Vehicle positioning accuracy evaluation method, device and electronic equipment

Abstract

The embodiments of the present application provide a method, device, and electronic device for evaluating vehicle positioning accuracy, wherein the method includes: extracting first lane centerline data from a vector map of a lane; collecting first positioning trajectory data of the vehicle regarding the lane in different positioning modes; converting the first lane centerline data and the first positioning trajectory data into data of the same format to obtain second lane centerline data and second positioning trajectory data; obtaining the deviation distance between the positioning trajectory in each positioning mode and the lane centerline of the lane based on the second positioning trajectory data and the second lane centerline data in each positioning mode; and evaluating the positioning accuracy of the vehicle in each positioning mode based on the deviation distance. The embodiments of the present application quantify the positioning accuracy of different positioning modes by obtaining the deviation distance between the positioning trajectory and the lane centerline in each positioning mode, thereby effectively improving the accuracy of vehicle positioning accuracy evaluation.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of autonomous driving technology, and in particular to a vehicle positioning accuracy evaluation method, device, and electronic equipment. Background Art

[0002] With the development and application of autonomous driving technology, the positioning accuracy of autonomous driving is a key part to ensure safe driving. Lane-level high-precision maps and centimeter-level high-precision positioning are often considered to be the basis for the realization of autonomous driving.

[0003] In autonomous driving scenarios, RTK (Real-time kinematic) signals may be affected by tall buildings on the ground, easily resulting in significant multipath effects. This will adversely affect the positioning results of positioning modes such as integrated navigation and SLAM (Simultaneous Localization and Mapping). Current accuracy evaluation schemes for autonomous vehicles are difficult to accurately evaluate the positioning accuracy of different positioning modes, resulting in the inability to effectively guarantee the positioning accuracy of autonomous vehicles.

[0004] Therefore, how to accurately evaluate the positioning accuracy of autonomous vehicles has become an urgent problem to be solved. Summary of the Invention

[0005] In view of the above problems, the embodiments of the present application provide a vehicle positioning accuracy evaluation method, device and electronic device to solve the problem that it is difficult for autonomous driving vehicles to accurately evaluate the positioning accuracy of different positioning modes.

[0006] In order to achieve the above objectives, this application provides the following technical solutions:

[0007] According to a first aspect of an embodiment of the present application, a vehicle positioning accuracy evaluation method is provided, comprising:

[0008] Extracting the centerline data of the first lane from the vector map of the lane;

[0009] Collecting first positioning trajectory data of the vehicle about the lane in different positioning modes;

[0010] Converting the first lane centerline data and the first positioning trajectory data into data of the same format to obtain second lane centerline data and second positioning trajectory data;

[0011] Based on the second positioning trajectory data and the second lane centerline data in each positioning mode, a deviation distance between the positioning trajectory in each positioning mode and the lane centerline of the lane is obtained; and based on the deviation distance, the positioning accuracy of the vehicle in each positioning mode is evaluated respectively.

[0012] In an embodiment of the present application, the above method is used to extract the lane centerline data of the vector map of the lane as the reference true value, and after converting the lane centerline data and the positioning trajectory data under each positioning mode into the same format, the deviation distance between the positioning trajectory and the lane centerline under each positioning mode is obtained, and the positioning accuracy of different positioning modes is quantified, thereby effectively improving the accuracy of vehicle positioning accuracy evaluation.

[0013] In one embodiment, the different positioning modes include any combination of combined navigation, laser synchronous positioning, visual synchronous positioning or fusion positioning modes.

[0014] In the embodiments of the present application, the above method is used to further illustrate the evaluation of the positioning accuracy of a specific positioning mode in an autonomous driving vehicle.

[0015] In one embodiment, before extracting the first lane centerline data in the lane vector map, the method further includes:

[0016] A digital orthophoto image of the lane is generated, and the digital orthophoto image is vectorized into a vector map of the lane.

[0017] In the embodiment of the present application, the above method can be used to truly restore the lane scene, improve the accuracy of the vector map, and further improve the evaluation accuracy of the positioning accuracy.

[0018] In one embodiment, converting the first lane centerline data and the first positioning trajectory data into data in the same format to obtain second lane centerline data and second positioning trajectory data includes:

[0019] Converting the first lane centerline data and the first positioning trajectory data into a preset format to form lane centerline discrete points and positioning trajectory discrete points respectively;

[0020] The obtaining, based on the second positioning trajectory data and the second lane centerline data in each positioning mode, respectively, the deviation distance between the positioning trajectory and the lane centerline in each positioning mode includes:

[0021] Obtain the deviation distance between the discrete points of the positioning trajectory and the discrete points of the lane centerline in each positioning mode; and

[0022] The deviation distance between the positioning trajectory and the lane centerline in each positioning mode is obtained based on the deviation distance between the discrete points of the positioning trajectory and the discrete points of the lane centerline in each positioning mode.

[0023] In the embodiment of the present application, the above method can be used to facilitate the identification of the deviation distance between the lane centerline data and the positioning trajectory data.

[0024] In one embodiment, obtaining the deviation distance between the discrete points of the positioning trajectory and the discrete points of the lane centerline in each positioning mode includes:

[0025] Using the vector map and lane centerline discrete points as a base map;

[0026] The positioning trajectory discrete points in each positioning mode are superimposed on the base map, and the deviation distances between the positioning trajectory discrete points in each positioning mode and the lane centerline discrete points are respectively obtained on the base map after the discrete points are superimposed.

[0027] In the embodiment of the present application, the above method can further improve the efficiency of obtaining the deviation distance between the discrete points of the positioning trajectory and the discrete points of the lane centerline in each positioning mode.

[0028] In one embodiment, after respectively evaluating the positioning accuracy of the vehicle in each positioning mode based on the deviation distance, the method further includes:

[0029] If the evaluation result of the positioning accuracy is lower than a preset threshold, a functional module causing a deviation error in the corresponding positioning mode is determined based on the evaluation result of the positioning accuracy.

[0030] In the embodiment of the present application, the above method can be used to determine the positioning function module where the deviation error occurs based on the evaluation result of the positioning accuracy, so as to improve the corresponding positioning algorithm in a targeted manner.

[0031] In one embodiment, determining the functional module causing the deviation error in the corresponding positioning mode based on the evaluation result of the positioning accuracy includes:

[0032] Randomly combine the positioning modes whose evaluation results are lower than the preset threshold;

[0033] Perform similarity judgment on the evaluation results of the positioning accuracy between the randomly combined positioning modes to obtain a similarity judgment result;

[0034] Based on the similarity judgment result, a positioning mode with the lowest similarity is selected from each positioning mode whose evaluation result is lower than a preset threshold, and a functional module causing a deviation error in the positioning mode is determined.

[0035] In the embodiment of the present application, the above method can be used to further refine the evaluation results of each positioning mode, and more specifically determine the positioning function module where the deviation error occurs.

[0036] According to a second aspect of an embodiment of the present application, a vehicle positioning accuracy evaluation device is provided, comprising:

[0037] an extraction module configured to extract first lane centerline data from a vector map of lanes;

[0038] A collection module configured to collect first positioning trajectory data of the vehicle about the lane in different positioning modes;

[0039] a conversion module configured to convert the first lane centerline data and the first positioning trajectory data into data of the same format to obtain second lane centerline data and second positioning trajectory data;

[0040] an acquisition module configured to acquire a deviation distance between the positioning trajectory in each positioning mode and the lane centerline of the lane based on the second positioning trajectory data and the second lane centerline data in each positioning mode; and

[0041] An evaluation module is configured to evaluate the positioning accuracy of the vehicle in each positioning mode based on the deviation distance.

[0042] In one embodiment, the positioning accuracy evaluation device is used to perform any possible implementation of the first aspect described above.

[0043] According to a third aspect of an embodiment of the present application, an electronic device is provided, comprising: a memory, a processor, and a computer program, wherein the computer program is stored in the memory, and the processor runs the computer program to execute the vehicle positioning accuracy evaluation method.

[0044] According to a fourth aspect of an embodiment of the present application, a computer-readable storage medium is provided, wherein the storage medium includes a computer program, and the computer program is used to implement the vehicle positioning accuracy evaluation method.

[0045] According to a fifth aspect of an embodiment of the present application, a computer program product is provided, which includes a computer program code. When the computer program code is run on a computer, the computer executes the vehicle positioning accuracy evaluation method.

[0046] According to the sixth aspect of an embodiment of the present application, a chip is provided, comprising a memory and a processor, wherein the memory is used to store a computer program, and the processor is used to call and run the computer program from the memory to execute the vehicle positioning accuracy evaluation method. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 A schematic diagram of a possible scenario provided in an embodiment of the present application;

[0048] Figure 2 A flow chart of a vehicle positioning accuracy evaluation method provided in an embodiment of the present application;

[0049] Figure 3This is a schematic diagram comparing the lane centerline and the manual driving combined navigation trajectory in an embodiment of the present application;

[0050] Figure 4 Schematic diagram of the comparison between the lane centerline and the Lidar-SLAM trajectory in the embodiment of the present application;

[0051] Figure 5 Schematic diagram of the comparison between the lane centerline and the V-SLAM trajectory in the embodiment of the present application;

[0052] Figure 6 This is a schematic diagram comparing the lane centerline and the fused positioning trajectory in an embodiment of the present application;

[0053] Figure 7 The lane centerline in the embodiment of this application is compared with the combined navigation, Lidar-SLAM, V-SLAM, and fusion positioning trajectory respectively;

[0054] Figure 8 A flow chart of another vehicle positioning accuracy evaluation method provided in an embodiment of the present application;

[0055] Figure 9 A flow chart of another vehicle positioning accuracy evaluation method provided in an embodiment of the present application;

[0056] Figure 10 A schematic diagram of the structure of a vehicle positioning accuracy evaluation device provided in an embodiment of the present application;

[0057] Figure 11 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0058] Common autonomous driving positioning solutions mainly include combined navigation, laser SLAM, visual SLAM, etc. In autonomous driving scenarios, RTK signals may be affected by tall buildings on the ground, resulting in obvious multipath effects, which have an adverse impact on the combined navigation and SLAM positioning results. As for the problem of positioning accuracy generated by RTK signals, the relevant technologies currently mainly propose to ensure RTK signal coverage or adjust the positioning solution according to the abnormality of RTK signals, but do not consider making an accurate evaluation of the specific positioning mode affected by the RTK signal. The positioning accuracy of different positioning modes is different, which will also lead to the positioning accuracy of autonomous driving vehicles not being effectively guaranteed. Therefore, there is an urgent need for a positioning true value that can be used as a reference to conduct qualitative and quantitative evaluation of the vehicle's ground positioning results.

[0059] In view of this, this embodiment provides a method for evaluating the positioning accuracy of a vehicle. By extracting the lane centerline data of the vector map of the lane as a high-precision true value, and relying on GIS (Geographic Information System) software and ROS (Robot Operating System) data exchange software, the vector map and positioning trajectory are converted into ShapeFile files of the same format, which can then be unified in the UTM (Universal Transverse Mercator Grid System) coordinate system to determine the deviation between the vector map lane centerline and the positioning trajectory of different topics, and qualitatively and quantitatively determine the accuracy and error distribution of different positioning schemes, thereby improving the positioning scheme in a targeted manner to promote the iterative development of autonomous driving positioning algorithms. It can be understood that different topics correspond to different positioning modes.

[0060] In order to make the purpose, technical solutions and advantages of the present application clearer, the technical solutions in the embodiments of the present application will be described in more detail below in conjunction with the drawings in the embodiments of the present application. In the drawings, the same or similar reference numerals throughout represent the same or similar parts or parts with the same or similar functions. The described embodiments are part of the embodiments of the present application, not all of the embodiments. The embodiments described below with reference to the drawings are exemplary and are intended to be used to explain the present application, and should not be understood as limitations on the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0061] Figure 1 A possible scenario diagram provided for an embodiment of the present application is as follows: Figure 1 As shown, it includes a moving autonomous driving vehicle, other vehicles and corresponding lane scenes, wherein the corresponding lane scene, for example, the lane in the port scene. In this embodiment, the autonomous driving vehicle is equipped with an on-board computing platform with processing performance, which can independently realize the positioning accuracy evaluation of the vehicle; the lane includes the lane centerline.

[0062] It is understandable that with the rapid rise of artificial intelligence, various artificial intelligence (AI) chips and software algorithms have also made repeated breakthroughs, and the research on autonomous vehicles has become more and more in-depth. Under normal circumstances, in order to ensure the normal operation of autonomous vehicles, autonomous vehicles will be equipped with an on-board computing platform with powerful processing performance and equipped with more peripherals, such as sensors, cameras, radars and other sensor devices. Moreover, when the autonomous driving function is turned on, the on-board computing platform and peripherals all work in full-speed mode to ensure the safety of vehicle operation as much as possible.

[0063] In some examples, autonomous vehicles can interact with terminal devices or the cloud in a way that allows the terminal devices or the cloud to undertake part of the data processing work, or the terminal devices or the cloud to undertake all the data processing work, where the terminal devices may include but are not limited to computers, smart phones, tablets, e-book readers, Moving Picture experts group audio layer III (MP3) players, Moving Picture experts group audio layer IV (MP4) players, portable computers, car computers, wearable devices, desktop computers, set-top boxes, smart TVs, etc.

[0064] The above is a brief description of the scenario diagram of this application. Figure 1 The autonomous driving vehicle in the embodiment is taken as an example to explain in detail the vehicle positioning accuracy evaluation method provided in the embodiment of the present application.

[0065] Please refer to Figure 2 , Figure 2 A vehicle positioning accuracy evaluation method provided in an embodiment of the present application includes steps S201-S205.

[0066] Step S201: extract the first lane centerline data from the lane vector map.

[0067] In this embodiment, the vector map of the lane can be vectorized by methods such as RTK manual field collection and drone aerial survey. After verification, it can be considered to have high accuracy at the centimeter level. It can be used as a relatively high-precision true value to judge the vehicle's driving positioning status, which can effectively improve the evaluation accuracy of the autonomous driving vehicle's positioning accuracy.

[0068] Combined with the actual application scenarios of ports, since ports are autonomous driving implementation scenarios that participate in actual production and operation operations in semi-open scenarios, there are many large-scale mechanical equipment, containers, fences, stone piers, etc. on the site. Due to the obstruction of images by ground objects, the RTK signal has obvious multipath effects, which will have a serious impact on the combined navigation and positioning results. It is necessary to use passive positioning methods such as Lidar-SLAM and V-SLAM, combined with the active positioning method of combined navigation, to form a fusion positioning strategy in which combined navigation, Lidar-SLAM, and V-SLAM are adaptively weighted according to the scene. However, the accuracy of each positioning method of vehicle autonomous driving is usually impossible to effectively determine. The vector map is vectorized by RTK manual field collection, drone aerial survey, etc., and after verification, it can be considered to have high accuracy at the centimeter level. It can be used as a relatively high-precision true value to judge the positioning status of the vehicle autonomous driving.

[0069] In a specific lane centerline extraction method, a GUI (Graphical User Interface) tool may be used in an operating system to extract lane centerline data of a vector map.

[0070] Vector images, also known as object-oriented images or drawing images, are defined as a series of points connected by lines. Graphic elements in a vector file are called objects. Each object is a self-contained entity, capable of possessing attributes such as color, shape, outline, size, and screen position. Vector images are drawn based on geometric properties. A vector can be a point or a line. These files take up less space, and because they contain independent, separate images, they can be freely reassembled.

[0071] In a preferred embodiment, a vector map of lanes is generated using photogrammetry technology, which can more realistically restore the lane scene and provide a reference for positioning truth. Specifically, before extracting the first lane centerline data from the vector map of lanes (step S201), the following steps are also included:

[0072] A digital orthophoto image of the lane is generated based on a photogrammetry technique, and the digital orthophoto image is vectorized into a vector map of the lane.

[0073] Photogrammetry, a type of engineering surveying technology, primarily relies on specialized photographic equipment, capturing the essential features of the measured object through digitized images. The entire photographic application process begins with sampling the object, upon which a data model is established. Data acquisition is crucial. To ensure the accuracy and comprehensiveness of the results, mainstream CCD (charge-coupled device) digital imaging technology is typically employed. After acquiring the data, it is categorized and digitally processed. In other embodiments, other technologies can also be employed to generate digital orthophotos of the lanes.

[0074] Digital orthophoto (DO) is based on aerial photographs or remote sensing images (monochrome / color), which are scanned, processed, and subjected to radiometric correction, differential correction, and mosaicking pixel by pixel. The image data is clipped according to the range of the topographic map, and the information of topographic elements is added to the image plane in the form of symbols, line drawings, annotations, kilometer grids, and outline (inside / outside) decoration, forming an image database stored in the form of raster data. It has the geometric accuracy and image characteristics of a topographic map.

[0075] Step S202: collecting first positioning trajectory data of the vehicle with respect to the lane in different positioning modes.

[0076] In this embodiment, the different positioning modes include any combination of combined navigation, laser synchronous positioning, visual synchronous positioning or fusion positioning modes.

[0077] Among them, laser synchronous positioning is specifically laser synchronous positioning and mapping Lidar-SLAM, and visual synchronous positioning is specifically visual synchronous positioning and mapping V-SLAM. It can be understood that integrated navigation is a navigation system that uses multiple sensors on the same platform to implement complementary, mutual verification, and mutual calibration. Fusion positioning mode is a navigation system that uses multiple sensors on the same platform to implement information fusion. Integrated navigation is commonly used in both manual and autonomous driving, and fusion positioning mode is more common in autonomous vehicles.

[0078] In this embodiment, the vehicle is an autonomous vehicle capable of switching between autonomous, semi-autonomous, and manual modes. It participates in actual ship loading and unloading operations in a real port environment, capturing vehicle positioning trajectories in typical scenarios such as the work yard and quay crane. These trajectories include combined navigation trajectories in manual mode, and Lidar-SLAM, V-SLAM, and fused positioning trajectories in autonomous mode. This trajectory recording process can be performed while meeting the requirements of actual container ship loading and unloading operations, ensuring the data is authentic and valid.

[0079] In one example, for the collection of the first positioning trajectory data of an autonomous vehicle or a semi-autonomous vehicle, the vehicle ROS positioning data collection module can be turned on to collect trajectory data bags such as combined navigation, Lidar-SLAM, V-SLAM, and fusion positioning while in motion to obtain the first positioning trajectory data of the autonomous vehicle driving on the lane.

[0080] In another example, for a vehicle in manual driving state, the vehicle ROS positioning data acquisition module is turned on, and the combined navigation trajectory data packet is collected in a moving state. For a vehicle in manual driving state, the driver can visually observe the lane lines during the collection process, and the vehicle should be driven in the center of the lane as much as possible, so as to facilitate the judgment of whether the positioning trajectory is near the center line of the vector map lane.

[0081] Step S203: Convert the first lane centerline data and the first positioning trajectory data into data of the same format to obtain second lane centerline data and second positioning trajectory data.

[0082] In this embodiment, the first lane centerline data is used as the true value reference object for vehicle positioning accuracy evaluation. Considering that in actual applications, the lane centerline data and the collected vehicle positioning trajectory data are in different formats, it is not convenient to perform trajectory comparison between the trajectory data and the lane centerline data, which in turn affects the evaluation effect of positioning accuracy. In view of this, this embodiment first converts the format of the lane centerline data and the positioning trajectory data before performing trajectory deviation judgment, and further performs trajectory deviation judgment based on the same format data.

[0083] Furthermore, by converting both the first lane centerline data and the first positioning trajectory data into a preset format, such as the shp format, to form discrete point data, it is convenient to identify the deviation distance between the lane centerline data and the positioning trajectory data. Specifically, step S203 includes the following steps:

[0084] The first lane centerline data and the first positioning trajectory data are respectively converted into shp format to form lane centerline discrete points and positioning trajectory discrete points.

[0085] It is understood that the shp format is a vector graphics format that can store the position and related attributes of geometric data. The shp format is not limited to this application. In some embodiments, the first positioning trajectory data can also be converted into other formats to form discrete point data.

[0086] Specifically, after the lane centerline data of the vector map is extracted using a GUI tool, it is stored in the form of discrete lane centerline points in the shp format; and, the positioning trajectory data is converted into discrete trajectory points in the shp format using a GUI tool and stored.

[0087] Regarding the format conversion of the first positioning trajectory data, in the manual driving state, use the GUI tool to convert the combined navigation trajectory data bag package collected by the vehicle in the manual driving state into discrete trajectory points in shp format and store them; in the automatic driving state or semi-automatic driving state, use the GUI tool to convert the collected combined navigation, Lidar-SLAM, V-SLAM, fusion positioning and other trajectory data bag packages into discrete trajectory points in shp format and store them.

[0088] Step S204: Obtain a deviation distance between the positioning trajectory in each positioning mode and the lane centerline of the lane based on the second positioning trajectory data and the second lane centerline data in each positioning mode.

[0089] In a specific implementation, taking discrete point data in shp format as an example, step S204 specifically includes the following steps:

[0090] Obtain the deviation distance between the discrete points of the positioning trajectory and the discrete points of the lane centerline in each positioning mode; and

[0091] The deviation distance between the positioning trajectory and the lane centerline in each positioning mode is obtained based on the deviation distance between the discrete points of the positioning trajectory and the discrete points of the lane centerline in each positioning mode.

[0092] In this embodiment, by comparing the deviation distances between the discrete points of the lane centerline and the discrete points of the positioning trajectory in each positioning mode, the deviation distances between the positioning trajectory and the lane centerline in each positioning mode can be obtained. Based on the deviation distances between the positioning trajectory and the lane centerline in each mode, the positioning accuracy of the vehicle in each mode can be obtained.

[0093] To further improve the efficiency of obtaining the deviation distance, the steps in this embodiment are to obtain the deviation distance between the discrete points of the positioning trajectory and the discrete points of the lane centerline in each positioning mode, specifically including the following steps:

[0094] Using the vector map and the lane centerline discrete points as a base map in a geographic information system (GIS);

[0095] The positioning trajectory discrete points in each positioning mode are superimposed on the base map, and the deviation distances between the positioning trajectory discrete points in each positioning mode and the lane centerline discrete points are respectively obtained on the base map after the discrete points are superimposed.

[0096] In this embodiment, the vector map and the discrete points of the lane centerline are loaded into the GIS system as a base map, and the discrete points of the manual driving trajectory are superimposed on the base map to determine the combined navigation positioning state when the actual vehicle is located on the lane centerline, and then determine the combined navigation positioning accuracy; and the discrete points of the automatic driving and semi-automatic driving trajectory are superimposed on the base map to determine the deviation of the vehicle's combined navigation, Lidar-SLAM, V-SLAM, fusion positioning and other trajectory discrete points from the lane centerline in the automatic driving state. By measuring the deviation distance between the combined navigation, Lidar-SLAM, V-SLAM, fusion positioning trajectory and the lane centerline, the positioning error distribution in the automatic driving state is quantitatively determined, and then the positioning accuracy of the combined navigation, Lidar-SLAM, V-SLAM, fusion positioning, etc. is qualitatively and quantitatively determined.

[0097] It should be noted that using the geographic information system GIS to obtain the deviation distance is only an example of this embodiment. In other examples, other methods can also be used to use the vector map and the lane centerline discrete points as the base map, and use the map to obtain the corresponding deviation distance.

[0098] Combine Figure 3-7 As shown, Figure 3-7 They are respectively the comparison between the lane centerline of the vector map and the combined navigation trajectory, the comparison between the lane centerline of the vector map and the Lidar-SLAM trajectory, the comparison between the lane centerline of the vector map and the V-SLAM trajectory, the comparison between the lane centerline of the vector map and the fusion positioning trajectory, and the comparison between the lane centerline of the vector map and the combined navigation, Lidar-SLAM, V-SLAM, fusion positioning trajectory and lane centerline. Among them, Figure 7 It is also a trajectory comparison in the scenario of vectorizing digital orthophotos into vector maps of lanes.

[0099] Step S205 : Evaluate the positioning accuracy of the vehicle in each positioning mode based on the deviation distance.

[0100] Compared with related technologies, the evaluation of vehicle positioning accuracy is usually an overall evaluation, and there is no lane line true value as a reference basis, which leads to the problem of low accuracy of vehicle positioning accuracy evaluation. This embodiment uses steps such as data collection, data conversion, data comparison, and data accuracy comparison to obtain (for example, by recording, measurement, and other collection methods) ROS estimation data bags with positioning-related topics in manual driving and autonomous driving states. The ROS and GIS data exchange GUI tool is used to unify the ROS and GIS data into a data format supported by the GIS system. By comparing the positioning trajectory with the vector map lane centerline as the true value, the lateral deviation between the vector map and the positioning trajectory is measured, and the positioning effects of different positioning strategies are qualitatively and quantitatively judged. Appropriate weights are assigned to different positioning strategies to guide the improvement of positioning and the fusion of positioning algorithm strategies, thereby promoting the iterative update of the autonomous driving positioning algorithm and achieving high-precision autonomous driving positioning.

[0101] Please refer to Figure 8 , Figure 8 This is a flow chart of another method for evaluating vehicle positioning accuracy provided by an embodiment of the present application. Compared to the previous embodiment, after completing the positioning accuracy evaluation, this embodiment further determines the positioning function module with deviation error based on the positioning accuracy evaluation results, thereby targetedly improving the corresponding positioning algorithm and providing strong support for autonomous driving positioning technology. Specifically, after evaluating the positioning accuracy of the vehicle in each positioning mode based on the deviation distance (step S105), step S801 is also included.

[0102] Step S801: If the evaluation result of the positioning accuracy is lower than a preset threshold, a functional module causing a deviation error in a corresponding positioning mode is determined based on the evaluation result of the positioning accuracy.

[0103] In this embodiment, the positioning accuracy evaluation results for each positioning mode are determined to be below a preset threshold. If the results are below the threshold, it indicates that the functional modules of that positioning mode may need to be optimized and improved. For example, in the case of manual driving, if the positioning trajectory deviation in the integrated navigation mode is large, it is assumed that the positioning deviation is caused by issues such as the on-board RTK hardware, the integrated navigation algorithm, or scene signal obstruction.

[0104] It should be noted that those skilled in the art can adaptively set the preset threshold value in combination with existing technologies and practical applications.

[0105] Please refer to Figure 9 , Figure 9A flow chart of another vehicle positioning accuracy evaluation method provided in an embodiment of the present application. Based on the previous embodiment, this embodiment further refines the evaluation results of each positioning mode. After performing threshold judgment on the evaluation results, this embodiment further refines the judgment of the positioning mode in which deviation errors occur in each positioning mode by random combination. Specifically, based on the evaluation results of the positioning accuracy, the functional module that causes the deviation error in the corresponding positioning mode is determined (step S801), including steps S801a-S801c.

[0106] S801a, randomly combining the positioning modes whose evaluation results are lower than a preset threshold;

[0107] S801b, performing similarity judgment on the evaluation results of the positioning accuracy between the randomly combined positioning modes to obtain a similarity judgment result;

[0108] S801c: Selecting a positioning mode with the lowest similarity from the positioning modes whose evaluation results are lower than a preset threshold based on the similarity judgment result, and determining a functional module that causes a deviation error in the positioning mode.

[0109] In actual applications, vehicles usually use multiple positioning modes for positioning at the same time. In order to further refine the evaluation results of each positioning mode, this embodiment further refines the positioning mode with deviation error among the positioning modes by random combination after performing threshold judgment on the evaluation results.

[0110] For example, the positioning modes with problems such as vehicle combined navigation, Lidar-SLAM, V-SLAM, and fusion positioning are comprehensively determined. If the combined navigation trajectory deviation is large under manual driving, it can be directly considered that the functional modules in the combined navigation mode, such as on-board RTK hardware, combined navigation algorithm, scene signal occlusion and other problems cause the positioning deviation; if the trajectory deviation of Lidar-SLAM, V-SLAM, fusion positioning and other problems is large under automatic driving or semi-automatic driving, and the fusion positioning deviation is consistent with the SLAM deviation trend and error distribution (or the similarity is greater than a certain value), it can be considered that the positioning deviation is caused by the Lidar, camera hardware, or SLAM algorithm of Lidar-SLAM and V-SLAM; if there is no obvious deviation between the above positioning trajectories, it can be considered that the vehicle control accuracy deviation exists in the automatic driving control system.

[0111] Please refer to Figure 10 , Figure 10 This is a schematic diagram of the structure of a vehicle positioning accuracy evaluation device provided in an embodiment of the present application. The device includes an extraction module 101, a collection module 102, a conversion module 103, an acquisition module 104, and an evaluation module 105, wherein:

[0112] An extraction module 101 is configured to extract first lane centerline data from a vector map of a lane;

[0113] A collection module 102 configured to collect first positioning trajectory data of the vehicle about the lane in different positioning modes;

[0114] A conversion module 103 is configured to convert the first lane centerline data and the first positioning trajectory data into data of the same format to obtain second lane centerline data and second positioning trajectory data;

[0115] An acquisition module 104 is configured to acquire a deviation distance between the positioning trajectory in each positioning mode and the lane centerline of the lane based on the second positioning trajectory data and the second lane centerline data in each positioning mode; and

[0116] The evaluation module 105 is configured to evaluate the positioning accuracy of the vehicle in each positioning mode based on the deviation distance.

[0117] In one embodiment, the different positioning modes include any combination of combined navigation, laser synchronous positioning, visual synchronous positioning or fusion positioning modes.

[0118] In one embodiment, the apparatus further comprises:

[0119] The generating module is configured to generate a digital orthophoto image of the lane and vectorize the digital orthophoto image into a vector map of the lane before the extracting module extracts the first lane centerline data in the vector map of the lane.

[0120] In one embodiment, the conversion module is specifically configured to convert the first lane centerline data and the first positioning trajectory data into a preset format to form lane centerline discrete points and positioning trajectory discrete points;

[0121] The acquisition module includes:

[0122] A first acquiring unit is configured to acquire a deviation distance between a discrete point of the positioning trajectory and a discrete point of the lane centerline in each positioning mode; and

[0123] The second acquisition unit is configured to respectively acquire the deviation distance between the positioning trajectory and the lane centerline in each positioning mode based on the deviation distance between the discrete points of the positioning trajectory and the lane centerline in each positioning mode.

[0124] In one embodiment, the first acquiring unit is specifically configured to:

[0125] a base map acquisition unit configured to use the vector map and lane centerline discrete points as a base map; and

[0126] The superposition acquisition unit is configured to superimpose the discrete points of the positioning trajectory in each positioning mode on the base map, and respectively obtain the deviation distance between the discrete points of the positioning trajectory in each positioning mode and the discrete points of the lane centerline on the base map after the discrete points are superimposed.

[0127] In one embodiment, the apparatus further comprises:

[0128] The error determination module is configured to determine, after the evaluation module evaluates the positioning accuracy of the vehicle in each positioning mode based on the deviation distance, a functional module that causes the deviation error in the corresponding positioning mode based on the evaluation result of the positioning accuracy when the evaluation result of the positioning accuracy is lower than a preset threshold.

[0129] In one embodiment, the error determination module includes:

[0130] a combining unit configured to randomly combine positioning modes whose evaluation results are lower than a preset threshold;

[0131] a judgment unit configured to perform similarity judgment on evaluation results of positioning accuracy between randomly combined positioning modes to obtain a similarity judgment result;

[0132] The determining unit is configured to select a positioning mode with the lowest similarity from each positioning mode whose evaluation result is lower than a preset threshold based on the similarity judgment result, and determine a functional module that causes a deviation error in the positioning mode.

[0133] The present application also provides an electronic device. Figure 11 As shown, it includes: a memory 111, a processor 112 and a computer program, the computer program is stored in the memory, and the processor runs the computer program to execute the vehicle positioning accuracy evaluation method.

[0134] An embodiment of the present application also provides a computer-readable storage medium, which includes a computer program, and the computer program is used to implement the vehicle positioning accuracy evaluation method.

[0135] The embodiment of the present application also provides a computer program product, which includes computer program code. When the computer program code is run on a computer, the computer executes the vehicle positioning accuracy evaluation method.

[0136] An embodiment of the present application also provides a chip, including a memory and a processor, wherein the memory is used to store a computer program, and the processor is used to call and run the computer program from the memory to execute the vehicle positioning accuracy evaluation method.

[0137] In this embodiment, by converting the data format, breaking through the data barriers between ROS and GIS, the vector map and vehicle trajectories of different positioning strategies are unified in the UTM coordinate system, which is convenient for intuitive and vivid visualization in the GIS system. By comparing the lateral deviation between the vehicle trajectories of different positioning modes and the center line of the lane of the vector map, the accuracy and error distribution of different positioning strategies can be further determined, thereby providing a reference basis for improving different positioning strategies such as combined navigation, Lidar-SLAM, V-SLAM, and fusion positioning, and providing an effective true value reference for the positioning algorithm, thereby improving the positioning algorithm in a targeted manner, providing strong support for autonomous driving positioning technology, and achieving good results.

[0138] The embodiments of the present application provide a method that a person of ordinary skill in the art can understand that all or some of the steps, systems, and functional modules / units in the device disclosed above can be implemented as software, firmware, hardware, and appropriate combinations thereof. In a hardware implementation, the division between the functional modules / units mentioned in the above description does not necessarily correspond to the division of physical components; for example, a physical component can have multiple functions, or a function or step can be performed by several physical components in cooperation. Some physical components or all physical components can be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or implemented as hardware, or implemented as an integrated circuit, such as an application-specific integrated circuit. Such software can be distributed on a computer-readable medium, which can include a computer storage medium (or non-transitory medium) and a communication medium (or temporary medium).

[0139] As is well known to those skilled in the art, the term computer storage media includes volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information (such as computer-readable instructions, data structures, program modules or other data). Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer.

[0140] Furthermore, as is well known to those skilled in the art, communication media typically embodies computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.

[0141] In the description of the embodiments of the present application, the term "and / or" merely represents a type of association relationship describing associated objects, indicating that three relationships may exist. For example, A and / or B may represent three situations: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the term "at least one" represents any combination of at least two of any one or more of a plurality of items. For example, at least one of A, B, and C may represent any one or more elements selected from a set including A, B, and C.

[0142] In the description of the embodiments of the present application, the terms "first", "second", "third", "fourth", etc. (if any) are used to distinguish similar objects and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0143] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some or all of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A vehicle positioning accuracy evaluation method, which is applied in a port scenario and is characterized by: include: Extracting the centerline data of the first lane from the vector map of the lane; Collecting first positioning trajectory data of the vehicle about the lane in different positioning modes, wherein the different positioning modes include any combination of integrated navigation, laser synchronous positioning, visual synchronous positioning or fusion positioning mode; Converting the first lane centerline data and the first positioning trajectory data into shp format respectively to form lane centerline discrete points and positioning trajectory discrete points; Using the vector map and lane centerline discrete points as a base map; superimposing the discrete points of the positioning trajectory in each positioning mode on the base map, and obtaining the deviation distances between the discrete points of the positioning trajectory in each positioning mode and the discrete points of the lane centerline on the base map after the discrete points are superimposed; and obtaining the deviation distances between the positioning trajectory in each positioning mode and the lane centerline based on the deviation distances between the discrete points of the positioning trajectory in each positioning mode and the discrete points of the lane centerline; and evaluating the positioning accuracy of the vehicle in each positioning mode based on the deviation distances; If the evaluation result of the positioning accuracy is lower than the preset threshold, randomly combining the positioning modes with the evaluation results lower than the preset threshold; Perform similarity judgment on the evaluation results of the positioning accuracy between the randomly combined positioning modes to obtain a similarity judgment result; Based on the similarity judgment result, a positioning mode with the lowest similarity is selected from each positioning mode whose evaluation result is lower than a preset threshold, and a functional module causing a deviation error in the positioning mode is determined.

2. The method according to claim 1, characterized in that Before extracting the first lane centerline data in the lane vector map, the following is also included: A digital orthophoto image of the lane is generated, and the digital orthophoto image is vectorized into a vector map of the lane.

3. A vehicle positioning accuracy evaluation device, which is applied in a port scenario and is characterized by: include: an extraction module configured to extract first lane centerline data from a vector map of lanes; A collection module configured to collect first positioning trajectory data of the vehicle about the lane in different positioning modes, wherein the different positioning modes include any combination of integrated navigation, laser synchronous positioning, visual synchronous positioning or fusion positioning mode; a conversion module configured to convert the first lane centerline data and the first positioning trajectory data into data of the same format to obtain second lane centerline data and second positioning trajectory data; an acquisition module configured to acquire a deviation distance between the positioning trajectory in each positioning mode and the lane centerline of the lane based on the second positioning trajectory data and the second lane centerline data in each positioning mode; and an evaluation module configured to evaluate the positioning accuracy of the vehicle in each positioning mode based on the deviation distance; The conversion module is specifically configured to convert the first lane centerline data and the first positioning trajectory data into shp format respectively to form lane centerline discrete points and positioning trajectory discrete points; The acquisition module includes: A first acquiring unit is configured to acquire a deviation distance between a discrete point of the positioning trajectory and a discrete point of the lane centerline in each positioning mode; and a second acquiring unit configured to acquire, based on the deviation distances between the discrete points of the positioning trajectory and the lane centerline in each positioning mode, the deviation distances between the discrete points of the positioning trajectory and the lane centerline in each positioning mode; The first acquisition unit is specifically configured to: a base map acquisition unit configured to use the vector map and lane centerline discrete points as a base map; and a superposition acquisition unit configured to superimpose the discrete points of the positioning trajectory in each positioning mode on the base map, and respectively obtain, on the base map after the superposition of the discrete points, the deviation distance between the discrete points of the positioning trajectory in each positioning mode and the discrete points of the lane centerline; The apparatus further includes: an error determination module configured to, after the evaluation module evaluates the positioning accuracy of the vehicle in each positioning mode based on the deviation distance, determine, if the evaluation result of the positioning accuracy is lower than a preset threshold, a functional module causing the deviation error in the corresponding positioning mode based on the evaluation result of the positioning accuracy; The error determination module includes: a combining unit configured to randomly combine positioning modes having evaluation results lower than a preset threshold; a judgment unit configured to perform similarity judgment on evaluation results of positioning accuracy between randomly combined positioning modes to obtain a similarity judgment result; The determining unit is configured to select a positioning mode with the lowest similarity from each positioning mode whose evaluation result is lower than a preset threshold based on the similarity judgment result, and determine a functional module that causes a deviation error in the positioning mode.

4. An electronic device, characterized in that: include: A memory, a processor and a computer program, wherein the computer program is stored in the memory, and the processor runs the computer program to execute the vehicle positioning accuracy evaluation method according to claim 1 or 2.

5. A computer-readable storage medium, characterized in that The storage medium includes a computer program, and the computer program is used to implement the vehicle positioning accuracy evaluation method according to claim 1 or 2.

Images