Automatic emergency braking control method and system based on inverse braking compensation model

Abstract

The application provides an automatic emergency braking control method and system based on a reverse braking compensation model, and relates to the technical field of road traffic safety. A collision time between a self-vehicle and a target object is calculated, and when the collision time is less than a preset trigger threshold, an automatic emergency braking function is triggered and a theoretical braking deceleration is obtained. The theoretical braking deceleration is input into a reverse braking compensation model, a quasi-state driving environment is constructed based on the current vehicle state, target object information and theoretical braking deceleration, a process of braking to a stop at the theoretical braking deceleration is simulated, and a simulation driving distance is obtained. The theoretical braking deceleration is optimized by adjusting a braking coefficient until the simulation driving distance is less than or equal to a theoretical braking distance, and a corrected deceleration is obtained. The quasi-state driving environment is built by the reverse braking compensation model and combined with unsupervised learning, dynamic checking and accurate compensation of the braking value are realized, the sensing and execution deviations of a traditional AEB system are effectively made up, and the reliability and safety of automatic emergency braking of a passenger car are improved.

Description

Technical Field

[0001] This invention relates to the field of road traffic safety technology, and in particular to an automatic emergency braking control method and system based on a reverse braking compensation model. Background Technology

[0002] Automatic Emergency Braking (AEB) is a core technology in the field of active safety for commercial buses. Its core logic involves collecting information about the vehicle's status and surrounding objects through sensors such as cameras and radar. The ADAS controller then calculates the necessary braking parameters, and the ABS-ESC / EBS system executes the braking operation, forming a closed loop of "perception-decision-execution" to avoid or mitigate collision risks. As a key component of driver assistance systems, this technology is crucial for improving bus driving safety and handling complex road scenarios.

[0003] With the continuous growth of car ownership, road traffic safety issues are becoming increasingly serious, and the complexity of commercial bus operation scenarios and safety requirements are increasing simultaneously. Compared with passenger cars, the application of intelligent driving technology in commercial buses started later, and the actual adaptability and reliability of existing AEB systems are no longer sufficient to meet the industry's high standards for active safety, urgently requiring technological optimization to make up for performance shortcomings.

[0004] The current mainstream AEB solutions in the industry have the following shortcomings: at the execution level, there is a lack of dynamic compensation mechanism. The braking performance of the ABS_ESC / EBS system is affected by factors such as differences in vehicle tuning, long-term performance degradation, changes in vehicle load, and slippery road surfaces, making it difficult to stably achieve the theoretical braking target. In addition, traditional solutions only calculate braking values ​​based on idealized models without fully considering actual execution errors, resulting in a disconnect between theoretical values ​​and actual braking effects. Ultimately, this often leads to safety hazards such as failure to stop, which restricts the improvement of the active safety performance of buses. Summary of the Invention

[0005] To address the aforementioned issues, this invention proposes an automatic emergency braking control method and system based on a reverse braking compensation model. By constructing a simulated driving environment through the reverse braking compensation model and combining it with unsupervised learning, dynamic verification and accurate compensation of braking values ​​are achieved, effectively compensating for the perception and execution deviations of traditional AEB systems and improving the reliability and safety of automatic emergency braking for buses.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: In a first aspect, the present invention provides an automatic emergency braking control method based on a reverse braking compensation model, comprising: Based on vehicle information and perception information, the collision time between the vehicle and the target object is calculated; when the collision time is less than the preset trigger threshold, the automatic emergency braking function is triggered, and the theoretical braking deceleration is calculated. The theoretical braking deceleration is input into the reverse braking compensation model to correct the deceleration: a simulated driving environment is constructed based on the current vehicle state, target information, and theoretical braking deceleration; inverse dynamics simulation is performed in the environment to simulate the process of braking to a stop at the theoretical braking deceleration, and the simulated driving distance is obtained; if the simulated driving distance is greater than the theoretical stopping distance, the theoretical braking deceleration is continuously optimized by adjusting the braking coefficient until the simulated driving distance is less than or equal to the theoretical stopping distance, and the corrected deceleration is obtained as the final braking command, and braking is executed based on the final braking command.

[0007] Secondly, the present invention provides an automatic emergency braking control system based on a reverse braking compensation model, comprising: The braking trigger module is configured to calculate the collision time between the vehicle and the target object based on the vehicle information and perception information; when the collision time is less than a preset trigger threshold, the automatic emergency braking function is triggered and the theoretical braking deceleration is calculated. The reverse braking compensation module is configured to input the theoretical braking deceleration into the reverse braking compensation model and correct the deceleration: a simulated driving environment is constructed based on the current vehicle state, target information, and theoretical braking deceleration; reverse dynamics simulation is performed in the environment to simulate the process of braking to a stop at the theoretical braking deceleration, and the simulated driving distance is obtained; if the simulated driving distance is greater than the theoretical stopping distance, the theoretical braking deceleration is continuously optimized by adjusting the braking coefficient until the simulated driving distance is less than or equal to the theoretical stopping distance, and the corrected deceleration is obtained as the final braking command, and braking is executed based on the final braking command.

[0008] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the automatic emergency braking control method based on a reverse braking compensation model as described in the first aspect.

[0009] Fourthly, the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the automatic emergency braking control method based on the reverse braking compensation model described in the first aspect.

[0010] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention integrates perception and multi-dimensional vehicle data to construct a simulated driving environment. Solver Core accurately predicts braking effects and dynamically outputs compensation values, addressing the pain point of the disconnect between theoretical braking values ​​and actual execution in traditional systems, significantly reducing the risk of collisions due to brake failure. Simultaneously, the model's integrated unsupervised learning function can autonomously optimize parameters based on each vehicle's actual operating data, adapting to complex conditions such as load changes and road surface variations, forming a customized braking compensation strategy. By incorporating braking execution capability into the decision-making closed loop, it breaks through the limitations of idealized models, improving the AEB system's adaptability to complex scenarios and significantly enhancing the active safety performance and intelligence level of commercial buses.

[0011] Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute a limitation thereof.

[0012] Figure 1 The main flowchart of an automatic emergency braking control method based on a reverse braking compensation model provided in an embodiment of the present invention; Figure 2 This is a flowchart of an automatic emergency braking control method based on a reverse braking compensation model, provided as an embodiment of the present invention. Detailed Implementation

[0013] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0014] Example 1 like Figure 1 As shown, this embodiment discloses an automatic emergency braking control method based on a reverse braking compensation model, including the following steps: S1: Based on the vehicle information and perception information, calculate the collision time between the vehicle and the target object; when the collision time is less than the preset trigger threshold, trigger the automatic emergency braking function and calculate the theoretical braking deceleration; S2: Input the theoretical braking deceleration into the reverse braking compensation model to correct the deceleration: Construct a simulated driving environment based on the current vehicle state, target information, and theoretical braking deceleration; Perform reverse dynamics simulation in the environment to simulate the process of braking to a stop at the theoretical braking deceleration and obtain the simulated driving distance; If the simulated driving distance is greater than the theoretical stopping distance, continuously optimize the theoretical braking deceleration by adjusting the braking coefficient until the simulated driving distance is less than or equal to the theoretical stopping distance, obtain the corrected deceleration as the final braking command, and execute braking based on the final braking command.

[0015] Next, combined Figure 2 This embodiment provides a detailed description of an automatic emergency braking control method based on a reverse braking compensation model.

[0016] (a) Information Input This stage mainly involves the input of vehicle information and perception information.

[0017] Vehicle information includes vehicle speed, accelerator pedal status, brake pedal status, acceleration, steering wheel angle, yaw rate, turn signal status, gear information, and function switch status.

[0018] The perceived information is the target information fused from the camera and radar of the sensing device, including target type, target width, target longitudinal distance, target lateral distance, target longitudinal relative velocity, target lateral relative velocity, target confidence level, etc.

[0019] The above information is transmitted to the ADAS controller via CAN signals, where it is then used for further prediction processing.

[0020] (ii) Collision prediction TTC is the collision time, calculated as follows:

[0021] In the above formula, , For the vehicle's speed and acceleration, , The longitudinal velocity and acceleration of the target object. This represents the longitudinal distance between the vehicle and the target object.

[0022] The TTC value is valid when the target vehicle is within the vehicle's travel path and its longitudinal movement speed is less than the vehicle's speed.

[0023] In the above formula, The lateral relative velocity between the vehicle and the target object. The time of longitudinal collision with the target object at the current vehicle speed. The distance between the camera and the edge of the vehicle.

[0024] If the above formula is met, a collision risk may occur. Therefore, a TTC threshold is set. When the TTC threshold is met, the automatic emergency braking function is triggered.

[0025] (III) Braking value output When the automatic emergency braking function is triggered, deceleration calculation (initial) is performed:

[0026] In the above formula, The deceleration that the vehicle should perform, For the vehicle's speed, The longitudinal velocity of the target object, The longitudinal distance between the vehicle and the target object. This is the theoretical braking distance.

[0027] The TTC threshold references are shown in Table 1 below: Table 1 TTC Threshold Reference

[0028] Table 1, the TTC threshold reference table, defines the time thresholds for triggering the Automatic Emergency Braking (AEB) function under different vehicle speed conditions. The table lists the TTC (Time to Collision) thresholds set by the system for the "initial braking" and "emergency braking" phases under different combinations of the vehicle's speed and the target speed (all target speeds in the table are 0, i.e., stationary targets). For example, when the vehicle is traveling at 60 km / h and the target ahead is stationary, the system will trigger an initial braking warning or light braking when the TTC drops to 2.5 seconds. If the danger has not passed and the TTC continues to drop to 2.3 seconds, then emergency full braking will be triggered.

[0029] This table is designed based on a comprehensive consideration of factors such as actual vehicle dynamics, driver reaction time, and system response delay, ensuring that the system can provide timely warnings and interventions at different speeds while avoiding false triggers. In the actual system, the controller determines the applicable TTC threshold based on the real-time vehicle speed and target speed using a lookup table or interpolation method, and compares it with the real-time calculated TTC. Once the TTC is detected to be below the corresponding threshold, a collision risk is determined, and the corresponding braking strategy is activated according to the threshold level, thereby achieving a phased and adjustable automatic emergency braking response.

[0030] (iv) Collision check This stage mainly involves the reverse braking compensation model's operation, where the braking value is checked to obtain the braking compensation value, i.e., optimizing the deceleration in stage c:

[0031] In the above formula, The corrected deceleration that the vehicle should perform. For theoretical deceleration, This is the braking compensation value. The braking coefficient is the first time the system enters the reverse braking compensation model after the function is triggered. , .

[0032] The specific process is as follows: First, input perception information, decision-making information, and vehicle information into the model.

[0033] The perceived information includes: the relative distance to the target object in the lateral direction, the relative distance to the target object in the longitudinal direction, the relative velocity in the lateral direction, and the relative velocity in the longitudinal direction. Decision information includes: theoretical braking value, i.e., deceleration in phase c when the function is triggered; Vehicle information includes: vehicle speed and deceleration during ABS_ESC / EBS operation.

[0034] Secondly, after obtaining the deceleration when the function is triggered, the model performs curve fitting on the deceleration to obtain the deceleration change curve. At this point, the speed change curve of the vehicle can be obtained by integration. The distance traveled can be obtained by integrating the vehicle's speed. :

[0035]

[0036] In the above formula, For the deceleration fitting curve, The initial velocity of the vehicle is denoted as .

[0037] Based on Input Features data and A simulated driving environment is generated, which is a high-precision simulation environment that can simulate vehicle and object information on real roads. Placed in a simulated driving environment, judgment Is it less than ,otherwise Increase by 0.1 increments until... The braking compensation value is obtained. .

[0038] Subsequently, the vehicle's braking is performed by the ABS_ESC / EBS components. The ABS_ESC / EBS force applied to each vehicle will be affected by the vehicle's load, resulting in errors. Due to inertia, the braking force under full load is less than the braking force under no-load conditions. Therefore, when the vehicle is under load, the calibration value should be multiplied by a coefficient of not less than 1. , The value increases with increasing load (for example, no-load condition). Full load condition Unsupervised is a self-learning function for the reverse braking compensation model. It analyzes the response of ABS_ESC / EBS in various vehicle operating environments. For example, the braking force differs when the vehicle is unloaded and fully loaded, thus optimizing and improving the Solver Core. The system also stores the vehicle and calculation data. Unsupervised can store data separately for each vehicle, ensuring that each vehicle's data is used only by that vehicle. Each trigger of the automatic emergency braking function is a self-learning process, continuously optimizing the accuracy of the calibration values.

[0039] Specifically, in the braking execution and verification phase of the automatic emergency braking system, the final output of the vehicle's braking force is completed by the ABS / ESC or EBS system. However, its actual execution effect is significantly affected by the vehicle's load condition.

[0040] Due to differences in inertia, a fully loaded vehicle requires greater braking force to achieve the same deceleration as an unloaded vehicle. If the system uses fixed parameters, braking distance errors will occur under different load conditions, affecting safety and comfort.

[0041] To address this issue, this embodiment integrates an unsupervised self-learning mechanism based on a reverse braking compensation model. This mechanism enables the system to continuously perceive its own braking performance and dynamically optimize key control parameters, thereby achieving personalized adaptive braking tailored to the unique characteristics of each vehicle.

[0042] The key to the self-learning mechanism lies in building a continuous "data-driven optimization" closed loop. Each trigger of the automatic emergency braking function is a complete data collection and learning opportunity. The system synchronously records multi-dimensional data throughout the entire braking event cycle, including the vehicle status at the time of triggering, control commands, and execution feedback. The vehicle status includes load conditions and initial vehicle speed inferred from indirect signals; control commands include theoretical deceleration and verification braking values; and execution feedback includes measured deceleration, wheel speed changes, and the final distance to the target.

[0043] This data is encrypted and stored in association with the vehicle's unique identifier, forming a private performance database specific to that vehicle. For example, the system will record: under a fully loaded condition, AEB is triggered at an initial speed of 60km / h, with an initial deceleration command of 7m / s², but the actual maximum deceleration measured by the IMU is 6.5m / s², thus matching the specific operating conditions with the response.

[0044] Based on accumulated historical data, the self-learning model will conduct in-depth analysis to quantify the deviation between the actual braking effect and the expected effect, and to trace the root cause of the deviation.

[0045] Taking the braking coefficient k, which compensates for the impact of load, as an example, the model performs cluster analysis on historical data under different load conditions. Assuming that in the fully loaded data set, the model uses statistical regression to find that the actual deceleration is consistently lower than the expected value by about 8% in multiple events, this indicates that the current full-load compensation coefficient used by the system, such as k=1.3, is insufficient to completely compensate for the braking efficiency degradation caused by the load. Based on this, the model's internal algorithm calculates a new, better coefficient estimate, for example, adjusting k from 1.3 to 1.35, in order to better match the commanded braking force with the actual demand under similar operating conditions in the future. This process is not a simple table lookup and replacement, but a refined calibration based on continuous updates of the data probability distribution. The model's internal algorithm can employ Bayesian inference or least squares fitting; the specific algorithm can be selected by those skilled in the art.

[0046] For example, suppose the initial load compensation coefficients of the system are set as follows: no-load baseline coefficient k=1, and full-load coefficient k=1.3. In a fully loaded state, AEB is triggered, and the system uses k=1.3 to calculate the braking command. However, data playback shows that the actual measured deceleration is still slightly lower than the expected value, and the vehicle's final stopping position slightly exceeds the expected safe distance. This indicates that under the current fully loaded state, even with a coefficient of 1.3, the braking force is still slightly insufficient. The self-learning model identifies this pattern. It calculates the average deviation between the actual measured deceleration and the expected value in multiple events within the "fully loaded" data subset. Through algorithmic calculation, it concludes that to eliminate this systematic deviation, the "fully loaded coefficient" k needs to be fine-tuned from 1.3 to 1.35. Meanwhile, in the no-load event data, the model finds that the actual measured deceleration is very close to, and sometimes even slightly exceeds, the expected value, indicating high braking efficiency. The current coefficient k=1.0 is appropriate and does not require adjustment, or more refined curve optimization may be possible.

[0047] A deeper level of optimization lies in the iterative upgrade of the reverse braking compensation model itself (i.e., the Solver Core). The self-learning mechanism is not limited to adjusting a few explicit parameters; it continuously trains the complex mapping relationships within the model using massive amounts of input-output-result data pairs. The model can learn the nonlinear laws governing the braking system response as it changes with implicit factors such as vehicle speed, temperature, and even mechanical wear. For example, through long-term data analysis, the model may discover a specific delay pattern in the initial stage of brake hydraulic buildup under low-temperature conditions, thus compensating for this characteristic in advance under similar conditions in the future. This allows the model to gradually evolve from a general-purpose computing engine into a dedicated intelligent core that deeply understands the unique characteristics of the vehicle's braking system.

[0048] All parameter adjustments and model updates are performed within a strict safety framework. Any suggested adjustments are not immediately applied to real-time control; instead, they undergo thorough verification testing in an offline, high-precision mimicry simulation environment to ensure that their safety and effectiveness surpass previous versions under various boundary scenarios. After verification, the update package can be written to the vehicle controller via secure OTA or offline channels. At this point, the system completes a full self-learning iteration, further improving the accuracy and adaptability of its braking commands. Through this iterative process, the vehicle continuously accumulates experience throughout its lifecycle, making the automatic emergency braking system not only an executor of a pre-set program but also a safety partner that grows and evolves alongside the vehicle itself.

[0049] During the ABS / ESC / EBS execution phase, as the vehicle's response and execution link, after the ADAS controller issues the initial braking command, the ABS / ESC or EBS system immediately takes over and is responsible for executing the required deceleration. This is converted into actual hydraulic or pneumatic pressure. The system not only executes commands, but more importantly, it transmits the actual braking system values ​​(such as actual deceleration and wheel speed changes) back in real time to the core "reverse braking compensation model" for verification. This model immediately calculates and outputs a compensated and optimized verification braking value. The ABS / ESC / EBS system then performs a secondary execution based on this value, ultimately ensuring that the vehicle can achieve a stable and reliable complete stop under the action of this dynamically adjusted, closed-loop automatic emergency braking function.

[0050] This specific embodiment utilizes a reverse braking compensation model to achieve real-time monitoring and dynamic compensation of the vehicle's braking performance during the function triggering process. It incorporates actual performance capabilities into the safety decision-making loop, breaking the limitations of traditional AEB systems that rely solely on idealized models for calculation, thus significantly improving the system's actual efficiency and vehicle safety. Furthermore, this reverse braking compensation model integrates an unsupervised learning component, enabling the model to autonomously discover patterns and optimize internal calculation parameters using a large amount of real-world operational data, achieving continuous self-learning and self-evolution. This not only enhances adaptability to complex and changing driving scenarios but also significantly improves the intelligence level of the bus's active safety system.

[0051] Example 2 This embodiment provides an automatic emergency braking control system based on a reverse braking compensation model, including: The braking trigger module is configured to calculate the collision time between the vehicle and the target object based on the vehicle information and perception information; when the collision time is less than a preset trigger threshold, the automatic emergency braking function is triggered and the theoretical braking deceleration is calculated. The reverse braking compensation module is configured to input the theoretical braking deceleration into the reverse braking compensation model and correct the deceleration: a simulated driving environment is constructed based on the current vehicle state, target information, and theoretical braking deceleration; reverse dynamics simulation is performed in the environment to simulate the process of braking to a stop at the theoretical braking deceleration, and the simulated driving distance is obtained; if the simulated driving distance is greater than the theoretical stopping distance, the theoretical braking deceleration is continuously optimized by adjusting the braking coefficient until the simulated driving distance is less than or equal to the theoretical stopping distance, and the corrected deceleration is obtained as the final braking command, and braking is executed based on the final braking command.

[0052] Example 3 This embodiment provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of an automatic emergency braking control method based on a reverse braking compensation model as described in Embodiment 1 above.

[0053] Example 4 This embodiment provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps in the automatic emergency braking control method based on the reverse braking compensation model described in Embodiment 1 above.

[0054] The steps or modules involved in Embodiments 2 to 4 above correspond to those in Embodiment 1. For specific implementation details, please refer to the relevant description section of Embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood as including any medium capable of storing, encoding, or carrying an instruction set for execution by a processor and enabling the processor to perform any of the methods in this invention.

[0055] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. An automatic emergency braking control method based on a reverse braking compensation model, characterized in that, include: Based on vehicle information and perception information, the collision time between the vehicle and the target object is calculated; When the collision time is less than the preset trigger threshold, the automatic emergency braking function is triggered, and the theoretical braking deceleration is calculated. The theoretical braking deceleration is input into the reverse braking compensation model to correct the deceleration: a simulated driving environment is constructed based on the current vehicle state, target information and theoretical braking deceleration; Inverse dynamics simulation is performed in the environment to simulate the process of braking to a stop at the theoretical braking deceleration, and the simulated driving distance is obtained. Specifically, this includes: using the theoretical braking deceleration as input, performing curve fitting to obtain the deceleration change curve over time; performing a first integration on the change curve to obtain the vehicle speed change curve, and then performing a second integration to obtain the simulated distance traveled by the vehicle from the start of braking to a complete stop; the simulated driving environment is a simulation platform that integrates the vehicle dynamics model and the road environment model. If the simulated driving distance is greater than the theoretical braking distance, the theoretical braking deceleration is continuously optimized by adjusting the braking coefficient until the simulated driving distance is less than or equal to the theoretical braking distance. The corrected deceleration is then used as the final braking command, and braking is executed based on this final braking command. Specifically, this includes: Compare the simulated driving distance with the theoretical braking distance; If the simulated driving distance is less than or equal to the theoretical braking distance, the theoretical braking deceleration will be output as the final braking command. Otherwise, the braking compensation mechanism is activated, which gradually increases the braking deceleration by increasing the braking coefficient in preset steps, and then re-simulates and compares the distance based on the increased braking deceleration until the simulated driving distance meets the condition of being less than or equal to the theoretical stopping distance, and outputs the corresponding braking deceleration as the corrected final braking command.

2. The automatic emergency braking control method based on a reverse braking compensation model as described in claim 1, characterized in that, The calculation of the collision time between the vehicle and the target object based on vehicle information and perception information specifically includes: The vehicle information includes vehicle speed, acceleration, and steering status; the perception information includes target object type, relative distance, and relative speed. The collision time The calculation is as follows: ； in, , For the vehicle's speed and acceleration, , The longitudinal velocity and acceleration of the target object. This represents the longitudinal distance between the vehicle and the target object.

3. The automatic emergency braking control method based on a reverse braking compensation model as described in claim 1, characterized in that, The preset trigger thresholds include graded primary braking thresholds and emergency braking thresholds, forming a two-stage triggering mechanism. Specifically, based on the real-time vehicle speed and the target vehicle speed, the threshold corresponding to the current operating condition is determined by looking up a table, and the calculated collision time is compared with it: when the collision time is first lower than the primary braking threshold, primary braking is triggered; if the danger continues and the collision time drops below the emergency braking threshold, emergency braking is triggered.

4. The automatic emergency braking control method based on the reverse braking compensation model as described in claim 1, characterized in that, When the collision time is less than a preset trigger threshold, the automatic emergency braking function is triggered, and the theoretical braking deceleration is calculated, specifically including: When the target vehicle is in the vehicle's driving path and its longitudinal movement speed is less than the vehicle's speed, the automatic emergency braking function is triggered: in, The lateral relative velocity between the vehicle and the target object. The time of longitudinal collision with the target object at the current vehicle speed. The distance between the camera and the edge of the vehicle; When the automatic emergency braking function is triggered, the theoretical braking deceleration is calculated. : in, For the vehicle's speed, The longitudinal velocity of the target object, The longitudinal distance between the vehicle and the target object. This is the theoretical braking distance.

5. The automatic emergency braking control method based on a reverse braking compensation model as described in claim 1, characterized in that, The optimization of the braking coefficient is achieved through an unsupervised self-learning mechanism. Specifically, during actual braking, the actual execution data of the braking system and the actual response data of the vehicle are collected. By comparing the deviation between the theoretical braking command and the actual braking effect of the vehicle, the braking characteristics under different vehicle loads, driving speeds and environmental conditions are analyzed autonomously. Based on this analysis result, the braking coefficient and the internal parameters in the reverse braking compensation model are continuously iteratively updated.

6. An automatic emergency braking control system based on a reverse braking compensation model, characterized in that, An automatic emergency braking control method based on a reverse braking compensation model as described in any one of claims 1-5 includes: The braking trigger module is configured to calculate the collision time between the vehicle and the target object based on the vehicle information and perception information; when the collision time is less than a preset trigger threshold, the automatic emergency braking function is triggered and the theoretical braking deceleration is calculated. The reverse braking compensation module is configured to input the theoretical braking deceleration into the reverse braking compensation model and correct the deceleration: a simulated driving environment is constructed based on the current vehicle state, target information, and theoretical braking deceleration; reverse dynamics simulation is performed in the environment to simulate the process of braking to a stop at the theoretical braking deceleration, and the simulated driving distance is obtained; if the simulated driving distance is greater than the theoretical stopping distance, the theoretical braking deceleration is continuously optimized by adjusting the braking coefficient until the simulated driving distance is less than or equal to the theoretical stopping distance, and the corrected deceleration is obtained as the final braking command, and braking is executed based on the final braking command.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps in the automatic emergency braking control method based on the reverse braking compensation model as described in any one of claims 1-5.

8. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps in the automatic emergency braking control method based on the reverse braking compensation model as described in any one of claims 1-5.

Images