[0025] During the road test of the vehicle under test, the vehicle must be accompanied by the evaluation team; the evaluation client corresponds one-to-one with the evaluation personnel, specifically a client program installed on the mobile evaluation device of the evaluation personnel; the mobile evaluation device includes mobile phone, tablet computer, laptop computer, and in-vehicle network computer; during the road test of the vehicle under test, the evaluation personnel use the evaluation client to evaluate, score, and comment on the current evaluation item according to the evaluation requirements of each evaluation item;

[0026] The questionnaire page is a web service page deployed on the Internet; the evaluator uses the mobile evaluation device to access the questionnaire page and answers the questions of the subjective experience questionnaire through the current page; the subjective experience questionnaire is a questionnaire for collecting information on the vehicle riding experience for the evaluator.

[0027] The subjective experience questionnaire consists of multiple sets of questions; the types of the question sets include at least vehicle safety issues, vehicle efficiency issues, vehicle comfort issues, vehicle credibility issues, purchase intention issues, and vehicle usability issues; each set of questions consists of multiple question items; each question item corresponds to a set of question identifiers and question text;

[0028] The basic version of the road test dataset includes the basic information, the test trajectory, the system dataset, the sensor dataset, the evaluation dataset, and the questionnaire dataset;

[0029] The basic information includes task identifier, test date, test vehicle identifier, and evaluation group; the evaluation group consists of one or more personnel identifiers.

[0030] When the test trajectory is not empty, it is formed by sequentially sorting one or more trajectory point data elements; the trajectory point data elements include trajectory point timestamps, latitude and longitude coordinates, road signs, road segment signs, and lane signs.

[0031] The system dataset consists of multiple acquisition sequences A i Composition, 1≤i≤Na, the total number Na is consistent with the total number of system data types in the system monitoring list; the collected sequence A i Each of the acquired sequences A corresponds one-to-one with the system data types. i When not empty, it is formed by sorting one or more of the corresponding system data elements in chronological order; the system data elements include the collection timestamp and the data value;

[0032] The sensor dataset consists of multiple acquisition sequences B j Composition, 1≤j≤Nb, the total number Nb is consistent with the total number of sensor types in the sensor monitoring list; the acquisition sequence B j Each of the sensor types corresponds one-to-one; each of the acquisition sequences B j When not empty, it is formed by sorting one or more corresponding sensor data elements in chronological order; the sensor data elements include acquisition timestamps and data files;

[0033] The evaluation dataset consists of multiple subject sets C k Composition, 1≤k≤Nc, the total number Nc is consistent with the total number of evaluation subjects in the current road test task; the subject set C k Each of the subject sets C corresponds one-to-one with the assessment subjects; k When not empty, it consists of one or more of the corresponding evaluation data elements; each subject set C kThe total number of evaluation data elements is consistent with the total number of evaluation items for the corresponding evaluation subject; the evaluation data elements include evaluation item identifier, personnel identifier, evaluation period, evaluation score, and comment text; the evaluation period consists of a set of corresponding evaluation start timestamps and evaluation end timestamps;

[0034] The questionnaire dataset consists of multiple question-and-answer sets D u Composition, 1≤u≤Nd, the total number Nd is consistent with the total number of questions in the subjective experience questionnaire; the question-and-answer set D u Each question and answer set D corresponds one-to-one with the question set; u When not empty, it consists of one or more of the corresponding question-and-answer data elements; each question-and-answer set D u The total number of question-and-answer data elements is consistent with the total number of question items in the corresponding question set; the question-and-answer data elements include question identifier, personnel identifier, response timestamp, and answer text;

[0035] The spatiotemporal version of the tag dataset includes the tag cluster, system tag set, and sensor tag set;

[0036] The tag cluster consists of a time period tag set and a region tag set; the time period tag set consists of multiple time period tags; the region tag set consists of multiple region tags;

[0037] The system label set corresponds to the system dataset, and consists of Na A's. i The tag set consists of; the A i Tag set and the collected sequence A i One-to-one correspondence; the A i The tag set consists of multiple tag data elements; the A i The tag data elements of the tag set and the corresponding acquisition sequence A i The system data elements correspond one-to-one; the A i Each tag data element in the tag set consists of a corresponding collection timestamp, the time period tag, and the region tag; the A i The collection timestamp of each tag data element in the tag set is matched with the collection timestamp of the corresponding system data element;

[0038] The sensor tag set corresponds to the sensor dataset, and consists of Nb B tags. j The tag set consists of; the B j Tag set and the collected sequence B j One-to-one correspondence; the B mentioned j The tag set consists of multiple tag data elements; the B j The tag data elements of the tag set and the corresponding acquisition sequence B jThe sensor data elements correspond one-to-one; the B j Each tag data element in the tag set also consists of a corresponding collection timestamp, the time period tag, and the region tag; the B j The acquisition timestamp of each tag data element in the tag set is matched with the acquisition timestamp of the corresponding sensor data element.

[0039] Preferably, the data processing service is specifically used to set the task identifier, test date, test vehicle identifier, and evaluation group corresponding to the basic information when setting the basic information of the basic version-road test dataset: based on the task identifier corresponding to the current road test task, the specified test date, the vehicle identifier of the vehicle under test, and the identifiers of all personnel in the evaluation personnel group.

[0040] Preferably, the data processing service is specifically used to: when updating the system dataset and sensor dataset of the basic version-road test dataset in real time based on the system data element set and sensor data element set received each time; use the currently received system data element set and sensor data element set as the corresponding first set and second set; and assign each system data element in the first set to the corresponding acquisition sequence A in the system dataset. i Add to the second set; and add each of the sensor data elements in the second set to the corresponding acquisition sequence B in the sensor dataset. j Add to;

[0041] The data processing service is specifically used to, when updating the evaluation dataset of the basic version-road test dataset in real time based on the evaluation data elements received each time: add the currently received evaluation data elements to the corresponding subject set C in the evaluation dataset. k Add to;

[0042] The data processing service is specifically used to add the currently received trajectory point data elements to the test trajectory when the test trajectory of the basic version-road test dataset is updated in real time based on the trajectory point data elements received each time.

[0043] The data processing service is specifically used to, when updating the questionnaire dataset of the basic version-road test dataset in real time based on the received question-and-answer data element set: add each question-and-answer data element in the currently received question-and-answer data element set to the corresponding question-and-answer set D in the questionnaire dataset. u Add to.

[0044] Preferably, the data processing service is specifically used when performing data element deduplication on the basic version-road test dataset:

[0045] The subject sets C of the evaluation dataset. k As the current subject set; and group one or more assessment data elements with the same assessment item identifier in the current subject set into a class to form a corresponding first set; and in each of the first sets, only the assessment data element with the most recent time is retained, and the remaining assessment data elements are deleted as duplicate data elements;

[0046] And the question and answer sets D of the questionnaire dataset u As the current question and answer set; and group one or more question and answer data elements with the same question identifier in the current question and answer set into a class to form a corresponding second set; and in each second set, only the question and answer data element with the most recent time is retained, and the remaining question and answer data elements are deleted as duplicate data elements.

[0047] Preferably, the data processing service is specifically used to: convert the timestamps of all data elements in the basic version-road test dataset to a preset first timestamp format when performing time standardization processing on the basic version-road test dataset; and convert the time information of all timestamps that have completed the format conversion to the time information of a preset first time zone.

[0048] Preferably, the data processing service is specifically used when the corresponding tag cluster is obtained by setting spatiotemporal tags based on the test trajectory:

[0049] The overall trajectory time period of the test trajectory is identified to obtain the corresponding overall test time period; and the overall test time period is divided into multiple test sub-time periods according to a preset test time period segmentation rule; and a unique label is assigned to each test sub-time period as the corresponding time period label; and all the obtained time period labels form the corresponding time period label set;

[0050] The high-precision map area along the test trajectory is divided into multiple test areas according to a preset test area segmentation rule; a unique label is assigned to each test area as the corresponding area label; and all the obtained area labels form the corresponding area label set.

[0051] The obtained time period label set and the region label set form the corresponding label cluster.

[0052] Preferably, the data processing service is specifically used when the data elements of the system dataset and the sensor dataset are labeled based on the label cluster to obtain the corresponding spatiotemporal version-label dataset, which is then stored in the data storage system:

[0053] Each of the acquisition sequences A in the system dataseti As the current sequence; and perform a round of traversal on all the system data elements of the current sequence; during this round of traversal, the system data element currently traversed is taken as the current system data element; the acquisition timestamp of the current system data element is taken as the current timestamp; the time period label of the test sub-time period to which the current timestamp belongs is taken as the current time period label; the trajectory point data element whose trajectory point timestamp is closest to the current timestamp in the test trajectory is taken as the current trajectory point; the region label of the test area to which the current trajectory point belongs is taken as the current region label; and the acquisition timestamp of the current system data element, the corresponding current time period label, and the current region label are combined to form a corresponding label data element; and at the end of this round of traversal, all the label data elements obtained in this round of traversal are combined to form an A corresponding to the current sequence. i Tag set;

[0054] Each of the acquisition sequences B in the sensor dataset j As the current sequence; and perform a round of traversal on all the sensor data elements of the current sequence; during this round of traversal, the sensor data element currently traversed is taken as the current sensor data element; the acquisition timestamp of the current sensor data element is taken as the current timestamp; the time period label of the test sub-time period to which the current timestamp belongs is taken as the current time period label; the trajectory point data element whose trajectory point timestamp is closest to the current timestamp is taken as the current trajectory point; the area label of the test area to which the current trajectory point belongs is taken as the current area label; and the acquisition timestamp of the current sensor data element, the corresponding current time period label, and the current area label are combined to form a corresponding label data element; and at the end of this round of traversal, all the label data elements obtained in this round of traversal are combined to form a B corresponding to the current sequence. j Tag set;

[0055] And from the obtained Na A i The tag set forms the corresponding system tag set; and the obtained Nb B j The tag set forms the corresponding sensor tag set; and the tag cluster, the system tag set, and the sensor tag set form the corresponding spatiotemporal version-tag dataset, which is stored in the data storage system.

[0056] Preferably, the data storage system is further configured to use a distributed file system to store the sensor dataset of the basic version-road test dataset for each road test task; and to use a distributed database to store the system dataset, the evaluation dataset, and the questionnaire dataset of the basic version-road test dataset for each road test task; and to use a distributed database to store the spatiotemporal version-label dataset for each road test task.

[0057] This invention provides a data acquisition system for autonomous vehicle testing. As described above, the system includes: a vehicle data acquisition device, an evaluation client, a questionnaire page, a data processing service, and a data storage system. The vehicle data acquisition device is used to acquire system data items and sensor data items of the autonomous driving system, and to acquire data on the test trajectory of the vehicle under test. The evaluation client is used to acquire evaluation data provided by evaluators. The questionnaire page is used to provide subjective experience questionnaires to evaluators and to acquire data on the evaluators' responses. The data processing service is used to update the basic version of the road test dataset for the road test task based on the real-time acquired data sent by the vehicle data acquisition device, the evaluation client, and the questionnaire page. A subset of this dataset contains the system dataset, sensor dataset, evaluation dataset, and questionnaire dataset. The data processing service is also used to set up label clusters based on the test trajectory and to label the system / sensor datasets based on the label clusters to obtain the corresponding spatiotemporal version of the label dataset. The data storage system is used to store the basic version of the road test dataset and the spatiotemporal version of the label dataset corresponding to each road test task. This invention increases the dimension of human factors data and improves the feature richness of road test data by adding evaluation and questionnaire datasets; and provides necessary data support for a retrieval mechanism that can locate data blocks based on tags by adding spatiotemporal label clusters and their corresponding spatiotemporal version-label datasets. Attached Figure Description

[0058] Figure 1 A module structure diagram of an autonomous vehicle test data acquisition system provided in an embodiment of the present invention;

[0059] Figure 2 This is a schematic diagram of the basic version - road test dataset and the spatiotemporal version - label dataset for each road test task provided in the embodiments of the present invention. Detailed Implementation

[0060] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0061] This invention provides an autonomous vehicle test data acquisition system, such as... Figure 1 The module structure diagram of an autonomous vehicle test data acquisition system provided in an embodiment of the present invention is shown, mainly including: vehicle acquisition device 1, evaluation client 2, questionnaire page 3, data processing service 4, and data storage system 5. The data processing service 4 is connected to the vehicle acquisition device 1, evaluation client 2, questionnaire page 3, and data storage system 5 respectively.

[0062] (I) Vehicle Data Collection Equipment 1:

[0063] In this embodiment of the invention, the vehicle data acquisition device 1 is installed on the vehicle under test. Each road test task for each vehicle under test corresponds to a task identifier.

[0064] Vehicle data acquisition device 1 is used to store system monitoring lists and sensor monitoring lists.

[0065] Here, the system monitoring list in this embodiment of the invention consists of system data types from multiple autonomous driving systems. When the sensor monitoring list in this embodiment of the invention is not empty, it consists of one or more sensor types; sensor types include at least cameras, LiDAR, millimeter-wave radar, ultrasonic sensors, and infrared sensors.

[0066] The vehicle data acquisition device 1 is also used to collect data values ​​of multiple types of system data specified in the system monitoring list in real time through the vehicle CAN bus during the road test of the vehicle under test, generate corresponding system data element sets, collect data files generated by multiple types of sensors specified in the sensor monitoring list in real time, generate corresponding sensor data element sets, and send the system data element sets and sensor data element sets to the data processing service 4.

[0067] Here, the system data element set in this embodiment of the invention includes multiple system data elements; the system data elements in the system data element set correspond one-to-one with the system data types in the system monitoring list. When the sensor data element set in this embodiment of the invention is not empty, it consists of one or more sensor data elements; the sensor data elements in the sensor data element set correspond one-to-one with the sensor types in the sensor monitoring list.

[0068] The vehicle data acquisition device 1 is also used to collect the latitude and longitude information of the vehicle under test, as well as the information of the road, road segment, lane and intersection where the vehicle is located, in real time based on the built-in positioning device and the high-precision road map during the road test of the vehicle under test, generate the corresponding trajectory point data elements and send them to the data processing service 4.

[0069] (II) Evaluation Client 2:

[0070] The evaluation client 2 is used to collect data on the evaluation scores and comment texts made by the evaluators on each evaluation item they are responsible for during the road test of the vehicle under test, generate corresponding evaluation data elements, and send them to the data processing service 4.

[0071] It should be noted that each road test task of the vehicle under test in this embodiment of the invention includes multiple evaluation subjects, and each road test task will be assigned a corresponding evaluation personnel group.

[0072] The evaluation subjects in this embodiment of the invention include at least the following subject types: illegal / irregular behavior evaluation subject, lateral driving behavior evaluation subject, longitudinal driving behavior evaluation subject, other vehicle game behavior evaluation subject, driving comfort evaluation subject, excellent handling behavior evaluation subject, and vehicle signal light control behavior evaluation subject; each evaluation subject consists of one or more evaluation items, and each evaluation item corresponds to an evaluation item identifier and a set of evaluation requirements.

[0073] The evaluation team in this embodiment of the invention consists of one or more evaluation personnel; each evaluation personnel corresponds to a personnel identifier and is responsible for one or more evaluation subjects.

[0074] In this embodiment of the invention, the road test of the vehicle under test requires a team of evaluators to be on board. The evaluation client 2 in this embodiment corresponds one-to-one with each evaluator, specifically a client program installed on the evaluator's mobile evaluation device; mobile evaluation devices include mobile phones, tablets, laptops, and in-vehicle network computers. During the road test of the vehicle under test, the evaluators use the evaluation client 2 to evaluate, score, and comment on each evaluation item according to its requirements.

[0075] (III) Questionnaire Page 3:

[0076] Questionnaire page 3 is used to provide subjective experience questionnaires to the evaluators, and to collect data on the text of the answers given by the current evaluators to each question item in the subjective experience questionnaire to generate corresponding question and answer data elements. The question and answer data element set composed of all the question and answer data elements corresponding to the subjective experience questionnaire is sent to the data processing service 4.

[0077] It should be noted that the questionnaire page 3 in this embodiment of the invention is a web service page deployed on the Internet; the evaluators use mobile evaluation devices to access the questionnaire page 3 and answer questions in the subjective experience questionnaire through the current page.

[0078] The subjective experience questionnaire in this invention is a questionnaire for collecting vehicle ride experience information from evaluators. The subjective experience questionnaire consists of multiple question sets; the question sets include at least the following types: vehicle safety, vehicle efficiency, vehicle comfort, vehicle reliability, purchase intention, and vehicle usability. Each question set consists of multiple question items; each question item corresponds to a set of question identifiers and question text.

[0079] (iv) Data Processing Service 4:

[0080] Data processing service 4 is used to plan the corresponding basic version of the road test dataset for the current task on the data storage system 5 before the road test task of the vehicle under test is executed; and to set the basic information of the basic version of the road test dataset.

[0081] Data processing service 4 is also used to update the system dataset and sensor dataset of the basic version-road test dataset in real time based on the system data element set and sensor data element set received each time during the road test of the vehicle under test; to update the evaluation dataset of the basic version-road test dataset in real time based on the evaluation data element received each time; to update the test trajectory of the basic version-road test dataset in real time based on the trajectory point data element received each time; and to update the questionnaire dataset of the basic version-road test dataset in real time based on the received question and answer data element set.

[0082] Data processing service 4 is also used to perform data element deduplication on the basic version-road test dataset after the road test task of the tested vehicle is completed; then perform time standardization on the basic version-road test dataset; then set spatiotemporal labels based on the test trajectory to obtain the corresponding label cluster; and then label the data elements of the system dataset and sensor dataset based on the label cluster to obtain the corresponding spatiotemporal version-label dataset, which is then stored in the data storage system 5.

[0083] like Figure 2 The diagram illustrates the basic version of the road test dataset and the spatiotemporal version of the label dataset for each road test task provided in this embodiment of the invention. The basic version of the road test dataset in this embodiment includes basic information, test trajectory, system dataset, sensor dataset, evaluation dataset, and questionnaire dataset. Specifically:

[0084] 1) Basic information includes task identifier, test date, test vehicle identifier, and evaluation group;

[0085] The evaluation team consists of one or more personnel identifiers;

[0086] 2) When the test trajectory is not empty, it is formed by sequentially sorting one or more trajectory point data elements; wherein, the trajectory point data elements include trajectory point timestamp, latitude and longitude coordinates, road signs, road segment signs, and lane signs;

[0087] 3) The system dataset consists of multiple acquisition sequences A i composition;

[0088] Where 1≤i≤Na, the total number Na is consistent with the total number of system data types in the system monitoring list; the collection sequence A i Each data acquisition sequence A corresponds one-to-one with the system data type. i When not empty, it is composed of one or more corresponding system data elements sorted in chronological order; system data elements include collection timestamp and data value;

[0089] 4) The sensor dataset consists of multiple acquisition sequences B j composition;

[0090] Where 1≤j≤Nb, the total number Nb is consistent with the total number of sensor types in the sensor monitoring list; acquisition sequence B j Each acquisition sequence B corresponds one-to-one with the sensor type. j When not empty, it is composed of one or more corresponding sensor data elements sorted in chronological order; sensor data elements include acquisition timestamp and data file;

[0091] 5) The assessment dataset consists of multiple subject sets C k composition;

[0092] Where 1≤k≤Nc, the total number Nc is consistent with the total number of evaluation subjects in the current road test task; subject set C k Each subject corresponds one-to-one with the assessment subjects; each subject set C k When not empty, it consists of one or more corresponding evaluation data elements; each subject set C k The total number of evaluation data elements is consistent with the total number of evaluation items for the corresponding evaluation subject; the evaluation data elements include evaluation item identifier, personnel identifier, evaluation period, evaluation score, and comment text; the evaluation period consists of a set of corresponding evaluation start timestamp and evaluation end timestamp;

[0093] 6) The questionnaire dataset consists of multiple question-and-answer sets D u composition;

[0094] Where 1≤u≤Nd, the total number Nd is consistent with the total number of questions in the subjective experience questionnaire; the question-and-answer set D u Each question and answer set D corresponds one-to-one with a question set; u When not empty, it consists of one or more corresponding question-and-answer data elements; each question-and-answer set Du The total number of question-and-answer data elements is consistent with the total number of question items in the corresponding question set; the question-and-answer data elements include question identifier, personnel identifier, response timestamp, and answer text.

[0095] like Figure 2 As shown, the spatiotemporal version of the tag dataset in this embodiment of the invention includes a tag cluster, a system tag set, and a sensor tag set. Specifically:

[0096] 1) The tag cluster consists of a time period tag set and a region tag set;

[0097] The time period label set consists of multiple time period labels; the region label set consists of multiple region labels.

[0098] 2) The system label set corresponds to the system dataset, consisting of Na A's. i Tag set composition;

[0099] Among them, A i Tag set and acquisition sequence A i One-to-one correspondence; A i A tag set consists of multiple tag data elements; A i The tag data elements of the tag set and the corresponding acquisition sequence A i The system data elements correspond one-to-one; A i Each tag data element in the tag set consists of a corresponding collection timestamp, time period tag, and region tag; A i The collection timestamp of each tag data element in the tag set is matched with the collection timestamp of the corresponding system data element; that is, the embodiments of the present invention complete the data association of a pair of mutually matching system data elements and tag data elements through the "collection timestamp".

[0100] 3) The sensor tag set corresponds to the sensor dataset, consisting of Nb B... j Tag set composition;

[0101] Among them, B j Tag set and acquisition sequence B j One-to-one correspondence; B j A tag set consists of multiple tag data elements; B j The tag data elements of the tag set and the corresponding acquisition sequence B j The sensor data elements correspond one-to-one; B j Each tag data element in the tag set also consists of a corresponding collection timestamp, time period tag, and region tag; B j The acquisition timestamp of each tag data element in the tag set is matched with the acquisition timestamp of the corresponding sensor data element; that is, the embodiments of the present invention complete the data association of a pair of matching sensor data elements and tag data elements through the "acquisition timestamp".

[0102] In a specific implementation of this invention, the data processing service 4 is specifically used to set the task identifier, test date, test vehicle identifier, and evaluation group corresponding to the basic information of the basic version-road test dataset when setting the basic information: based on the task identifier corresponding to the current road test task, the specified test date, the vehicle identifier of the tested vehicle, and the identifiers of all personnel in the evaluation group.

[0103] In another specific implementation of this invention, the data processing service 4 is specifically used to: when updating the system dataset and sensor dataset of the basic version-road test dataset in real time based on the system data element set and sensor data element set received each time; use the currently received system data element set and sensor data element set as the corresponding first set and second set; and add each system data element in the first set to the corresponding acquisition sequence A in the system dataset. i Add to the second set; and add each sensor data element in the second set to the corresponding acquisition sequence B in the sensor dataset. j Add to.

[0104] In another specific implementation of this invention, the data processing service 4 is specifically used to, when updating the evaluation dataset of the basic version-road test dataset in real time based on each received evaluation data element: add the currently received evaluation data element to the corresponding subject set C in the evaluation dataset. k Add to.

[0105] In another specific implementation of this invention, the data processing service 4 is specifically used to add the currently received trajectory point data elements to the test trajectory when updating the test trajectory of the basic version-road test dataset in real time based on the trajectory point data elements received each time.

[0106] In another specific implementation of this invention, the data processing service 4 is specifically used to, when updating the questionnaire dataset of the basic version-road test dataset in real time based on the received question-and-answer data element set: to add each question-and-answer data element in the currently received question-and-answer data element set to the corresponding question-and-answer set D in the questionnaire dataset. u Add to.

[0107] In another specific implementation of this invention, the data processing service 4 is specifically used when performing data element deduplication processing on the basic version-road test dataset:

[0108] Step A1, combine the subject sets C of the assessment dataset. kThis is the current subject set; one or more assessment data elements with the same assessment item identifier in the current subject set are grouped into a class to form the corresponding first set; and in each first set, only the assessment data element with the most recent time is retained, and the remaining assessment data elements are deleted as duplicate data elements;

[0109] Step A2, and then combine the question and answer sets D of the questionnaire dataset. u This serves as the current question and answer set; one or more question and answer data elements with the same question identifier in the current question and answer set are grouped into a class to form a corresponding second set; and in each second set, only the most recent question and answer data element is retained, and the remaining question and answer data elements are deleted as duplicate data elements.

[0110] In another specific implementation of this invention, the data processing service 4 is specifically used to: convert the timestamps of all data elements in the basic version-road test dataset according to a preset first timestamp format when performing time standardization processing on the basic version-road test dataset; and convert the time information of all timestamps that have completed the format conversion into the time information of a preset first time zone.

[0111] Here, the first timestamp format of this embodiment of the invention is a pre-set timestamp data format, such as the Unix timestamp format; the first time zone is a pre-set time zone, such as the UTC+8 time zone.

[0112] In another specific implementation of this invention, the data processing service 4 is specifically used when obtaining the corresponding tag cluster by setting spatiotemporal tags based on the test trajectory:

[0113] Step B1: Identify the overall test time period of the test trajectory to obtain the corresponding overall test time period; divide the overall test time period into multiple test sub-time periods according to the preset test time period segmentation rules; assign a unique label to each test sub-time period as the corresponding time period label; and form the corresponding time period label set by all the obtained time period labels.

[0114] Here, the test time segmentation rule in this embodiment of the invention is a pre-set and dynamically configurable time segmentation rule that can be customized based on actual application needs; for example, dividing the overall test time into multiple test sub-times with a time segment length of 20 minutes.

[0115] Step B2 involves dividing the high-precision map area along the test trajectory into multiple test areas according to the preset test area segmentation rules; assigning a unique label to each test area as the corresponding area label; and forming a corresponding area label set from all the obtained area labels.

[0116] Here, the test area segmentation rule of this embodiment of the invention is a pre-set and dynamically configurable spatial segmentation rule that can be customized based on actual application needs; for example, all road segments and intersections along the test trajectory are connected in series to form a continuous road segment / intersection chain space, and then the road segment / intersection chain space is divided into multiple continuous subspaces in the manner that each road segment is a sub-region, and each subspace is regarded as a test area;

[0117] Step B3, and the obtained time period label set and area label set are combined to form the corresponding label cluster.

[0118] In another specific implementation of this invention, the data processing service 4 is specifically used to store the corresponding spatiotemporal version-labeled dataset into the data storage system 5 when the data elements of the system dataset and sensor dataset are labeled based on the label cluster.

[0119] Step C1, collect each acquisition sequence A from the system dataset. i As the current sequence; and perform a round of traversal on all system data elements of the current sequence; during this round of traversal, the system data element currently traversed is taken as the current system data element; the acquisition timestamp of the current system data element is taken as the current timestamp; the time period label of the test sub-time period to which the current timestamp belongs is taken as the current time period label; the trajectory point data element whose trajectory point timestamp is closest to the current timestamp in the test trajectory is taken as the current trajectory point; the region label of the test area to which the current trajectory point belongs is taken as the current region label; and a corresponding label data element is formed by the acquisition timestamp of the current system data element, the corresponding current time period label, and the current region label; and at the end of this round of traversal, all the label data elements obtained in this round of traversal form an A corresponding to the current sequence. i Tag set;

[0120] Step C2, convert each acquisition sequence B of the sensor dataset j As the current sequence; and perform a round of traversal on all sensor data elements in the current sequence; during this round of traversal, the sensor data element currently traversed is taken as the current sensor data element; the acquisition timestamp of the current sensor data element is taken as the current timestamp; the time period label of the test sub-time period to which the current timestamp belongs is taken as the current time period label; the trajectory point data element whose trajectory point timestamp is closest to the current timestamp is taken as the current trajectory point; the region label of the test area to which the current trajectory point belongs is taken as the current region label; and a corresponding label data element is formed by the acquisition timestamp of the current sensor data element, the corresponding current time period label, and the current region label; and at the end of this round of traversal, all the label data elements obtained in this round of traversal form a B corresponding to the current sequence. j Tag set;

[0121] Step C2, and from the obtained Na A i The tag set forms the corresponding system tag set; and the obtained Nb B j The tag set forms the corresponding sensor tag set; and the tag cluster, system tag set and sensor tag set form the corresponding spatiotemporal version of the tag dataset, which is stored in the data storage system 5.

[0122] (V) Data Storage System 5:

[0123] Data storage system 5 is used to store the basic version - road test dataset and the spatiotemporal version - label dataset corresponding to each road test task.

[0124] Data storage system 5 is also used to store the sensor dataset of the base version of the road test dataset for each road test task using a distributed file system (DFS); to store the system dataset, evaluation dataset, and questionnaire dataset of the base version of the road test dataset for each road test task using a distributed database (DDB); and to store the spatiotemporal version of the label dataset for each road test task using a distributed database.

[0125] Here, the distributed file system used in this embodiment of the invention includes the Hadoop Distributed File System; the distributed database used in this embodiment of the invention includes the Apache ShardingSphere system.

[0126] This invention provides a data acquisition system for autonomous vehicle testing. As described above, the system includes: a vehicle data acquisition device, an evaluation client, a questionnaire page, a data processing service, and a data storage system. The vehicle data acquisition device is used to acquire system data items and sensor data items of the autonomous driving system, and to acquire data on the test trajectory of the vehicle under test. The evaluation client is used to acquire evaluation data provided by evaluators. The questionnaire page is used to provide subjective experience questionnaires to evaluators and to acquire data on the evaluators' responses. The data processing service is used to update the basic version of the road test dataset for the road test task based on the real-time acquired data sent by the vehicle data acquisition device, the evaluation client, and the questionnaire page. A subset of this dataset contains the system dataset, sensor dataset, evaluation dataset, and questionnaire dataset. The data processing service is also used to set up label clusters based on the test trajectory and to label the system / sensor datasets based on the label clusters to obtain the corresponding spatiotemporal version of the label dataset. The data storage system is used to store the basic version of the road test dataset and the spatiotemporal version of the label dataset corresponding to each road test task. This invention increases the dimension of human factors data and improves the feature richness of road test data by adding evaluation and questionnaire datasets; and provides necessary data support for a retrieval mechanism that can locate data blocks based on tags by adding spatiotemporal label clusters and their corresponding spatiotemporal version-label datasets.

[0127] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0128] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented in hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0129] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A test data acquisition system for autonomous vehicles, characterized in that, The system includes: vehicle data collection equipment, evaluation client, questionnaire page, data processing service, and data storage system; The data processing service is connected to the vehicle data collection device, the evaluation client, the questionnaire page, and the data storage system, respectively. The vehicle data acquisition equipment is installed on the vehicle being tested; The vehicle data acquisition equipment is used to store the system monitoring list and the sensor monitoring list; The vehicle data acquisition device is also used to collect data values ​​of multiple types of system data specified in the system monitoring list in real time through the vehicle CAN bus during the road test of the vehicle under test, generate corresponding system data element sets, collect data files generated by multiple types of sensors specified in the sensor monitoring list in real time, generate corresponding sensor data element sets, and send the system data element sets and the sensor data element sets to the data processing service. The vehicle data acquisition device is also used to collect the latitude and longitude information of the vehicle under test, as well as the information of the road, road segment, lane and intersection where the vehicle is located, in real time based on the built-in positioning device and the high-precision road map, and generate corresponding trajectory point data elements to send to the data processing service. The evaluation client is used to collect data on the evaluation scores and comment texts made by the evaluators on each evaluation item they are responsible for during the road test of the vehicle under test, generate corresponding evaluation data elements, and send them to the data processing service. The questionnaire page is used to provide the evaluator with a subjective experience questionnaire, and to collect data on the answer text of the current evaluator on each question item of the subjective experience questionnaire to generate corresponding question and answer data elements. The question and answer data elements corresponding to the subjective experience questionnaire are combined into a question and answer data element set and sent to the data processing service. The data processing service is used to plan a corresponding basic version of the road test dataset for the current task on the data storage system before the road test task of the vehicle under test is executed; and to set the basic information of the basic version of the road test dataset. The data processing service is also used to, during the road test of the vehicle under test, update the system dataset and sensor dataset of the basic version-road test dataset in real time based on the system data element set and sensor data element set received each time; update the evaluation dataset of the basic version-road test dataset in real time based on the evaluation data element received each time; update the test trajectory of the basic version-road test dataset in real time based on the trajectory point data element received each time; and update the questionnaire dataset of the basic version-road test dataset in real time based on the received question and answer data element set. The data processing service is also used to: first, perform data element deduplication on the basic version-road test dataset after the road test task of the vehicle under test is completed; then, perform time standardization on the basic version-road test dataset; then, set spatiotemporal labels based on the test trajectory to obtain the corresponding label cluster; and then, label the data elements of the system dataset and the sensor dataset based on the label cluster to obtain the corresponding spatiotemporal version-label dataset and store it in the data storage system. The data storage system is used to store the basic version-road test dataset and the spatiotemporal version-label dataset corresponding to each road test task.

2. The autonomous vehicle test data acquisition system according to claim 1, characterized in that, The system monitoring list consists of system data types from multiple autonomous driving systems; When the sensor monitoring list is not empty, it consists of one or more sensor types; the sensor types include cameras, lidar, millimeter-wave radar, ultrasonic sensors, and infrared sensors; The system data element set includes multiple system data elements; the system data elements in the system data element set correspond one-to-one with the system data types in the system monitoring list; When the sensor data element set is not empty, it consists of one or more sensor data elements; the sensor data elements in the sensor data element set correspond one-to-one with the sensor types in the sensor monitoring list; Each road test task of each of the aforementioned vehicles corresponds to a task identifier, and each road test task includes multiple evaluation subjects, and each road test task will be assigned a corresponding evaluation personnel group. The assessment subjects include at least the following subject types: illegal / irregular behavior assessment subject, lateral driving behavior assessment subject, longitudinal driving behavior assessment subject, other vehicle interaction behavior assessment subject, driving comfort assessment subject, excellent handling behavior assessment subject, and vehicle signal light control behavior assessment subject; each assessment subject consists of one or more assessment items, and each assessment item corresponds to an assessment item identifier and a set of assessment requirements; The evaluation team consists of one or more of the evaluation personnel; each evaluation personnel corresponds to a personnel identifier and is responsible for one or more of the evaluation subjects; During the road test of the vehicle under test, the vehicle must be accompanied by the evaluation team; the evaluation client corresponds one-to-one with the evaluation personnel, specifically a client program installed on the mobile evaluation device of the evaluation personnel; the mobile evaluation device includes mobile phone, tablet computer, laptop computer, and in-vehicle network computer; during the road test of the vehicle under test, the evaluation personnel use the evaluation client to evaluate, score, and comment on the current evaluation item according to the evaluation requirements of each evaluation item; The questionnaire page is a web service page deployed on the Internet; the evaluator uses the mobile evaluation device to access the questionnaire page and answers the questions of the subjective experience questionnaire through the current page; the subjective experience questionnaire is a questionnaire for collecting information on the vehicle riding experience for the evaluator. The subjective experience questionnaire consists of multiple sets of questions; the types of the question sets include at least vehicle safety issues, vehicle efficiency issues, vehicle comfort issues, vehicle credibility issues, purchase intention issues, and vehicle usability issues; each set of questions consists of multiple question items; each question item corresponds to a set of question identifiers and question text; The basic version of the road test dataset includes the basic information, the test trajectory, the system dataset, the sensor dataset, the evaluation dataset, and the questionnaire dataset; The basic information includes task identifier, test date, test vehicle identifier, and evaluation group; the evaluation group consists of one or more personnel identifiers. When the test trajectory is not empty, it is formed by sequentially sorting one or more trajectory point data elements; the trajectory point data elements include trajectory point timestamps, latitude and longitude coordinates, road signs, road segment signs, and lane signs. The system dataset consists of multiple acquisition sequences A i Composition, 1≤i≤Na, the total number Na is consistent with the total number of system data types in the system monitoring list; the collected sequence A i Each of the acquired sequences A corresponds one-to-one with the system data types. i When not empty, it is formed by sorting one or more of the corresponding system data elements in chronological order; the system data elements include the collection timestamp and the data value; The sensor dataset consists of multiple acquisition sequences B j Composition, 1≤j≤Nb, the total number Nb is consistent with the total number of sensor types in the sensor monitoring list; the acquisition sequence B j Each of the sensor types corresponds one-to-one; each of the acquisition sequences B j When not empty, it is formed by sorting one or more corresponding sensor data elements in chronological order; the sensor data elements include acquisition timestamp and data file; The evaluation dataset consists of multiple subject sets C k Composition, 1≤k≤Nc, the total number Nc is consistent with the total number of evaluation subjects in the current road test task; the subject set C k Each of the subject sets C corresponds one-to-one with the assessment subjects; k When not empty, it consists of one or more of the corresponding evaluation data elements; each subject set C k The total number of evaluation data elements is consistent with the total number of evaluation items for the corresponding evaluation subject; the evaluation data elements include evaluation item identifier, personnel identifier, evaluation period, evaluation score, and comment text; the evaluation period consists of a set of corresponding evaluation start timestamps and evaluation end timestamps; The questionnaire dataset consists of multiple question-and-answer sets D u Composition, 1≤u≤Nd, the total number Nd is consistent with the total number of questions in the subjective experience questionnaire; the question-and-answer set D u Each question and answer set D corresponds one-to-one with the question set; u When not empty, it consists of one or more of the corresponding question-and-answer data elements; each question-and-answer set D u The total number of question-and-answer data elements is consistent with the total number of question items in the corresponding question set; the question-and-answer data elements include question identifier, personnel identifier, response timestamp, and answer text. The spatiotemporal version of the tag dataset includes the tag cluster, system tag set, and sensor tag set; The tag cluster consists of a time period tag set and a region tag set; the time period tag set consists of multiple time period tags; the region tag set consists of multiple region tags; The system label set corresponds to the system dataset, and consists of Na A's. i The tag set consists of; the A i Tag set and the collected sequence A i One-to-one correspondence; the A i The tag set consists of multiple tag data elements; the A i The tag data elements of the tag set and the corresponding acquisition sequence A i The system data elements correspond one-to-one; the A i Each tag data element in the tag set consists of a corresponding collection timestamp, the time period tag, and the region tag; the A i The collection timestamp of each tag data element in the tag set is matched with the collection timestamp of the corresponding system data element; The sensor tag set corresponds to the sensor dataset, and consists of Nb B tags. j The tag set consists of; the B j Tag set and the acquired sequence B j One-to-one correspondence; the B mentioned j The tag set consists of multiple tag data elements; the B j The tag data elements of the tag set and the corresponding acquisition sequence B j The sensor data elements correspond one-to-one; the B j Each tag data element in the tag set also consists of a corresponding collection timestamp, the time period tag, and the region tag; the B j The acquisition timestamp of each tag data element in the tag set is matched with the acquisition timestamp of the corresponding sensor data element.

3. The autonomous vehicle test data acquisition system according to claim 2, characterized in that, The data processing service is specifically used to set the task identifier, test date, test vehicle identifier, and evaluation group corresponding to the basic information when setting the basic information of the basic version-road test dataset: based on the task identifier corresponding to the current road test task, the specified test date, the vehicle identifier of the vehicle under test, and the identifiers of all personnel in the evaluation personnel group.

4. The autonomous vehicle test data acquisition system according to claim 2, characterized in that, The data processing service is specifically used to update the system dataset and sensor dataset of the basic version-road test dataset in real time based on the system data element set and sensor data element set received each time: the system data element set and sensor data element set received at that time are used as the corresponding first set and second set; And each of the system data elements in the first set is assigned to the corresponding acquisition sequence A in the system dataset. i Add to the second set; and add each of the sensor data elements in the second set to the corresponding acquisition sequence B in the sensor dataset. j Add to; The data processing service is specifically used to, when updating the evaluation dataset of the basic version-road test dataset in real time based on the evaluation data elements received each time: add the currently received evaluation data elements to the corresponding subject set C in the evaluation dataset. k Add to; The data processing service is specifically used to add the currently received trajectory point data elements to the test trajectory when the test trajectory of the basic version-road test dataset is updated in real time based on the trajectory point data elements received each time. The data processing service is specifically used to, when updating the questionnaire dataset of the basic version-road test dataset in real time based on the received question-and-answer data element set: add each question-and-answer data element in the currently received question-and-answer data element set to the corresponding question-and-answer set D in the questionnaire dataset. u Add to.

5. The autonomous vehicle test data acquisition system according to claim 2, characterized in that, The data processing service is specifically used when performing data element deduplication on the basic version-road test dataset: The subject sets C of the evaluation dataset. k As the current subject set; and group one or more assessment data elements with the same assessment item identifier in the current subject set into a class to form a corresponding first set; and in each of the first sets, only the assessment data element with the most recent time is retained, and the remaining assessment data elements are deleted as duplicate data elements; And the question and answer sets D of the questionnaire dataset u As the current question and answer set; and group one or more question and answer data elements with the same question identifier in the current question and answer set into a class to form a corresponding second set; and in each second set, only the question and answer data element with the most recent time is retained, and the remaining question and answer data elements are deleted as duplicate data elements.

6. The autonomous vehicle test data acquisition system according to claim 2, characterized in that, The data processing service is specifically used to: convert the timestamps of all data elements in the basic version-road test dataset according to a preset first timestamp format when performing time standardization processing on the basic version-road test dataset; It also converts the time information of all timestamps that have completed the format conversion into the time information of the preset first time zone.

7. The autonomous vehicle test data acquisition system according to claim 2, characterized in that, The data processing service is specifically used when obtaining the corresponding tag cluster by setting spatiotemporal tags based on the test trajectory: The overall test time period is obtained by identifying the overall trajectory time period of the test trajectory. The overall test period is divided into multiple sub-test periods according to a preset test period segmentation rule; Each test sub-period is assigned a unique label as the corresponding period label; The obtained time period labels constitute the corresponding time period label set; The high-precision map area along the test trajectory is divided into multiple test areas according to the preset test area segmentation rules. Each test region is assigned a unique label as its corresponding region label; The obtained region labels constitute the corresponding region label set; The obtained time period label set and the region label set form the corresponding label cluster.

8. The autonomous vehicle test data acquisition system according to claim 7, characterized in that, The data processing service is specifically used when the data elements of the system dataset and the sensor dataset are labeled based on the label cluster to obtain the corresponding spatiotemporal version-label dataset, which is then stored in the data storage system: Each of the acquisition sequences A in the system dataset i As the current sequence; and perform a traversal of all the system data elements in the current sequence; During this round of traversal, the system data element currently being traversed is taken as the current system data element; The timestamp of the current system data element collection is used as the current timestamp; And the time period label of the test sub-time period to which the current timestamp belongs is used as the current time period label; The trajectory point data element whose timestamp is closest to the current timestamp in the test trajectory is taken as the current trajectory point; and the region label of the test area to which the current trajectory point belongs is taken as the current region label; The current system data element is composed of the collection timestamp, the corresponding current time period label, and the current region label, forming a corresponding label data element. At the end of this round of traversal, all the label data elements obtained in this round of traversal are combined to form an A corresponding to the current sequence. i Tag set; Each of the acquisition sequences B in the sensor dataset j As the current sequence; and perform a round of traversal on all the sensor data elements in the current sequence; During this round of traversal, the sensor data element currently being traversed is taken as the current sensor data element; The timestamp of the current sensor data element acquisition is used as the current timestamp; And the time period label of the test sub-time period to which the current timestamp belongs is used as the current time period label; The trajectory point data element whose timestamp is closest to the current timestamp in the test trajectory is taken as the current trajectory point; and the region label of the test area to which the current trajectory point belongs is taken as the current region label; A corresponding tag data element is formed by the acquisition timestamp of the current sensor data element, the corresponding current time period tag, and the current area tag; At the end of this round of traversal, all the label data elements obtained in this round of traversal are combined to form a B corresponding to the current sequence. j Tag set; And from the obtained Na A i The tag set constitutes the corresponding system tag set; and the obtained Nb B j The tag set forms the corresponding sensor tag set; and the tag cluster, the system tag set, and the sensor tag set form the corresponding spatiotemporal version-tag dataset, which is stored in the data storage system.

9. The autonomous vehicle test data acquisition system according to claim 2, characterized in that, The data storage system is also used to store the sensor dataset of the base version-road test dataset for each road test task using a distributed file system; and to store the system dataset, the evaluation dataset, and the questionnaire dataset of the base version-road test dataset for each road test task using a distributed database. A distributed database is used to store the spatiotemporal version-label datasets for each road test task.

Images