Automatic driving simulation method and device, electronic equipment and storage medium
By combining open-loop and closed-loop simulation modes, along with interactive behavior judgment and key time point optimization, the problem of scene distortion in autonomous driving simulation testing was solved, achieving efficient and accurate simulation verification.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-23
- Publication Date
- 2026-03-24
AI Technical Summary
Existing autonomous driving simulation tests suffer from scene distortion, resulting in low verification efficiency and poor reliability, and making it impossible to effectively obtain vehicle test data.
By adopting a combination of open-loop and closed-loop simulation, and by judging interactive behavior and determining key time points, the autonomous driving algorithm is optimized. Open-loop and closed-loop simulations are performed to accurately locate separation time points and reduce scene distortion.
It improves the accuracy and efficiency of autonomous driving simulation, reduces system consumption and labor costs, enhances the robustness of simulation algorithms, and increases the utilization rate of complex scenarios.
Smart Images
Figure CN121721935A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of automatic driving, and in particular to an automatic driving simulation method and device, an electronic device, and a storage medium. BACKGROUND
[0002] With the development of artificial intelligence technology and modern manufacturing industry, automatic driving technology has gradually entered people's daily life and subtly changed people's travel mode. Automatic driving technology has great application prospects in military and civilian use. In military use, it can not only be used as an unmanned transportation platform, but also as an unmanned demolition vehicle, an unmanned combat platform, an unmanned patrol and surveillance vehicle, etc. In civilian use, in addition to bringing convenience to human life, it can also reduce the incidence of traffic accidents and improve road traffic efficiency, but it also increases energy consumption and causes frequent traffic accidents. The update and iteration of automatic driving technology is the key to solving the above problems. Therefore, automatic driving technology has become the focus of attention in the field of automobile industry and academia.
[0003] At present, the most important thing is to ensure that the automatic driving system of the vehicle can run safely and reliably on the road. This requires that before the vehicle is put into production, the automatic driving system installed on the vehicle needs to be verified in multiple dimensions, in all directions, and for multiple miles. However, due to the limitations of real vehicle testing, the testing work is not only inefficient, but also greatly limited by the site and has poor repeatability, resulting in the technical problem that effective vehicle test data cannot be obtained. Therefore, we usually want to verify whether the problem in a certain scene is solved or whether it occurs back through simulation, but since the algorithm duration will change, the mode of closed-loop switching at the time point when the scene is first established may cause scene distortion when the algorithm version changes greatly, resulting in invalid verification.
[0004] In summary, how to provide an automatic driving simulation strategy to avoid scene distortion and improve the accuracy and reliability of the automatic driving simulation environment has become a technical problem to be solved in the field of automatic driving technology. SUMMARY
[0005] The embodiments of the present application provide an automatic driving simulation method, device, electronic device and storage medium to partially or completely solve the above technical problems.
[0006] In a first aspect, the embodiments of the present application provide an automatic driving simulation method, comprising:
[0007] Obtaining to-be-simulated data, the to-be-simulated data comprising an original data structure, a to-be-simulated automatic driving algorithm and a node that needs to be re-executed, wherein the node that needs to be re-executed is used to optimize the algorithm;
[0008] The original data structure is used to determine an interactive behavior of an automatic driving simulation environment and determine a key time point. In a case where the automatic driving simulation environment includes the interactive behavior, the automatic driving simulation environment is classified into an automatic interactive scene, and open-loop simulation is performed on the to-be-simulated data and the node that needs to be executed and does not include a chassis, to obtain first simulation data and an interactive time period set. The key time point includes a time point at which automatic driving is switched to manual driving or a time point at which an important interactive behavior occurs.
[0009] Based on the interactive time period set and the key time point, a separation time point is determined according to a separation time point definition rule, where the separation time point represents a time point at which a position of a vehicle starts to change following a new algorithm change.
[0010] According to the separation time point, closed-loop simulation is performed on the first simulation data and the node that needs to be re-executed and includes the chassis, to obtain second simulation data.
[0011] In a second aspect, an embodiment of the present application provides an automatic driving simulation system, including:
[0012] A to-be-simulated data acquisition module is configured to acquire to-be-simulated data, where the to-be-simulated data includes an original data structure, a to-be-simulated automatic driving algorithm, and a node that needs to be re-executed, and the node that needs to be re-executed is used to optimize the algorithm.
[0013] An open-loop simulation module is configured to determine an interactive behavior of an automatic driving simulation environment based on an original data structure and determine a key time point. In a case where the automatic driving simulation environment includes the interactive behavior, the automatic driving simulation environment is classified into an automatic interactive scene, and open-loop simulation is performed on the to-be-simulated data and the node that needs to be executed and does not include a chassis, to obtain first simulation data and an interactive time period set. The key time point includes a time point at which automatic driving is switched to manual driving or a time point at which an important interactive behavior occurs.
[0014] A separation time point confirmation module is configured to determine a separation time point based on the interactive time period set and the key time point according to a separation time point definition rule. The separation time point represents a time point at which a position of a vehicle starts to change following a new algorithm change.
[0015] A closed-loop simulation module is configured to perform closed-loop simulation on the first simulation data and the node that needs to be re-executed and includes a chassis according to the determined separation time point, to obtain second simulation data.
[0016] In a third aspect, an embodiment of the present application provides an electronic device including a memory, a processor, and a computer program stored in the memory. The processor implements the method described in any of the preceding embodiments when executing the computer program.
[0017] In a fourth aspect, the embodiments of the present application provide a computer readable storage medium, wherein the computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement the method of any one of the above.
[0018] Compared with the prior art, the present application has the following advantages:
[0019] According to the embodiments of the present application, the to-be-simulated data is obtained, which can include an original data structure, a to-be-simulated automatic driving algorithm, and a node that needs to be re-executed, wherein the node that needs to be re-executed contains optimization of the algorithm; based on the original data structure, an interactive behavior judgment is made on the automatic driving simulation environment, and a time point at which the automatic driving is switched to manual driving or a key time point at which an important interactive behavior occurs is determined, in a case where the automatic driving simulation environment contains interactive behaviors, an automatic interactive scenario classification is made on the automatic driving simulation environment, and open-loop simulation is performed on the to-be-simulated data and the node that needs to be executed and does not include a chassis, to obtain first simulation data and an interactive time period set; based on the interactive time period set and the key time point, a separation time point is determined according to a separation time point definition rule, wherein the separation time point represents a time point at which the ego vehicle position starts to change following a new algorithm variation; based on the separation time point, closed-loop simulation is performed on the first simulation data and the node that needs to be re-executed and includes a chassis, to obtain second simulation data. As can be seen from the above process, by introducing a combination mode of open-loop simulation and closed-loop simulation and a judgment for interactive behaviors, the embodiments first perform open-loop simulation to mine key time points of interactive behaviors, and then confirm a chassis closed-loop timing (i.e., a separation time point) of closed-loop simulation based on the key time points, so that accurate simulation under a complex interactive scenario can be achieved, the variation of the simulated automatic driving algorithm is more robust, the separation point is accurately positioned, scene distortion is reduced, the utilization rate of important complex scenarios can be improved, and the efficiency and accuracy of simulation are improved while system consumption and manual cost are significantly reduced.
[0020] The above description is only a summary of the technical solutions of the present application. In order to more clearly understand the technical means of the present application, the embodiments of the present application can be implemented in accordance with the contents of the description, and in order to make the above and other purposes, features and advantages of the present application more apparent and easy to understand, the specific embodiments of the present application are described below. BRIEF DESCRIPTION OF DRAWINGS
[0021] In the drawings, identical or similar components or elements are denoted by identical reference numerals throughout the several views, unless otherwise specified. The drawings are not necessarily to scale. It is to be understood that the drawings only depict some embodiments in accordance with the present application and should not be considered as limiting the scope of the present application.
[0022] Figure 1 A scene schematic diagram of the automatic driving simulation method provided in the present application is shown in the following figure;
[0023] Figure 2 A flow chart of the automatic driving simulation method of an embodiment of the present application is shown in the following figure;
[0024] Figure 3 A comparison schematic diagram of the adjustment process of the separation time point of the simulation verification stage of another embodiment of the present application is shown in the following figure;
[0025] Figure 4 A structural block diagram of the automatic driving simulation device of an embodiment of the present application is shown in the following figure; and
[0026] Figure 5 A block diagram of an electronic device for implementing an embodiment of the present application is shown in the following figure. DETAILED DESCRIPTION
[0027] In the following, only some exemplary embodiments are simply described. As those skilled in the art can recognize, the described embodiments can be modified in various different ways without departing from the concept or scope of the present application. Therefore, the drawings and the description are considered to be exemplary in nature, rather than limiting.
[0028] In order to facilitate understanding of the technical solutions of the embodiments of the present application, the related technologies of the embodiments of the present application are described below. The following related technologies can be combined with the technical solutions of the embodiments of the present application in any way as optional solutions, which all belong to the protection scope of the embodiments of the present application.
[0029] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiment description will be simply introduced below. Obviously, the drawings in the following description are only some examples or embodiments of the present application, and for those skilled in the art, without paying creative labor, the present application can also be applied to other similar scenarios according to these drawings. Unless it is obvious from the language environment or otherwise stated, the same reference numbers in the figures represent the same structures or operations.
[0030] It should be understood that the "system", "device", "unit" and / or "module" used herein is a method for distinguishing different components, elements, parts, portions or assemblies at different levels. However, if other words can achieve the same purpose, the words can be replaced by other expressions.
[0031] As shown in the specification and claims, unless the context clearly indicates otherwise, the words "comprise", "comprising", "consist of" and "consisting of" do not preclude the inclusion of additional steps or elements. In general, the terms "comprise" and "comprising" are used in the sense of "including" rather than "consisting of".
[0032] Flowcharts are used in the specification to illustrate the operation of systems in accordance with embodiments of the specification. It should be understood that the operations in the flowcharts need not necessarily be performed in the order shown. Rather, various steps can be handled in different order or simultaneously. Additionally, other operations can be added or removed from the flowcharts.
[0033] First, before specifically introducing the technical solutions of the embodiments of the disclosure, the technical background or technical evolution context based on which the embodiments of the disclosure are introduced.
[0034] Nowadays, simulation testing is a crucial step in the development process of automatic driving systems on the market. However, due to various reasons, the number of problems in the simulation testing stage of automatic driving systems has increased significantly in the past decade. Many automatic driving research and development teams have encountered various challenges in simulation testing, such as inaccurate scene restoration, imperfect sensor models, and insufficient algorithm verification, which have led to malfunctions or safety hazards in the actual application of automatic driving systems. A large number of car companies, technology companies, and other companies are facing technical difficulties due to the shortcomings of automatic driving systems in the simulation testing stage, and have to retest and improve, incurring huge time and economic losses. Therefore, it is necessary to provide a simulation method for vehicle automatic driving to effectively improve the scene with less resource consumption, meet the accurate simulation under complex scenes, provide a suitable verification period for automatic evaluation, and eliminate the distortion of the scene caused by the too long continuous effect of the new algorithm at the original separation time point.
[0035] When the technical personnel corrects the original automatic driving algorithm through a simulation, if there is an interactive behavior after the automatic driving algorithm, the problem of too early or too late separation time point may occur, which will cause the behavior pattern of the algorithm to change, and the accuracy of the original simulation will be greatly reduced.
[0036] In the related art before the present application, there are generally two kinds. One is to use a parameterized model to dynamically adjust the simulation scene to adapt to changes in different algorithms and environments. The other is to add real-time monitoring simulation of key indicators in simulation and dynamically adjust simulation parameters to avoid distortion. The above technologies often rely on complex scene design, real-time monitoring systems or a large number of parameter adjustments, and require highly complex algorithms and a large amount of computing resources to process and analyze data, increasing the number of simulations that need to be run, the resources consumed, and the cost. In addition, the above solutions cannot effectively ensure the accuracy of the verification at the beginning of the simulation, but are post-adjustment, checking problems through conclusions.
[0037] Therefore, the embodiments of the present application provide an automatic driving simulation method and device, electronic equipment and storage medium to solve all or part of the above technical problems.
[0038] In order to more clearly show the method flow for automatic driving simulation provided in the embodiments of the present application, the following gives an example of the embodiments of the present application. Figure 1 The schematic diagram for the method for implementing the embodiments of the present application is shown in FIG. 1. As shown in FIG. 1, the method provided by the embodiments of the present application includes the following steps. Figure 1As shown, the embodiments of the present application first make a judgment on interactive behavior of the autonomous driving simulation environment based on the original data results, when it is determined that the autonomous driving simulation environment contains interactive behavior, the autonomous driving simulation environment is classified into an automatic interactive scenario, and open-loop simulation is performed on the to-be-simulated data and the nodes that need to be executed without the chassis to obtain first simulation data and an interactive time period set; when it is determined that the autonomous driving simulation environment does not contain interactive behavior, it is considered that the separation time point is unchanged, and closed-loop simulation is performed on the to-be-simulated data and the nodes that need to be re-executed including the chassis according to the original separation time point. The first simulation data is judged whether there is a key time point of the original problem, the original problem refers to the problem found in real world driving or early simulation test, these problems may involve performance defects of autonomous driving vehicles, system failure, or inappropriate response in a specific traffic scenario, when performing simulation test, researchers will try to reproduce these problems in order to analyze the reasons in depth and develop corresponding solutions; when the first simulation data after open-loop simulation does not contain the key time point, closed-loop simulation is performed on the first simulation data and the nodes that need to be re-executed including the chassis to obtain second simulation data; when the first simulation data after open-loop simulation contains the key time point, the separation time point is determined according to the separation time point definition rule based on the interactive time period set and the key time point; when the key time point is the third type of time point, the separation time point is defined as the later time point of the interaction start time point and the key buffer point; when the key time point is not the third type of time point, the separation time point is defined as the key time point; based on the change of the separation time point, closed-loop simulation is performed on the first simulation data and the nodes that need to be re-executed including the chassis to obtain second simulation data. Wherein, open-loop simulation is a simulation mode without feedback mechanism. In this mode, the decision and behavior of the autonomous driving system will not affect the state of the simulation environment. That is, the simulation environment will not respond or change according to the behavior of the autonomous driving system; closed-loop simulation includes feedback mechanism, that is, the decision and behavior of the autonomous driving system will affect the state of the simulation environment in real time, and the change of the environment will in turn affect the input of the system, forming a continuous interactive cycle.
[0039] In the above process, before conducting open-loop simulation, it is determined whether there is interactive behavior. Open-loop simulation is only performed for situations where interactive behavior may lead to premature or late separation points to identify key time points and confirm the separation time point. Interactive behavior refers to the occurrence of interaction between two vehicles. Specific interaction scenarios can be divided into scenarios with inconsistent specific behaviors, such as lane changes (other vehicles changing lanes and inserting into one's own lane queue and one vehicle changing lanes and inserting into another vehicle's lane queue), overtaking, meeting oncoming traffic, merging, and obstacle avoidance. In addition, after determining the existence of interactive behavior and conducting open-loop simulation, it is further determined whether the simulation data obtained from the open-loop simulation contains key time points, and a second screening is performed. Cases that do not contain key time points are directly subjected to closed-loop simulation. Through the combined model of open-loop and closed-loop simulation, and after multiple screenings, accurate simulation of complex scenarios is achieved, improving the robustness of simulation algorithm changes while achieving effective scenario improvement with relatively low resource consumption, and without incurring excessive costs. This solution can complete automated parameter selection during the first open-loop simulation, that is, accurately identifying the latest interaction time point of the new algorithm in the early stage of simulation, dynamically determining the separation time point, and reducing scenario distortion caused by the algorithm.
[0040] This application provides an autonomous driving simulation method, such as... Figure 2 The flowchart shown is a representation of an autonomous driving simulation method according to an embodiment of this application, which may include:
[0041] Step S201: Obtain the simulation data, which includes the original data structure, the autonomous driving algorithm to be simulated, and the nodes that need to be re-executed. The nodes that need to be re-executed are used to optimize the algorithm.
[0042] The raw data structures involved in this application include sensor data, map data, environmental conditions, and intermediate data for autonomous driving transportation. Sensor data includes raw data from sensors used to simulate vehicles in the autonomous driving simulation system, including but not limited to radar data, lidar data, and camera data. Map data includes road and geographic information used to construct the virtual environment. Environmental conditions include weather conditions, lighting conditions, and traffic flow described in the simulation environment. Intermediate data for autonomous driving includes perception fusion results, prediction and planning data, and control commands. These raw data structures constitute a comprehensive simulation environment for testing and verifying the performance of autonomous driving algorithms under various conditions. The above are only some possible data types; in specific implementations, data can be added or removed according to the needs of the simulation environment, and this application does not impose any limitations on this.
[0043] Furthermore, the nodes requiring re-execution in the embodiments of this application include key aspects that require special attention and may need repeated verification during the reverse testing process. These nodes are categorized in this application into re-execution nodes that include the chassis and re-execution nodes that do not. Re-execution nodes that include the chassis include vehicle dynamics simulation, chassis control, and durability testing; re-execution nodes that do not include the chassis include perception system testing, decision-making evaluation, and traffic flow simulation.
[0044] It is understandable that, since the re-execution node involving the chassis involves physical conditions, such as the friction coefficient between the vehicle and the ground, and the wind resistance caused by the weather, this application only includes the chassis in the closed-loop simulation when optimizing the algorithm. This means that in this embodiment, the open-loop simulation will not cause the entire simulation to be distorted due to the error caused by the chassis.
[0045] Step S202: Based on the original data structure, the interactive behavior of the autonomous driving simulation environment is judged and key time points are determined. If it is determined that the autonomous driving simulation environment contains interactive behavior, the autonomous driving simulation environment is classified into automated interactive scenarios, and open-loop simulation is performed on the simulation data and the nodes to be executed excluding the chassis to obtain the first simulation data and the set of interactive time periods. The key time points include the time point when autonomous driving is switched to manual driving or the time point when important interactive behavior occurs.
[0046] To better reduce the amount of data processed and analyzed, and to achieve effective improvement in scenarios with lower resource consumption, in one optional implementation, interactive behavior judgment is performed on the simulation environment based on the original data structure. This also includes judging cases where the simulation scenario does not contain interactive behavior, including performing closed-loop simulation on the data to be simulated and nodes including the chassis that need to be re-executed, to obtain second simulation data. Here, automated interactive scenario classification refers to classifying scenarios with interactive behavior; interactive behavior refers to the occurrence of interaction between two vehicles. Specific interactive scenarios can be divided into scenarios with inconsistent specific behaviors, such as lane changing (other vehicles changing lanes and inserting into the vehicle's queue, and the vehicle changing lanes and inserting into another vehicle's queue), overtaking, meeting oncoming traffic, merging, and obstacle avoidance. In specific implementations, behavioral interaction categories can be added or removed according to the actual interactive scenarios, and this application does not impose any restrictions on this.
[0047] To intercept issues before road testing, a full simulation is performed to observe whether new problems will be introduced. This approach is similar to classifying problems and verifying the effectiveness of solutions within each category, both requiring the identification of critical time points related to the original problem. In one optional implementation, an open-loop simulation is performed on the data to be simulated and the nodes requiring re-execution (excluding the chassis). If the first simulation data after the open-loop simulation does not contain critical time points, a closed-loop simulation can be performed on the first simulation data and the nodes requiring re-execution (including the chassis) to obtain second simulation data. By filtering this scenario, data that does not require subsequent separation point determination is directly subjected to closed-loop simulation, effectively reducing computational resource consumption.
[0048] The aforementioned critical time points refer to the takeover time point or critical interaction time point, denoted as K. The takeover time point refers to the point at which autonomous driving switches to manual driving, while the critical interaction time point refers to the point at which significant interactive behaviors occur. Since both the takeover time point and the critical interaction time point represent when the original problem occurred, their priority needs to be defined. It is understandable that after the takeover time point, the location relationship data of the scene is manipulated manually and has fallen outside the scope of autonomous driving; therefore, a separation time point must be selected at least before the takeover time point. In summary, the priority for determining critical time points is: critical interaction time point before the takeover point > takeover time point > critical interaction point after takeover.
[0049] In one optional implementation, the first simulation data and the set of interaction time periods obtained through open-loop simulation are used. The set of interaction time periods represents the latest time segment at the original position where an interaction must be initiated, i.e., the time point at which interaction must occur. This can be expressed by the formula (F...). s1 F e1 )…(F sN F eN ).
[0050] Step S203: When the first simulation data contains key time points, based on the set of interaction time periods and the key time points, determine the separation time point according to the separation time point definition rule, wherein the separation time point represents the time point at which the vehicle position begins to change following the new algorithm.
[0051] To accurately locate the separation point, improve the utilization rate of important and complex scenarios, provide a suitable verification time period for automated evaluation, and prevent scenario distortion caused by the excessive duration of the new algorithm due to the original separation time point, an optional implementation provides a separation time point definition rule. This rule includes defining the separation time point based on the relationship between key time points and the interaction time period set in the original data structure. Key time points include a first type, a second type, and a third type. The first type includes key interaction points after the takeover point; the second type includes the takeover point; and the third type includes key interaction points before the takeover point. The interaction time period set includes the interaction start time point and the interaction end time point. When the key time point is a first type or a second type, the separation time point is defined as the key time point. When the key time point is a third type, the separation time point is defined as the later time point between the interaction start time point and the key buffer time point. The key buffer time point is defined as the shortest buffer time point set according to the automated interaction scenario classification before the key time point. It is understandable that, since the takeover point is the point at which autonomous driving switches to manual driving, if the separation is only performed at the first type of time point, the original data will be used in the time period after the second type of time point, resulting in distortion.
[0052] In one possible computational instance for autonomous driving scenarios, the separation time point is denoted as D, and its selection rules are as follows:
[0053] Determine whether a certain interaction time period contains a key time point K. If a certain interaction time period contains a key time point K, the separation time point D is defined as max(km, F). s The critical buffer time (K) is the time point later than the interaction start time point and the critical buffer time point. When no interaction time period contains the critical time point K, the separation time point D is defined as the critical time point K, meaning separation occurs from K. Here, m represents a correction value set according to the classification of automated interaction scenarios, used to adaptively adjust the critical buffer time based on different interaction scenarios. For example, vehicle merging scenarios (cut-in) are short, often occurring within 8 seconds; lane changing scenarios are longer, requiring a 20-second buffer, etc. By adaptively setting the critical buffer time, the integrity of the simulation scenario is maintained without introducing longer, useless buffers that could lead to distortion.
[0054] To reduce scene distortion caused by algorithm variations and improve verification accuracy, after dynamically determining the separation time point, an evaluable interaction reference range is provided. Different separation time points D have different evaluable interaction reference ranges. When the separation time point D is defined as max(km, F... sWhen ), the evaluation range is (max(km, F) s ), F e When the separation time point D is defined as the critical time point K, the evaluation range is (K, inf).
[0055] To better understand the role of automatically and accurately determining the separation time point in the simulation verification phase... Figure 3 This is a comparative schematic diagram of the process of adjusting the separation time point in the simulation verification stage according to an embodiment of this application.
[0056] like Figure 3 As shown, during the open-loop simulation reproduction phase, under the same positional relationships, the new algorithm will make a decision on whether interaction is needed. Depending on the algorithm, the decision time may be earlier or later, and the decision-making process may also be interrupted. If a decision requiring interaction occurs before the original takeover time point (i.e., the time when autonomous driving switches to manual driving) or the key interaction time point, and this decision covers the takeover time point or the key interaction time point, it means that the new algorithm must take interactive action at the new time to complete the scenario solution.
[0057] Step S204: Based on the separation time point, perform closed-loop simulation on the first simulation data and the nodes that need to be re-executed, including the chassis, to obtain the second simulation data.
[0058] In one alternative implementation, after obtaining the second simulation data, the second simulation data is automatically evaluated using a function-based evaluation method or a scenario-based evaluation method to obtain one or more of the following: autonomous vehicle performance, driving coordination, standard matching, and learning evolution.
[0059] Corresponding to the application scenarios and methods provided in the embodiments of this application, the embodiments of this application also provide an autonomous driving simulation device. For example... Figure 4 The diagram shown is a structural block diagram of an autonomous driving simulation device according to an embodiment of this application. The autonomous driving simulation device may include:
[0060] The simulation data acquisition module 401 is used to acquire simulation data, which includes the original data structure, the autonomous driving algorithm to be simulated, and the nodes that need to be re-executed, wherein the nodes that need to be re-executed are used to optimize the algorithm.
[0061] The open-loop simulation module 402 is used to judge the interactive behavior of the autonomous driving simulation environment and determine key time points based on the original data structure. When it is determined that the autonomous driving simulation environment contains interactive behavior, the module classifies the autonomous driving simulation environment into automated interactive scenarios and performs open-loop simulation on the data to be simulated and the nodes to be executed excluding the chassis, to obtain the first simulation data and the set of interactive time periods. The key time points include the time point when autonomous driving is switched to manual driving or the time point when important interactive behavior occurs.
[0062] The separation time point confirmation module 403 is used to determine the separation time point based on the interaction time period set and the key time point according to the separation time point definition rules; wherein, the separation time point is the time point at which the vehicle position begins to change following the new algorithm.
[0063] The closed-loop simulation module 404 is used to perform closed-loop simulation on the first simulation data and the nodes that need to be re-executed, including the chassis, according to the determined separation time point, to obtain the second simulation data.
[0064] In one optional embodiment, the apparatus further includes an evaluation module for automatically evaluating the second simulation data using a function-based evaluation method or a scenario-based evaluation method to obtain one or more of the following: autonomous vehicle performance, driving coordination, standard matching, and learning evolution.
[0065] The functions of each module in each device in the embodiments of this application can be found in the corresponding description in the above method, and they have corresponding beneficial effects, which will not be repeated here.
[0066] Figure 5 This is a block diagram of an electronic device used to implement embodiments of this application. Figure 5 As shown, the electronic device includes a memory 510 and a processor 520. The memory 510 stores a computer program that can run on the processor 520. When the processor 520 executes the computer program, it implements the method described in the above embodiments. The number of memories 510 and processors 520 can be one or more.
[0067] The electronic device also includes:
[0068] The communication interface 530 is used to communicate with external devices and exchange and transmit data.
[0069] If the memory 510, processor 520, and communication interface 530 are implemented independently, they can be interconnected via a bus to communicate with each other. This bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. This bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 5 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.
[0070] Optionally, in a specific implementation, if the memory 510, processor 520, and communication interface 530 are integrated on a single chip, then the memory 510, processor 520, and communication interface 530 can communicate with each other through an internal interface.
[0071] This application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method provided in this application.
[0072] This application also provides a chip including a processor for calling and executing instructions stored in a memory, causing a communication device with the chip installed to perform the method provided in this application.
[0073] This application also provides a chip, including: an input interface, an output interface, a processor, and a memory. The input interface, output interface, processor, and memory are connected through an internal connection path. The processor is used to execute code in the memory. When the code is executed, the processor is used to execute the method provided in the application embodiment.
[0074] It should be understood that the aforementioned processor can be a Central Processing Unit (CPU), or other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. General-purpose processors can be microprocessors or any conventional processor. It is worth noting that the processor can be a processor supporting Advanced Reduced Instruction Set Machines (ARM) architecture.
[0075] Further, optionally, the aforementioned memory may include read-only memory and random access memory. The memory may be volatile memory or non-volatile memory, or may include both. Non-volatile memory may include read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory may include random access memory (RAM), which serves as an external cache. By way of example, but not limitation, many forms of RAM are available. Examples include Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDR SDRAM), Enhanced Synchronous DRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), and Direct Rambus RAM (DR RAM).
[0076] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions according to this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transferred from one computer-readable storage medium to another.
[0077] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of those different embodiments or examples.
[0078] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "a plurality of" means two or more, unless otherwise explicitly specified.
[0079] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing a particular logical function or process. Furthermore, the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functionality involved.
[0080] The logic and / or steps described in the flowchart or otherwise herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus or device (such as a computer-based system, a processor-included system or other system that can fetch and execute instructions from, an instruction execution system, apparatus or device).
[0081] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. All or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware, the program being stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiments.
[0082] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. This storage medium can be a read-only memory, a disk, or an optical disk, etc.
[0083] The above description is merely an exemplary embodiment of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various variations or substitutions within the technical scope described in this application, and these should all be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.
Claims
1. An autonomous driving simulation method, wherein, The method includes: Obtain the simulation data, which includes the original data structure, the autonomous driving algorithm to be simulated, and the nodes that need to be re-executed, wherein the nodes that need to be re-executed are used to optimize the algorithm; Based on the original data structure, interactive behavior judgment is performed in the autonomous driving simulation environment and key time points are determined. If interactive behavior is determined to be included in the autonomous driving simulation environment, the autonomous driving simulation environment is classified into automated interactive scenarios, and open-loop simulation is performed on the simulation data and the nodes to be executed excluding the chassis to obtain the first simulation data and the set of interactive time periods. The key time points include the time point when autonomous driving is switched to manual driving or the time point when important interactive behavior occurs. When the first simulation data contains key time points, the separation time point is determined based on the set of interaction time periods and the key time points, according to the separation time point definition rules. The separation time point represents the time point at which the vehicle position begins to change following the new algorithm. Based on the separation time point, a closed-loop simulation is performed on the first simulation data and the nodes that need to be re-executed, including the chassis, to obtain the second simulation data.
2. The method according to claim 1, wherein, The method for determining interactive behavior in the simulation environment based on the original data structure also includes determining cases where the simulation scene does not contain interactive behavior. The method includes: A closed-loop simulation is performed on the data to be simulated and the nodes that need to be re-executed, including the chassis, to obtain the second simulation data.
3. The method according to claim 1, wherein, The method of performing open-loop simulation on the data to be simulated and the nodes that need to be re-executed (excluding the chassis) also includes cases where the first simulation data after the open-loop simulation does not contain key time points. A closed-loop simulation is performed on the first simulation data and the nodes that need to be re-executed, including the chassis, to obtain the second simulation data.
4. The method according to claim 1, wherein, The rules for defining the separation time point specifically include: Based on the relationship between key time points and interaction time period sets in the original data structure, key time points include a first type of key time point, a second type of key time point, and a third type of key time point; the first type of key time point includes key interaction points after the takeover point; the second type of key time point includes the takeover point; the third type of key time point includes key interaction points before the takeover point; the interaction time period set includes the interaction start time point and the interaction end time point; When the critical time point is a first-type critical time point or a second-type critical time point, the separation time point is defined as the critical time point; When the critical time point is a third type of critical time point, the separation time point is defined as the later time point between the interaction start time point and the critical buffer time point; wherein, the critical buffer time point is defined as the shortest buffer time point set before the critical time point according to the classification of automated interaction scenarios.
5. The method according to claim 1, wherein, The original data structure includes: Sensor data, map data, environmental conditions, and intermediate data for autonomous driving transportation.
6. The method according to any one of claims 1-5, wherein, The method further includes: The second simulation data is automatically evaluated using function-based or scenario-based evaluation methods to obtain one or more of the following: autonomous vehicle performance, driving coordination, standard matching, and learning evolution.
7. An autonomous driving simulation device, wherein, The device includes: The simulation data acquisition module is used to acquire the simulation data, which includes the original data structure, the autonomous driving algorithm to be simulated, and the nodes that need to be re-executed, wherein the nodes that need to be re-executed are used to optimize the algorithm. The open-loop simulation module is used to judge interactive behavior and determine key time points in the autonomous driving simulation environment based on the original data structure. When it is determined that the autonomous driving simulation environment contains interactive behavior, the module classifies the autonomous driving simulation environment into automated interactive scenarios and performs open-loop simulation on the data to be simulated and the nodes to be executed excluding the chassis, to obtain the first simulation data and the set of interactive time periods. The key time points include the time point when autonomous driving switches to manual driving or the time point when important interactive behavior occurs. The separation time point confirmation module is used to determine the separation time point based on the interaction time period set and the key time points according to the separation time point definition rules; wherein, the separation time point is the time point at which the vehicle position begins to change following the new algorithm. The closed-loop simulation module is used to perform closed-loop simulation on the first simulation data and the nodes that need to be re-executed, including the chassis, according to the determined separation time point, to obtain the second simulation data.
8. The apparatus according to claim 7, wherein, The device further includes: The evaluation module is used to automatically evaluate the second simulation data using a function-based evaluation method or a scenario-based evaluation method, and to obtain one or more of the following: autonomous vehicle performance, driving coordination, standard matching, and learning evolution.
9. An electronic device comprising a memory, a processor, and a computer program stored in the memory, wherein the processor, when executing the computer program, implements the method of any one of claims 1-6.
10. A computer-readable storage medium storing a computer program that, when executed by a processor, implements the method of any one of claims 1-6.