Accelerator mistaken stepping prevention method and device, medium and product

Abstract

The embodiment of the invention provides an accelerator mistaken stepping prevention method and device, a medium and a product. The method comprises the following steps: acquiring multi-dimensional sensing data of a vehicle, and determining an environmental risk index of the vehicle based on the multi-dimensional sensing data; if the environmental risk index is greater than a preset safety threshold value and the accelerator opening value of the vehicle is greater than a first preset accelerator opening threshold value, determining a mistaken accelerator stepping probability coefficient of the driver based on the multi-dimensional sensing data; if the accelerator mistaken stepping probability coefficient is smaller than or equal to a preset probability threshold value, real-time state information and real-time environment information of the driver are determined, and the vehicle is controlled to execute a first braking control strategy based on the real-time state information and the real-time environment information; and if the accelerator mistaken stepping probability coefficient is larger than the preset probability threshold value, the vehicle is controlled to execute a second braking control strategy, and the braking strength of the second braking control strategy is larger than that of the first braking control strategy. The method is used for achieving the effect of effectively balancing the braking strength and the driving comfort while ensuring safe braking.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of vehicle control, in particular to a throttle pedal anti-misoperation method, device, medium and product. BACKGROUND

[0002] Vehicle driving safety has always been the core issue of the development of the automobile industry. Among all kinds of traffic accident causes, driver human operation errors account for a dominant position. Among them, throttle pedal misoperation is a typical and serious human error, which refers to that when the driver intends to brake, the accelerator pedal is mistakenly stepped on as the brake pedal, causing the vehicle to unexpectedly accelerate instead of decelerate, thereby causing a collision.

[0003] In related technologies, the distance of the preceding vehicle or the driver's behavior is detected by a single sensor (such as a camera or radar), and the misoperation risk is judged by an algorithm to trigger electromagnetic braking, but the monocular vision system has a high misjudgment rate in complex environments, and the response delay may affect the intervention effect, affecting the user's experience. SUMMARY

[0004] The present application provides a throttle pedal anti-misoperation method, device, medium and product, which realizes coherent and efficient decision-making from environmental risk assessment to driver operation judgment and accurate braking control, greatly improving the safety and reliability of vehicle driving.

[0005] In a first aspect, the present application provides a throttle pedal anti-misoperation method, comprising: acquiring multi-dimensional sensing data of a vehicle, and determining an environmental risk index of the vehicle based on the multi-dimensional sensing data; if the environmental risk index is greater than a preset safety threshold, and the throttle opening value of the vehicle is greater than a first preset throttle opening threshold, then determining a misoperation probability coefficient of the driver based on the multi-dimensional sensing data; if the misoperation probability coefficient is less than or equal to a preset probability threshold, then determining real-time state information and real-time environmental information of the driver, and controlling the vehicle to execute a first braking control strategy based on the real-time state information and the real-time environmental information; if the misoperation probability coefficient is greater than a preset probability threshold, then controlling the vehicle to execute a second braking control strategy, the braking strength of the second braking control strategy being greater than the braking strength of the first braking control strategy.

[0006] In a possible implementation, the real-time state information includes foot movement parameters, and the controlling the vehicle to execute the first brake control strategy based on the real-time state information and the real-time environment information includes: determining foot movement parameters of the driver based on the multi-dimensional sensing data, and determining foot acceleration of the driver based on the foot movement parameters; determining a relative distance between the vehicle and a preceding vehicle based on the real-time environment information; and if the foot acceleration is greater than a preset acceleration threshold and the relative distance is less than or equal to a preset distance threshold, controlling the vehicle to execute a motor pre-warning strategy in the first brake control strategy.

[0007] In a possible implementation, the real-time state information further includes lower limb trajectory parameters, and the controlling the vehicle to execute the first brake control strategy based on the real-time state information further includes: if the foot acceleration is less than or equal to the preset acceleration threshold and / or the relative distance is greater than a preset distance threshold, determining real-time position information of the driver's foot based on the lower limb trajectory parameters; if the real-time position information deviates from an effective action area of a brake pedal of the vehicle, obtaining a real-time throttle opening value of the vehicle; and if the real-time throttle opening value is greater than a second preset throttle opening threshold, controlling the vehicle to execute a smooth brake strategy in the first brake control strategy, the second preset throttle opening threshold being less than the first preset throttle opening threshold.

[0008] In a possible implementation, the controlling the vehicle to execute the second brake control strategy includes: determining a staged torque attenuation instruction of a torque of the vehicle based on the second brake control strategy, so as to attenuate the driving torque of the vehicle to an idle torque threshold based on the staged torque attenuation instruction; and determining a hydraulic control instruction based on the second brake control strategy, so as to trigger a preset electromagnetic valve group to perform full-pressure braking on the vehicle based on the hydraulic control instruction.

[0009] In a possible implementation, the multi-dimensional sensing data includes radar data, and the determining the environment risk index of the vehicle based on the multi-dimensional sensing data includes: determining a relative relationship between the vehicle and a preceding vehicle based on the radar data, the relative relationship including a relative distance, a relative speed, and a relative heading angle; and performing weighted calculation on the relative distance, the relative speed, and the relative heading angle based on a preset weighting strategy, so as to take a calculation result as the environment risk index of the vehicle.

[0010] In a possible implementation, the determining the accelerator misstep probability coefficient of the driver based on the multi-dimensional sensing data comprises: determining an environmental risk index of the vehicle and foot movement parameters and lower limb trajectory parameters of the driver based on the multi-dimensional sensing data; determining a brake intention probability parameter of the driver based on the environmental risk index, the foot movement parameters and the lower limb trajectory parameters; obtaining historical accelerator misstep data of the driver, the historical accelerator misstep data comprising a historical accelerator misstep frequency; and determining the accelerator misstep probability parameter of the driver based on the historical accelerator misstep frequency and the brake intention probability parameter.

[0011] In a possible implementation, the method further comprises: obtaining real-time multi-dimensional sensing data of the vehicle after the vehicle executes the first brake control strategy or the second brake control strategy; determining a reinforcement learning reward value based on the real-time multi-dimensional sensing data, and optimizing the first brake control strategy and the second brake control strategy based on the reward value.

[0012] In a second aspect, an embodiment of the present application provides an accelerator misstep prevention device, the device comprising: an obtaining module configured to obtain multi-dimensional sensing data of a vehicle and determine an environmental risk index of the vehicle based on the multi-dimensional sensing data; a judging module configured to, if the environmental risk index is greater than a preset safety threshold value and an accelerator opening value of the vehicle is greater than a first preset accelerator opening threshold value, determine an accelerator misstep probability coefficient of a driver based on the multi-dimensional sensing data; a first control module configured to, if the accelerator misstep probability coefficient is less than or equal to a preset probability threshold value, determine real-time state information and real-time environmental information of the driver and control the vehicle to execute a first brake control strategy based on the real-time state information and the real-time environmental information; and a second control module configured to, if the accelerator misstep probability coefficient is greater than the preset probability threshold value, control the vehicle to execute a second brake control strategy, the brake strength of the second brake control strategy being greater than the brake strength of the first brake control strategy.

[0013] In a third aspect, an embodiment of the present application provides an electronic device, comprising: a memory, a processor;

[0014] The memory stores computer execution instructions;

[0015] The processor executes the computer execution instructions stored in the memory, so that the processor executes the first aspect and / or various possible implementations of the first aspect as described above.

[0016] In a fourth aspect, an embodiment of the present application provides a computer readable storage medium, the computer readable storage medium storing computer execution instructions, the computer execution instructions being executed by a processor to implement the first aspect and / or various possible implementations of the first aspect as described above.

[0017] Fifthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the first aspect and / or various possible implementations of the first aspect.

[0018] The accelerator pedal misoperation prevention method, device, medium, and product provided in this application, by acquiring multi-dimensional sensor data of the vehicle, can comprehensively and accurately capture information on the vehicle's own state, surrounding environment, and driver operation. Based on the environmental risk index determined by the multi-dimensional sensor data, it can objectively reflect the degree of danger in the vehicle's environment. When the environmental risk index exceeds a preset threshold and the accelerator pedal opening value is large, the probability coefficient of the driver's accidental accelerator pedal operation is determined based on the multi-dimensional sensor data. This allows for a more accurate judgment of whether the driver's operational intention is reasonable. If the probability coefficient of accidental accelerator pedal operation reaches a preset probability threshold, the driver's real-time state information is first determined, and a first braking control strategy with lower braking intensity is executed accordingly, ensuring both safety and driving stability. If the probability coefficient of accidental accelerator pedal operation is higher, a second braking control strategy with greater braking intensity is executed, ensuring the vehicle can quickly respond to sudden dangers. This achieves a coherent and efficient decision-making process from environmental risk assessment to driver operation judgment and precise braking control, greatly improving the safety and reliability of vehicle operation. Attached Figure Description

[0019] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0020] Figure 1 A schematic diagram illustrating the scenario for preventing accidental accelerator pedal presses provided in this application;

[0021] Figure 2 Flowchart of the accelerator pedal anti-accidental pressing method provided in this application Figure 1 ;

[0022] Figure 3 Flowchart of the accelerator pedal anti-accidental pressing method provided in this application Figure 2 ;

[0023] Figure 4 A schematic diagram of the accelerator pedal anti-accidental pressing device provided in this application;

[0024] Figure 5 A schematic diagram of the structure of the electronic device provided in this application.

[0025] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0026] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0027] First, let me explain the terms used in this application:

[0028] Solenoid valve hydraulic braking is a braking method that combines solenoid valve control with hydraulic transmission technology. Its core is the precise control of the flow direction, pressure, and volume of hydraulic oil through a solenoid valve, thereby achieving rapid response and precise distribution of braking force. Specifically, when the solenoid valve coil is energized, the electromagnetic force drives the valve core to move, changing the hydraulic oil flow path and allowing high-pressure oil to enter the brake wheel cylinder, pushing the piston to press against the brake disc or brake drum, generating braking force. When de-energized, the valve core returns to its original position under the action of a spring, the hydraulic oil flows back, and the brake is released. This technology, with its millisecond-level response speed of the solenoid valve and the high power density of the hydraulic system, is widely used in active safety systems such as ABS (Anti-lock Braking System) and ESP (Electronic Stability Program) in automobiles. It can achieve dynamic braking pressure regulation such as "pressure increase," "pressure maintenance," and "pressure reduction," significantly improving braking stability and safety.

[0029] Figure 1 This is a schematic diagram illustrating the scenario for preventing accidental accelerator pedal presses provided in this application, such as... Figure 1 As shown, the specific application scenario of this application includes a vehicle-side 110 and a server-side 120. The server-side 120 acquires multi-dimensional sensor data of the vehicle, including radar data, pressure sensor data, and motion trajectory sensor data, and then determines the vehicle's environmental risk index based on this data. If the environmental risk index is greater than a preset safety threshold, and the vehicle's throttle opening value is greater than a preset throttle opening threshold, the probability coefficient of the driver accidentally pressing the accelerator can be determined based on the multi-dimensional sensor data. If the server-side 120 determines that the probability coefficient of the driver accidentally pressing the accelerator is less than or equal to a preset probability threshold, it determines the driver's real-time status information and real-time environmental information, and controls the vehicle to execute a first braking control strategy based on this information. Conversely, if the server-side 120 determines that the probability coefficient of accidentally pressing the accelerator is greater than the preset probability threshold, it controls the vehicle to execute a second braking control strategy, the braking intensity of which is greater than that of the first braking control strategy. This achieves precise control of the vehicle's accelerator to prevent accidental pressing.

[0030] Figure 1 The server 120 shown can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms. No restrictions are placed on this. The vehicle 110 can communicate with the server 120 via wireless networks such as 3G (third-generation mobile information technology), 4G (fourth-generation mobile information technology), and 5G (fifth-generation mobile information technology). No restrictions are placed on this as well.

[0031] In related technologies, a single sensor (such as a camera or radar) is used to detect the distance to the vehicle in front or the driver's behavior. An algorithm is used to determine the risk of accidental braking and trigger electromagnetic braking. However, monocular vision systems have a high misjudgment rate in complex environments, and the response delay may affect the intervention effect and the user experience.

[0032] The accelerator pedal misoperation prevention method provided in this application acquires multi-dimensional sensor data of the vehicle and determines an environmental risk index, which can accurately assess the vehicle's condition in complex environments. When the environmental risk index exceeds a preset threshold and the accelerator pedal opening value is large, the probability coefficient of the driver accidentally pressing the accelerator pedal is further determined based on the multi-dimensional sensor data. If the coefficient reaches a preset probability threshold, the driver's real-time status information is first determined and a first braking control strategy is executed accordingly. If the probability coefficient of accidentally pressing the accelerator pedal is higher, a second braking control strategy with greater braking intensity is executed. This progressive and flexible judgment and control mechanism can effectively avoid dangers caused by the driver accidentally pressing the accelerator pedal and improve the vehicle's driving safety in complex and dangerous environments.

[0033] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0034] Figure 2 Flowchart of the accelerator pedal anti-accidental pressing method provided in this application Figure 1 ,like Figure 2 As shown, the data migration method includes at least steps S201 to S204, which are described in detail below:

[0035] Step S201: Obtain multi-dimensional sensor data of the vehicle and determine the environmental risk index of the vehicle based on the multi-dimensional sensor data.

[0036] For example, a data acquisition network encompassing multiple sensor types is constructed. This network should integrate devices including, but not limited to, forward-facing millimeter-wave radar, ultrasonic sensors, cameras, inertial measurement units, and wheel speed sensors. The forward-facing millimeter-wave radar can detect the distance, speed, and angle of obstacles within a 120-meter range in front of the vehicle in real time. The ultrasonic sensors focus on near-field parking environment perception at 0.2-5 meters. The cameras capture visual information such as road signs, pedestrians, and vehicles through image recognition technology. The inertial measurement unit records the vehicle's acceleration, angular velocity, and other motion state data. The wheel speed sensors accurately measure the rotational speed of each wheel. These sensors transmit data to the central processing unit in real time via a high-speed bus. The central processing unit then preprocesses the acquired raw data, including filtering and noise reduction, time synchronization, and coordinate unification, ensuring consistency of data from different sensors in the spatiotemporal dimensions. Based on this, a multi-source data fusion algorithm is used to deeply integrate the data from various sensors. For example, methods such as Kalman filtering or Bayesian networks are used to correlate and verify the ranging data from the millimeter-wave radar with the visual recognition results from the cameras, improving the accuracy and robustness of obstacle detection.

[0037] Then, the fused data is analyzed using a pre-trained environmental risk assessment model. This model is typically built based on machine learning or deep learning techniques and trained on a large amount of real-world driving scenario data. It can dynamically calculate the environmental risk index based on parameters such as the current vehicle position, speed, acceleration, and the distribution and motion of surrounding obstacles. This index comprehensively reflects the probability of a collision or other dangerous event occurring in the current environment. Its calculation process needs to consider multiple factors, such as whether the safe distance to the vehicle in front is less than a threshold, whether a vehicle on the side suddenly changes lanes, and whether a pedestrian suddenly enters the driving path.

[0038] In step S202, if the environmental risk index is greater than the preset safety threshold and the throttle opening value of the vehicle is greater than the first preset throttle opening threshold, then the probability coefficient of the driver accidentally pressing the accelerator is determined based on the multi-dimensional sensor data.

[0039] For example, when the environmental risk index exceeds a preset safety threshold and the vehicle throttle opening value is greater than a first preset threshold, the calculation process for the probability coefficient of accidental throttle pedal application is initiated. This process relies on the deep integration of vehicle state data, environmental perception information, and driving behavior patterns. Specifically, it can integrate the vehicle's own dynamic state data (such as vehicle speed, acceleration, steering angle, brake pedal status, wheel speed differences, etc.) with real-time scene information provided by the environmental perception unit (such as the distance / relative speed of obstacles ahead detected by forward millimeter-wave radar, the position of nearby obstacles monitored by ultrasonic sensors, the dynamics of road signs / pedestrians / vehicles identified by cameras, and the changes in vehicle attitude recorded by the inertial measurement unit, etc.), and ensure the consistency of various data in the spatiotemporal dimensions through a multi-source data fusion algorithm. Subsequently, a machine learning-based pattern recognition model is used to perform multi-dimensional analysis on the integrated data, for example, comparing the characteristic differences between normal acceleration scenarios and potential accidental throttle pedal application scenarios. For example, during normal acceleration, changes in throttle opening are usually accompanied by a smooth increase in vehicle speed, coordination of steering operations, and a reasonable match of environmental risks. In scenarios of accidental accelerator pedal presses, abnormal patterns may emerge, such as a sudden increase in throttle opening with an abnormal increase in vehicle speed, a contradiction between steering input and throttle opening (e.g., sharp steering while simultaneously pressing the accelerator hard), and a negative correlation between environmental risk (e.g., approaching obstacle) and throttle opening (i.e., the higher the risk, the greater the throttle opening). By extracting these features and inputting them into a pre-trained classification model (e.g., random forest, neural network), the probability coefficient of a driver accidentally pressing the accelerator can be quantitatively calculated. This coefficient comprehensively reflects the degree of matching between the current driving behavior and typical accidental accelerator pedal press patterns.

[0040] In step S203, if the probability coefficient of accidentally pressing the accelerator is less than or equal to the preset probability threshold, the driver's real-time status information and real-time environmental information are determined, and the vehicle is controlled to execute the first braking control strategy based on the real-time status information and real-time environmental information.

[0041] For example, when the probability coefficient of accidentally pressing the accelerator is less than or equal to a preset probability threshold, the driver's real-time state information can be determined by combining vehicle state data and environmental perception information. For instance, the driver's facial expressions, gaze direction, and head posture can be monitored by an onboard camera, combined with steering wheel grip force sensor data to determine whether the driver is distracted, fatigued, or tense. At the same time, forward-facing millimeter-wave radar and ultrasonic sensors continuously scan the surrounding environment to obtain the current implementation environment information, which includes real-time scene information such as the distance to obstacles ahead, relative speed, and the dynamics of vehicles to the side. Then, the driver's real-time state information and real-time environmental information are cross-validated with the vehicle's own dynamic data such as vehicle speed, acceleration, steering angle, brake pedal status, and wheel speed differences to ensure the spatiotemporal consistency of the state information. Subsequently, a first braking control strategy is formulated based on the driver's real-time state information. This strategy emphasizes gentle intervention and adaptive adjustment. For example, when the driver's gaze deviates from the road or the steering wheel grip is abnormal, progressive braking (such as gradually increasing the braking force to below a safe threshold) combined with audible and visual warnings will be prioritized to guide the driver to refocus on the road conditions. If environmental risks continue to increase (such as an obstacle rapidly approaching from the front), the braking strategy will be dynamically adjusted. For example, it will optimize the distribution of braking force to all four wheels in conjunction with vehicle dynamic stability control (such as ESP) to avoid loss of vehicle control due to sudden braking. At the same time, multimodal reminders such as ambient lighting and seat vibration will enhance the driver's perception and feedback.

[0042] In step S204, if the probability coefficient of accidentally pressing the accelerator is greater than the preset probability threshold, the vehicle is controlled to execute a second braking control strategy, and the braking intensity of the second braking control strategy is greater than the braking intensity of the first braking control strategy.

[0043] For example, when the probability coefficient of accidentally pressing the accelerator exceeds a preset probability threshold, the vehicle immediately initiates a high-intensity second braking control strategy. This strategy relies on deep collaborative analysis of vehicle state data and environmental perception information. It can quickly integrate the vehicle's own dynamic state (such as vehicle speed, acceleration, steering angle, brake pedal state, wheel speed differences, and changes in vehicle posture) with real-time scene data from the environmental perception unit (such as the distance / relative speed of obstacles ahead detected by forward millimeter-wave radar, the position of nearby obstacles monitored by ultrasonic sensors, the dynamics of road signs / pedestrians / vehicles identified by cameras, and the vehicle's trajectory recorded by the inertial measurement unit). It ensures spatiotemporal consistency through multi-source data fusion algorithms and assesses the urgency of the current risk based on a pre-trained emergency braking model. Subsequently, a second braking measure with significantly higher braking intensity than the first strategy is executed. For example, the electronic stability program (ESP) dynamically adjusts the four-wheel braking force distribution, prioritizing the application of greater braking force to the high-risk side wheels to shorten the braking distance, while combining anti-lock braking (ABS) to avoid the risk of loss of control due to wheel lock-up. During braking, the system continuously monitors the driver's real-time status (such as facial expression, gaze direction, head posture, and steering wheel operation). If any abnormal reaction is detected in the driver (such as fright, panic, or operational error), the braking strategy will be dynamically adjusted. For example, the braking force will be gradually increased instead of sudden braking to avoid causing secondary accidents.

[0044] Furthermore, in some feasible embodiments, multimodal warning mechanisms can be triggered, such as high-frequency audible and visual alarms, seat vibration, and seatbelt pretensioning, enhancing the driver's perception of emergencies. Simultaneously, emergency braking signals are sent to surrounding vehicles and infrastructure via vehicle-to-everything (V2X) technology, improving overall traffic safety. The execution of the entire second braking strategy emphasizes "rapid response, precise intervention, and dynamic adjustment." Through real-time interaction of vehicle status and environmental information, when the risk of accidental accelerator pedal press is confirmed as a high-probability event, stronger braking intensity is used to achieve proactive safety control. At the same time, the smoothness of the braking process and the driver's acceptability are considered, achieving precise responses to safety risks in complex driving scenarios through multi-dimensional data fusion and intelligent algorithms.

[0045] In the embodiments provided in this application, by integrating multi-dimensional sensor data to dynamically assess environmental risks and the probability of accidental accelerator pedal press, graded braking intervention is achieved for the driver's accidental accelerator pedal press, effectively balancing braking intensity and driving comfort while ensuring safe braking.

[0046] Based on the above embodiments, in one exemplary embodiment provided in this application, the real-time status information includes foot movement parameters, and the specific implementation process of controlling the vehicle to execute the first braking control strategy based on the real-time status information and real-time environmental information may further include steps S301 to S303, which are described in detail below:

[0047] Step S301: Determine the driver's foot motion parameters based on multidimensional sensor data, and determine the driver's foot acceleration based on the foot motion parameters.

[0048] Step S302: Determine the relative distance to the vehicle in front based on real-time environmental information.

[0049] Step S303: If the foot acceleration is greater than a preset acceleration threshold and the relative distance is less than or equal to a preset distance threshold, then control the vehicle to execute the motor warning strategy in the first braking control strategy.

[0050] For example, by fusing foot sensor data (such as foot pressure distribution collected by a plantar pressure sensor array and foot spatial displacement data acquired by a foot motion trajectory tracking module) with multidimensional sensor data (such as the relative distance to the vehicle ahead measured by millimeter-wave radar, the outline of the vehicle ahead captured by camera visual recognition, and road marking information), the foot sensor data can be analyzed in real time to determine the driver's foot motion parameters. Specifically, foot acceleration characteristics are calculated using the rate of change of plantar pressure gradient. For instance, when the pressure sensor detects a sudden increase in pressure in the heel area and a rapid rise in pressure in the toe area, the foot acceleration value is calculated by combining it with foot motion trajectory tracking data. Simultaneously, the relative distance to the vehicle ahead is determined using multidimensional sensor data collected from real-time environmental information. Specifically, the Doppler effect of millimeter-wave radar and the camera visual ranging algorithm can be used to collaboratively calculate the relative distance between the vehicle and the vehicle ahead. The radar data provides high-precision distance and speed information, while the camera compensates for the limitations of radar in measuring heading angle through image recognition technology, forming a precise measurement of relative distance through multi-source data fusion. Subsequently, the calculated foot acceleration is compared with a preset acceleration threshold, and the relative distance is simultaneously verified to be less than or equal to a preset distance threshold. When the foot acceleration exceeds the threshold (indicating that the driver has a tendency to rapidly depress the accelerator pedal) and the relative distance is less than or equal to the safety threshold, it is determined that the motor warning strategy in the first braking control strategy needs to be triggered. This strategy can generate a reverse torque pulse through the motor controller to apply slight resistance feedback the moment the driver's foot contacts the accelerator pedal, which serves as a warning while avoiding excessive intervention in normal driving operations.

[0051] Optionally, in some feasible embodiments, there is a close correlation between the dynamic monitoring of foot acceleration and the change in throttle opening value. Specifically, when the driver presses the accelerator pedal, foot sensors (such as high-sensitivity capacitive arrays and miniature millimeter-wave radars) capture in real time the gradient changes in plantar pressure distribution and the acceleration characteristics of foot spatial displacement. Simultaneously, the vehicle electronic control unit (ECU) records the numerical changes of the throttle opening sensor. By learning from historical driving data through machine learning algorithms, a mapping model between foot acceleration and the rate of change of throttle opening can be established. For example, in normal acceleration scenarios, foot acceleration is usually positively correlated with the linear increase of throttle opening (i.e., the faster the opening increases, the greater the foot acceleration), and the peak acceleration gradually decays after the opening stabilizes. However, in accidental pressing scenarios, foot acceleration may exhibit an abnormal surge (such as accelerating from 0 to the peak value in a short period of time) and does not match the actual demand for throttle opening (such as high acceleration pressing when the vehicle is already traveling at high speed). By comparing the dynamic relationship between the current foot acceleration and the rate of change of throttle opening in real time, if the acceleration value is detected to be significantly beyond the threshold range of the normal acceleration mode (such as exceeding the theoretical acceleration limit calculated based on the rate of change of throttle opening), it is determined to be a potential risk of accidental pedaling. In this case, the vehicle status (such as vehicle speed and RPM) and environmental information (such as the distance to obstacles in front) will be combined for further verification, and the corresponding braking control strategy will be triggered.

[0052] In the embodiments provided in this application, by real-time monitoring of the driver's foot acceleration and setting a threshold to trigger a motor warning strategy, early identification and low-intensity active intervention of accidental accelerator pedal press are achieved, effectively reducing the risk of collision while avoiding excessive braking that interferes with normal driving.

[0053] Based on the above embodiments, in one exemplary embodiment provided in this application, the real-time status information further includes lower limb trajectory parameters, and the specific implementation process of controlling the vehicle to execute the first braking control strategy based on the real-time status information may further include steps S401 to S403, which are described in detail below:

[0054] Step S401: If the foot acceleration is less than or equal to a preset acceleration threshold and / or the relative distance is greater than a preset distance threshold, then determine the real-time position information of the driver's feet based on the lower limb trajectory parameters.

[0055] Step S402: If the real-time position information deviates from the effective area of ​​the vehicle's brake pedal, then obtain the real-time throttle opening value of the vehicle.

[0056] Step S403: If the real-time throttle opening value is greater than the second preset throttle opening threshold, then control the vehicle to execute the smooth braking strategy in the first braking control strategy, where the second preset throttle opening threshold is less than the first preset throttle opening threshold.

[0057] For example, when the foot acceleration is less than or equal to a preset acceleration threshold, the lower limb motion capture unit, based on an infrared time-of-flight sensor, constructs a 100Hz high-sampling-rate 3D motion model of the lower limbs. This model tracks changes in lower limb joint angles, foot spatial displacement trajectory, and motion velocity vectors in real time to accurately analyze the driver's real-time foot position information. Then, it determines whether this real-time position deviates from the effective operating area of ​​the vehicle's brake pedal. If the foot position exceeds the reasonable operating range of the brake pedal (e.g., not contacting the pedal or excessive position deviation), it indicates that the driver may have made an operational error or is not prepared to brake. In this case, it is necessary to further obtain the vehicle's real-time throttle opening value and compare it with a second preset throttle opening threshold (which is less than the first preset throttle opening threshold). Specifically, if the real-time throttle opening exceeds the second threshold, it indicates that although the driver did not press the accelerator suddenly, the throttle opening is still at a high level, potentially indicating a risk of misoperation. Under these operating conditions, the smooth braking strategy in the first braking control strategy will be triggered. This strategy can dynamically adjust the distribution of braking force to the four wheels through the electronic stability program (ESP), and combine it with anti-lock braking (ABS) to prevent wheel lock-up. At the same time, it adopts a progressive braking force increase method (such as gradually increasing the braking force from the initial braking force to the safety threshold), and is accompanied by multimodal reminders (such as seat vibration, seat belt pretensioning and gentle audible and visual reminders) to ensure braking effect while minimizing the impact on vehicle stability and passenger comfort.

[0058] Optionally, when the vehicle's sensors detect that the relative distance between the vehicle and the vehicle in front exceeds a preset distance threshold, the real-time position information of the driver's feet can be determined based on a pre-built lower limb motion model and the vehicle's internal spatial layout parameters. This real-time position information is then carefully compared with the pre-defined effective operating area of ​​the brake pedal to determine if the driver's feet are currently within the reasonable range for operating the brake pedal. If the comparison shows that the driver's real-time foot position deviates from the reasonable operating range of the brake pedal, it means that the driver may not be able to operate the brake pedal in a timely and effective manner. In this case, the vehicle's real-time throttle opening value can be used, and this throttle opening value can be rigorously compared with a second preset throttle opening threshold. If the comparison result shows that the real-time throttle opening value is greater than the second preset throttle opening threshold, this fully demonstrates the potential risk of the driver accidentally pressing the accelerator. Therefore, to avoid danger caused by accidental throttle pressing, the smooth braking strategy in the first braking control strategy can be implemented to reduce the vehicle speed smoothly and safely, ensuring driving safety.

[0059] Optionally, in some feasible embodiments, the driver's foot acceleration and the relative distance between the vehicle and the vehicle in front are detected in real time through a data acquisition network of multiple types of sensors in the vehicle. When the driver's foot acceleration is detected to be less than or equal to a preset acceleration threshold and the relative distance between the vehicle and the vehicle in front is greater than a preset distance threshold, the real-time position information of the driver's feet is calculated. Subsequently, this real-time position information is compared with a reasonable operating range predefined for the vehicle's brake pedal. If the comparison result shows that the real-time position of the driver's feet exceeds this reasonable operating range, and the real-time throttle opening value of the vehicle is detected to be greater than a preset second preset throttle opening threshold, this series of situations indicates that the vehicle is very likely to be in an acceleration state, and the driver's foot position is abnormal, posing a great potential danger. To ensure driving safety, the vehicle can be controlled to execute the smooth braking strategy in the first braking control strategy.

[0060] In the embodiments provided in this application, by combining foot acceleration, lower limb trajectory and throttle opening multi-level judgment, accurate recognition of the driver's intention and smooth braking intervention are achieved in non-accidental pedaling scenarios, effectively reducing potential safety risks under high throttle opening while avoiding accidental intervention.

[0061] Based on the above embodiments, in one exemplary embodiment provided in this application, the specific implementation process of controlling the vehicle to execute the second braking control strategy may further include the following steps, which are described in detail below:

[0062] Based on the second braking control strategy, a graded torque decay command for the vehicle's torque is determined, so as to reduce the vehicle's drive torque to the idle torque threshold based on the graded torque decay command; and,

[0063] The hydraulic control command is determined based on the second braking control strategy, and the preset solenoid valve group is triggered to apply full-pressure braking to the vehicle based on the hydraulic control command.

[0064] For example, based on the vehicle's current dynamic state and environmental risk index, a multi-source data fusion algorithm dynamically generates a graded torque decay command. This command, based on multiple parameters such as the probability coefficient of accidental accelerator pedal press, real-time vehicle speed, engine speed, and drive torque, gradually decays the drive torque from its current value to the idle torque threshold according to a preset gradient. For instance, the torque is first reduced to the half-threshold at a relatively fast rate to quickly reduce the vehicle's acceleration capability, and then smoothly transitions to the idle level at a slower rate, thereby avoiding the risk of vehicle loss of control due to a sudden drop in torque. Simultaneously, a hydraulic control command is generated. This command, combined with real-time feedback from the brake pedal travel sensor, master cylinder pressure sensor, and wheel speed sensor, achieves full-pressure braking through the dynamic control of a preset solenoid valve group.

[0065] Specifically, the solenoid valve assembly rapidly adjusts the brake fluid pressure distribution according to commands, prioritizing the application of maximum available braking force to the high-risk wheels to shorten the braking distance. Simultaneously, the intervention of the anti-lock braking system (ABS) prevents wheel lock-up, ensuring directional stability during braking. Furthermore, during execution, the system continuously monitors vehicle status (such as lateral acceleration and changes in vehicle attitude) and environmental information (such as distance to obstacles ahead and the dynamics of vehicles to the side). If abnormal conditions are detected (such as vehicle yaw or slippery road surfaces), the system dynamically adjusts the torque attenuation gradient and hydraulic pressure distribution strategy. For example, it increases the braking pressure on the high-risk side or slows down the torque attenuation rate to balance braking intensity and vehicle controllability. Ultimately, through the synergistic effect of graded torque attenuation and full-pressure braking, while ensuring safe braking, it effectively reduces the risk of acceleration due to accidental accelerator pedal application, and achieves the shortest braking distance and optimal directional stability through precise hydraulic control, forming a high-intensity active safety intervention closed loop from torque management to hydraulic braking.

[0066] In the embodiments provided in this application, the second braking control strategy synchronously triggers graded torque decay and full-pressure braking, thereby achieving strong coordinated intervention of power output and braking system in high-risk scenarios, significantly shortening braking distance and improving active safety protection capabilities in emergency situations.

[0067] Based on the above embodiments, in one exemplary embodiment provided in this application, the multi-dimensional sensing data further includes radar data, and the specific implementation process of determining the environmental risk index of a vehicle based on multi-dimensional sensing data may further include steps S501 and S502, which are described in detail below:

[0068] Step S501: Determine the relative relationship between the vehicle and the vehicle in front based on radar data. The relative relationship includes relative distance, relative speed and relative heading angle.

[0069] Step S502: The relative distance, relative speed, and relative heading angle are weighted and calculated based on a preset weighting strategy, and the calculation result is used as the vehicle's environmental risk index.

[0070] For example, in vehicle active safety, the process of determining the relative relationship between a vehicle and the vehicle in front and calculating the environmental risk index based on radar data relies on the synergistic effect of multi-dimensional perception data fusion and dynamic weighting algorithms. Specifically, forward-facing millimeter-wave radar collects reflected signals from the vehicle in front in real time. After analysis by the signal processing unit, three core parameters are extracted: relative distance (the straight-line distance between the vehicle and the object in front), relative speed (the rate at which they approach or move away from each other), and relative heading angle (the angle between the vehicle's direction of motion and the object's direction of motion). The relative distance is accurately calculated using a time-of-flight (ToF) algorithm, the relative speed is directly measured by the Doppler frequency shift effect, and the relative heading angle is derived from the azimuth and pitch angle data of the radar beam combined with the vehicle's own attitude information. Subsequently, a preset weighting strategy can be used to dynamically weight and calculate these three parameters. This preset weighting strategy can be built based on historical collision data and an expert knowledge base. It assigns the highest weight to relative distance (e.g., 0.6) because it directly reflects the time to collision (TTC), the second highest weight to relative speed (e.g., 0.3) because it affects collision energy, and the lowest weight to relative heading angle (e.g., 0.1). However, when the heading angle deviates from the threshold (e.g., 30 degrees), a weight multiplication mechanism is triggered.

[0071] Optionally, during the calculation, each parameter is first normalized to the 0-1 range (e.g., relative distance divided by the safe distance threshold, relative speed divided by the maximum allowable approach speed). To adapt to different driving scenarios (e.g., highway / city / congestion), the weight allocation is dynamically adjusted. For example, in a highway scenario, due to high vehicle speed and long braking distance, the weight of relative speed is increased to 0.4, while the weight of relative heading angle remains at 0.1; in a congested scenario, the weight of relative distance may be increased to 0.7 to enhance the ability to prevent rear-end collisions at close range. Furthermore, the entire weighted calculation process is driven by real-time data streams, updating the risk index every preset time interval to ensure rapid response to environmental changes. Ultimately, this environmental risk index serves as the core input of the second braking control strategy, guiding the selection of corresponding intervention measures (e.g., torque attenuation intensity, hydraulic braking pressure).

[0072] Optionally, in some feasible embodiments, the environmental risk index expression for a vehicle can be as follows:

[0073]

[0074] In the formula, This is the relative position distance. For relative velocity, Relative heading angle , , All are weighting coefficients.

[0075] In the embodiments provided in this application, by fusing relative distance, speed and heading angle data sensed by radar and dynamically weighting the calculation, the environmental risk index is accurately quantified, providing a highly timely basis for collision threat assessment for vehicle active safety systems.

[0076] Based on the above embodiments, in one exemplary embodiment provided in this application, the specific implementation process of determining the probability coefficient of a driver's accidental acceleration based on multi-dimensional sensing data may further include steps S601 to S604, which are described in detail below:

[0077] Step S601: Determine the vehicle's environmental risk index, as well as the driver's foot movement parameters and lower limb trajectory parameters, based on multi-dimensional sensor data.

[0078] Step S602: Determine the driver's braking intention probability parameters based on the environmental risk index, foot movement parameters, and lower limb trajectory parameters.

[0079] For example, environmental perception units such as forward-facing millimeter-wave radar, lateral ultrasonic sensors, and panoramic cameras are used to collect key parameters around the vehicle in real time, such as the distance to obstacles ahead, relative speed, heading angle, road curvature, and weather lighting conditions. After calculation using a dynamic weighting strategy, a continuous environmental risk index (0-100 points) is generated, which reflects the potential threat level of the current environment to vehicle safety. At the same time, foot movement parameters are captured by foot sensors (such as high-sensitivity capacitive arrays and miniature millimeter-wave radar) to obtain the gradient of plantar pressure distribution, foot spatial displacement acceleration, and time-series characteristics of contact with the pedal. Combined with real-time tracking of lower limb trajectory parameters (such as changes in lower limb joint angles, foot spatial displacement trajectory, and motion velocity vector), a precise characterization of the driver's foot and lower limb dynamic behavior is formed. Subsequently, a machine learning-based pattern recognition model can be used to fuse environmental risk index, foot movement parameters, and lower limb trajectory parameters. By comparing the typical feature differences between normal braking scenarios (such as progressive pedaling, pedal travel matching vehicle speed) and accidental pedaling scenarios (such as rapid pedaling, pedal travel contradicting vehicle speed), the probability parameters of braking intent can be quantitatively calculated.

[0080] Optionally, when the environmental risk index is high (such as an obstacle rapidly approaching from the front), if the foot acceleration suddenly increases and the lower limb trajectory shows that the foot moves rapidly toward the accelerator pedal, the probability parameter of braking intention will be significantly reduced (indicating a high risk of accidental pedaling); conversely, if the foot movement is stable and the lower limb trajectory points toward the brake pedal, the probability parameter will be increased (indicating an active braking intention).

[0081] Step S603: Obtain the driver's historical accelerator pedal mis-pressing data, which includes the historical accelerator pedal mis-pressing frequency.

[0082] Step S604: Determine the probability parameters of the driver's accidental acceleration based on the historical frequency of accidental accelerator pedal presses and the probability parameters of braking intention.

[0083] For example, historical accelerator pedal mis-press data of the driver can be retrieved through the vehicle's storage unit or a cloud-based historical database, with a focus on extracting the frequency of historical accelerator pedal mis-presses. This historical accelerator pedal mis-press frequency can be calculated based on mis-press event records within a past period (e.g., 6 months) using a time decay algorithm (e.g., recent mis-press events have higher weights) to form a basic indicator reflecting the driver's long-term tendency to mis-press pedals. Subsequently, combined with the currently collected braking intention probability parameters, a Bayesian probability fusion model is used to jointly calculate the probability of historical accelerator pedal mis-presses and braking intention probability parameters.

[0084] Specifically, the historical frequency of accidental accelerator pedal presses serves as the prior probability, reflecting the driver's inherent tendency to mis-press the pedal, while the braking intention probability parameter, as the likelihood probability, reflects the probability of mis-pressing the pedal in the current scenario. Both are dynamically updated using a Bayesian formula to determine the posterior probability, i.e., the accidental accelerator pedal press probability parameter. Furthermore, to adapt to different driving scenarios (such as urban congestion and highway cruising), the weighting of the prior probabilities is dynamically adjusted. For example, in highway scenarios, due to the longer braking distance and shorter reaction time, the weight of the historical frequency of accidental accelerator pedal presses is appropriately reduced, while the weight of the real-time braking intention probability is increased.

[0085] In the embodiments provided in this application, by integrating multi-dimensional data such as environmental risk index, driver foot movement parameters, lower limb trajectory parameters and historical frequency of accidental pedal presses, a precise quantitative assessment of the probability of accidental accelerator pedal presses is achieved, providing a dynamic decision-making basis for graded braking strategies, and effectively improving the adaptability of active safety systems to complex driving scenarios and the accuracy of risk prevention and control.

[0086] Based on the above embodiments, in one exemplary embodiment provided in this application, the specific implementation process of the above-mentioned accelerator pedal anti-accidental pressing method may further include steps S701 and S702, which are described in detail below:

[0087] Step S701: After the vehicle executes the first braking control strategy or the second braking control strategy, acquire the vehicle's real-time multi-dimensional sensing data.

[0088] Step S702: Determine the reinforcement learning reward value based on real-time multidimensional sensing data, and optimize the first braking control strategy and the second braking control strategy based on the reward value.

[0089] For example, after the vehicle executes the first or second braking control strategy, a dynamic feedback loop is constructed using real-time collected multi-dimensional sensor data (such as radar echo signals, foot pressure distribution, lower limb joint angles, vehicle speed, and throttle opening). Specifically, a reinforcement learning reward value is calculated in real time based on this data. This reward value is generated through weighted fusion of multi-dimensional indicators, such as collision avoidance rate (reflecting whether the strategy successfully avoids a collision), braking smoothness (measuring passenger comfort), false intervention rate (assessing whether the strategy excessively interferes with normal driving), and energy recovery efficiency (the effect of kinetic energy recovery for electric vehicles). The weights of each indicator are dynamically adjusted according to the current driving scenario (such as urban congestion or highway cruising). Subsequently, a deep reinforcement learning algorithm (such as DQN or PPO) is used to associate the reward value with the corresponding state and action pairs, and the parameters of the braking control strategy are updated through a neural network model. For example, if the reward value indicates that the braking intensity of a certain second braking strategy is too high, causing passenger discomfort, the algorithm will lower the trigger threshold of full-pressure braking in that scenario or adjust the gradient of hydraulic braking pressure. Meanwhile, historical strategy execution data is stored in an experience replay pool for offline training and model iteration, ensuring that strategy optimization is based not only on current real-time feedback but also on long-term driving experience. The entire optimization process is achieved through a hybrid online and offline learning mechanism, guaranteeing both real-time response and long-term strategy evolution. Ultimately, this allows the braking control strategy to maintain high safety while continuously improving its adaptability to complex driving scenarios (such as rainy and snowy weather, and sudden obstacles), forming a complete intelligent evolutionary closed loop from strategy execution to effect evaluation to strategy optimization, significantly enhancing the robustness and intelligence of vehicle active safety.

[0090] In the embodiments provided in this application, a reinforcement learning mechanism driven by real-time sensor data is used to achieve dynamic adaptive optimization of the braking control strategy, thereby improving collision avoidance capabilities while continuously enhancing the system's intelligent adaptability to complex driving scenarios and robustness of safety decisions.

[0091] Please see Figure 3 , Figure 3 This is a flowchart illustrating the accelerator pedal anti-accidental pressing method provided in this application. Figure 2 Please see Figure 3The system acquires multi-dimensional sensor data of the vehicle and determines the vehicle's environmental risk index based on this data. If the environmental risk index is greater than a preset safety threshold and the vehicle's throttle opening value is greater than a first preset throttle opening threshold, the system determines the probability coefficient of the driver accidentally pressing the accelerator based on the multi-dimensional sensor data. If the probability coefficient of accidentally pressing the accelerator is less than or equal to a preset probability threshold, the system determines the driver's foot movement parameters based on the multi-dimensional sensor data and determines the driver's foot acceleration based on these parameters. Simultaneously, the system determines the relative distance between the vehicle and the vehicle in front using real-time environmental information. If the foot acceleration is greater than a preset acceleration threshold and the relative distance is less than or equal to a preset distance threshold, the system controls the vehicle to execute the motor warning strategy in the first braking control strategy. If the foot acceleration is less than or equal to a preset acceleration threshold and / or the relative distance is greater than a preset distance threshold, the system determines the real-time position information of the driver's feet based on lower limb trajectory parameters. If the real-time position information deviates from the effective operating area of ​​the vehicle's brake pedal, the system acquires the vehicle's real-time throttle opening value. If the real-time throttle opening value is greater than the second preset throttle opening threshold, the vehicle is controlled to execute the smooth braking strategy in the first braking control strategy, where the second preset throttle opening threshold is less than the first preset throttle opening threshold. If the probability coefficient of accidental throttle pedal application is greater than a preset probability threshold, a graded torque attenuation command for the vehicle's torque is determined based on the second braking control strategy to attenuate the vehicle's drive torque to the idle torque threshold; and a hydraulic control command is determined based on the second braking control strategy to trigger a preset solenoid valve group to apply full-pressure braking to the vehicle. Furthermore, after the vehicle executes the first or second braking control strategy, real-time multi-dimensional sensor data of the vehicle is acquired; a reinforcement learning reward value is determined based on the real-time multi-dimensional sensor data, and the first and second braking control strategies are optimized based on the reward value. For detailed implementation processes, please refer to the descriptions in the foregoing embodiments; these will not be repeated here.

[0092] Figure 4 This is a structural schematic diagram of the accelerator pedal anti-accidental pressing device provided in this application, as shown below. Figure 4As shown, the accelerator pedal misoperation prevention device 40 provided in this embodiment includes: an acquisition module 410, used to acquire multi-dimensional sensor data of the vehicle and determine the environmental risk index of the vehicle based on the multi-dimensional sensor data; a judgment module 420, used to determine the probability coefficient of the driver's accidental accelerator pedal operation based on the multi-dimensional sensor data if the environmental risk index is greater than a preset safety threshold and the accelerator pedal opening value of the vehicle is greater than a first preset accelerator pedal opening threshold; a first control module 430, used to determine the driver's real-time status information and real-time environmental information if the probability coefficient of accidental accelerator pedal operation is less than or equal to a preset probability threshold, and control the vehicle to execute a first braking control strategy based on the real-time status information and real-time environmental information; and a second control module 440, used to control the vehicle to execute a second braking control strategy if the probability coefficient of accidental accelerator pedal operation is greater than a preset probability threshold, wherein the braking intensity of the second braking control strategy is greater than the braking intensity of the first braking control strategy.

[0093] In one possible implementation, the first control module 430 is further configured to: determine the driver's foot movement parameters based on multi-dimensional sensing data, and determine the driver's foot acceleration based on the foot movement parameters; determine the relative distance with the vehicle in front based on real-time environmental information; and if the foot acceleration is greater than a preset acceleration threshold and the relative distance is less than or equal to a preset distance threshold, control the vehicle to execute the motor warning strategy in the first braking control strategy.

[0094] In one possible implementation, the first control module 430 is further configured to: determine the real-time position information of the driver's feet based on the lower limb trajectory parameters if the foot acceleration is less than or equal to a preset acceleration threshold and / or the relative distance is greater than a preset distance threshold; obtain the real-time throttle opening value of the vehicle if the real-time position information deviates from the effective action area of ​​the vehicle's brake pedal; and control the vehicle to execute the smooth braking strategy in the first braking control strategy if the real-time throttle opening value is greater than a second preset throttle opening threshold, wherein the second preset throttle opening threshold is less than the first preset throttle opening threshold.

[0095] In one possible implementation, the second control module 440 is further configured to: determine a graded torque attenuation command for the vehicle's torque based on the second braking control strategy, so as to attenuate the vehicle's driving torque to an idle torque threshold based on the graded torque attenuation command; and determine a hydraulic control command based on the second braking control strategy, so as to trigger a preset solenoid valve group to perform full-pressure braking on the vehicle based on the hydraulic control command.

[0096] In one possible implementation, the acquisition module 410 is further configured to determine the relative relationship between the vehicle and the vehicle in front based on radar data, the relative relationship including relative distance, relative speed and relative heading angle; and to perform weighted calculation on the relative distance, relative speed and relative heading angle based on a preset weighting strategy, so as to use the calculation result as the vehicle's environmental risk index.

[0097] In one possible embodiment, the acquisition module 410 is further configured to: determine the vehicle's environmental risk index and the driver's foot movement parameters and lower limb trajectory parameters based on multi-dimensional sensing data; determine the driver's braking intention probability parameters based on the environmental risk index, foot movement parameters, and lower limb trajectory parameters; acquire the driver's historical accelerator pedal mis-pressing data, including the historical accelerator pedal mis-pressing frequency; and determine the driver's accidental accelerator pedal pressing probability parameters based on the historical accelerator pedal mis-pressing frequency and the braking intention probability parameters.

[0098] In one possible embodiment, the acquisition module 410 is further configured to acquire real-time multi-dimensional sensing data of the vehicle after the vehicle executes the first braking control strategy or the second braking control strategy; determine the reinforcement learning reward value based on the real-time multi-dimensional sensing data; and optimize the first braking control strategy and the second braking control strategy based on the reward value.

[0099] The accelerator pedal anti-accidental pressing device provided in this embodiment can perform the method provided in the above method embodiment. Its implementation principle and technical effect are similar, and will not be described in detail here.

[0100] Figure 5 A schematic diagram of the structure of the electronic device provided in this application. Figure 5 As shown, the electronic device 50 provided in this embodiment includes at least one processor 510 and a memory 520. Optionally, the device 50 further includes a communication component 530. The processor 510, memory 520, and communication component 530 are connected via a bus 540.

[0101] In a specific implementation, at least one processor 510 executes computer execution instructions stored in memory 520, causing at least one processor 510 to perform the above-described method.

[0102] The specific implementation process of processor 510 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0103] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0104] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0105] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0106] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0107] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described method.

[0108] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0109] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.

[0110] The division of units is merely a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0111] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0112] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0113] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0114] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0115] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A method for preventing accidental accelerator pedal press, characterized in that, include: Acquire multidimensional sensor data of the vehicle and determine the environmental risk index of the vehicle based on the multidimensional sensor data; If the environmental risk index is greater than a preset safety threshold and the throttle opening value of the vehicle is greater than a first preset throttle opening threshold, then the probability coefficient of the driver accidentally pressing the accelerator is determined based on the multi-dimensional sensing data. If the probability coefficient of accidentally pressing the accelerator is less than or equal to a preset probability threshold, then the driver's real-time status information and real-time environment information are determined, and the vehicle is controlled to execute a first braking control strategy based on the real-time status information and the real-time environment information. If the probability coefficient of accidentally pressing the accelerator is greater than a preset probability threshold, the vehicle is controlled to execute a second braking control strategy, the braking intensity of the second braking control strategy being greater than the braking intensity of the first braking control strategy.

2. The method as described in claim 1, characterized in that, The real-time status information includes foot movement parameters, and the step of controlling the vehicle to execute a first braking control strategy based on the real-time status information and the real-time environmental information includes: The driver's foot motion parameters are determined based on the multidimensional sensing data, and the driver's foot acceleration is determined based on the foot motion parameters. The relative distance to the vehicle in front is determined based on the real-time environmental information. If the foot acceleration is greater than a preset acceleration threshold and the relative distance is less than or equal to a preset distance threshold, then the vehicle is controlled to execute the motor warning strategy in the first braking control strategy.

3. The method as described in claim 2, characterized in that, The real-time status information also includes lower limb trajectory parameters, and the step of controlling the vehicle to execute a first braking control strategy based on the real-time status information further includes: If the foot acceleration is less than or equal to the preset acceleration threshold and / or the relative distance is greater than the preset distance threshold, then the real-time position information of the driver's foot is determined based on the lower limb trajectory parameters; If the real-time position information deviates from the effective operating area of ​​the vehicle's brake pedal, then the real-time throttle opening value of the vehicle is obtained. If the real-time throttle opening value is greater than the second preset throttle opening threshold, then the vehicle is controlled to execute the smooth braking strategy in the first braking control strategy, where the second preset throttle opening threshold is less than the first preset throttle opening threshold.

4. The method according to any one of claims 1 to 3, characterized in that, The control of the vehicle to execute the second braking control strategy includes: Based on the second braking control strategy, a graded torque attenuation command for the vehicle's torque is determined, so as to attenuate the vehicle's drive torque to an idle torque threshold based on the graded torque attenuation command; and, Based on the second braking control strategy, a hydraulic control command is determined to trigger a preset solenoid valve group to apply full-pressure braking to the vehicle.

5. The method according to any one of claims 1 to 3, characterized in that, The multidimensional sensing data also includes radar data, and determining the environmental risk index of the vehicle based on the multidimensional sensing data includes: The relative relationship between the vehicle and the vehicle in front is determined based on the radar data. The relative relationship includes the relative distance, relative speed, and relative heading angle. The relative distance, relative speed, and relative heading angle are weighted and calculated based on a preset weighting strategy, and the calculation result is used as the environmental risk index of the vehicle.

6. The method as described in claim 5, characterized in that, The determination of the probability coefficient of the driver accidentally pressing the accelerator based on the multi-dimensional sensor data includes: The environmental risk index of the vehicle and the driver's foot movement parameters and lower limb trajectory parameters are determined based on the multidimensional sensing data. The probability parameters of the driver's braking intention are determined based on the environmental risk index, the foot movement parameters, and the lower limb trajectory parameters. Acquire the driver's historical accelerator pedal mis-press data, which includes the frequency of historical accelerator pedal mis-presses. The probability parameter of the driver's accidental acceleration is determined based on the historical frequency of accidental accelerator pedal presses and the probability parameter of braking intent.

7. The method according to any one of claims 1 to 3, characterized in that, The method further includes: After the vehicle executes the first braking control strategy or the second braking control strategy, real-time multi-dimensional sensing data of the vehicle is acquired. The reinforcement learning reward value is determined based on the real-time multidimensional sensing data, and the first braking control strategy and the second braking control strategy are optimized based on the reward value.

8. An electronic device, characterized in that, include: Memory, processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1 to 7.

10. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the method described in any one of claims 1 to 7.

Images