Intelligent emergency braking system and method based on dynamic driving scene understanding and prediction
By processing multi-source heterogeneous data and using graph neural network prediction, a dynamic scenario model is constructed to generate an adaptive emergency braking strategy. This solves the safety and response lag problems of existing systems in complex scenarios and realizes intelligent and adaptive emergency braking control.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-01-07
- Publication Date
- 2026-04-07
AI Technical Summary
Existing automatic emergency braking systems suffer from insufficient safety, high false triggering rate, and delayed response when faced with complex, dynamic, and non-standardized driving scenarios due to a lack of deep scenario understanding, interactive behavior prediction, and situation-adaptive decision-making capabilities.
By acquiring multi-source heterogeneous data, preprocessing data, and synchronizing it in time and space, a dynamic scene model is constructed. Graph neural networks are used for scene understanding and prediction to generate an adaptive emergency braking strategy. The strategy is then combined with the driver's state to make decisions and execute control.
It achieves in-depth understanding of complex scenarios and forward-looking risk assessment, generating accurate and reliable adaptive emergency braking strategies, thereby improving the system's safety and response efficiency.
Smart Images
Figure CN121799381A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of vehicle emergency braking control technology, and in particular to an intelligent emergency braking system and method based on dynamic driving scenario understanding and prediction. Background Technology
[0002] Automatic Emergency Braking (AEB) systems are central to modern vehicle safety, but current technologies suffer from systemic bottlenecks when handling complex and dynamic real-world scenarios. At the perception level, these systems rely on traditional sensors and shallow fusion schemes, exhibiting weak recognition capabilities for "long-tail" targets such as irregularly shaped obstacles in road construction zones. Furthermore, information sources are unreliable and incomplete in adverse weather conditions. A deeper bottleneck lies in understanding and prediction: existing systems are essentially "event-driven," lacking a global contextual understanding of driving scenarios (e.g., unable to integrate discrete cones and signs into a "construction zone"). Their analysis is limited to static snapshots, making it difficult to anticipate chain reactions triggered by multi-vehicle interactions. Moreover, their predictive capabilities are isolated, failing to comprehensively assess the interactive intentions of multiple agents and the driver's real-time behavioral tendencies. Ultimately, these deficiencies propagate to the decision-making level, causing AEB to become a lagging, rigid threshold trigger. It either misses the intervention opportunity due to a failure to anticipate risks or triggers unnecessary emergency braking due to misjudgment, severely impacting safety and user experience. Therefore, there is an urgent need in this field for a new AEB technology solution that can deeply understand dynamic scenarios, accurately predict their evolution, and make adaptive intelligent decisions accordingly. Summary of the Invention
[0003] This invention provides an intelligent emergency braking system and method based on dynamic driving scenario understanding and prediction, which is used to solve the fundamental problems of insufficient safety, high false trigger rate and delayed response caused by the lack of deep scenario understanding, interactive behavior prediction and situation adaptive decision-making capabilities of existing automatic emergency braking systems when facing complex, dynamic and non-standard driving scenarios.
[0004] The objective of this invention can be achieved through the following technical solutions: The first aspect of this invention is to provide an intelligent emergency braking system based on dynamic driving scenario understanding and prediction, comprising: The multi-source heterogeneous data acquisition module is used to collect real-time vehicle internal status information data, vehicle external environment information data, and collaborative perception information data, and form multi-source heterogeneous data. The data preprocessing and time alignment module is used to clean, correct, and spatiotemporally synchronize the multi-source heterogeneous data. The Scene Understanding and Evolution Prediction module is used to build dynamic scene models based on preprocessed data, and to perform scene context understanding, future multimodal evolution prediction, and quantitative risk assessment. The context-aware AEB decision and planning module is used to generate an adaptive emergency braking control strategy based on the quantitative risk assessment results and the scenario context. The vehicle execution control interface module is used to convert the control strategy into control commands for the vehicle's underlying actuators.
[0005] Furthermore, the scene understanding and evolution prediction module includes: The feature fusion and representation unit is used to project multi-sensor data onto a unified bird's-eye view space centered on the vehicle itself to obtain a bird's-eye view representation. A dynamic scene graph construction unit is used to construct a dynamic scene graph containing nodes and edges based on the bird's-eye view representation, wherein nodes represent scene entities and edges represent the interaction relationships between entities; The graph neural network inference unit is used to perform multiple rounds of information transmission and feature updates on the dynamic scene graph, output node feature vectors containing global context, and obtain scene semantic classification and entity interaction intent through the decoder. The temporal evolution prediction unit is used to predict multiple possible evolution paths of the scene and their probability of occurrence within multiple future time steps based on the scene graph feature sequence of continuous temporal sequence. The risk assessment unit is used to calculate future risk indicators for each predicted future evolution path and combine them with the driver's state vector to output a comprehensive risk score.
[0006] Furthermore, the graph neural network inference unit uses a message passing mechanism to aggregate information between nodes. After K rounds of iteration, the feature vector of each node incorporates all the information in its K-hop neighborhood.
[0007] Furthermore, the temporal evolution prediction unit includes a long short-term memory network or a gated recurrent unit for encoding the temporal dynamics of the scene, and employs a conditional variational autoencoder or a multi-head attention mechanism as a decoder to generate multimodal future scene predictions.
[0008] Furthermore, the risk assessment unit's inputs include: the minimum predicted collision time, minimum predicted distance, and path conflict identifier calculated based on the predicted trajectory, as well as the driver state vector obtained from the driver's operation signals and driver state encoding.
[0009] Furthermore, the decision logic of the context-aware AEB decision and planning module is as follows: when the future risk score predicted by the system exceeds the preset dynamic threshold, intervention is triggered, and the type and intensity of the intervention strategy are adaptively adjusted according to the risk score, scenario classification and predicted interaction intent.
[0010] A second aspect of the present invention is to provide an intelligent emergency braking method based on dynamic driving scenario understanding and prediction, comprising: S1: Real-time acquisition and preprocessing of multi-source heterogeneous data, and performance of data preprocessing and spatiotemporal alignment; S2: Construct a unified bird's-eye view representation based on preprocessed data, and build a dynamic scene map in real time based on this; S3: Use graph neural networks to reason about the dynamic scene graph to understand the global context of the scene and the interaction intentions of entities; S4: Based on a time-series dynamic scene graph, predict multiple possible future scene evolution paths, perform quantitative risk assessment on each path, and output a comprehensive risk score; S5: Generate a situation-adaptive emergency braking decision strategy based on the comprehensive risk score and scenario context information for each path; S6: Transform the decision-making strategy into specific vehicle control commands and execute them.
[0011] Furthermore, the quantitative risk assessment process integrates a driver state vector generated from vehicle internal state information data, which is used to dynamically correct the risk indicators calculated based on the predicted trajectory.
[0012] A third aspect of the present invention is to provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the intelligent emergency braking method based on dynamic driving scenario understanding and prediction.
[0013] A fourth aspect of the present invention is to provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the intelligent emergency braking method based on dynamic driving scenario understanding and prediction.
[0014] Compared with existing technologies, the beneficial effects of this invention are: real-time acquisition of vehicle internal state information data, vehicle external environment information data, and collaborative perception information data, forming multi-source heterogeneous data; providing the system with a comprehensive, real-time, and multi-layered driving scenario information foundation; cleaning, correcting, and spatiotemporally synchronizing the multi-source heterogeneous data; ensuring the consistency and accuracy of data used in subsequent processing and eliminating information misalignment; constructing a dynamic scenario model based on preprocessed data to perform scenario context understanding, future multimodal evolution prediction, and quantitative risk assessment; enabling the system to possess deep cognition and forward-looking risk assessment capabilities for complex scenarios; generating an adaptive emergency braking control strategy based on the quantitative risk assessment results and scenario context; realizing intelligent and adaptive emergency braking strategy generation based on predicted risk; converting the control strategy into control commands for the vehicle's underlying actuators; ensuring that intelligent decisions can be accurately and reliably converted into actual braking actions of the vehicle; and solving the fundamental problems of insufficient safety, high false triggering rate, and delayed response caused by the lack of deep scenario understanding, interactive behavior prediction, and situational adaptive decision-making capabilities in existing automatic emergency braking systems when facing complex, dynamic, and non-standardized driving scenarios. Attached Figure Description
[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0016] Figure 1 This invention provides a flowchart illustrating an intelligent emergency braking system based on dynamic driving scenario understanding and prediction. Figure 2 This invention provides a flowchart illustrating an intelligent emergency braking method based on dynamic driving scenario understanding and prediction. Detailed Implementation
[0017] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.
[0018] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover 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.
[0019] To address the problems existing in the background technology, this study designs an intelligent emergency braking system and method based on dynamic driving scenario understanding and prediction, which has significant practical implications.
[0020] like Figure 1 As shown, the first aspect of the present invention is to provide an intelligent emergency braking system based on dynamic driving scenario understanding and prediction, comprising: The multi-source heterogeneous data acquisition module 101 is used to collect vehicle internal status information data, vehicle external environment information data and collaborative perception information data in real time, and form multi-source heterogeneous data. The data preprocessing and time alignment module 102 is used to clean, correct and spatiotemporally synchronize the multi-source heterogeneous data. The scene understanding and evolution prediction module 103 is used to build a dynamic scene model based on preprocessed data, and to perform scene context understanding, future multimodal evolution prediction and quantitative risk assessment. The context-aware AEB decision and planning module 104 is used to generate an adaptive emergency braking control strategy based on the quantitative risk assessment results and the scenario context. The vehicle execution control interface module 105 is used to convert the control strategy into control commands for the vehicle's underlying actuators.
[0021] The scene understanding and evolution prediction module 103 includes: The feature fusion and representation unit is used to project multi-sensor data onto a unified bird's-eye view space centered on the vehicle itself to obtain a bird's-eye view representation. A dynamic scene graph construction unit is used to construct a dynamic scene graph containing nodes and edges based on the bird's-eye view representation, wherein nodes represent scene entities and edges represent the interaction relationships between entities; The graph neural network inference unit is used to perform multiple rounds of information transmission and feature updates on the dynamic scene graph, output node feature vectors containing global context, and obtain scene semantic classification and entity interaction intent through the decoder. The temporal evolution prediction unit is used to predict multiple possible evolution paths of the scene and their probability of occurrence within multiple future time steps based on the scene graph feature sequence of continuous temporal sequence. The risk assessment unit is used to calculate future risk indicators for each predicted future evolution path and combine them with the driver's state vector to output a comprehensive risk score.
[0022] The graph neural network inference unit uses a message passing mechanism to aggregate information between nodes. After K iterations, the feature vector of each node incorporates all the information in its K-hop neighborhood.
[0023] The temporal evolution prediction unit includes a long short-term memory network or a gated recurrent unit to encode the temporal dynamics of the scene, and uses a conditional variational autoencoder or a multi-head attention mechanism as a decoder to generate multimodal future scene predictions.
[0024] The risk assessment unit takes as input the following: the minimum predicted collision time, minimum predicted distance, and path conflict identifier calculated based on the predicted trajectory, as well as the driver state vector obtained from the driver's operation signals and driver state encoding.
[0025] The decision logic of the context-aware AEB decision and planning module 104 is as follows: when the future risk score predicted by the system exceeds the preset dynamic threshold, intervention is triggered, and the type and intensity of the intervention strategy are adaptively adjusted according to the risk score, scenario classification and predicted interaction intent.
[0026] like Figure 2 As shown, a second aspect of the present invention is to provide an intelligent emergency braking method based on dynamic driving scenario understanding and prediction, comprising: Step S1: Acquire and preprocess multi-source heterogeneous data in real time, and perform data preprocessing and spatiotemporal alignment.
[0027] Specifically, multi-source heterogeneous data is collected; the multi-source heterogeneous data includes vehicle internal status information data, vehicle external environment information data, and collaborative perception information data.
[0028] This includes vehicle internal status information data (refined CAN bus data); specifically: power (engine speed, vehicle speed, etc.), chassis (yaw rate and vehicle height, etc.), steering (steering wheel angle, steering wheel angular velocity, and steering torque, etc.), and driver operation signals (brake pedal opening / position, brake pedal pressure, brake switch status, accelerator pedal opening, and throttle pedal opening, etc.). CAN stands for Controller Area Network.
[0029] The vehicle's external environment information data is high-resolution video / image data collected by the ADS camera; a dashcam can also be selected here. ADS stands for Advanced Driving-assistance System.
[0030] Among them, collaborative perception information data is V2X communication data collected by radar from vehicles, roads, and networks, including traffic information and early warning information; V2X stands for Vehicle-to-Everything, and V2X is also known as vehicle-to-everything wireless communication technology.
[0031] Preprocessing of multi-source heterogeneous data includes distortion correction, calibration, noise reduction, and image enhancement for image or video data; filtering out abnormal transition values or error frames from CAN bus data; extracting valid information fields from V2X communication data and checking the integrity and rationality of the data.
[0032] Spatiotemporal alignment is performed on the preprocessed multi-source heterogeneous data.
[0033] Step S2: Construct a unified bird's-eye view representation based on the preprocessed data, and build a dynamic scene map in real time based on this.
[0034] The vehicle's external environment information data is mapped onto the bird's-eye view (BEV) space by learning through neural networks, establishing a mapping relationship between image pixels and BEV grids. The collaborative perception information data is directly mapped onto the bird's-eye view space (top view) through coordinate transformation.
[0035] Thus, by mapping vehicle external environment information data and collaborative perception information data onto the bird's-eye view space, a BEV feature map is obtained.
[0036] The specific steps for constructing a dynamic scene graph based on the BEV feature map are as follows: The first step is to detect and identify nodes from the BEV feature map to obtain the node set (V). The second step is to define the relationships (edges) between the computation nodes and obtain the edge set (E). The third step is to assemble a dynamic scene graph using node sets and edge sets.
[0037] It's important to note that nodes (V) represent key entities in the scene, such as dynamic targets like vehicles, other motor vehicles, pedestrians, and non-motorized vehicles (VRUs), as well as static or semi-static targets like traffic lights, traffic signs, cones in road construction zones, guardrails, and irregularly shaped obstacles (like excavators). Each node contains its attribute features, such as position, speed, acceleration, size, and category (e.g., unknown obstacles identified by an anomaly detection model). Edges (E) represent relationships between nodes. Edges can be multi-dimensional, representing spatial proximity, motion interaction (e.g., following, lane merging intentions), or semantic subordination (e.g., cones being part of a temporary lane boundary). Edge attributes can include relative distance and relative speed.
[0038] It should be noted that one frame corresponds to one dynamic scene image.
[0039] At this point, a dynamic scene diagram is obtained.
[0040] Step S3: Use a graph neural network to reason about the dynamic scene graph to understand the global context of the scene and the entity interaction intentions.
[0041] The reasoning and scene understanding of the dynamic scene graph are performed using a graph neural network (GNN); when using a graph neural network, the input is the dynamic scene graph, and the output is scene context information.
[0042] The scene context information includes global scene classification and interaction intent of key entities; and prediction of interaction intent for global scene classification (e.g., construction zone, congested intersection, highway cruise) and each key node (e.g., vehicle, pedestrian) (e.g., probability of the vehicle in front intending to change lanes to the left is 85%).
[0043] Step S4: Based on the time-series dynamic scene graph, predict multiple possible future scene evolution paths, perform quantitative risk assessment on each path, and output a comprehensive risk score.
[0044] Specifically, based on the dynamic scene graphs of the past few frames, a series encoder is used to obtain several possible future scene evolution paths; based on the several possible future scene evolution paths and the risk assessment network, a comprehensive risk score is obtained for each path.
[0045] Among them, the temporal encoder (such as LSTM / GRU) is used to capture the dynamic evolution of the scene; the multimodal predictive decoder (such as conditional variational autoencoder CVAE) is used to generate several different future trajectory sequences; and the risk assessment network combines the predicted trajectory with the driver's state vector to calculate the risk score.
[0046] The driver state vector is composed of vehicle internal state information data.
[0047] The risk assessment network is a multilayer perceptron. Its inputs are the driver's state vector and risk indicators, including the minimum predicted TTC (Time to Collision) between the driver and other vehicles, the minimum predicted distance, and whether path crossing conflicts exist.
[0048] Among them, LSTM stands for Long Short-Term Memory; GRU stands for Gated Recurrent Unit; and CVAE stands for Conditional Variational Auto-Encoder.
[0049] Step S5: Generate a situation-adaptive emergency braking decision strategy based on the comprehensive risk score and scenario context information for each path.
[0050] The emergency braking decision strategy is generated based on a pre-defined contextualized rule base or policy network.
[0051] Step S6: Transform the decision strategy into specific vehicle control commands and execute them.
[0052] This concludes the embodiment.
[0053] A third aspect of the present invention is to provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement an intelligent emergency braking method based on dynamic driving scenario understanding and prediction.
[0054] A fourth aspect of the present invention is to provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements an intelligent emergency braking method based on dynamic driving scenario understanding and prediction.
[0055] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, optical storage, etc.) containing computer-usable program code.
[0056] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, systems, and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0057] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.
[0058] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.
[0059] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the present invention.
Claims
1. An intelligent emergency braking system based on dynamic driving scenario understanding and prediction, characterized in that, include: The multi-source heterogeneous data acquisition module is used to collect real-time vehicle internal status information data, vehicle external environment information data, and collaborative perception information data, and form multi-source heterogeneous data. The data preprocessing and time alignment module is used to clean, correct, and spatiotemporally synchronize the multi-source heterogeneous data. The Scene Understanding and Evolution Prediction module is used to build dynamic scene models based on preprocessed data, and to perform scene context understanding, future multimodal evolution prediction, and quantitative risk assessment. The context-aware AEB decision and planning module is used to generate an adaptive emergency braking control strategy based on the quantitative risk assessment results and the scenario context. The vehicle execution control interface module is used to convert the control strategy into control commands for the vehicle's underlying actuators.
2. The intelligent emergency braking system based on dynamic driving scenario understanding and prediction according to claim 1, characterized in that, The scene understanding and evolution prediction module includes: The feature fusion and representation unit is used to project multi-sensor data onto a unified bird's-eye view space centered on the vehicle itself to obtain a bird's-eye view representation. A dynamic scene graph construction unit is used to construct a dynamic scene graph containing nodes and edges based on the bird's-eye view representation, wherein nodes represent scene entities and edges represent the interaction relationships between entities; The graph neural network inference unit is used to perform multiple rounds of information transmission and feature updates on the dynamic scene graph, output node feature vectors containing global context, and obtain scene semantic classification and entity interaction intent through the decoder. The temporal evolution prediction unit is used to predict multiple possible evolution paths of the scene and their probability of occurrence within multiple future time steps based on the scene graph feature sequence of continuous temporal sequence. The risk assessment unit is used to calculate future risk indicators for each predicted future evolution path and combine them with the driver's state vector to output a comprehensive risk score.
3. The intelligent emergency braking system based on dynamic driving scenario understanding and prediction according to claim 2, characterized in that, The graph neural network inference unit uses a message passing mechanism to aggregate information between nodes. After K rounds of iteration, the feature vector of each node incorporates all the information in its K-hop neighborhood.
4. The intelligent emergency braking system based on dynamic driving scenario understanding and prediction according to claim 2, characterized in that, The temporal evolution prediction unit includes a long short-term memory network or a gated recurrent unit for encoding the temporal dynamics of the scene, and uses a conditional variational autoencoder or a multi-head attention mechanism as a decoder to generate multimodal future scene predictions.
5. The intelligent emergency braking system based on dynamic driving scenario understanding and prediction according to claim 2, characterized in that, The risk assessment unit's inputs include: the minimum predicted collision time, minimum predicted distance, and path conflict identifier calculated based on the predicted trajectory, as well as the driver state vector obtained from the driver's operation signals and driver state encoding.
6. The intelligent emergency braking system based on dynamic driving scenario understanding and prediction according to claim 1, characterized in that, The decision logic of the context-aware AEB decision and planning module is as follows: when the future risk score predicted by the system exceeds the preset dynamic threshold, intervention is triggered, and the type and intensity of the intervention strategy are adaptively adjusted according to the risk score, scenario classification and predicted interaction intent.
7. An intelligent emergency braking method based on dynamic driving scenario understanding and prediction, characterized in that, An intelligent emergency braking system based on dynamic driving scenario understanding and prediction as described in any one of claims 1-6, comprising: S1: Real-time acquisition and preprocessing of multi-source heterogeneous data, and performance of data preprocessing and spatiotemporal alignment; S2: Construct a unified bird's-eye view representation based on preprocessed data, and build a dynamic scene map in real time based on this; S3: Use graph neural networks to reason about the dynamic scene graph to understand the global context of the scene and the interaction intentions of entities; S4: Based on a time-series dynamic scene graph, predict multiple possible future scene evolution paths, perform quantitative risk assessment on each path, and output a comprehensive risk score; S5: Generate a situation-adaptive emergency braking decision strategy based on the comprehensive risk score and scenario context information for each path; S6: Transform the decision-making strategy into specific vehicle control commands and execute them.
8. The intelligent emergency braking method based on dynamic driving scenario understanding and prediction according to claim 7, characterized in that, The quantitative risk assessment process integrates a driver state vector generated from vehicle internal state information data, which is used to dynamically correct risk indicators calculated based on the predicted trajectory.
9. An electronic device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the intelligent emergency braking method based on dynamic driving scenario understanding and prediction as described in any one of claims 7-8.
10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the intelligent emergency braking method based on dynamic driving scenario understanding and prediction as described in any one of claims 7-8.