Method for generating automatic driving decision data and related device

Abstract

The present disclosure provides a method for generating automatic driving decision data and related equipment; wherein the method is applied to an automatic driving system, and the method comprises: in the frame-by-frame operation process of the automatic driving decision data, receiving a first message sent by at least one first data source through a mailbox; wherein the first data source is part / whole of original data sources subscribed by the mailbox; in response to an operation instruction of a new frame of automatic driving decision, judging whether the mailbox contains target data meeting the operation condition of the frame to be operated based on the time information of the first message; if yes, performing batch processing on the target data to obtain batch data of the frame to be operated; performing operation process of the frame to be operated on the batch data to obtain the automatic driving decision data of the new frame. The present disclosure can improve the stability of automatic driving decision generation.

Description

Technical Field

[0001] This disclosure relates to the field of autonomous driving technology, and in particular to a method and related equipment for generating autonomous driving decision data. Background Technology

[0002] In the field of autonomous driving technology, it is usually necessary to integrate data from multiple sources to generate driving strategies. However, data from different sources is often difficult to control. The algorithm process for generating autonomous driving decisions can only consider various uncontrollable factors of the data source as much as possible and take different coping methods. Even so, it is difficult to achieve global optimization of the data, resulting in autonomous driving decision generation being too passive and having poor stability, which brings unsafe factors to autonomous driving. Summary of the Invention

[0003] In view of this, the purpose of this disclosure is to provide a method and related equipment for generating autonomous driving decision data, so as to improve the stability of autonomous driving decision generation.

[0004] In a first aspect, embodiments of this disclosure provide a method for generating autonomous driving decision data, applied to an autonomous driving system. The method includes: during frame-by-frame computation of autonomous driving decision data, receiving a first message sent by at least one first data source via an email address; wherein the first data source is part / all of the original data sources subscribed to by the email address; in response to a computation instruction for a new frame of autonomous driving decision, determining, based on the time information of the first message, whether the email address contains target data that meets the computation conditions of the frame to be computed; if so, performing batch processing on the target data to obtain batched data of the frame to be computed; and performing computational flow processing on the batched data to obtain a new frame of autonomous driving decision data.

[0005] Secondly, embodiments of this disclosure provide an apparatus for generating autonomous driving decision data, applied to an autonomous driving system. The apparatus includes: a receiving module, configured to receive a first message sent by at least one first data source via a mailbox during frame-by-frame computation of the autonomous driving decision data; a judging module, configured to, in response to a computation instruction for a new frame of autonomous driving decision, determine, based on the time information of the first message, whether the mailbox contains target data that meets the computation conditions of the frame to be computed; a batching module, configured to, if the target data is contained, perform batching processing on the target data to obtain batched data of the frame to be computed; and a computation module, configured to perform computational processing on the batched data of the frame to be computed to obtain a new frame of autonomous driving decision data.

[0006] Thirdly, embodiments of this disclosure provide an electronic device, including a processor and a memory, wherein the memory stores machine-executable instructions that can be executed by the processor, and the processor executes the machine-executable instructions to implement the above-described method for generating autonomous driving decision data.

[0007] Fourthly, embodiments of this disclosure provide a computer-readable storage medium storing computer-executable instructions. When the computer-executable instructions are invoked and executed by a processor, the computer-executable instructions cause the processor to implement the above-described method for generating autonomous driving decision data.

[0008] The embodiments disclosed herein bring the following beneficial effects: The aforementioned method and related equipment for generating autonomous driving decision data centrally send and receive data collected by the autonomous driving system through an email component. When a new frame of decision data needs to be calculated, the data that meets the conditions in the email are batched and processed to generate decision data. This improves the controllability of data from different data sources and enhances the overall stability of the autonomous driving system.

[0009] Other features and advantages of this disclosure will be set forth in the following description and will be apparent in part from the description or may be learned by practicing the disclosure. The objects and other advantages of this disclosure are realized and obtained through the structures particularly pointed out in the description, claims and drawings.

[0010] To make the above-mentioned objects, features and advantages of this disclosure more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the specific embodiments of this disclosure or the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this disclosure. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0012] Figure 1 This is a flowchart of one embodiment of the method for generating autonomous driving decision data in this disclosure; Figure 2 A schematic diagram of an apparatus for generating autonomous driving decision data provided in an embodiment of this disclosure; Figure 3 This is a schematic diagram of an electronic device provided in an embodiment of the present disclosure. Detailed Implementation

[0013] To make the objectives, technical solutions, and advantages of the embodiments of this disclosure clearer, the technical solutions of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this disclosure, and not all embodiments. Based on the embodiments of this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0014] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in this disclosure, claims, and accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a particular order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” or “having,” and any variations thereof, are intended to cover a 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.

[0015] For ease of understanding, the specific process of the embodiments of this disclosure is described below. The embodiments of this disclosure are applied to autonomous driving systems. Please refer to [link / reference]. Figure 1 One embodiment of the method for generating autonomous driving decision data in this disclosure includes: Step S10: During the frame-by-frame processing of autonomous driving decision data, receive a first message sent by at least one first data source via email; wherein, the first data source is part / all of the original data source subscribed to by the email. An autonomous driving vehicle (ADV) system, also known as an autonomous driving system, is a system used to control a vehicle to drive autonomously. Autonomous driving decision data is the control data used to control the vehicle to drive autonomously, which may include vehicle control commands such as acceleration, deceleration, and steering, etc., without being limited here.

[0016] Autonomous driving decision data is generated frame by frame. Specifically, it can be generated frame by frame according to a preset frame rate. For example, every 100 milliseconds, a new autonomous driving decision data is generated based on a frame of data collected, thus obtaining a new frame of autonomous driving decision data.

[0017] A single frame of data can include datasets captured or generated at the same point in time or over a period of time. In autonomous driving systems, a single frame can contain observations from all sensors (such as cameras, LiDAR, etc.) at the same moment / period. Taking a time period as an example, assuming a preset frame rate of 10Hz, the time window for the first frame is 0-100ms, the time window for the second frame is 101-200ms, and so on for the frames to be processed. The specific time windows are not limited here.

[0018] The autonomous driving system can operate based on a preset computing framework. When generating autonomous driving decision data, the computing process can be based on a Directed Acyclic Graph (DAG) execution framework. In other words, when processing the batch of data into the computing process of the frame to be computed, the batch of data of the frame to be computed is input into the DAG execution framework for processing, thereby obtaining a new frame of autonomous driving decision data.

[0019] In computer science, a graph is a data structure composed of nodes and edges. In a Directed Acyclic Graph (DAG), nodes represent operations or tasks in computation, and edges represent dependencies or data flows between operations. "Directed" means that edges have a direction, indicating that the output of one operation is the input of another operation. "Acyclic" means that there are no cycles in the graph, ensuring that operations can be executed in a certain order without infinite loops.

[0020] In the pre-defined computational framework of an autonomous driving system, DAG is a data structure and execution model used to indicate the computational flow of an algorithm. By decomposing the algorithm into a series of operations and using DAG to indicate the dependencies between these operations, the pre-defined computational framework can effectively manage and schedule the execution of the algorithm, making the autonomous driving system portable, lightweight (minimized dependencies, focused functions), efficient and safe.

[0021] In one implementation, a DAG (Directed Acyclic Graph) is used to indicate the computational flow for processing a frame of data. The execution of the DAG is periodic; that is, every certain period of time, the DAG needs to process a new frame of data to generate a new frame of autonomous driving decision data to maintain the real-time performance of the system. When a new frame of autonomous driving decision data needs to be generated, the DAG requests the corresponding message batch from the Mailbox. Only at this time will the Mailbox deliver the previously collected message batch (i.e., the combined batch data of the frames to be processed) to the DAG, so that the DAG can process the combined batch data according to the computational flow of the frames to be processed to obtain a new frame of autonomous driving decision data.

[0022] The mailbox is a component in the pre-defined computing framework. Through the mailbox, messages can be sent, received, and processed in a centralized manner, enabling centralized message management. This allows for better control over the arrival and processing timing of messages, creating conditions for system optimization and solving the problem of poor system stability caused by inconsistent data sources.

[0023] Email can obtain messages from different data sources by subscribing to data sources. The data sources that the email subscribes to are called raw data sources. Raw data sources can include any data source required for generating autonomous driving decision data, such as various sensors (e.g., cameras, LiDAR), user commands, and other systems of autonomous vehicles. Specific data sources are not limited here.

[0024] Within the time window of the frame to be processed, if the mailbox receives data from at least one source from different original data sources, then the source of the data received by the mailbox within that time window is considered the first data source. For example, suppose the mailbox subscribes to three data sources: original data source A, original data source B, and original data source C. If, within the time window of the frame to be processed, the mailbox only receives data from original data source B and original data source C, then original data source B and original data source C are considered the first data sources. The specific source is not limited here.

[0025] Within the time window of the frame to be processed, any data source can send the data it has collected or specified according to its own data collection frequency or the frame rate of the autonomous driving decision data generation. In other words, within the time window of the frame to be processed, any data source may not send any data to the mailbox, may send data to the mailbox once, or may send data to the mailbox more than once. The specifics are not limited here.

[0026] Regardless of the number of times the data source sends data, within the time window of the frame to be processed, the data received by the mailbox from any original data source is the first message. The first message contains data sent at least once by at least one data source. The first data source is the original data source that sends data to the mailbox at least once within the time window of the frame to be processed. The first message contains at least the latest data collected and sent by the first data source, but the specifics are not limited here.

[0027] For example, suppose the autonomous driving system generates autonomous driving decision data at a frequency of 10Hz, and suppose a camera of the autonomous driving system sends the latest acquired image data to the mailbox at a data acquisition frequency of 30Hz. Then, within the time window of the waiting frame, three data transmissions from the camera can be received, which are included in the first message. The specific details are not limited here.

[0028] Similarly, vehicle configuration information, as a data source, may send data to the mailbox at a low frequency. Therefore, during the time window of the waiting frame, the mailbox may not receive any data from the vehicle configuration information data source, or it may only receive data from the vehicle configuration information data source once. The specifics are not limited here.

[0029] Step S20: In response to the calculation instruction of the new frame of autonomous driving decision, based on the time information of the first message, determine whether the mailbox contains target data that meets the calculation conditions of the frame to be calculated. The computational instructions for the new frame of autonomous driving decision-making can be automatically triggered according to a preset frame rate. For example, when the time window for the next frame arrives, the computational instructions for the new frame of autonomous driving decision-making will be automatically triggered, so as to obtain autonomous driving decision data by performing frame-by-frame computation.

[0030] After the operation instruction is triggered, it can first be determined whether the mailbox contains target data that meets the operation conditions of the frame to be operated. If it does not contain the target data or does not contain it completely, it continues to wait until the mailbox receives all the target data that meets the operation conditions of the frame to be operated. Only then will the subsequent batch processing be triggered, so that the operation process of the frame to be operated can be based on the complete target data, avoiding problems caused by incomplete data. At the same time, batch processing can reduce the overhead of repeated execution and improve the throughput of the system.

[0031] When determining whether the mailbox contains target data that meets the conditions for the operation of the frame to be processed, the judgment can be made based on the latest data sent by each data source in the mailbox, or based on the time information of the latest data sent by each data source in the mailbox. The time information includes the time information of the first message, but the specific time information is not limited here.

[0032] When determining whether the mailbox contains target data that meets the operation conditions of the frame to be operated, it is also possible to determine whether the data source corresponding to the latest data in the mailbox is required by the operation process of the frame to be operated. If so, the operation conditions of the frame to be operated can be determined according to the latest data or the time information of the latest data. This allows the mailbox to increase the certainty of data from different sources, making the execution mode of the autonomous driving system more controllable and predictable, thereby simplifying the difficulty of system optimization.

[0033] Step S30: If included, perform batch processing on the target data to obtain the batch data of the frame to be processed; Understandably, the core function of Mailbox is to receive messages from the outside and organize them into "batches". The advantage of doing so is that the autonomous driving system can process multiple messages at once, rather than executing each message individually, and can better control the arrival and processing time of messages, creating conditions for system optimization.

[0034] In one implementation, the mailbox can wait for the delivery of external messages. Only when the messages from the data source necessary for the operation process of the current operation frame arrive will a message batch be ready to execute the above operation process. Introducing the Mailbox component into the preset operation framework of the autonomous driving system is a major improvement to the message processing method in the traditional operation framework. Through centralized message processing, it achieves a more controllable execution mode, higher system performance, stronger system determinism and better scalability, laying the foundation for building an efficient and reliable operation framework.

[0035] When batch processing target data, each piece of data that arrives in the mailbox within the time window of the frame to be processed, or whose acquisition time falls within the time window of the frame to be processed, can be assigned the same frame identifier (ID). During batch processing, the target data can be distinguished based on the frame ID, and the target data can be filtered and aggregated in a certain way to obtain the batch data of the frames to be processed. The frame ID can be the timestamp of the message arriving in the mailbox. The frame ID can be maintained at a frequency of 10Hz, and messages arriving within the same frame's time window can share the same frame ID.

[0036] When batch processing target data, if the target data contains data sent more than once from the same data source, the data from the same data source in the target data can be aggregated in a preset way, such as selecting the latest data, the most recently sent data, calculating the average, weighted average, etc. The specific method is not limited here.

[0037] For example, if Mailbox assigns frame IDs at a frequency of 10Hz, then three consecutive image messages from a 30Hz camera will be assigned the same frame ID. After batch processing, the data from that camera that enters the processing flow of the frame to be processed can be the latest of the three image messages, which is not limited here.

[0038] Step S40: Process the batch data according to the operation process of the frame to be processed to obtain the new frame of autonomous driving decision data.

[0039] The computation flow of the frame to be processed is an algorithmic computation flow defined by a DAG (Directed Acyclic Graph). Nodes in the DAG represent computational operations, and edges represent data dependencies. The DAG executes these operations in topological order. When all the input data (i.e., messages or computation results from previous nodes) of a node is ready, that node is triggered to execute. In this embodiment, the Mailbox provides the prepared message batches (i.e., batched data) to the DAG. The DAG processes the messages according to their content and its own computational flow to obtain the autonomous driving decision data for the new frame.

[0040] For example, in the process of processing batch data into frames based on DAG, the first node can be to preprocess the LiDAR data (e.g., remove noise), the second node can be to process the camera images (e.g., identify objects), the subsequent nodes can be to fuse data from different sensors and then perform path planning, and finally, the output of DAG can be the vehicle control commands (e.g., acceleration, deceleration, steering), which is the autonomous driving decision data of the new frame.

[0041] The method for generating autonomous driving decision data provided by the above implementation method centrally sends and receives data collected by the autonomous driving system through an email component. When a new frame of decision data needs to be calculated, the data that meets the conditions in the email are batched and processed to generate decision data. This improves the controllability of data from different data sources and enhances the overall stability of the autonomous driving system.

[0042] Next, we will explain in detail how autonomous driving decision data is generated.

[0043] In one implementation, the step of receiving a first message sent by at least one first data source via an email address during the frame-by-frame computation of autonomous driving decision data includes: during the frame-by-frame computation of autonomous driving decision data, all original data sources subscribed to by the email address send the collected data to the email address according to the corresponding data collection frequency; within the time window of the frame to be computed, the original data source of the collected data is sent to the email address and identified as the first data source; within the time window of the frame to be computed, the data received by the email address and sent at least once by the first data source is identified as the first message.

[0044] Different data sources can define their own data collection frequency. Since data sources generally do not have storage functions, the data they collect is usually sent immediately to a module with storage functions. In this embodiment, the mailbox, as a module with storage functions, can receive data from different sources collected at different data collection frequencies for centralized processing.

[0045] Within the time window of the frame to be processed, all original data sources that send the collected data to the mailbox are defined as the first data source. Data received by the mailbox that is sent at least once by the first data source within the time window of the frame to be processed is defined as the first message. It should be noted that, due to the different data collection frequencies of different data sources, the first data source may not necessarily include all data sources required for the processing flow of the frame to be processed. Data from data sources not included in the first data source can be obtained from the cache or other methods, which are not specifically limited here.

[0046] In one implementation, after the step of receiving a first message sent by at least one first data source via an email address during the frame-by-frame computation of autonomous driving decision data, the method further includes: deleting data sent by the first data source that belongs to the previous frame time window from the email cache; and caching the first message.

[0047] Understandably, Mailbox provides a message caching mechanism to prevent message loss. The cache can store one or more frames of messages received by the mailbox. This implementation describes saving one frame as a specific implementation method. In this implementation, after the mailbox receives the first message, data in the mailbox cache that was collected at a time or arrived at a time that belongs to the previous frame's time window and was sent by the first data source is deleted, thereby freeing up cache space and reducing the amount of data that needs to be processed.

[0048] After deletion, the first message is saved in the mailbox's cache space. Since the first message contains data sent by the first data source, the data collection time or the time of arrival in the mailbox belongs to the time window of the frame to be processed, this implementation method is equivalent to updating the data corresponding to the first data source in the mailbox's cache space to the time window of the frame to be processed, so as to achieve the purpose of replacing the old message with the new message, thereby releasing the cache space and reducing the amount of data to be processed.

[0049] It should be noted that in the email cache, data not sent by the primary data source is not deleted or updated. The original data sources other than the primary data source still retain the state of the data they sent last time. This allows both frequently updated and infrequently updated data to be updated in the cache at the required frequency. The generation of autonomous driving decision data is no longer limited by the different update frequencies of data from different data sources, thus improving the controllability and stability of the autonomous driving system.

[0050] In one implementation, the step of receiving a first message from at least one first data source via an email during the frame-by-frame computation of autonomous driving decision data includes: receiving a first message from at least one first data source transmitted by a vehicle communication system in airborne mode via a first subscription interface of the email during the frame-by-frame computation of autonomous driving decision data; or reading a first message from at least one first data source from simulation data of an autonomous driving simulator via a second subscription interface of the email.

[0051] Understandably, autonomous driving systems can operate in different environments. The operating environment in which an autonomous driving system is deployed on an actual autonomous vehicle is called onboard mode, while the operating environment in which it is deployed on an autonomous driving simulator is called simulation mode. Simulation mode can be used for the development and testing of autonomous driving systems.

[0052] In airborne mode, data is usually transmitted in real time through the vehicle's communication system (e.g., CAN bus, Ethernet). Therefore, Mailbox can subscribe to and receive data sources such as sensor data and control commands from the vehicle's communication system through the first subscription interface based on communication, thereby obtaining the first message.

[0053] In simulation mode, the data is typically pre-recorded and stored in a designated file. This file can be in a format used to store large amounts of sensor data; the data stored in this file is the simulation data from the autonomous driving simulator. Mailbox can obtain the first message by reading the simulated sensor data from the designated file through a second subscription interface based on a data reader.

[0054] In this embodiment, different subscription interfaces of Mailbox can be flexibly implemented according to actual needs to adapt to different data sources and operating environments. The simulation environment can also use email, providing a flexible and efficient virtual environment for the development and testing of autonomous driving systems.

[0055] In one implementation, the step of determining whether the mailbox contains target data that meets the operation conditions of the frame to be operated, based on the time information of the first message, in response to the operation instruction of the new frame of autonomous driving decision, includes: determining the target data source to which the data required for the operation of the frame to be operated belongs in response to the operation instruction of the new frame of autonomous driving decision; determining whether the freshness of the data corresponding to the target data source in the cache of the mailbox meets the preset operation conditions of the frame to be operated, based on the time information of the first message; if it meets the preset operation conditions of the frame to be operated, determining that the mailbox contains target data that meets the operation conditions of the frame to be operated; wherein, the target data includes data in the cache of the mailbox whose freshness meets the preset operation conditions of the frame to be operated and which has been sent at least once by the target data source.

[0056] In the process of generating autonomous driving decision data, the data source to which the data required for the operation of the frame to be operated belongs is called the target data source. Before entering the operation process of the frame to be operated, it can be determined whether the target data source exists in the cache of the mailbox. If it exists, it can be further determined whether the freshness of the target data source in the cache meets certain conditions. If it does, it can be determined that the mailbox contains target data that meets the operation conditions of the frame to be operated. It is considered that the data of the frame to be operated in the mailbox is ready and can be processed in subsequent batch processing to send the data of this batch to the operation process of autonomous driving decision data.

[0057] Different data sources may have different requirements for data freshness. Data freshness can be determined based on the time difference between the data collection time or the time it arrives in the mailbox and the current time. In one implementation, the time difference can be calculated based on the time in timestamp format. The timestamp can be recorded in the frame ID of the corresponding data in the mailbox to reduce the required cache space.

[0058] If the freshness meets the preset operation conditions of the operation frame, it can contain data sent more than once from the target data source. In other words, the target data may contain multiple data from the same data source (collected or sent at different times). These data can be aggregated through subsequent batch processing to reduce the pressure on the operation process of the operation frame and improve the efficiency of decision data generation.

[0059] In one implementation, the time information includes: the acquisition time or arrival time of data sent at least once by the target data source in the mailbox cache; the step of determining whether the freshness of the data corresponding to the target data source in the mailbox cache meets the preset operation conditions of the frame to be operated based on the time information of the first message includes: determining whether the time difference between the acquisition time or arrival time of the data sent at least once by the target data source in the mailbox cache and the current time is less than or equal to the preset tolerable time difference corresponding to the target data source; if so, it is determined that the mailbox contains target data that meets the operation conditions of the frame to be operated; wherein, the target data includes data in the mailbox cache whose time difference between the acquisition time or arrival time of the data sent at least once by the target data source and the current time is less than or equal to the preset tolerable time difference corresponding to the target data source.

[0060] In some algorithms, using slightly outdated data can still yield acceptable results. For example, in autonomous driving, vehicle speed information may not need to be precisely mapped to the current moment. The algorithm can use speed information delayed by a few milliseconds for path planning without significantly impacting safety. In this case, if, for some reason (e.g., network latency, sensor malfunction), the latest message from a data source does not arrive in a timely manner, Mailbox can use the older message stored in its cache instead of waiting for the new message to arrive.

[0061] This approach improves system robustness, ensuring that the computation process of frames awaiting computation can still obtain data for processing even in the event of message loss or delay, thereby enhancing system reliability. It also reduces latency; in some cases, using cached old messages can avoid waiting for new messages to arrive, thus reducing system response latency.

[0062] Different tolerable time differences can be preset for different target data sources. As long as the data corresponding to the target data source in the mailbox cache is within the time range of the tolerable time difference, it can be used as target data and has the opportunity to enter the calculation process of the frame to be calculated and participate in the generation of autonomous driving decision data.

[0063] For example, in autonomous driving systems, environmental data acquisition sensors, such as cameras, are usually data sources with a high acquisition frequency (e.g., 30 frames per second). When generating autonomous driving decision data, the real-time requirements for environmental data are also high. Therefore, the tolerable time difference corresponding to the data source from the environmental data acquisition sensor is relatively small, such as 10-30ms, but no specific limit is made here.

[0064] In autonomous driving systems, since the vehicle's configuration information generally does not change, its collection / update frequency is much lower than that of environmental data collection sensors. The tolerable time difference corresponding to the data source of vehicle configuration information is relatively large, for example, it can be more than 1 hour, but no specific limit is made here.

[0065] In one implementation, the step of batch processing the target data to obtain batched data of the frame to be processed includes: if the data in the target data whose freshness meets the preset operation conditions of the frame to be processed and which is sent by the target data source more than once, then the data sent at least once by the second data source in the cache is aggregated in a specified manner to obtain the aggregated data corresponding to the second data source; wherein, the second data source refers to the target data source in the target data whose freshness meets the preset operation conditions of the frame to be processed and which is sent more than once within the time window of the frame to be processed; and assigning the same frame identifier to the data corresponding to the target data source and the aggregated data corresponding to the second data source to obtain the batched data of the frame to be processed.

[0066] Within the time window of the frame to be processed, the data source that sends the collected data to the email address more than once is the second data source. In other words, the target data that is included in the target data, sent more than once by any target data source, and whose freshness meets the preset processing conditions of the frame to be processed, corresponds to the target data source of the second data source.

[0067] In the mailbox cache, the data corresponding to the second data source, that is, the data sent more than once by the second data source in the time window of the frame to be processed, can be aggregated in a specified way so that there is only one data corresponding to each second data source, so as to facilitate the processing of the frame to be processed and reduce the amount of messages that need to be processed.

[0068] After obtaining the aggregated data corresponding to the second data source, the second data source required for the operation process of the frame to be operated can be batch processed, and the target data source other than the second data source can be batch processed. The data to be operated includes the aggregated data corresponding to the second data source and the data in the mailbox cache corresponding to the target data source other than the second data source. By assigning the same frame identifier, the batch processing can be completed and the batch data of the frame to be operated can be obtained, making the batch processing more efficient and convenient.

[0069] In one implementation, the step of performing a computational process on the batched data to obtain a new frame of autonomous driving decision data includes: the computational framework of the frame to be processed responding to the batched data preparation command, performing a computational process on the batched data to obtain a new frame of autonomous driving decision data.

[0070] In one implementation, the data structure and execution model of the autonomous driving system (taking DAG as an example) requests the corresponding message batch from the Mailbox when it needs to process a new frame of data. Only at this time will the Mailbox hand over the previously collected message batch to the DAG. The timing of the Mailbox's preparation of message batches is driven by the DAG's needs, rather than by the time the messages arrive. This reduces the overhead of repeated execution of the DAG, increases the system's throughput, and ensures that the DAG obtains all the data required to process a frame of data, avoiding problems caused by incomplete data.

[0071] The computation framework of the frame to be processed can be a preset computation framework of the autonomous driving system. Its output structure and execution model can be based on DAG. When the preset computation framework executes to the node that needs to process a new frame of data, it automatically triggers the computation instruction for the new frame of autonomous driving decision. After waiting for the mailbox to collect the batch data of the frame to be processed, it triggers the batch data preparation instruction, thereby triggering the computation process of the new frame of autonomous driving decision, obtaining the autonomous driving decision data of the new frame. This reduces the overhead of repeated execution of the computation framework, increases the throughput of the system, and ensures that the computation framework obtains all the data required to process a frame of data, avoiding problems caused by incomplete data.

[0072] For the corresponding method embodiments described above, see [link to relevant documentation]. Figure 2 The diagram illustrates a device for generating autonomous driving decision data, applied to an autonomous driving system. The device includes: a receiving module 22, used to receive a first message from at least one first data source via a mailbox during frame-by-frame computation of the autonomous driving decision data; a judging module 24, used to, in response to a computation instruction for a new frame of autonomous driving decision, determine whether the mailbox contains target data that meets the computation conditions of the frame to be computed, based on the time information of the first message; a batching module 26, used to, if the target data is contained, perform batch processing on the target data to obtain batched data for the frame to be computed; and a computation module 28, used to process the batched data according to the computation flow of the frame to be computed to obtain a new frame of autonomous driving decision data.

[0073] The aforementioned autonomous driving decision data generation device centrally sends and receives data collected by the autonomous driving system through an email component. When a new frame of decision data needs to be calculated, the data that meets the conditions in the email is batched and processed to generate decision data. This improves the controllability of data from different data sources and enhances the overall stability of the autonomous driving system.

[0074] Optionally, the receiving module 22 is further configured to: during the frame-by-frame computation of autonomous driving decision data, send the collected data to the mailbox according to the corresponding data collection frequency from all original data sources subscribed to by the mailbox; send the original data source of the collected data to the mailbox within the time window of the frame to be computed, and determine it as the first data source; within the time window of the frame to be computed, determine the data received by the mailbox that has been sent at least once by the first data source as the first message.

[0075] Optionally, the above device further includes: a deletion module for deleting data sent by the first data source that belongs to the previous frame time window from the cache of the mailbox; and a caching module for caching the first message.

[0076] Optionally, the receiving module 22 is further configured to: receive a first message from at least one first data source transmitted by the vehicle communication system in airborne mode through the first subscription interface of the mailbox during the frame-by-frame calculation of the autonomous driving decision data; or, read a first message from at least one first data source from the simulation data of the autonomous driving simulator through the second subscription interface of the mailbox.

[0077] Optionally, the aforementioned judgment module 24 includes: a response unit, configured to determine the target data source to which the data required for the operation of the frame to be operated belongs in response to the operation instruction of a new frame of autonomous driving decision; a judgment unit, configured to determine whether the freshness of the data corresponding to the target data source in the cache of the mailbox meets the preset operation conditions of the frame to be operated based on the time information of the first message; and a determination unit, configured to determine, if it meets the conditions, that the mailbox contains target data that meets the operation conditions of the frame to be operated; wherein the target data includes data in the cache of the mailbox whose freshness meets the preset operation conditions of the frame to be operated and which has been sent at least once by the target data source.

[0078] Optionally, the time information includes: the acquisition time or arrival time of data sent at least once by the target data source in the mailbox's cache; the aforementioned judgment unit is specifically used to: determine, based on the time information, whether the time difference between the acquisition time or arrival time of data sent at least once by the target data source in the mailbox's cache and the current time is less than or equal to the preset tolerable time difference corresponding to the target data source; if so, then determine that the mailbox contains target data that meets the operation conditions of the frame to be operated; wherein, the target data includes data in the mailbox's cache where the time difference between the acquisition time or arrival time of data sent at least once by the target data source and the current time is less than or equal to the preset tolerable time difference corresponding to the target data source.

[0079] Optionally, the batching module 26 is specifically used for: if the data in the target data whose freshness meets the preset operation conditions of the frame to be processed, and which is sent by the target data source more than once, then the data in the cache that is sent at least once by the second data source is aggregated in a specified manner to obtain the aggregated data corresponding to the second data source; wherein, the second data source refers to the target data source in the target data whose freshness meets the preset operation conditions of the frame to be processed, and which sends data more than once within the time window of the frame to be processed; and assigning the same frame identifier to the data corresponding to the target data source and the aggregated data corresponding to the second data source to obtain the batched data of the frame to be processed.

[0080] Optionally, the above-mentioned calculation module 28 is specifically used for: the calculation framework of the frame to be calculated responds to the preparation and readiness instruction of the batch data, performs the calculation process of the frame to be calculated on the batch data, and obtains a new frame of autonomous driving decision data.

[0081] This embodiment also provides an electronic device, including a processor and a memory. The memory stores machine-executable instructions that can be executed by the processor. The processor executes the machine-executable instructions to implement the above-described method for generating autonomous driving decision data. This electronic device can be a server or a terminal device.

[0082] See Figure 3 As shown, the electronic device includes a processor 100 and a memory 101. The memory 101 stores machine-executable instructions that can be executed by the processor 100. The processor 100 executes the machine-executable instructions to implement the above-described method for generating autonomous driving decision data.

[0083] Furthermore, Figure 3 The electronic device shown also includes a bus 102 and a communication interface 103, with the processor 100, the communication interface 103 and the memory 101 connected via the bus 102.

[0084] The memory 101 may include high-speed random access memory (RAM) and may also include non-volatile memory, such as at least one disk storage device. Communication between this system network element and at least one other network element is achieved through at least one communication interface 103 (which can be wired or wireless), such as the Internet, wide area network, local area network, metropolitan area network, etc. The bus 102 may be an ISA bus, PCI bus, or EISA bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 3 The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0085] The processor 100 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of the processor 100 or by instructions in software form. The processor 100 may be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it may also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this disclosure. The general-purpose processor may be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this disclosure can be directly manifested as execution by a hardware decoding processor, or execution by a combination of hardware and software modules in the decoding processor. The software module can reside in a readily available storage medium in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, or registers. This storage medium is located in memory 101. The processor 100 reads information from memory 101 and, in conjunction with its hardware, completes the steps of the method described in the foregoing embodiments, for example: During the frame-by-frame computation of autonomous driving decision data, a first message is received via an email address from at least one first data source; wherein, the first data source is part or all of the original data sources subscribed to by the email address; in response to the computation instruction for a new frame of autonomous driving decision, based on the time information of the first message, it is determined whether the email address contains target data that meets the computation conditions of the frame to be computed; if it does, the target data is batch-processed to obtain batch data of the frame to be computed; the batch data is processed according to the computation flow of the frame to be computed to obtain the new frame of autonomous driving decision data.

[0086] In this approach, data collected by the autonomous driving system is centrally sent and received via an email component. When a new frame of decision data needs to be calculated, the data in the email that meets the conditions is batched together and processed to generate decision data. This improves the controllability of data from different data sources and enhances the overall stability of the autonomous driving system.

[0087] Optionally, the step of receiving a first message sent by at least one first data source via email during the frame-by-frame computation of autonomous driving decision data includes: during the frame-by-frame computation of autonomous driving decision data, all original data sources subscribed to by the email send the collected data to the email according to the corresponding data collection frequency; within the time window of the frame to be computed, the original data source of the collected data is sent to the email and identified as the first data source; within the time window of the frame to be computed, the data received by the email and sent at least once by the first data source is identified as the first message.

[0088] Optionally, after the step of receiving a first message sent by at least one first data source via email during the frame-by-frame computation of autonomous driving decision data, the method further includes: deleting data sent by the first data source that belongs to the previous frame time window from the cache of the email; and caching the first message.

[0089] Optionally, the step of receiving a first message from at least one first data source via an email during the frame-by-frame computation of autonomous driving decision data includes: receiving a first message from at least one first data source transmitted by the vehicle communication system in airborne mode via a first subscription interface of the email during the frame-by-frame computation of autonomous driving decision data; or reading a first message from at least one first data source from the simulation data of the autonomous driving simulator via a second subscription interface of the email.

[0090] Optionally, the step of determining whether the mailbox contains target data that meets the operation conditions of the frame to be operated based on the time information of the first message in response to the operation instruction of the new frame of autonomous driving decision includes: determining the target data source to which the data required for the operation of the frame to be operated belongs in response to the operation instruction of the new frame of autonomous driving decision; determining whether the freshness of the data corresponding to the target data source in the cache of the mailbox meets the preset operation conditions of the frame to be operated based on the time information of the first message; if it meets the preset operation conditions of the frame to be operated, determining that the mailbox contains target data that meets the operation conditions of the frame to be operated; wherein, the target data includes data in the cache of the mailbox whose freshness meets the preset operation conditions of the frame to be operated and which has been sent at least once by the target data source.

[0091] Optionally, the time information includes: the collection time or arrival time of data sent at least once by the target data source in the mailbox cache; the step of determining whether the freshness of the data corresponding to the target data source in the mailbox cache meets the preset operation conditions of the frame to be processed based on the time information of the first message includes: determining whether the time difference between the collection time or arrival time of the data sent at least once by the target data source in the mailbox cache and the current time is less than or equal to the preset tolerable time difference corresponding to the target data source; if so, it is determined that the mailbox contains target data that meets the operation conditions of the frame to be processed; wherein, the target data includes data in the mailbox cache whose time difference between the collection time or arrival time of the data sent at least once by the target data source and the current time is less than or equal to the preset tolerable time difference corresponding to the target data source.

[0092] Optionally, the step of batch processing the target data to obtain batched data of the frame to be processed includes: if the data in the target data whose freshness meets the preset operation conditions of the frame to be processed and which is sent by the target data source more than once, then the data sent at least once by the second data source in the cache is aggregated in a specified manner to obtain the aggregated data corresponding to the second data source; wherein, the second data source refers to the target data source in the target data whose freshness meets the preset operation conditions of the frame to be processed and which sends data more than once within the time window of the frame to be processed; and assigning the same frame identifier to the data corresponding to the target data source and the aggregated data corresponding to the second data source to obtain the batched data of the frame to be processed.

[0093] Optionally, the step of performing a calculation process on the batched data to obtain a new frame of autonomous driving decision data includes: the calculation framework of the frame to be calculated responds to the preparation command of the batched data, performs a calculation process on the batched data to obtain a new frame of autonomous driving decision data.

[0094] This embodiment also provides a computer-readable storage medium storing computer-executable instructions. When these computer-executable instructions are invoked and executed by a processor, they cause the processor to implement the aforementioned method for generating autonomous driving decision data. For example: During the frame-by-frame computation of autonomous driving decision data, a first message is received via an email address from at least one first data source; wherein, the first data source is part or all of the original data sources subscribed to by the email address; in response to the computation instruction for a new frame of autonomous driving decision, based on the time information of the first message, it is determined whether the email address contains target data that meets the computation conditions of the frame to be computed; if it does, the target data is batch-processed to obtain batch data of the frame to be computed; the batch data is processed according to the computation flow of the frame to be computed to obtain the new frame of autonomous driving decision data.

[0095] In this approach, data collected by the autonomous driving system is centrally sent and received via an email component. When a new frame of decision data needs to be calculated, the data in the email that meets the conditions is batched together and processed to generate decision data. This improves the controllability of data from different data sources and enhances the overall stability of the autonomous driving system.

[0096] Optionally, the step of receiving a first message sent by at least one first data source via email during the frame-by-frame computation of autonomous driving decision data includes: during the frame-by-frame computation of autonomous driving decision data, all original data sources subscribed to by the email send the collected data to the email according to the corresponding data collection frequency; within the time window of the frame to be computed, the original data source of the collected data is sent to the email and identified as the first data source; within the time window of the frame to be computed, the data received by the email and sent at least once by the first data source is identified as the first message.

[0097] Optionally, after the step of receiving a first message sent by at least one first data source via email during the frame-by-frame computation of autonomous driving decision data, the method further includes: deleting data sent by the first data source that belongs to the previous frame time window from the cache of the email; and caching the first message.

[0098] Optionally, the step of receiving a first message from at least one first data source via an email during the frame-by-frame computation of autonomous driving decision data includes: receiving a first message from at least one first data source transmitted by the vehicle communication system in airborne mode via a first subscription interface of the email during the frame-by-frame computation of autonomous driving decision data; or reading a first message from at least one first data source from the simulation data of the autonomous driving simulator via a second subscription interface of the email.

[0099] Optionally, the step of determining whether the mailbox contains target data that meets the operation conditions of the frame to be operated based on the time information of the first message in response to the operation instruction of the new frame of autonomous driving decision includes: determining the target data source to which the data required for the operation of the frame to be operated belongs in response to the operation instruction of the new frame of autonomous driving decision; determining whether the freshness of the data corresponding to the target data source in the cache of the mailbox meets the preset operation conditions of the frame to be operated based on the time information of the first message; if it meets the preset operation conditions of the frame to be operated, determining that the mailbox contains target data that meets the operation conditions of the frame to be operated; wherein, the target data includes data in the cache of the mailbox whose freshness meets the preset operation conditions of the frame to be operated and which has been sent at least once by the target data source.

[0100] Optionally, the time information includes: the collection time or arrival time of data sent at least once by the target data source in the mailbox cache; the step of determining whether the freshness of the data corresponding to the target data source in the mailbox cache meets the preset operation conditions of the frame to be processed based on the time information of the first message includes: determining whether the time difference between the collection time or arrival time of the data sent at least once by the target data source in the mailbox cache and the current time is less than or equal to the preset tolerable time difference corresponding to the target data source; if so, it is determined that the mailbox contains target data that meets the operation conditions of the frame to be processed; wherein, the target data includes data in the mailbox cache whose time difference between the collection time or arrival time of the data sent at least once by the target data source and the current time is less than or equal to the preset tolerable time difference corresponding to the target data source.

[0101] Optionally, the step of batch processing the target data to obtain batched data of the frame to be processed includes: if the data in the target data whose freshness meets the preset operation conditions of the frame to be processed and which is sent by the target data source more than once, then the data sent at least once by the second data source in the cache is aggregated in a specified manner to obtain the aggregated data corresponding to the second data source; wherein, the second data source refers to the target data source in the target data whose freshness meets the preset operation conditions of the frame to be processed and which sends data more than once within the time window of the frame to be processed; and assigning the same frame identifier to the data corresponding to the target data source and the aggregated data corresponding to the second data source to obtain the batched data of the frame to be processed.

[0102] Optionally, the step of performing a calculation process on the batched data to obtain a new frame of autonomous driving decision data includes: the calculation framework of the frame to be calculated responds to the preparation command of the batched data, performs a calculation process on the batched data to obtain a new frame of autonomous driving decision data.

[0103] The computer program product of the method, apparatus, electronic device and storage medium for generating autonomous driving decision data provided in this disclosure includes a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the methods described in the preceding method embodiments. For specific implementation, please refer to the method embodiments, which will not be repeated here.

[0104] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the system and apparatus described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0105] Furthermore, in the description of the embodiments of this disclosure, unless otherwise expressly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this disclosure based on the specific circumstances.

[0106] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this disclosure, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this disclosure. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0107] In the description of this disclosure, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this disclosure and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this disclosure. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0108] Finally, it should be noted that the above embodiments are merely specific implementations of this disclosure, used to illustrate the technical solutions of this disclosure, and not to limit it. The protection scope of this disclosure is not limited thereto. Although this disclosure has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments, or make equivalent substitutions for some of the technical features, within the scope of the technology disclosed in this disclosure. Such modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this disclosure, and should all be covered within the protection scope of this disclosure. Therefore, the protection scope of this disclosure should be determined by the protection scope of the claims.

Claims

1. A method for generating autonomous driving decision data, applied to an autonomous driving system, characterized in that, The method includes: During the frame-by-frame processing of autonomous driving decision data, a first message is received via email from at least one first data source; wherein, the first data source is part / all of the original data sources subscribed to by the email address. In response to the computational instructions for a new frame of autonomous driving decision, based on the time information of the first message, it is determined whether the mailbox contains target data that meets the computational conditions of the frame to be computed; If it is included, the target data is batch processed to obtain batch data of the frame to be processed; The batch data is processed using the processing flow of the frame to be processed to obtain a new frame of autonomous driving decision data.

2. The method according to claim 1, characterized in that, The step of receiving a first message from at least one first data source via email during the frame-by-frame processing of autonomous driving decision data includes: During the frame-by-frame processing of autonomous driving decision data, all original data sources subscribed to by the mailbox send the collected data to the mailbox according to the corresponding data collection frequency; Within the time window of the frame to be processed, the original data source of the collected data is sent to the mailbox and identified as the first data source; Within the time window of the frame to be processed, data received through the mailbox and sent at least once by the first data source is identified as the first message.

3. The method according to claim 1, characterized in that, In the frame-by-frame processing of autonomous driving decision data, after the step of receiving a first message sent by at least one first data source via email, the method further includes: Delete the data sent by the first data source that belongs to the previous frame time window from the cache of the mailbox; The first message is cached.

4. The method according to claim 1, characterized in that, The step of receiving a first message from at least one first data source via email during the frame-by-frame processing of autonomous driving decision data includes: During the frame-by-frame computation of autonomous driving decision data, a first message from at least one first data source transmitted by the vehicle communication system in airborne mode is received via the first subscription interface of the email; or, The first message from at least one first data source is read from the simulation data of the autonomous driving simulator through the second subscription interface of the mailbox.

5. The method according to claim 3, characterized in that, The step of determining whether the mailbox contains target data that meets the calculation conditions of the frame to be calculated, based on the time information of the first message, in response to the calculation instruction for a new frame of autonomous driving decision, includes: In response to the computational instructions for a new frame of autonomous driving decision, determine the target data source to which the data required for the computation of the frame to be computed belongs; Based on the time information of the first message, determine whether the freshness of the data corresponding to the target data source in the cache of the mailbox meets the preset operation conditions of the frame to be operated; If the conditions are met, it is determined that the mailbox contains target data that meets the conditions for the operation of the frame to be processed; wherein, the target data includes data in the cache of the mailbox whose freshness meets the preset conditions for the operation of the frame to be processed, and which has been sent at least once by the target data source.

6. The method according to claim 5, characterized in that, The time information includes: the collection time or the time when the data sent at least once by the target data source in the mailbox's cache is received; The step of determining whether the freshness of the data corresponding to the target data source in the mailbox cache meets the preset calculation conditions of the frame to be calculated based on the time information of the first message includes: Based on the time information, determine whether the time difference between the collection time or the arrival time of the data sent at least once by the target data source in the mailbox cache and the current time is less than or equal to the preset tolerable time difference corresponding to the target data source; If so, it is determined that the mailbox contains target data that meets the operation conditions of the frame to be operated; wherein, the target data includes data in the cache of the mailbox whose time difference between the acquisition time or the time of arrival of the data sent by the target data source at least once and the current time is less than or equal to the preset tolerable time difference corresponding to the target data source.

7. The method according to claim 5, characterized in that, The step of batch processing the target data to obtain batched data of the frame to be processed includes: If the target data contains data sent by the target data source that meets the preset operation conditions of the frame to be processed and has been sent more than once, then the data sent at least once by the second data source in the cache is aggregated in a specified way to obtain the aggregated data corresponding to the second data source; wherein, the second data source refers to the target data source that contains data sent more than once within the time window of the frame to be processed and has been sent more than once. Assign the same frame identifier to the data corresponding to the target data source and the aggregated data corresponding to the second data source to obtain the batch data of the frames to be processed.

8. The method according to claim 1, characterized in that, The step of processing the batched data into a new frame of autonomous driving decision data to obtain the processing flow of the next frame includes: The computation framework of the frame to be computed responds to the preparation-ready instruction of the batch data, performs the computation process of the frame to be computed on the batch data, and obtains a new frame of autonomous driving decision data.

9. An apparatus for generating autonomous driving decision data, applied to an autonomous driving system, characterized in that, The device includes: The receiving module is used to receive a first message sent by at least one first data source via an email during the frame-by-frame processing of autonomous driving decision data. The judgment module is used to respond to the calculation instructions of the new frame of autonomous driving decision and, based on the time information of the first message, determine whether the mailbox contains target data that meets the calculation conditions of the frame to be calculated. The batching module is used to batch the target data if it contains data, to obtain the batched data of the frame to be processed. The calculation module is used to process the batch data into a calculation process to obtain a new frame of autonomous driving decision data.

10. An electronic device, characterized in that, The system includes a processor and a memory, the memory storing machine-executable instructions that can be executed by the processor, the processor executing the machine-executable instructions to implement the method for generating autonomous driving decision data according to any one of claims 1-8.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when invoked and executed by a processor, cause the processor to implement the method for generating autonomous driving decision data according to any one of claims 1-8.

Images