A simulation scene set generation method, device, equipment and medium

Abstract

This invention discloses a method, apparatus, device, and medium for generating simulation scene sets. The method includes: determining a road test dataset to be processed from an original road test dataset based on at least one filtering condition; wherein the road test dataset to be processed includes at least one set of data to be processed; constructing a multi-dimensional spatial key data search structure tree based on the road test dataset to be processed; and determining a target road test simulation set corresponding to the simulation scene generation conditions based on pre-set simulation scene generation conditions and the multi-dimensional spatial key data search structure tree. This invention solves the technical problems of low efficiency, low adaptability, and inability to effectively filter simulation scene sets that meet the training requirements of autonomous driving algorithms when determining simulation scene sets. It achieves efficient and stable homogenization of the geographical distribution of simulation scene sets, improves the adaptability of simulation scene sets to the training requirements of autonomous driving algorithms, and increases the efficiency of generating simulation scene sets.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method, apparatus, device, and medium for generating simulation scene sets. Background Technology

[0002] Autonomous driving relies on computer and artificial intelligence technologies to complete safe and efficient driving operations without human intervention. In practical applications, this is achieved through autonomous driving algorithms. Before being applied to autonomous vehicles, these algorithms need to be trained based on driving scenario data until they meet the requirements. Therefore, training autonomous driving algorithms requires driving scenario data.

[0003] Currently, a large amount of road test data can be collected using sensors and other devices on vehicles, and this collected road test data can be used as driving scenario data. In addition, important scenarios in the driving scenario data can be manually labeled, and then simulation training can be carried out based on the labeled important driving scenario data.

[0004] However, the driving scenario data obtained through the above methods contains a large amount of redundant data, and much of this data involves relatively simple scenarios that are not useful for training autonomous driving algorithms. Simulation testing based on this data would lead to a significant waste of resources. Furthermore, manually selecting driving scenario data is not only inefficient but also subjective, resulting in low accuracy for the simulation system. Summary of the Invention

[0005] This invention provides a method, apparatus, device, and medium for generating simulation scene sets, which achieves efficient and stable homogenization of the geographical distribution of road test datasets, improves the driving scene data coverage efficiency of the simulation system, enhances the adaptability of simulation scene sets to the training requirements of autonomous driving algorithms, and improves the efficiency of generating simulation scene sets.

[0006] In a first aspect, the present invention provides a method for generating a simulation scene set, the method comprising:

[0007] Based on at least one filtering criterion, a road test dataset to be processed is determined from the original road test dataset; wherein the road test dataset to be processed includes at least one set of data to be processed;

[0008] Based on the road test dataset to be processed, a multi-dimensional spatial key data search structure tree is constructed.

[0009] Based on the pre-set simulation scenario generation conditions and the multi-dimensional spatial key data search structure tree, the target road test simulation set corresponding to the simulation scenario generation conditions is determined.

[0010] Secondly, the present invention provides a simulation scene set generation device, the device comprising:

[0011] A dataset determination module is used to determine a road test dataset to be processed from the original road test dataset based on at least one filtering condition; wherein the road test dataset to be processed includes at least one set of data to be processed;

[0012] The search tree construction module is used to construct a multi-dimensional spatial key data search structure tree based on the road test dataset to be processed.

[0013] The simulation set determination module is used to determine the target road test simulation set corresponding to the simulation scene generation conditions based on the pre-set simulation scene generation conditions and the multi-dimensional spatial key data search structure tree.

[0014] Thirdly, the present invention provides a data processing electronic device, comprising:

[0015] At least one processor; and

[0016] A memory that is communicatively connected to at least one processor; wherein,

[0017] The memory stores a computer program that can be executed by at least one processor, such that the at least one processor is able to execute the simulation scene set generation method of any embodiment of the present invention.

[0018] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions that, when executed by a processor, implement the simulation scene set generation method of any embodiment of the present invention.

[0019] Fifthly, the present invention provides a computer program product, which includes a computer program that, when executed by a processor, implements the simulation scene set generation method of any embodiment of the present invention.

[0020] The technical solution provided by this invention determines a road test dataset to be processed from the original road test dataset based on at least one filtering condition. This road test dataset includes at least one set of data to be processed. Then, based on the road test dataset to be processed, a multi-dimensional spatial key data search structure tree is constructed. Subsequently, based on pre-set simulation scenario generation conditions and the multi-dimensional spatial key data search structure tree, a target road test simulation set corresponding to the simulation scenario generation conditions is determined. This technical solution solves the technical problems of low efficiency, low adaptability, and inability to effectively filter simulation scenario sets that meet the training requirements of autonomous driving algorithms when determining simulation scenario sets. It achieves efficient and stable homogenization of the geographical distribution of the road test dataset, improves the driving scenario data coverage efficiency of the simulation system, enhances the adaptability of the simulation scenario set to the training requirements of autonomous driving algorithms, and improves the efficiency of generating simulation scenario sets.

[0021] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is a flowchart of a simulation scene set generation method provided in Embodiment 1 of the present invention;

[0024] Figure 2 This is a flowchart of a simulation scene set generation method provided in Embodiment 2 of the present invention;

[0025] Figure 3 This is a schematic diagram showing the position of the coordinate data involved in Embodiment 2 of the present invention in a Cartesian coordinate system;

[0026] Figure 4 This is a schematic diagram of the multidimensional spatial key data search structure tree involved in Embodiment 2 of the present invention;

[0027] Figure 5 This is a schematic diagram of a simulation scene set generation device provided in Embodiment 3 of the present invention;

[0028] Figure 6 This is a schematic diagram of the structure of an electronic device provided in Embodiment 4 of the present invention. Detailed Implementation

[0029] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0030] It should be noted that the terms "first preset condition," "second preset condition," etc., used in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0031] Example 1

[0032] Before introducing this technical solution, an illustrative application scenario can be provided. Autonomous driving relies on autonomous driving algorithms. Before autonomous driving is formally applied to real-world driving scenarios, large-scale simulation tasks are needed to improve the performance of the autonomous driving algorithm model. Therefore, training autonomous driving algorithms requires highly suitable driving scenario data to improve the performance of the autonomous driving algorithm model. This invention provides a scheme for automatically selecting scenario sets that match the requirements of autonomous driving algorithms from multiple sets of scenario data, which helps to quickly and efficiently complete the simulation tasks of autonomous driving algorithms.

[0033] Figure 1 This is a flowchart of a simulation scene set generation method provided in Embodiment 1 of the present invention. This embodiment is applicable to situations where a scene set matching the requirements of an autonomous driving algorithm is selected from a large amount of scene set data. This method can be executed by a simulation scene set generation device, which can be implemented in hardware and / or software. The device can be configured on a computer device, such as a laptop, desktop computer, or smart tablet. Figure 1 As shown, the method includes:

[0034] S110. Determine the road test dataset to be processed from the original road test dataset based on at least one filtering condition.

[0035] The filtering criteria are pre-defined conditions used to select road test datasets that match the filtering criteria from a large number of raw road test datasets. The number of filtering criteria can be one or more.

[0036] In this embodiment, at least one filtering condition includes at least one of the following: driving state factor, start point and / or end point factor, obstacle quantity factor, vehicle trajectory factor, and vehicle speed factor. The driving state factor characterizes whether the vehicle is in autonomous driving mode; the start point and / or end point factor characterizes whether the start and end points of a route have been marked; the obstacle quantity factor characterizes the number of obstacles around the vehicle and their specific locations; the vehicle trajectory factor characterizes the path trajectory corresponding to a certain distance traveled by the vehicle; and the vehicle speed factor characterizes the vehicle's speed value and speed changes during driving.

[0037] The raw road test dataset consists of road test data pre-collected and stored in a scenario database. In practical applications, raw road test datasets for various road test scenarios can be obtained using road test equipment. This equipment can be integrated into vehicles to collect various road test data, or it can be dedicated equipment for collecting various road test data placed on the vehicle, thus achieving the goal of collecting a large amount of raw road test data. The raw road test dataset includes road test data corresponding to all roads within a specific area. The data content in the raw road test dataset includes data labels corresponding to the filtering criteria, allowing the identification of road test datasets corresponding to the filtering criteria based on the data labels.

[0038] The road test dataset to be processed is a dataset selected from the original road test dataset based on filtering criteria, which meets the requirements of the simulation task. The road test dataset to be processed includes one or more sets of data to be processed. During the acquisition of the original road test dataset, the road test equipment can collect data at fixed intervals. Therefore, the road test data to be processed is a segment of data with a preset duration, for example, a segment of road test data with a duration of 10 seconds.

[0039] In practical applications, a database for storing raw road test datasets can be pre-established. The raw road test datasets are then collected and stored in the database. During the data collection process, data collection vehicles can be pre-configured. These vehicles travel on various roads within a defined area, collecting measured data from road test equipment throughout the journey.

[0040] Because the collected road test data contains a large amount of redundant data, directly storing it into the database would waste database resources, and subsequently filtering out the road test dataset to be processed from the raw data would be time-consuming. Therefore, invalid data can be filtered out first. Invalid data generally originates from data collected normally by the vehicle's messaging system, but appears after the driver has switched the vehicle out of autonomous driving mode. For example, the driver may want to make a temporary stop, change routes, or fail to set a new destination after arriving at the original destination. Pre-filtering this invalid data and preventing it from entering subsequent processing stages can improve system efficiency. In addition, recording recoverable key data points can help filter out invalid data. In autonomous driving, the system's own path service can generate reasonable trajectories that reach the destination. Therefore, for the collected road test data, only the start point, end point, and stops along the way need to be saved. The original trajectory can be reconstructed using the same version of the path service for subsequent data processing. Based on this, waypoints along the route can be filtered out.

[0041] It is worth noting that the original road test data can record the time or location of scenarios that are important to autonomous driving algorithms, such as meeting oncoming traffic and making unprotected left turns, so as to generate scenario sets later based on manually set conditions.

[0042] Based on the above embodiments, determining the road test dataset to be processed from the original road test dataset includes: determining the valid road test dataset from the original road test dataset based on at least one filtering condition; and determining the road test dataset to be processed from the valid road test dataset based on at least one pre-set scenario condition.

[0043] In this context, valid road test data can be understood as road test data that meets the testing requirements of the simulation task. Scene conditions are pre-defined conditions. For example, data with certain scene characteristics in the original road test dataset can be pre-labeled with corresponding scene tags, and a mapping relationship between scene tags and corresponding datasets can be established. Based on this, the scene set corresponding to the scene tag can be determined. The road test dataset to be processed includes multiple road test data sets, each including corresponding coordinate data. Coordinate data can represent the spatial geographical location corresponding to the current road test data to be processed; for example, it can be represented in the form of coordinate pairs.

[0044] In this embodiment, test users have specific needs when training autonomous driving algorithms based on simulation test sets. To determine the set of scenarios that meet the test users' simulation task requirements, they can select conditions corresponding to the current simulation task requirements from at least one filtering condition. After determining the conditions corresponding to the current simulation task requirements, valid road test data is determined from the original road test data based on the determined filtering conditions. Then, based on the selected scenario labels, the set of scenarios corresponding to the scenario labels can be determined from the valid road test data as the road test dataset to be processed.

[0045] S120. Based on the road test dataset to be processed, construct a multi-dimensional spatial key data search structure tree.

[0046] The search structure tree is a hierarchical structure determined based on the coordinate data of each road test data point in the road test dataset. The search structure tree is used to identify spatially uniform road test data points from multiple road test data points in the road test dataset.

[0047] In this embodiment, the multidimensional spatial key data search structure tree can be constructed using a k-dimensional tree (kd-tree). A kd-tree is a tree-like data structure used for spatial partitioning. This data structure continuously adjusts the partitioning axis of each point, ensuring that each node divides the space as evenly as possible. In this solution, a kd-tree is used to ensure that each transit point bisects the real-world ground plane as evenly as possible, allowing us to extract any number of scenes from the road test dataset and ensuring that the physical locations of the scenes are approximately evenly distributed.

[0048] Specifically, after determining the road test dataset to be processed, which may contain a large number of data points, each with corresponding coordinate data, a multidimensional spatial key data search structure tree can be constructed based on this coordinate data. In constructing this structure tree, the coordinate data corresponding to each road test data point can be placed in a plane. Then, based on the x-coordinate of each data point, the plane is divided into two hyperplanes. Further, based on the y-coordinate, the two hyperplanes are divided into multiple subplanes. This process continues, alternating between dividing the planes based on both the x-coordinate and y-coordinate of the data points, until all coordinate data corresponding to all road test data points has been divided into planes. Finally, based on the coordinate data within these multiple planes, the root and leaf nodes of the multidimensional spatial key data search structure tree are determined.

[0049] S130. Based on the pre-set simulation scenario generation conditions and the multi-dimensional key data search structure tree, determine the target road test simulation set corresponding to the simulation scenario generation conditions.

[0050] The simulation scenario generation condition can be understood as the number of simulation scenarios required to complete the simulation task. For example, the simulation scenario generation condition could be "the number of simulation scenario sets required to complete the simulation task is 50 scenario data sets". The target road test simulation set is the final set of scenario test data determined for the simulation task.

[0051] In this embodiment, the multi-dimensional spatial key data search structure tree constructed in the above embodiments includes a large amount of scene data. However, not every scene data in the search structure tree is used as the target road test simulation set. The number of simulation scene sets required to complete the simulation task can be preset, and then a sufficient number of scene data can be searched sequentially from the root node of the multi-dimensional spatial key data search structure tree to the leaf nodes as the target road test simulation set. In practical applications, the search is performed sequentially from the root node to the leaf nodes of the multi-dimensional spatial key data search structure tree, and the result is a series of coordinate points. The mapping relationship between the coordinate points and the corresponding scene data can be preset. After determining the coordinate points, the scene data corresponding to the coordinate points can be determined as the target road test simulation set based on the preset mapping relationship.

[0052] For example, if the multidimensional key data search structure tree includes 1000 coordinate points, and the simulation scene generation condition could be "the number of simulation scene sets required to complete the simulation task is 50 scene data points," then the search proceeds sequentially from the root node to the leaf nodes of the multidimensional key data search structure tree until 50 coordinate data points are found. After determining the 50 coordinate data points, the 50 scene data points corresponding to these 50 coordinate points can be determined as the target road test simulation set based on a pre-defined mapping relationship.

[0053] It should be noted that some manually input conditions can be added during the selection process for the target road test simulation set. These conditions can include, for example, "road test data during morning and evening rush hours" or "road test data belonging to area A".

[0054] The technical solution provided by this invention determines a road test dataset to be processed from the original road test dataset based on at least one filtering condition. This road test dataset includes at least one set of data to be processed. Then, based on the road test dataset to be processed, a multi-dimensional spatial key data search structure tree is constructed. Subsequently, based on pre-set simulation scenario generation conditions and the multi-dimensional spatial key data search structure tree, a target road test simulation set corresponding to the simulation scenario generation conditions is determined. This technical solution solves the technical problems of low efficiency, low adaptability, and inability to effectively filter simulation scenario sets that meet the training requirements of autonomous driving algorithms when determining simulation scenario sets. It achieves efficient and stable homogenization of the geographical distribution of the road test dataset, improves the driving scenario data coverage efficiency of the simulation system, enhances the adaptability of the simulation scenario set to the training requirements of autonomous driving algorithms, and improves the efficiency of generating simulation scenario sets.

[0055] Example 2

[0056] Figure 2 This is a flowchart of a simulation scene set generation method provided in Embodiment 2 of the present invention. Based on the above embodiments, this embodiment further refines steps S120 and S130. This embodiment can be combined with various optional solutions from one or more of the above embodiments. For example... Figure 2 As shown, the method includes:

[0057] S210. Determine the road test dataset to be processed from the original road test dataset based on at least one filtering condition.

[0058] S220. Obtain the coordinate data corresponding to each road test data in the road test dataset to be processed.

[0059] In this embodiment, the coordinate data corresponding to each road test data point in the road test dataset to be processed can be determined, and the coordinate data corresponding to each road test data point to be processed can be stored in a database. During the collection of the original road test dataset, the road test equipment can collect data at fixed intervals, so the road test data to be processed is a segment of data with a preset duration. For example, if the road test equipment can collect data once every 1 second, and the road test data to be processed is a segment of road test data with a duration of 10 seconds, then the road test data to be processed includes 10 data point values. Each of the 10 data point values ​​corresponds to coordinate data. When determining the coordinate data corresponding to a particular road test data point in the road test dataset to be processed, the abscissa of the coordinate data corresponding to the particular road test data point can be obtained by averaging the abscissas of the coordinate data of the 10 data point values; the ordinate of the coordinate data corresponding to the particular road test data point can be obtained by averaging the ordinates of the coordinate data of the 10 data point values, thereby obtaining the coordinate data corresponding to each road test data point in the road test dataset to be processed.

[0060] For example, if the coordinate data corresponding to the 7 road test data points in the obtained road test dataset are (1,6), (2,7), (3,2), (4,8), (5,4), (6,8.5), and (7,1.8), the position of each coordinate data point in the Cartesian coordinate system can be found in [reference needed]. Figure 3 .

[0061] S230. Determine intermediate values ​​based on the x-coordinate of the coordinate data.

[0062] Based on the above embodiment, the abscissas of the seven road test data coordinates to be processed are 1, 2, 3, 4, 5, 6, and 7, with the middle value being 4.

[0063] S240. Divide the plane into at least two planes based on the intermediate values, and divide the at least two planes into multiple planes based on the ordinates of the coordinate data within the two planes.

[0064] For example, such as Figure 3 As shown, based on the determined intermediate value of 4, a straight line perpendicular to the x-axis can be drawn through the coordinate point (4,8) in the Cartesian coordinate system, dividing the Cartesian coordinate system into two planes: the left plane and the right plane. Figure 3 In the left plane, there are three coordinate points: (1,6), (2,7), and (3,2); Figure 3 The right plane includes three coordinate points: (5,4), (6,8.5), and (7,1.8). Further processing is needed first. Figure 3 Given three points in the left plane with ordinates of 2, 6, and 7, and a midpoint of 6, we can draw a line perpendicular to the y-axis through the point (1,6) in a Cartesian coordinate system. This divides the left plane into an upper and lower plane. In the upper plane, there is only one point (2,7), so we can draw a line perpendicular to the x-axis through this point. In the lower plane, there is only one point (3,2), so we can draw a line perpendicular to the x-axis through this point. Based on this... Figure 3 The left-middle plane is divided into four planes. For Figure 3Given three points in the right plane with ordinates of 1.8, 4, and 8.5, where the middle ordinate is 4, we can draw a line perpendicular to the y-axis through the point (5,4) in a Cartesian coordinate system. This divides the right plane into an upper and lower plane. In the upper plane, there is only one point (6,8.5), so we can draw a line perpendicular to the x-axis through this point. In the lower plane, there is only one point (7,1.8), so we can draw a line perpendicular to the x-axis through this point. Based on this... Figure 3 The right-center plane is divided into four planes.

[0065] S250. Based on coordinate data in multiple planes, determine the multidimensional spatial key data search structure tree.

[0066] Based on the example above, the multidimensional spatial key data search structure tree determined by the coordinate data (1,6), (2,7), (3,2), (4,8), (5,4), (6,8.5), and (7,1.8) is shown below. Figure 4 .like Figure 4 As shown, the coordinates of the first layer of the multidimensional key data search structure tree are (4,8), the coordinates of the second layer are (1,6) and (5,4), and the coordinates of the third layer are (2,7), (3,2), (6,8.5) and (7,1.8).

[0067] Based on the above embodiments, the method for determining the multidimensional spatial key data search structure tree may include: determining the root node and leaf nodes of the multidimensional spatial key data search structure tree based on coordinate data in multiple planes.

[0068] The root node and leaf nodes correspond to the respective drive test data to be processed.

[0069] For example, such as Figure 4 As shown, for the multidimensional key data search structure tree, the first-level coordinates (4,8) can be used as the root node, the second-level coordinates (1,6) and (5,4) can be used as leaf nodes, and the third-level coordinates (2,7), (3,2), (6,8.5), and (7,1.8) can be used as leaf nodes. The root and leaf nodes can be represented by coordinate data. Based on the mapping relationship between the coordinate data and the road test data to be processed, the corresponding road test data to be processed for the root and leaf nodes of the multidimensional key data search structure tree can be determined.

[0070] S260. Based on the number of scenarios, search for at least one node's data by starting from the root node of the multidimensional key data search structure tree.

[0071] Based on the above example, the simulation scene generation conditions include the number of scenes, which can be customized. For example, if the number of scenes is set to 3, after determining the number of scenes, at least one node data can be searched downwards from the root node of the multidimensional space key data search structure tree. Based on this, such as... Figure 4 As shown, the search can proceed downwards from the root node (4,8) to the two leaf nodes (1,6) and (5,4). The final search will find at least one node with data including (4,8), (1,6) and (5,4).

[0072] S270. Generate a target road test simulation set based on the road test data to be processed corresponding to at least one node data.

[0073] In this embodiment, based on the mapping relationship between at least one node data coordinate data and the road test data to be processed, the road test data to be processed corresponding to the node data in the multi-dimensional spatial key data search structure tree can be determined, and then the road test data to be processed corresponding to the node data can be used as the target road test simulation set. After determining the target road test simulation set, these data contents can be packaged into a whole simulation set to generate the target road test simulation set. Subsequently, the target road test simulation set can be directly fed back to the simulation system for testing and training of autonomous driving algorithms.

[0074] The technical solution provided by this invention determines the road test dataset to be processed from the original road test dataset based on at least one filtering condition. Then, it obtains the coordinate data corresponding to each road test data point in the dataset, determines an intermediate value based on the abscissa of the coordinate data, and divides the dataset into at least two planes based on the intermediate value. Based on the ordinate of the coordinate data within each of the two planes, it further divides the at least two planes into multiple planes. Finally, based on the coordinate data within the multiple planes, it determines a multi-dimensional spatial key data search structure tree. Subsequently, based on the number of scenarios, it searches downwards from the root node of the multi-dimensional spatial key data search structure tree for at least one node data point, and generates a target road test simulation set based on the road test data to be processed corresponding to the at least one node data point. This technical solution solves the technical problems of low efficiency, low adaptability, and inability to effectively filter out simulation scenario sets that meet the training requirements of autonomous driving algorithms when generating simulation scenario sets. It achieves efficient and stable homogenization of the geographical distribution of the road test dataset, improves the driving scenario data coverage efficiency of the simulation system, enhances the adaptability of the simulation scenario set to the training requirements of autonomous driving algorithms, and improves the efficiency of generating simulation scenario sets.

[0075] Example 3

[0076] Figure 5 This is a schematic diagram of a simulation scene set generation device provided in Embodiment 3 of the present invention. This device can execute the simulation scene set generation method provided in the embodiments of the present invention. The device includes: a dataset determination module 310, a search tree construction module 320, and a simulation set determination module 330.

[0077] The dataset determination module 310 is used to determine the road test dataset to be processed from the original road test dataset based on at least one filtering condition; wherein the road test dataset to be processed includes at least one set of data to be processed;

[0078] The search tree construction module 320 is used to construct a multi-dimensional spatial key data search structure tree based on the road test dataset to be processed.

[0079] The simulation set determination module 330 is used to determine the target road test simulation set corresponding to the simulation scene generation conditions based on the pre-set simulation scene generation conditions and the multi-dimensional spatial key data search structure tree.

[0080] Based on the above technical solutions, raw road test datasets for various road test scenarios are obtained using road test equipment.

[0081] Based on the above technical solutions, at least one screening condition includes at least one of the following: driving state factor, starting point and / or ending point factor, number of obstacles factor, vehicle driving trajectory factor, and vehicle driving speed factor.

[0082] Based on the above technical solutions, the dataset determination module 310 also includes: an effective dataset determination unit and a dataset determination unit.

[0083] The effective data set determination unit is used to determine the effective road test data set from the original road test data set based on at least one filtering condition.

[0084] The dataset determination unit is used to determine the road test dataset to be processed from the valid road test dataset based on at least one pre-set scenario condition;

[0085] The unprocessed road test dataset includes multiple unprocessed road test datasets, each containing corresponding coordinate data.

[0086] Based on the above technical solutions, the search tree construction module 320 also includes: a coordinate data acquisition unit, an intermediate data determination unit, a plane partitioning unit, and a search tree determination unit.

[0087] The coordinate data acquisition unit is used to acquire the coordinate data corresponding to each road test data in the road test dataset to be processed.

[0088] The intermediate data determination unit is used to determine intermediate values ​​based on the horizontal coordinate of the coordinate data.

[0089] A plane partitioning unit is used to divide at least two planes based on intermediate values, and to divide at least two planes into multiple planes based on the ordinates of the coordinate data within the two planes.

[0090] The search tree determination unit is used to determine the multidimensional spatial key data search tree structure based on coordinate data in multiple planes.

[0091] Based on the above technical solutions, the search tree determination unit is also used to determine the root node and leaf node of the multidimensional spatial key data search tree structure based on coordinate data in multiple planes; wherein the root node and leaf node correspond to the corresponding road test data to be processed.

[0092] Based on the above technical solutions, the simulation set determination module 330 also includes: a node data determination unit and a simulation set determination unit.

[0093] The node data determination unit is used to search for at least one node data by starting from the root node of the multidimensional key data search tree structure based on the number of scenarios.

[0094] The simulation set determination unit is used to generate a target road test simulation set based on the road test data to be processed corresponding to at least one node data.

[0095] The technical solution provided by this invention determines a road test dataset to be processed from the original road test dataset based on at least one filtering condition. This road test dataset includes at least one set of data to be processed. Then, based on the road test dataset to be processed, a multi-dimensional spatial key data search structure tree is constructed. Subsequently, based on pre-set simulation scenario generation conditions and the multi-dimensional spatial key data search structure tree, a target road test simulation set corresponding to the simulation scenario generation conditions is determined. This technical solution solves the technical problems of low efficiency, low adaptability, and inability to effectively filter simulation scenario sets that meet the training requirements of autonomous driving algorithms when determining simulation scenario sets. It achieves efficient and stable homogenization of the geographical distribution of the road test dataset, improves the driving scenario data coverage efficiency of the simulation system, enhances the adaptability of the simulation scenario set to the training requirements of autonomous driving algorithms, and improves the efficiency of generating simulation scenario sets.

[0096] The simulation scene set generation device provided in this disclosure can execute the simulation scene set generation method provided in any embodiment of this disclosure, and has the corresponding functional modules and beneficial effects of the execution method.

[0097] It is worth noting that the various units and modules included in the above-mentioned device are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be realized; in addition, the specific names of each functional unit are only for easy differentiation and are not used to limit the protection scope of the embodiments of this disclosure.

[0098] Example 4

[0099] Figure 6 This is a schematic diagram of an electronic device provided in Embodiment 4 of the present invention. The electronic device 10 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0100] like Figure 6 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0101] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0102] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as simulation scene set generation methods.

[0103] In some embodiments, the simulation scene set generation method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the simulation scene set generation method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the simulation scene set generation method by any other suitable means (e.g., by means of firmware).

[0104] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0105] Computer programs used to implement the methods of the present invention can be written in any combination of one or more programming languages. These computer programs can be provided to the processor of a general-purpose computer, a special-purpose computer, or other programmable simulation scene generation device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The computer programs can be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0106] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0107] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0108] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0109] A computing system may include clients and servers. Clients and servers are generally geographically separated and typically interact via communication networks. The client-server relationship is established by computer programs running on the respective computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, a host product within the cloud computing service system, addressing the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability. It should be understood that various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired results of the technical solution of this invention are achieved, and this is not limited herein. The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for generating a simulation scene set, characterized in that, include: Based on at least one filtering criterion, a road test dataset to be processed is determined from the original road test dataset; wherein the road test dataset to be processed includes at least one set of data to be processed; Based on the road test dataset to be processed, a multi-dimensional spatial key data search structure tree is constructed. Based on the pre-set simulation scenario generation conditions and the multi-dimensional key data search structure tree, the target road test simulation set corresponding to the simulation scenario generation conditions is determined. The search structure tree is used to identify spatially uniformly distributed road test data from multiple road test data in the road test dataset to be processed. The construction of a multi-dimensional spatial key data search structure tree based on the road test dataset to be processed includes: Obtain the coordinate data corresponding to each road test data in the road test dataset to be processed; Based on the x-coordinate of the coordinate data, determine the intermediate value; Based on the intermediate values, the planes are divided into at least two planes, and based on the ordinates of the coordinate data within the two planes, the at least two planes are further divided into multiple planes. Based on the coordinate data in the multiple planes, the root node and leaf node of the multidimensional spatial key data search structure tree are determined; wherein, the root node and the leaf node correspond to the corresponding road test data to be processed; The simulation scenario generation conditions include the number of scenarios. The determination of the target road test simulation set corresponding to the simulation scenario generation conditions, based on the pre-set simulation scenario generation conditions and the multi-dimensional spatial key data search structure tree, includes: Based on the number of scenarios, search for at least one node data by descending from the root node of the multidimensional key data search structure tree. The target road test simulation set is generated based on the road test data to be processed corresponding to the at least one node data.

2. The method according to claim 1, characterized in that, Also includes: Obtain raw road test datasets for various road test scenarios based on road test equipment.

3. The method according to claim 1, characterized in that, The at least one screening condition includes at least one of the following: driving state factor, starting point and / or ending point factor, number of obstacles factor, vehicle driving trajectory factor, and vehicle driving speed factor.

4. The method according to claim 1, characterized in that, The process of determining the road test dataset to be processed from the original road test dataset based on at least one filtering condition includes: Based on the at least one filtering condition, a valid road test dataset is determined from the original road test dataset; Based on at least one pre-set scenario condition, the road test dataset to be processed is determined from the valid road test dataset; The road test dataset to be processed includes multiple road test datasets, each of which includes corresponding coordinate data.

5. A simulation scene set generation device, characterized in that, include: A dataset determination module is used to determine a road test dataset to be processed from the original road test dataset based on at least one filtering condition; wherein the road test dataset to be processed includes at least one set of data to be processed; The search tree construction module is used to construct a multi-dimensional spatial key data search structure tree based on the road test dataset to be processed; The simulation set determination module is used to determine the target road test simulation set corresponding to the simulation scene generation conditions based on the pre-set simulation scene generation conditions and the multi-dimensional spatial key data search structure tree. The search tree structure is used to identify spatially uniformly distributed road test data from multiple road test data in the road test dataset. The search tree construction module includes: The coordinate data acquisition unit is used to acquire the coordinate data corresponding to each road test data in the road test dataset to be processed; An intermediate data determination unit is used to determine intermediate values ​​based on the horizontal coordinate of the coordinate data; A plane partitioning unit is used to divide the intermediate value into at least two planes, and to divide the at least two planes into multiple planes based on the ordinate of the coordinate data within the two planes. The search tree determination unit is used to determine the root node and leaf node of the multidimensional spatial key data search structure tree based on the coordinate data in the multiple planes; wherein the root node and the leaf node correspond to the corresponding road test data to be processed. The simulation scene generation conditions include the number of scenes, and the simulation set determination module includes: The node data determination unit is used to search for at least one node data by starting from the root node of the multidimensional space key data search structure tree based on the number of scenarios. The simulation set determination unit is used to generate the target road test simulation set based on the road test data to be processed corresponding to the at least one node data.

6. An electronic device, characterized in that, Electronic devices include: One or more processors; Storage device for storing one or more programs. When one or more programs are executed by one or more processors, the one or more processors implement the simulation scene set generation method as described in any one of claims 1-4.

7. A storage medium containing computer-executable instructions, which, when executed by a computer processor, are used to perform the simulation scene set generation method as claimed in any one of claims 1-4.

Images