Design method, device and equipment for cloud collaboration and storage medium

Abstract

The invention relates to the field of automatic driving simulation design, in particular to a cloud collaboration design method and device, equipment and a storage medium. The method comprises the following steps: acquiring a driving video sample set and vehicle multi-mode driving parameters; vehicle behavior detail extraction is carried out on the driving video sample set, implicit behavior feature probability learning is carried out, and a driving behavior probability transition rule is generated; obtaining an automatic driving test requirement; performing simulation scene three-dimensional simulation on the automatic driving test requirement to generate a three-dimensional simulation driving scene model; carrying out elastic capacity expansion processing on the vehicle multi-mode driving parameters so as to generate a cloud driving data stream; generating a cloud distributed architecture data stream based on the cloud driving data stream; and performing distributed micro-service function abstraction on the three-dimensional simulation driving scene model based on a driving behavior probability transition rule and a cloud distributed architecture data stream, and constructing a cloud distributed simulation engine. According to the invention, an efficient and accurate automatic driving simulation design method is realized.

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving simulation design, and in particular to a cloud-based collaborative design method, device, equipment and storage medium. Background Art

[0002] With the rapid development of autonomous driving technology, future cars will integrate more emerging technologies such as artificial intelligence and the Internet of Things to achieve vehicle intelligence and autonomy. However, in practical applications, there are still many challenges in the design and debugging of autonomous driving systems. The core algorithms and decision-making mechanisms of autonomous driving engines are extremely complex and require a lot of simulation tests and iterative optimization. Traditional stand-alone simulation systems are difficult to meet the needs of massive data processing and algorithm iteration. The efficiency of autonomous driving simulation data is low and the accuracy of simulation is poor. Autonomous driving systems need to integrate multi-source heterogeneous data from on-board sensors, maps, traffic information, etc. to achieve perception and decision-making in complex driving environments. How to effectively integrate and process these scattered data sources is another major difficulty in the design of autonomous driving systems. The verification of autonomous driving systems needs to run through the entire R&D life cycle, from algorithm design to final vehicle inspection. According to the needs of different stages, a flexible simulation test environment needs to be built to meet the requirements of rapid iteration and full-process verification. In order to solve the above problems, an intelligent autonomous driving engine simulation design method is needed. Summary of the invention

[0003] In order to solve the above technical problems, the present invention proposes a cloud-based collaborative design method, device, equipment and storage medium to solve at least one of the above technical problems.

[0004] To achieve the above object, the present invention provides a cloud-based collaborative design method, comprising the following steps: Step S1: Obtain a driving video sample set and vehicle multimodal driving parameters; extract vehicle behavior details from the driving video sample set, and perform implicit behavior feature probability learning to generate a driving behavior probability transfer law; Step S2: Obtaining the autonomous driving test requirements; performing a three-dimensional simulation of the simulation scene for the autonomous driving test requirements, thereby generating a three-dimensional simulation driving scene model; Step S3: elastically expand the vehicle multimodal driving parameters to generate a cloud driving data stream; and generate a cloud distributed architecture data stream based on the cloud driving data stream; Step S4: Based on the driving behavior probability transfer law and the cloud distributed architecture data flow, the three-dimensional simulation driving scene model is abstracted by distributed microservice functions, and distributed simulation calculations are performed to build a cloud distributed simulation engine; Step S5: Perform cloud driving simulation based on the cloud distributed simulation engine to generate a visualization view of the autonomous driving simulation; Step S6: Dynamically learn the autonomous driving details of the autonomous driving simulation visualization view and optimize the autonomous driving decision, thereby building a distributed simulation optimization model to execute the autonomous driving simulation engine design task.

[0005] The present invention also provides a cloud-based collaborative design device, comprising: The driving behavior analysis module is used to obtain driving video sample sets and vehicle multimodal driving parameters; extract vehicle behavior details from the driving video sample sets, and perform implicit behavior feature probability learning to generate driving behavior probability transfer rules; The three-dimensional scene module is used to obtain the autonomous driving test requirements; perform three-dimensional simulation of the simulation scene for the autonomous driving test requirements, thereby generating a three-dimensional simulation driving scene model; The elastic expansion module is used to elastically expand the vehicle's multi-modal driving parameters to generate a cloud-based driving data stream; based on the cloud-based driving data stream, a cloud-based distributed architecture data stream is generated; The distributed engine module is used to abstract the distributed microservice functions of the three-dimensional simulation driving scene model based on the driving behavior probability transfer law and the cloud distributed architecture data flow, and to perform distributed simulation calculations to build a cloud distributed simulation engine; The simulation visualization module is used to perform cloud-based driving simulation based on the cloud-based distributed simulation engine, thereby generating a visualization view of the autonomous driving simulation; The simulation optimization design module is used to dynamically learn the details of autonomous driving simulation visualization and optimize autonomous driving decisions, thereby building a distributed simulation optimization model to perform autonomous driving simulation engine design operations.

[0006] The present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of the cloud-based collaborative design method described in any one of the above items are implemented.

[0007] The present invention also provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the cloud-based collaborative design method described in any one of the above items are implemented.

[0008] The cloud-based collaborative design method, device, equipment and storage medium provided by the present invention have the following beneficial effects: by extracting vehicle behavior details and learning the probability of implicit behavior features from a driving video sample set, a deeper understanding of driving behavior is achieved, and a driving behavior probability transfer law is generated, thereby improving the prediction capability and accuracy of driving behavior; by performing three-dimensional simulation of simulation scenes for autonomous driving test requirements, a better understanding of test requirements is achieved, and a three-dimensional simulated driving scene model is generated to provide accurate scene data for subsequent simulation tests; elastic expansion processing makes driving data more comprehensive and diversified, providing more diverse data support for subsequent data processing and analysis; cloud-based driving data streams and cloud-based distributed architecture data streams are generated to achieve efficient data management and sharing, thereby improving data According to data processing efficiency and scalability, distributed microservice function abstraction and distributed simulation computing improve the flexibility and performance of the system, optimize the simulation calculation process, improve computing efficiency and response speed, build a cloud-based distributed simulation engine to achieve efficient simulation and calculation of driving scenarios, and provide strong support for the design and testing of autonomous driving systems. The autonomous driving simulation visualization view generated by cloud-based driving simulation can intuitively display the simulation results, help analyze driving scenarios and system performance, and dynamically learn autonomous driving details and optimize autonomous driving decisions to improve the decision-making process of the autonomous driving system, improve the safety and reliability of the system, and build a distributed simulation optimization model to optimize the design of the simulation engine, improve the efficiency and accuracy of the simulation, and better verify the performance and stability of the autonomous driving system. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1 A schematic diagram of the steps of a cloud-based collaborative design method of the present invention; Figure 2 Detailed implementation flow chart of step S1; Figure 3 Detailed implementation flow chart of step S2; Figure 4 Detailed implementation flow chart of step S3. DETAILED DESCRIPTION

[0010] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0011] The present application example provides a cloud-based collaborative design method, device, equipment and storage medium. The execution subjects of the cloud-based collaborative design method, device, equipment and storage medium include but are not limited to the following: mechanical equipment, data processing platform, cloud server node, network upload equipment, etc. that are equipped with the system can be regarded as the general computing nodes of the present application, and the data processing platform includes but is not limited to: at least one of an audio and image management system, an information management system, and a cloud-based data management system.

[0012] In the embodiment of the present invention, refer to Figure 1 , is a schematic flow chart of the steps of a cloud-based collaborative design method of the present invention. In this example, the steps of the cloud-based collaborative design method include: Step S1: Obtain a driving video sample set and vehicle multimodal driving parameters; extract vehicle behavior details from the driving video sample set, and perform implicit behavior feature probability learning to generate a driving behavior probability transfer law; In this embodiment, driving video data from various scenes and environments are collected, covering different driving behaviors and states, ensuring the representativeness and diversity of video samples, laying a foundation for subsequent behavioral feature extraction and analysis. Various driving-related data collected by on-board sensors, cameras and other devices, such as speed, acceleration, steering angle, etc., are simultaneously collected to ensure the correspondence between these multimodal driving parameters and the collected video samples in time and space. Computer vision technology, such as target detection and tracking, is used to extract detailed behavioral features such as the vehicle's motion trajectory and posture changes from the driving video. Combined with multimodal driving parameters, these behavioral features are further enriched and refined. For feature description, implicit behavior features are extracted from the extracted vehicle behavior details, such as the driver's driving style, the complexity of the road environment, etc., and a suitable probability model is selected, such as a Markov chain model, a hidden Markov model, a Bayesian network, etc., to learn the probability distribution of implicit behavior features. The selected probability model is trained using the collected driving video sample set and multimodal driving parameters to learn the probability transfer law of implicit behavior features and generate a driving behavior probability transfer matrix, which describes the transition probability between different driving behaviors, such as the probability of the driver switching from an acceleration state to a braking state under certain circumstances.

[0013] Step S2: Obtaining the autonomous driving test requirements; performing a three-dimensional simulation of the simulation scene for the autonomous driving test requirements, thereby generating a three-dimensional simulation driving scene model; In this embodiment, we work closely with the autonomous driving system R&D team to collect and sort out the scene requirements required for various types of autonomous driving tests, including different road types, traffic conditions, weather conditions, obstacle types and other factors. These demand data provide basic input for subsequent scene modeling and simulation design. Using the scene knowledge graph constructed above, various scene attributes (such as road type, traffic conditions, weather conditions, etc.) in the autonomous driving test requirements are mapped to the corresponding concepts and entities in the knowledge graph to obtain relevant semantic information and associations. Combined with the scene attribute information obtained from the knowledge graph, three-dimensional modeling and physical engine simulation technology are used to automatically construct a three-dimensional simulated driving scene that highly restores the actual environment. The scene contains various elements such as road networks, traffic lights, vehicles, pedestrians, weather effects, etc., and simulates real physical interaction behaviors. Using the powerful distributed computing capabilities of the cloud, the constructed three-dimensional simulation scene is deployed to the cloud, supporting multiple autonomous driving vehicles to perform real-time interactive simulation in the scene. The simulation process can highly restore various dynamic factors in the actual traffic environment, providing a reliable simulation basis for the verification of the autonomous driving algorithm.

[0014] Step S3: elastically expand the vehicle multimodal driving parameters to generate a cloud driving data stream; and generate a cloud distributed architecture data stream based on the cloud driving data stream; In this embodiment, the data is divided into multiple data slices, each of which contains a part of the data, which is convenient for distributed storage and processing. The data is compressed, for example, using compression algorithms such as gzip and zip to reduce data storage space and improve data transmission efficiency. Select a suitable cloud data flow platform, such as Apache Kafka, Apache Flink, Amazon Kinesis, etc., to build a cloud driving data flow. According to the characteristics and requirements of the data, configure the parameters of the data flow platform, such as data topics, data formats, data throughput, etc. Publish the data after elastic expansion processing to the cloud data flow platform to form a cloud driving data flow. Select a suitable cloud distributed architecture, such as microservice architecture, message queue architecture, data warehouse architecture, etc. According to the architecture selection, select the corresponding architecture components, such as message queues, databases, data processing engines, etc. Deploy the architecture components to the cloud environment, for example, using cloud service platforms such as AWS, Azure, GCP, etc.

[0015] Step S4: Based on the driving behavior probability transfer law and the cloud distributed architecture data flow, the three-dimensional simulation driving scene model is abstracted by distributed microservice functions, and distributed simulation calculations are performed to build a cloud distributed simulation engine; In this embodiment, the three-dimensional simulation driving scene model is decomposed into multiple functional modules, such as a road model, a vehicle model, a traffic light model, a pedestrian model, etc. Each functional module is abstracted into a microservice, such as a road model microservice, a vehicle model microservice, a traffic light model microservice, a pedestrian model microservice, etc. A service interface is defined for each microservice for communication and data interaction between other microservices. The driving behavior probability transfer law is used to simulate the driver's driving behavior, such as steering, acceleration, braking, etc. The interaction between the vehicle and the environment is simulated, such as the friction between the vehicle and the road, the collision between the vehicle and the pedestrian, etc. The data of the vehicle sensors is simulated, such as camera data, radar data, GPS data, etc. The cloud distributed architecture data stream is mapped to different microservices, such as sending vehicle sensor data to the vehicle model microservice, sending road information to the road model microservice, etc. Ensure that the data streams between different microservices are synchronized, for example, using technologies such as message queues and databases for data synchronization. Select a suitable distributed simulation framework, such as SimPy, Mesa, Repast, etc., to build a cloud distributed simulation engine. Configure the parameters of the distributed simulation framework according to the simulation requirements, such as simulation time step, simulation accuracy, simulation data storage, etc. Assign simulation tasks to different microservices, for example, assign the simulation task of the vehicle model to the vehicle model microservice, assign the simulation task of the road model to the road model microservice, etc. Use the distributed simulation framework to achieve parallel computing between different microservices and improve simulation efficiency. Ensure data synchronization between different microservices, for example, use technologies such as message queues and databases for data synchronization. Integrate all microservices and the distributed simulation framework to build a cloud-based distributed simulation engine. Deploy the cloud-based distributed simulation engine to the cloud environment, for example, using cloud service platforms such as AWS, Azure, and GCP for deployment.

[0016] Step S5: Perform cloud driving simulation based on the cloud distributed simulation engine to generate a visualization view of the autonomous driving simulation; In this embodiment, a simulation operating environment is constructed in the cloud, and a distributed simulation engine is deployed. Virtual machine technology is used here to run simulation computing tasks in a public cloud or a private cloud. Prepare basic maps and traffic environment resources. Including road network diagrams, traffic light locations, vehicle driving rules, etc. These data need to be preprocessed into a computer-readable format. Define simulation scenarios and participating objects. Such as simulation area, time, participating vehicle type and number and other scenario settings. Configure the autonomous driving algorithm module. It includes two main parts, perception and decision-making, and is simulated based on a real algorithm model. Run the simulation task and perform interactive simulation on the participating objects using a perception-decision-control cycle. Process and parse the simulation results. For example, vehicle driving trajectory, control instructions, and following vehicle interval distribution. Generate 3D visualization results. Package the simulation results into a video or roaming scene in chronological order for display. Compare and verify the simulation effect. Match and verify with the actual test results.

[0017] Step S6: Dynamically learn the autonomous driving details of the autonomous driving simulation visualization view and optimize the autonomous driving decision, thereby building a distributed simulation optimization model to execute the autonomous driving simulation engine design task.

[0018] In this embodiment, the simulation visualization results are labeled to obtain action labels such as lane change and tracking. Vehicle-related behavior features are extracted from them, such as speed, acceleration and deceleration change trends, etc. A deep learning algorithm is applied to train the labeled data set to learn the detailed knowledge of driving dynamics. A dynamic deep neural network model is built in the cloud, and the real-time simulation output is connected to the model input. When the model is running, the vehicle perception input is collected and the optimal decision behavior in the frame is output. The decision behavior is implemented as a control signal and input into the simulation engine to perform directional simulation for verification. Compare the actual results with the model output, and the calculated error is reflected to the model update parameters. After the model parameters are iteratively converged, the output is a distributed optimization model framework. The model is directly connected to the simulation platform to realize a parallel intelligent driving decision execution system. Continuously learn new samples for optimization and continuously improve the intelligence level of autonomous driving.

[0019] In this embodiment, refer to Figure 2 , is a flowchart of detailed implementation steps of step S1. In this embodiment, the detailed implementation steps of step S1 include: Step S11: Obtaining a driving video sample set and vehicle multimodal driving parameters; Step S12: extracting vehicle behavior details from the driving video sample set, thereby obtaining vehicle behavior detail data; Step S13: performing convolutional spatial coding on the vehicle behavior detail data to generate vehicle spatial coding features; Step S14: Tracking the spatial coding features of the vehicle in continuous frames in the time domain to generate a trajectory of vehicle behavior changes; Step S15: performing deep behavior feature mining on the vehicle behavior change trajectory to obtain driving behavior feature data; Step S16: Perform implicit behavior feature probability learning on the driving behavior feature data to generate a driving behavior probability transfer rule.

[0020] In this embodiment, after obtaining authorization from the user and the cloud platform, a driving video sample set is collected and obtained from public data sets, vehicle cameras, simulated driving environments, etc. Multimodal driving parameters of the vehicle are collected synchronously, for example: Video data: The driving video itself contains information such as road environment, vehicle status, driver behavior, etc. Sensor data: Vehicle sensor data, such as speed, steering wheel angle, accelerator pedal, brake pedal, etc. GPS data: Vehicle location information, such as latitude and longitude, altitude, etc. Other data: Other related data, such as timestamp, weather conditions, road conditions, etc. Preprocess the driving video, such as removing noise, adjusting brightness, cropping the screen, etc. Use target detection algorithms, such as YOLO, FasterR-CNN, etc., to identify targets such as vehicles, roads, pedestrians, traffic lights, etc. in the video screen. Use behavior recognition algorithms, such as LSTM, CNN, etc., to identify vehicle behaviors, such as steering, acceleration, deceleration, braking, etc. Extract detailed information about vehicle behavior, such as steering angle, acceleration time, deceleration time, braking distance, etc. Generate vehicle behavior detail data, including vehicle behavior detail information of each video frame, as input data for subsequent steps. Use convolutional neural networks, such as ResNet, VGG, etc., to perform convolutional spatial encoding on vehicle behavior detail data. The convolutional neural network extracts spatial features of vehicle behavior detail data, such as vehicle position, shape, color, and other information. Generate vehicle spatial encoding features, including vehicle spatial feature information of each video frame, as input data for subsequent steps. Use time domain feature extraction algorithms, such as LSTM, RNN, etc., to extract time domain features of vehicle spatial encoding features, such as vehicle motion trajectory, speed change, and other information. Generate vehicle behavior change trajectory based on the extracted time domain features, and record vehicle behavior change information of each video frame as input data for subsequent steps. Use deep learning models, such as LSTM, RNN, Transformer, etc., to perform deep behavior feature mining on vehicle behavior change trajectory. The deep learning model extracts driving behavior features of vehicle behavior change trajectory, such as driving style, driving habits, driving risk, and other information. Generate driving behavior feature data, including driving behavior feature information of each video frame, as input data for subsequent steps. Use Hidden Markov Model (HMM) or other probability models to learn the implicit behavior feature probability of driving behavior feature data. The HMM model generates a driving behavior probability transfer matrix to record the probability transfer relationship between different driving behaviors, such as the probability of transferring from an acceleration state to a deceleration state, the probability of transferring from a normal driving state to a dangerous driving state, etc. Generate a driving behavior probability transfer law, including the probability transfer relationship between driving behaviors, to predict the future change trend of driving behavior.

[0021] In this embodiment, refer to Figure 3, is a flowchart of detailed implementation steps of step S2. In this embodiment, the detailed implementation steps of step S2 include: Step S21: Obtaining autonomous driving test requirements; Step S22: performing scene attribute feature recognition on the autonomous driving test requirements, thereby generating driving scene attribute data; Step S23: performing cross-feature learning on driving scene attribute data to construct a scene knowledge graph; Step S24: Analyze the scenario requirements and use cases for the autonomous driving test requirements, and generate customized scenario use cases; Step S25: Perform three-dimensional simulation of the customized scenario use case based on the scenario knowledge graph to generate a three-dimensional simulated driving scenario model.

[0022] In this embodiment, test requirements proposed by the autonomous driving system development team, road management departments, etc. are collected, and the requirements include test scenarios, test objectives, test indicators and other requirements. The scenario descriptions in the test requirements are analyzed, and attribute features such as road types, weather conditions, and traffic conditions are identified. The identified scene attribute features are organized into structured scene attribute data, and machine learning and other methods are used to analyze the correlation and dependency between the scene attribute data. Based on the association between the attribute features, a knowledge graph describing different driving scenarios is constructed. According to the test requirements, scenario test cases that meet specific goals are designed. The use cases contain information such as scenario attributes, test steps, and expected results. The attribute feature information in the scenario knowledge graph is used to construct a three-dimensional simulation environment, and the customized scenario use cases are mapped to the three-dimensional simulation model to achieve visual simulation of the scene.

[0023] In this embodiment, refer to Figure 4 , is a flowchart of detailed implementation steps of step S3. In this embodiment, the detailed implementation steps of step S3 include: Step S31: performing heterogeneous data fusion on the vehicle multi-modal driving parameters to generate heterogeneous fusion driving parameters; Step S32: Building a cloud data transmission framework based on the cloud platform; Step S33: using a cloud data transmission framework to elastically expand the heterogeneous fusion driving parameters, thereby generating a cloud driving data stream; Step S34: Aggregate the cloud driving data stream in the cloud storage structure to generate a cloud distributed architecture data stream.

[0024] In this embodiment, driving-related data collected by various devices such as on-board sensors, cameras, GPS, etc., such as vehicle speed, location, direction, etc., are collected. Through pre-processing such as data cleaning and format conversion, these heterogeneous data are effectively integrated to generate unified heterogeneous fusion driving parameters. Functional modules such as data reception, transmission, and processing are built on the cloud computing platform, and a cloud data transmission channel based on technologies such as message queues and data buses is designed to achieve two-way data interaction between on-board devices and the cloud. Elastically scalable cloud computing resources are used to cope with high-concurrency driving data generated by large-scale vehicles. The heterogeneous fusion driving parameters are converted into cloud driving data streams that can be transmitted sustainably and stably through a cloud data transmission framework. The cloud driving data streams are stored in a distributed database or object storage system, and a distributed storage architecture is used to achieve high reliability, high scalability, and high-performance access to data.

[0025] In this embodiment, step S4 includes the following steps: Step S41: performing deep behavior coupling on the three-dimensional simulated driving scene model based on the driving behavior probability transfer law to generate an initial driving simulation model; Step S42: performing dynamic driving behavior mapping simulation on the initial driving simulation model to construct a dynamic driving three-dimensional twin model; Step S43: abstracting the distributed microservice functions of the dynamic driving 3D twin model to obtain multiple microservice modules; Step S44: Perform distributed simulation calculations on multiple microservice modules according to the cloud distributed architecture data flow to build a cloud distributed simulation engine.

[0026] In this embodiment, a deep learning algorithm, such as reinforcement learning, deep neural network, etc., is used to couple the driving behavior probability transfer matrix with the three-dimensional simulation driving scene model. Generate an initial driving simulation model, which can simulate the driver's driving behavior in the three-dimensional simulation driving scene model according to the driving behavior probability transfer law. Define the driving behavior mapping rules, and map the different driving behaviors in the driving behavior probability transfer matrix to the specific operations in the three-dimensional simulation driving scene model, for example, mapping the acceleration state to the acceleration operation of the vehicle, and mapping the braking state to the braking operation of the vehicle. Use a dynamic mapping algorithm, such as a mapping based on a time step, a mapping based on event triggering, etc., to dynamically map the driving behavior in the driving behavior probability transfer matrix to the three-dimensional simulation driving scene model. Select a suitable dynamic driving three-dimensional twin model framework, such as a framework based on a physical model, a framework based on machine learning, etc. According to the initial driving simulation model, the driving behavior mapping rules and the dynamic mapping algorithm, set the parameters of the dynamic driving three-dimensional twin model, such as the dynamic parameters of the vehicle, the environmental parameters, etc. Use training data to train the dynamic driving three-dimensional twin model, for example, using historical driving data, simulation data, etc. Build a dynamic driving 3D twin model that can simulate the driver's driving behavior in the 3D simulation driving scene model and predict driving results according to different driving conditions. Divide the functions of the dynamic driving 3D twin model, for example, abstract the functions such as vehicle dynamics simulation, environmental perception simulation, and driving decision simulation into independent modules. Define a microservice interface for each functional module for other modules or external system calls. According to the functional division and microservice interface definition, develop multiple microservice modules, each responsible for a specific function. Test each microservice module to ensure that it functions properly and meets performance requirements. Select a suitable cloud platform, such as AWS, Azure, Google Cloud, etc., for deploying a distributed simulation engine. Design the data flow of the cloud distributed architecture, define the data interaction method between each microservice module, and the data transmission path on the cloud platform. Deploy multiple microservice modules on the cloud platform and connect them according to the data flow design. Build a cloud distributed simulation engine that can coordinate multiple microservice modules for distributed simulation calculations and interact with data according to the data flow.

[0027] In this embodiment, the specific steps of step S5 are: Step S51: performing cloud driving simulation based on a cloud distributed simulation engine, thereby generating autonomous driving simulation data; Step S52: performing interactive simulation response analysis on the autonomous driving simulation data to obtain autonomous driving interactive response data; Step S53: Perform simulation visualization rendering on the autonomous driving interactive response data to generate an autonomous driving simulation visualization view.

[0028] In this embodiment, the cloud-based distributed simulation engine is started and initialized according to the simulation scenario settings. The simulation engine is run to simulate the driving process of the autonomous driving vehicle in the simulation scenario, and relevant data, such as the vehicle's position, speed, steering angle, sensor data, etc., are recorded. Autonomous driving simulation data is generated, including vehicle status data, environmental data, sensor data, etc. An autonomous driving interactive simulation model is constructed, for example, using a neural network model, a decision tree model, etc. to simulate the response of the autonomous driving system to different environments and sensor data. The autonomous driving simulation data is input into the interactive simulation model to simulate the response process of the autonomous driving system. Autonomous driving interactive response data is generated, including the decisions, control instructions, vehicle actions, etc. of the autonomous driving system. Data analysis methods, such as statistical analysis, machine learning, etc., are used to analyze the autonomous driving interactive response data, for example, to analyze the impact of different environments and sensor data on the decision of the autonomous driving system.

[0029] Evaluate the performance of the autonomous driving system based on evaluation indicators, such as safety performance, efficiency performance, and comfort performance. Select a suitable simulation visualization rendering engine, such as Unity, Unreal Engine, etc., to render the autonomous driving simulation visualization view. Set the parameters of the rendering engine as required, such as rendering quality, rendering effect, etc. Input the autonomous driving simulation data and interactive response data into the rendering engine. Use the rendering engine to render the autonomous driving simulation visualization view based on the input data, such as displaying the vehicle's driving trajectory, sensor data, decision-making process, etc. Output the autonomous driving simulation visualization view, such as in the form of video, image, animation, etc.

[0030] In this embodiment, the specific steps of step S6 are: Step S61: performing dynamic autonomous driving detail learning on the autonomous driving simulation visualization view to obtain autonomous driving detail features; Step S62: performing behavior feature evaluation on the autonomous driving detail features to generate an autonomous driving behavior performance evaluation value; Step S63: Optimize the autonomous driving decision of the cloud-based distributed simulation engine based on the autonomous driving behavior performance evaluation value, thereby building a distributed simulation optimization model to execute the autonomous driving simulation engine design task.

[0031] In this embodiment, image processing techniques, such as edge detection, feature point extraction, and target recognition, are used to extract autonomous driving detail features from the autonomous driving simulation visualization view. The extracted feature types include: vehicle driving trajectory, steering angle, vehicle speed change, sensor data change, environmental change, etc. Data analysis methods, such as statistical analysis and machine learning, are used to analyze the extracted autonomous driving detail features, such as analyzing the relationship between different detail features, the impact of different detail features on the autonomous driving system, etc. According to the analysis results, the autonomous driving detail features are clustered, for example, similar features are classified together to facilitate subsequent evaluation and optimization. According to the performance requirements of the autonomous driving system, behavioral feature evaluation indicators are defined, such as safety performance, efficiency performance, and comfort performance. The calculation method of each evaluation indicator is defined, for example, safety performance is measured using indicators such as the number of collisions and the severity of the collision, efficiency performance is measured using indicators such as driving time and driving distance, and comfort performance is measured using indicators such as acceleration change and steering angle change. An autonomous driving behavior performance evaluation model is constructed, for example, using a neural network model, a decision tree model, etc., and the autonomous driving detail features are evaluated according to the evaluation indicators and calculation methods. Use the evaluation model to evaluate each autonomous driving detail feature and generate an autonomous driving behavior performance evaluation value. Define optimization goals according to the performance requirements of the autonomous driving system, such as improving safety performance, improving efficiency performance, improving comfort performance, etc. Select appropriate optimization strategies according to the optimization goals, such as optimization based on reinforcement learning, optimization based on genetic algorithms, etc. Select an appropriate distributed simulation optimization model framework, such as a framework based on an agent model, a framework based on multi-objective optimization, etc. Set the parameters of the distributed simulation optimization model according to the optimization goals, optimization strategies, and autonomous driving behavior performance evaluation values. Use training data to train the distributed simulation optimization model, such as using historical simulation data, simulated data, etc. Assign optimization tasks to multiple microservice modules in the cloud-based distributed simulation engine. Each microservice module optimizes the autonomous driving decision according to the optimization strategy, such as adjusting vehicle control parameters, optimizing path planning, etc. Rerun the autonomous driving simulation based on the optimization results, and use the autonomous driving behavior performance evaluation model to evaluate the optimization effect.

[0032] In this embodiment, the present invention further provides a cloud-based collaborative design device, including: The driving behavior analysis module is used to obtain driving video sample sets and vehicle multimodal driving parameters; extract vehicle behavior details from the driving video sample sets, and perform implicit behavior feature probability learning to generate driving behavior probability transfer rules; The three-dimensional scene module is used to obtain the autonomous driving test requirements; perform three-dimensional simulation of the simulation scene for the autonomous driving test requirements, thereby generating a three-dimensional simulation driving scene model; The elastic expansion module is used to elastically expand the vehicle's multi-modal driving parameters to generate a cloud-based driving data stream; based on the cloud-based driving data stream, a cloud-based distributed architecture data stream is generated; The distributed engine module is used to abstract the distributed microservice functions of the three-dimensional simulation driving scene model based on the driving behavior probability transfer law and the cloud distributed architecture data flow, and to perform distributed simulation calculations to build a cloud distributed simulation engine; The simulation visualization module is used to perform cloud-based driving simulation based on the cloud-based distributed simulation engine, thereby generating a visualization view of the autonomous driving simulation; The simulation optimization design module is used to dynamically learn the details of autonomous driving simulation visualization and optimize autonomous driving decisions, thereby building a distributed simulation optimization model to perform autonomous driving simulation engine design operations.

[0033] The present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor implements any of the above-mentioned cloud-based collaborative design methods when executing the computer program.

[0034] The present invention also provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the method steps for implementing any of the above-mentioned cloud-based collaborative design methods.

[0035] Those skilled in the art clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0036] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it is stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or partly or all or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions for a computer device (a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media for storing program codes such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks or optical disks. Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0037] The above is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A cloud-based collaborative design method, characterized in that: The following steps are involved: Step S1: Obtain a driving video sample set and vehicle multimodal driving parameters; extract vehicle behavior details from the driving video sample set, and perform implicit behavior feature probability learning to generate a driving behavior probability transfer law; Step S2: Obtaining the autonomous driving test requirements; performing a three-dimensional simulation of the simulation scene for the autonomous driving test requirements, thereby generating a three-dimensional simulation driving scene model; Step S3: elastically expand the vehicle multimodal driving parameters to generate a cloud driving data stream; and generate a cloud distributed architecture data stream based on the cloud driving data stream; Step S4: Based on the driving behavior probability transfer law and the cloud distributed architecture data flow, the three-dimensional simulation driving scene model is abstracted by distributed microservice functions, and distributed simulation calculations are performed to build a cloud distributed simulation engine; Step S5: Perform cloud driving simulation based on the cloud distributed simulation engine to generate a visualization view of the autonomous driving simulation; Step S6: Dynamically learn the autonomous driving details of the autonomous driving simulation visualization view and optimize the autonomous driving decision, thereby building a distributed simulation optimization model to execute the autonomous driving simulation engine design task.

2. The cloud-based collaborative design method according to claim 1, characterized in that: The specific steps of step S1 are: Step S11: Obtaining a driving video sample set and vehicle multimodal driving parameters; Step S12: extracting vehicle behavior details from the driving video sample set, thereby obtaining vehicle behavior detail data; Step S13: performing convolutional spatial coding on the vehicle behavior detail data to generate vehicle spatial coding features; Step S14: Tracking the spatial coding features of the vehicle in continuous frames in the time domain to generate a trajectory of vehicle behavior changes; Step S15: performing deep behavior feature mining on the vehicle behavior change trajectory to obtain driving behavior feature data; Step S16: Perform implicit behavior feature probability learning on the driving behavior feature data to generate a driving behavior probability transfer rule.

3. The cloud-based collaborative design method according to claim 1, characterized in that: The specific steps of step S2 are: Step S21: Obtaining autonomous driving test requirements; Step S22: performing scene attribute feature recognition on the autonomous driving test requirements, thereby generating driving scene attribute data; Step S23: performing cross-feature learning on the driving scene attribute data to construct a scene knowledge graph; Step S24: Analyze the scenario requirements and use cases for the autonomous driving test requirements, and generate customized scenario use cases; Step S25: Perform three-dimensional simulation of the customized scenario use case based on the scenario knowledge graph to generate a three-dimensional simulated driving scenario model.

4. The cloud-based collaborative design method according to claim 1, characterized in that: The specific steps of step S3 are: Step S31: performing heterogeneous data fusion on the vehicle multimodal driving parameters to generate heterogeneous fusion driving parameters; Step S32: Building a cloud data transmission framework based on the cloud platform; Step S33: using a cloud data transmission framework to elastically expand the heterogeneous fusion driving parameters, thereby generating a cloud driving data stream; Step S34: Aggregate the cloud driving data stream in the cloud storage structure to generate a cloud distributed architecture data stream.

5. The cloud-based collaborative design method according to claim 1, characterized in that: The specific steps of step S4 are: Step S41: performing deep behavior coupling on the three-dimensional simulated driving scene model based on the driving behavior probability transfer law to generate an initial driving simulation model; Step S42: performing dynamic driving behavior mapping simulation on the initial driving simulation model to construct a dynamic driving three-dimensional twin model; Step S43: abstracting the distributed microservice functions of the dynamic driving 3D twin model to obtain multiple microservice modules; Step S44: Perform distributed simulation calculations on multiple microservice modules according to the cloud distributed architecture data flow to build a cloud distributed simulation engine.

6. The cloud-based collaborative design method according to claim 1, characterized in that: The specific steps of step S5 are: Step S51: performing cloud driving simulation based on a cloud distributed simulation engine, thereby generating autonomous driving simulation data; Step S52: performing interactive simulation response analysis on the autonomous driving simulation data to obtain autonomous driving interactive response data; Step S53: Perform simulation visualization rendering on the autonomous driving interactive response data to generate an autonomous driving simulation visualization view.

7. The cloud-based collaborative design method according to claim 1, characterized in that: The specific steps of step S6 are: Step S61: performing dynamic autonomous driving detail learning on the autonomous driving simulation visualization view to obtain autonomous driving detail features; Step S62: performing behavior feature evaluation on the autonomous driving detail features to generate an autonomous driving behavior performance evaluation value; Step S63: Optimize the autonomous driving decision of the cloud-based distributed simulation engine based on the autonomous driving behavior performance evaluation value, thereby building a distributed simulation optimization model to execute the autonomous driving simulation engine design task.

8. A cloud-based collaborative design device, characterized in that: include: Driving behavior analysis module, used to obtain driving video sample sets and vehicle multimodal driving parameters; Extract vehicle behavior details from driving video sample sets, learn implicit behavior feature probabilities, and generate driving behavior probability transfer rules; The three-dimensional scene module is used to obtain the autonomous driving test requirements; perform three-dimensional simulation of the simulation scene for the autonomous driving test requirements, thereby generating a three-dimensional simulation driving scene model; The elastic expansion module is used to elastically expand the vehicle's multi-modal driving parameters to generate a cloud-based driving data stream; based on the cloud-based driving data stream, a cloud-based distributed architecture data stream is generated; The distributed engine module is used to abstract the distributed microservice functions of the three-dimensional simulation driving scene model based on the driving behavior probability transfer law and the cloud distributed architecture data flow, and to perform distributed simulation calculations to build a cloud distributed simulation engine; The simulation visualization module is used to perform cloud-based driving simulation based on the cloud-based distributed simulation engine, thereby generating a visualization view of the autonomous driving simulation; The simulation optimization design module is used to dynamically learn the details of autonomous driving simulation visualization and optimize autonomous driving decisions, thereby building a distributed simulation optimization model to perform autonomous driving simulation engine design operations.

9. A computer device comprising a memory and a processor, wherein a computer program is stored in the memory, wherein: When the processor executes the computer program, the steps of the cloud-based collaborative design method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the cloud-based collaborative design method described in any one of claims 1 to 7 are implemented.

Images