A passing control system and method

Abstract

The application discloses an overtaking control system and method, which comprises a scene discrimination module, a decision control module and an execution module. The scene discrimination module discriminates whether the vehicle is in a right side vehicle cutting-in scene or an overtaking scene of the vehicle according to driver operation data and vehicle environment information around the vehicle. The decision control module correspondingly outputs a monitoring instruction, a braking instruction or an acceleration instruction. The execution module responds to the instruction action. The application solves the problems that the existing auxiliary driving system is prone to cause man-machine confrontation, control logic is rigid and safety hidden troubles are prominent under the overtaking working condition. The system adopts a differentiated scene coping strategy. When the right side vehicle cuts in, the system selects monitoring and alarm or active braking according to the driver state. When the vehicle overtakes, the system realizes rapid escape through early warning and acceleration assistance. The system realizes man-machine cooperation by adopting a control right distribution based on confidence and an S-shaped smooth transition curve. The system significantly improves the lateral collision protection capability and driving comfort without intervening in the steering control.

Description

Technical Field

[0001] This invention relates to the field of vehicle control technology, specifically to an overtaking control system and method. Background Technology

[0002] Currently, in the field of active safety technology for passenger vehicles, the prevention and control of lateral collision risks during vehicle overtaking mainly relies on traditional blind spot detection (BSD) systems and automatic emergency braking (AEB) systems. The mainstream implementation of existing technologies is usually based on a single environmental perception data source (such as relying solely on lateral radar to monitor blind spots). When the system detects a vehicle entering the blind spot from an adjacent lane, it only provides a passive warning by illuminating the rearview mirror warning lights; or, when a rear-end collision is detected to be imminent, it directly triggers emergency braking with maximum deceleration.

[0003] However, the aforementioned existing technologies have the following significant limitations in practical applications: The system lacks a coupled judgment of driver intent and operational status. When detecting lateral risks, existing systems often neglect to recognize the driver's current operational behavior. If the system still forcibly intervenes when the driver has already perceived the risk and performed emergency steering or braking, it is very likely to conflict with the driver's operation, leading to human-machine conflict and causing vehicle instability or unexpected steering. Conversely, if the driver is unaware and does not operate, the system's intervention is often delayed, failing to leave sufficient room for avoidance.

[0004] The existing technology suffers from poor adaptability to different scenarios and a simplistic control strategy. It typically fails to differentiate between two distinct driving scenarios: a vehicle cutting in from the right and the vehicle actively overtaking. It employs a one-size-fits-all warning or braking logic. Particularly in dangerous situations where the vehicle forcibly merges into another lane before completing its overtaking maneuver, the current system lacks effective tiered warning and acceleration intervention mechanisms, failing to assist the driver in quickly escaping the danger zone.

[0005] The abrupt switching of control logic negatively impacts the driving experience. Existing human-machine co-driving systems typically employ abrupt command switching when the system takes over or returns control to the driver. This results in sudden changes in vehicle acceleration or deceleration, creating a noticeable jerkiness and even causing panic among occupants, thus reducing the acceptance of driver assistance systems.

[0006] There are functional safety hazards. Existing driver assistance systems often intervene in steering control to prevent lane departure. However, in high-risk lateral movement scenarios such as overtaking, the system's active steering control is prone to conflict with the driver's steering input, posing a significant safety hazard. Summary of the Invention

[0007] The purpose of this invention is to provide an overtaking control system that can accurately identify overtaking scenarios, deeply integrate driver's operating intentions, achieve smooth human-machine collaboration, and not intervene in steering control.

[0008] To achieve this objective, the present invention provides an overtaking control system, which includes: The scene discrimination module is used to determine the driver's current operating status based on the driver's operation data. Based on the vehicle environment information around the vehicle and the driver's current operating status, it determines whether the vehicle is currently in a scenario where a vehicle cuts in from the right or in a scenario where the vehicle is overtaking. The decision control module is used to output monitoring instructions when the vehicle is in a scenario where a vehicle cuts in from the right and the driver's current operation state is an emergency operation state. When the vehicle is in a scenario where a vehicle cuts in from the right and the driver is currently operating normally, and the predicted collision time is less than a preset threshold, the braking acceleration required for braking is obtained based on the initial front-to-back distance between the vehicle and the vehicle on the right, the vehicle's speed, the distance between the vehicle and the vehicle on the right, the steering angle of the vehicle on the right, and the speed of the vehicle on the right, and a braking command is output based on the braking acceleration required for braking. When the vehicle is in an overtaking scenario and the driver starts to steer before the vehicle has completely overtaken the overtaken vehicle, an alarm command is output. If no correction operation by the driver is detected within a preset time, the acceleration for the safe overtaking time is output based on the time required for safe overtaking, the initial front-to-back distance between the vehicle and the overtaken vehicle, and the speed of the overtaken vehicle. An acceleration command is then output based on the acceleration for the safe overtaking time.

[0009] Preferably, when the decision control module determines that a vehicle is cutting in from the right and the driver is in a normal operating state, it calculates the predicted collision time using the following formula. ; ; in, This is the distance between this vehicle and the vehicle on its right. This represents the steering angle of the vehicle on the right. The speed of the vehicle on the right; When the calculated predicted collision time When the braking force is less than a preset threshold, the decision control module outputs a braking command and calculates the required braking acceleration according to the following formula. : ; in, This represents the initial front-to-back distance between this vehicle and the vehicle on its right. This is the speed of the vehicle.

[0010] Preferably, if no driver correction operation is detected within the specified time, the acceleration required for safe overtaking is output, and the time required for safe overtaking is first calculated using the following formula. : ; in, For the speed of this vehicle, The distance between the two vehicles. The steering angle of the vehicle being overtaken; Calculate the acceleration required for safe overtaking based on the time required for safe overtaking. : ; in, This represents the initial front-to-back distance between this vehicle and the vehicle being overtaken. The speed of the vehicle being overtaken; Preferably, the specific method for determining whether the situation is a scenario where a vehicle is cutting in from the right or an overtaking scenario based on the surrounding vehicle environment information and the driver's current operating status is as follows: If the vehicle environment information around the vehicle shows that there is a vehicle on the right side of the vehicle that has turned on its turn signal to the left or is actually turning to the left, it is determined to be in a scenario where a vehicle cuts in from the right; if the vehicle environment information around the vehicle shows that the vehicle and the vehicle on the right overlap longitudinally, and the driver's current operation includes increasing the accelerator pedal opening or turning the steering wheel to the left, it is determined to be in a scenario where the vehicle is overtaking.

[0011] Preferably, the monitoring command is used to increase the monitoring frequency of the left and rear areas defined by the vehicle to a set frequency, and to collect the relative speed and travel distance parameters of target vehicles in the area in real time; when a high-speed approaching vehicle with a relative speed greater than a set speed threshold and a distance less than a set distance threshold is detected in the left or rear areas defined by the vehicle, an audible and visual alarm is triggered to prompt the driver to avoid the risk; if no abnormal approaching vehicle is detected in the left or rear areas defined by the vehicle, no warning or control intervention is triggered.

[0012] Preferably, when the decision control module determines that the vehicle is overtaking, if it detects that the driver starts to turn before the vehicle has completely overtaken the overtaken vehicle, it controls the execution module to output an audible and visual alarm command at a set frequency. If no driver correction is detected within the preset time period, an acceleration command will be output. If, during the output of the acceleration command, the increase in accelerator pedal opening is not greater than the set percentage or the steering wheel return angle is not greater than the set angle, it is determined that no driver correction operation has been detected.

[0013] Preferably, the decision control module is further configured with a control authority allocation strategy based on confidence level: When the confidence level of the driver's operating intention is higher than the first set confidence threshold, the driver has complete control. When the confidence level is between the second set confidence level threshold and the first set confidence level threshold, the vehicle assistance control mode is adopted; When the confidence level is lower than the second set confidence threshold and the collision risk level is high, the vehicle is completely taken over by the decision control module.

[0014] Preferably, when the driver corrects his operation during a period when the vehicle is completely under the control of the decision control module, an S-shaped smooth transition curve is used to return control to the driver within a preset time, while controlling the vehicle's acceleration change rate to not exceed a set change rate threshold. Among them, the S-shaped smooth transition curve is a time-domain control quantity change curve generated based on the Sigmoid function. The slope of this curve approaches zero in the initial and final stages and monotonically increases in the middle stage, so that the output acceleration, deceleration or steering torque exhibits a smooth S-shaped change.

[0015] The beneficial effects of this invention are as follows: This invention proposes an overtaking control system and method. By constructing a hierarchical control allocation mechanism based on the confidence level of the driver's operating intention, it clearly defines the switching boundaries between three modes: complete system takeover, assisted control, and driver-led operation. Specifically, when the decision control module intervenes in braking or acceleration, by monitoring driver operation data in real time, once the driver's intervention intention is detected, the system immediately and seamlessly returns control within 100ms to 150ms via a smooth S-shaped transition curve. This mechanism completely avoids the common phenomena of grabbing the steering wheel or human-machine confrontation in existing technologies, ensuring driving safety and the driver's subjective sense of control.

[0016] Through deep integration of the environmental perception module and the driver behavior recognition module, differentiated proactive intervention strategies have been developed for two high-risk scenarios: right-hand vehicle cut-in and overtaking. In the right-hand vehicle cut-in scenario, the system can intelligently switch between pure monitoring alarm and active braking modes based on whether the driver is in an emergency operation state, effectively compensating for the driver's delayed reaction. In the overtaking scenario, by monitoring the steering behavior of the vehicle before it has completely passed, the system can promptly output a three-level audible and visual alarm and acceleration control command to assist the driver in quickly leaving the danger zone, greatly reducing the probability of side scrapes and collisions.

[0017] When switching control commands, the traditional step-type control output is abandoned, and an S-shaped smooth transition curve with the Sigmoid function as its core is adopted. This curve ensures that the slope of acceleration and deceleration approaches zero in the initial and final stages, and the intermediate stage is a smooth transition, strictly limiting the rate of change of vehicle acceleration to 2.5 m / s². 3Within this range. This technology effectively eliminates the dynamic shock and jerking sensation caused by the transfer of control, significantly improves the comfort experience of passengers in assisted driving mode, and solves the problem of abrupt intervention in existing systems. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the structure of the present invention; Figure 2 Control logic diagram for the right-side vehicle switching scenario; Figure 3 This is the control logic diagram for the overtaking scenario of this vehicle. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of them. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to represent selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0020] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments: Example 1 An overtaking control system, such as Figure 1 As shown, it includes: The scene discrimination module is used to determine the driver's current operating status based on the driver's operation data. Based on the vehicle environment information around the vehicle and the driver's current operating status, it determines whether the vehicle is currently in a scenario where a vehicle cuts in from the right or in a scenario where the vehicle is overtaking. The decision control module is used to output monitoring instructions when the vehicle is in a scenario where a vehicle cuts in from the right and the driver's current operation state is an emergency operation state. When the vehicle is in a scenario where a vehicle cuts in from the right and the driver is currently operating normally, and the predicted collision time is less than a preset threshold, the braking acceleration required for braking is obtained based on the initial front-to-back distance between the vehicle and the vehicle on the right, the vehicle's speed, the distance between the vehicle and the vehicle on the right, the steering angle of the vehicle on the right, and the speed of the vehicle on the right, and a braking command is output based on the braking acceleration required for braking. When the vehicle is in an overtaking scenario and the driver starts to steer before the vehicle has completely overtaken the overtaken vehicle, an alarm command is output. If no correction operation by the driver is detected within a preset time, the acceleration for the safe overtaking time is output based on the time required for safe overtaking, the initial front-to-back distance between the vehicle and the overtaken vehicle, and the speed of the overtaken vehicle. An acceleration command is then output based on the acceleration for the safe overtaking time.

[0021] In some preferred embodiments, an execution module is also included, which performs corresponding control according to monitoring instructions, braking instructions, alarm instructions or acceleration instructions.

[0022] In some preferred embodiments, a scene discrimination module is also included, which is used to determine the driver's operation status based on the driver's operation data, and to determine whether the current situation is a scenario where a vehicle cuts in from the right or a scenario where the vehicle is overtaking, based on the vehicle environment information around the vehicle and the driver's operation status, and to obtain a discrimination result.

[0023] For the scene discrimination module, the environmental perception module, which collects information about the surrounding vehicle environment, consists of a heterogeneous sensor network, specifically including a forward-facing monocular camera, a side-facing millimeter-wave radar, and a rear-facing ultrasonic sensor. The forward-facing monocular camera has a 120° horizontal field of view and is primarily responsible for identifying vehicle outlines and turn signal signals; the side-facing millimeter-wave radar has a detection range of up to 150m and a ranging accuracy of ±0.1m, used to accurately obtain the relative speed and distance of target vehicles; the rear-facing ultrasonic sensor has a detection range of 8m and is used to fill in blind spots with nearby obstacles. Before entering the scene discrimination module, the raw data collected by each sensor undergoes spatiotemporal synchronization and filtering by a preprocessing unit to remove outliers caused by sensor jitter or electromagnetic interference. The fused information about the surrounding vehicle environment is then output at a frequency of 100Hz, providing a highly reliable data foundation for subsequent decision-making and control.

[0024] Regarding the driver's current operating state, in some preferred embodiments, the driver's current operating state is a quantitative representation of the urgency of the driver's driving behavior based on driver operation data. Specifically, it is determined by the rate of change of the steering wheel angle collected by the steering wheel angle sensor and the brake pedal depressing speed collected by the brake pedal travel sensor. In optional embodiments, if the driver's operation data meets the conditions that the rate of change of the steering wheel angle is greater than 150° / s or the brake pedal depressing speed is greater than 300mm / s, then the driver's current operating state is determined to be an emergency operating state, indicating that the driver has actively identified the risk and taken emergency avoidance actions. If the driver's operation data meets the conditions that the brake pedal depressing speed is less than 100mm / s and the rate of change of the steering wheel angle is less than 50° / s, then the driver's current operating state is determined to be a normal operating state, indicating that the driver has not made an emergency response to the surrounding risks. This definition distinguishes the driver's intention to actively intervene through objective physical thresholds, providing a core basis for the selection of subsequent differentiated control strategies of the system. It avoids redundant intervention by the system when the driver has actively avoided risks, leading to human-machine confrontation, and also ensures that the system can promptly activate active protection when the driver does not respond.

[0025] In some preferred embodiments, the determination of the right-side vehicle cutting-in scenario and the overtaking scenario is based on the deep fusion of environmental perception data and driver's operational intentions: if the vehicle environment information around the vehicle shows that there is a vehicle on the right side of the vehicle that has turned on its turn signal to the left or actually turns to the left, it is determined to be in the right-side vehicle cutting-in scenario. This determination logic can accurately capture the precursor features of the neighboring vehicle's intention to encroach on the vehicle's driving path, reserving critical pre-aiming response time for the system; if the vehicle environment information around the vehicle shows that the vehicle and the vehicle on the right have longitudinal overlap, and the driver's current operation includes increasing the accelerator pedal opening or turning the steering wheel to the left, it is determined to be in the overtaking scenario. This determination logic effectively distinguishes between normal following and active overtaking and lane changing behavior by combining the relative position relationship of the vehicles and the overtaking operation command actively issued by the driver, avoiding misjudgments that may occur if only relying on environmental perception data, and ensuring that the system only intervenes in a targeted manner when the driver has a clear intention to overtake and there is a potential risk.

[0026] In the above technical solution, the core decision-making logic framework of the overtaking control system of this invention is constructed. It takes scene discrimination and hierarchical response mechanism as the basis for system operation. The scene discrimination module first identifies the two most common side collision risk scenarios: vehicles cutting in from the right and overtaking by the vehicle itself. At the same time, it locks whether the driver is in an emergency operation or a normal operation state, thus providing a basis for subsequent differentiated control command output. It also breaks the one-size-fits-all response mode of traditional driver assistance systems. For example, when a vehicle cuts in from the right, if the driver is already in emergency avoidance, only monitoring and alarm are issued. If the driver does not react, the system will actively brake. When the vehicle is overtaking, the system will decide whether to issue a warning or accelerate based on the overtaking status of the vehicle. This avoids human-machine confrontation at the source and ensures the accuracy of system intervention and the driver's dominant control.

[0027] In some preferred embodiments, such as Figure 2 As shown, when the decision control module determines that a vehicle is cutting in from the right and the driver is in a normal operating state, it calculates the predicted collision time using the following formula. ; ; in, This is the distance between this vehicle and the vehicle on its right. This represents the steering angle of the vehicle on the right. The speed of the vehicle on the right; When the calculated predicted collision time When the braking force is less than a preset threshold, the decision control module outputs a braking command and calculates the required braking acceleration according to the following formula. : ; in, This represents the initial front-to-back distance between this vehicle and the vehicle on its right. This is the speed of the vehicle.

[0028] In some preferred embodiments, after the decision control module calculates the theoretical minimum braking acceleration required to avoid a collision based on the aforementioned formula, it does not directly issue the theoretical value. Instead, it combines the current road surface adhesion coefficient, vehicle load status, and braking system response characteristics, and corrects the theoretical acceleration using a pre-calibrated big data correction coefficient to generate the final braking command. This command is sent to the brake-by-wire system actuator in the execution module in the form of a pulse width modulation signal or a bus message, controlling the brake fluid pressure or motor feedback torque so that the actual deceleration generated by the vehicle approximates the braking acceleration required for braking, while limiting the braking force control error to within 5%. This process achieves closed-loop control from "risk prediction" to "physical execution," ensuring that the braking force output by the system is sufficient to avoid a collision, but without causing vehicle instability or rear-end collisions due to excessive braking intensity, thus balancing safety and ride comfort.

[0029] The aforementioned technical solution specifies a collision risk assessment and braking decision algorithm for scenarios where a vehicle cuts in from the right. It uses a specific mathematical formula to convert physical parameters such as the distance between the vehicle and the vehicle on the right, the other vehicle's turning angle, and the speeds of both vehicles into quantifiable predicted collision times, and further derives the required braking acceleration. This transforms the previously vague perception of danger into precise physical calculations, enabling the system to judge risk based on real vehicle dynamics data rather than empirical thresholds. This significantly improves the timeliness and accuracy of braking intervention, avoiding premature intervention that could interfere with the driver's normal driving, and outputting just the right amount of braking force to effectively prevent a collision when the TTC (Traction Time Tolerance) is less than a preset threshold. Simultaneously, the large-scale data correction coefficient implicit in the formula ensures that the braking intensity meets the actual road conditions.

[0030] In some preferred embodiments, such as Figure 3 As shown, if no driver correction is detected within the specified time, the acceleration required for safe overtaking is output. The time required for safe overtaking is first calculated using the following formula. : ; in, For the speed of this vehicle, The distance between the two vehicles. The steering angle of the vehicle being overtaken; Calculate the acceleration required for safe overtaking based on the time required for safe overtaking. : ; in, This represents the initial front-to-back distance between this vehicle and the vehicle being overtaken. The speed of the vehicle being overtaken; The aforementioned technical solution specifies the calculation logic for safe overtaking acceleration in overtaking scenarios. It first calculates the safe time required to complete the overtaking maneuver using a formula, and then, combined with the speed of the overtaken vehicle, the initial distance between the two vehicles, and their relative positions, it calculates the required overtaking acceleration for the vehicle. This gives the system the ability to actively avoid danger, rather than being limited to passive braking. When the driver is detected to be swerving dangerously before the vehicle has fully overtaken and fails to correct it in time, the system can output a reasonable acceleration command through precise calculation to help the vehicle quickly increase the safe distance from the overtaken vehicle. This effectively solves the risk of side scraping caused by insufficient power or hesitation during overtaking. By scientifically planning the acceleration trajectory, it ensures the decisive completion of the overtaking maneuver while avoiding new safety risks caused by excessive acceleration.

[0031] In some preferred embodiments, the braking acceleration calculated by the above formula Acceleration time required for safe overtaking These are all theoretical physical values. In actual control command generation, the decision control module also needs to incorporate correction coefficients derived from large-sample data statistics. This coefficient is used to compensate for differences in vehicle mass, tire friction coefficient, and road adhesion conditions, i.e., the acceleration corresponding to the final executed command. By dynamically correcting this coefficient, the present invention can limit the braking force control error and throttle opening control error, ensuring that the theoretical calculation results are highly consistent with the actual vehicle dynamic response, and avoiding over-intervention or under-intervention caused by model distortion.

[0032] In some preferred embodiments, the specific method for determining the driver's current operating state based on driver operation data is as follows: if the driver operation data satisfies a steering wheel angle change rate greater than 150° / s, or a brake pedal depressing speed greater than 300mm / s, then the driver's current operating state is determined to be an emergency operating state; if the driver operation data satisfies a brake pedal depressing speed less than 100mm / s and a steering wheel angle change rate less than 50° / s, then the driver's current operating state is determined to be a normal operating state.

[0033] In some preferred embodiments, the specific method for determining whether the vehicle is in a right-hand vehicle cutting in or an overtaking scenario based on the surrounding vehicle environment information and the driver's current operating state is as follows: If the vehicle environment information around the vehicle shows that there is a vehicle on the right side of the vehicle that has turned on its turn signal to the left or is actually turning to the left, it is determined to be in a scenario where a vehicle cuts in from the right; if the vehicle environment information around the vehicle shows that the vehicle and the vehicle on the right overlap longitudinally, and the driver's current operation includes increasing the accelerator pedal opening or turning the steering wheel to the left, it is determined to be in a scenario where the vehicle is overtaking.

[0034] The aforementioned technical solution defines in detail the specific execution standards for scenario discrimination and driver state judgment. It concretizes abstract scenarios and states through quantitative indicators, enabling the system to have objective perception and understanding capabilities. For example, it clearly defines turning on the turn signal of a vehicle on the right as an entry signal, and longitudinal overlap and increased throttle as overtaking intentions. This eliminates the ambiguity in the system's judgment and ensures that the system can stably and consistently identify high-risk conditions under different driving styles and complex traffic environments, providing a solid data foundation for the correct instruction output of the subsequent decision control module.

[0035] In some preferred embodiments, the monitoring command is used to increase the monitoring frequency of the left and rear areas defined by the vehicle to a set frequency, and to collect the relative speed and travel distance parameters of target vehicles in the area in real time; when a high-speed approaching vehicle with a relative speed greater than a set speed threshold and a distance less than a set distance threshold is detected in the left or rear areas defined by the vehicle, an audible and visual alarm is triggered to prompt the driver to avoid the risk; if no abnormal approaching vehicle is detected in the left or rear areas defined by the vehicle, no warning or control intervention is triggered.

[0036] The aforementioned technical solution refines the execution strategy of monitoring commands. Its core lies in dynamically adjusting the monitoring frequency and alarm strategy according to the risk level. This means that the system maintains routine monitoring in non-emergency situations to save computing power, while immediately increasing the monitoring frequency of the left and rear areas to 200Hz when a potential hazard is detected, and focusing on monitoring vehicles approaching at high speed. This achieves optimized allocation of system resources and provides high-precision environmental perception at critical moments. Once a high-speed vehicle is detected from the side or rear, an audible and visual alarm is triggered immediately, acting as an "electronic lookout." Without interfering with vehicle control, it maximizes the driver's perception boundaries and reduces the probability of blind spot accidents.

[0037] In some preferred embodiments, when the decision control module determines that the vehicle is overtaking, if it detects that the driver starts to turn before the vehicle has completely overtaken the overtaken vehicle, it controls the execution module to output an audible and visual alarm command at a set frequency. If no driver correction is detected within the preset time period, an acceleration command will be output. If, during the output of the acceleration command, the increase in accelerator pedal opening is not greater than the set percentage or the steering wheel return angle is not greater than the set angle, it is determined that no driver correction operation has been detected.

[0038] In some preferred embodiments, the set-frequency audible and visual alarm command specifically manifests as a three-level warning mode: In the auditory dimension, the system controls the in-vehicle speakers to output an intermittent warning sound at a frequency of 3Hz. This frequency is close to the sensitive frequency band of human instinctive reactions, effectively waking the driver's attention. In the visual dimension, the system controls the central display screen of the instrument panel to flash a bright red light, simultaneously illuminating the blind spot monitoring indicator on the corresponding side rearview mirror. This combined audiovisual and visual strong alert method can maximize the driver's awareness of the current side collision risk without startling them, prompting them to take timely corrective action.

[0039] In some preferred embodiments, the specific thresholds for monitoring driver correction actions are set as follows: For braking intervention scenarios, if the brake pedal depress speed is detected to be greater than 50 mm / s or the steering wheel angle change rate is greater than 30° / s, it is determined that the driver has begun to intervene; for overtaking acceleration assist scenarios, if the accelerator pedal opening is detected to increase by more than 10% or the steering wheel return angle is greater than 20°, it is determined that the driver has taken effective corrective measures. These thresholds are set based on extensive real-vehicle test data, aiming to distinguish between the driver's active risk avoidance intentions and unconscious limb tremors, ensuring that the system can smoothly exit control within 100ms to 150ms after the driver begins effective operation, achieving true human-machine collaboration.

[0040] The aforementioned technical solution improves the early warning and tiered intervention mechanism in overtaking scenarios, clarifying the complete closed-loop process from alarm to assisted acceleration; it gives the driver priority to make corrections, only activating acceleration assistance when the driver fails to respond to the warning, and continuously monitoring the driver's intention to take over during acceleration; it fully embodies the "human-centered" design concept, both deterring the driver's dangerous overtaking impulses in a timely manner through three-level audible and visual alarms, and providing a safety net when the driver fails to react in time. At the same time, it judges the driver's corrective behavior by monitoring the increase in throttle opening and the steering wheel return angle, ensuring a high degree of coordination between system intervention and driver intention, and avoiding false intervention.

[0041] In some preferred embodiments, the decision control module is further configured with a confidence-based control allocation strategy: When the confidence level of the driver's operating intention is higher than the first set confidence threshold, the driver has complete control. When the confidence level is between the second set confidence level threshold and the first set confidence level threshold, the vehicle auxiliary control mode is adopted; When the confidence level is lower than the second set confidence threshold and the collision risk level is high, the vehicle is completely taken over by the decision control module.

[0042] Regarding the confidence level of driver intention, in some preferred embodiments, the confidence level is a probability value quantified by the decision control module after multi-dimensional fusion analysis of driver operation data and the current vehicle motion state, representing the clarity of the driver's control intention. Its value ranges from 0 to 100%. The calculation of this confidence level comprehensively considers the changing trends of the steering wheel angle change rate, braking or accelerator pedal depressing speed, and displacement: when the driver operation data shows high consistency (e.g., the steering wheel turns rapidly to the left and the accelerator pedal opening increases simultaneously) and matches the actual kinematic response of the vehicle, the confidence level approaches 100%, indicating that the driver's intention is clear; when the operation data is contradictory (e.g., the brake and accelerator pedals are pressed simultaneously) or the signal is weak and difficult to identify, the confidence level approaches 0%, indicating that the driver's intention is ambiguous or missing. This confidence level, serving as the core decision-making basis for human-machine collaborative control strategies, directly determines the attribution of control. A higher confidence level indicates a stronger driver's intention to control the vehicle, and the system is more inclined to maintain the driver's dominant position. Conversely, a lower confidence level indicates a weaker ability or less awareness of risk among drivers, and the system is more inclined to proactively intervene to ensure safety. By introducing this quantitative indicator, the system can accurately identify what the driver intends to do and whether the driver is effectively controlling the vehicle, thereby achieving an optimal balance between respecting human driving intentions and ensuring driving safety.

[0043] In an optional embodiment, the first confidence threshold is set to 80%, and the second confidence threshold is set to 30%. This value setting is based on statistical analysis of a large amount of driver-in-the-loop test data: when the confidence level of the driver's operational intent is higher than 80%, it indicates that the driver's operational behavior is highly consistent and clear, the driver has absolute control over the vehicle and a clear intention to avoid risks, therefore the system completely relinquishes control, retaining only background monitoring; when the confidence level is lower than 30%, it indicates that the driver may be distracted, fatigued, or unconscious, and the operational data shows chaotic or weak characteristics, unable to effectively cope with the impending collision risk. At this time, the system determines that the driver has lost effective control ability, thus forcibly taking over control completely to prevent the accident from occurring; setting the 30% to 80% range as an auxiliary control mode provides a buffer zone for the system, allowing the system to assist the driver in correcting driving deviations through pre-pressurization braking or slight torque intervention without depriving the driver of authority. By setting the above specific thresholds, this invention transforms abstract decision-making logic into executable and calibrable engineering parameters, ensuring the stability and reliability of the human-machine collaborative control strategy under different vehicle models and road conditions.

[0044] In some preferred embodiments, the vehicle assistance control mode refers to a transitional control strategy adopted by the system when the driver's operational intention confidence is between a first set confidence threshold of 30% to 80% and a second set confidence threshold, which is between fully manual driving and complete system takeover. In this mode, the decision control module does not generate independent braking or acceleration commands to override the driver's operation. Instead, the execution module fine-tunes the vehicle's longitudinal slip dynamics. Specifically, the system can pre-charge the brake lines with hydraulic pressure to shorten the response lag in emergency braking, or provide an additional driving torque of no more than 0.2g when the driver's throttle opening is insufficient, assisting the vehicle in maintaining a stable overtaking trajectory. Simultaneously, the intensity of control intervention is strictly limited, ensuring that the driver only needs to apply a small reverse operating torque to offset the system's assistance. The core of this mode is assistance rather than replacement. It utilizes the system's forward-looking perception capabilities to compensate for the driver's lag in hazard prediction while always retaining the driver's ultimate control over the vehicle. This effectively solves the risk control problem when the driver's intention is ambiguous, achieving efficient dynamic matching of human-machine driving weights.

[0045] Regarding collision risk levels, in some preferred embodiments, a "high" collision risk level refers to a critical state where the decision control module predicts and determines that a substantial side collision is imminent, based on real-time vehicle kinematic parameters obtained by the environmental perception module. The specific criteria are: when the predicted time to collision (TTC) is less than 2.5 seconds and the longitudinal overlap between the vehicle and the potentially at-risk vehicle is greater than 30%, the collision risk level is defined as high. A TTC less than 2.5 seconds means that without timely intervention, the collision will occur outside the driver's typical physiological reaction time window, requiring immediate system intervention to gain avoidance time. A longitudinal overlap greater than 30% indicates a substantial physical intrusion area between the two vehicles on their respective paths, excluding the natural avoidance space that might arise from vehicle misalignment, indicating an extremely high and irreversible probability of a scrape or collision. This definition provides an objective and quantifiable trigger for the system to activate its highest-level safety protection mechanism, ensuring that the system only takes strong intervention measures under extreme conditions that truly endanger driving safety, balancing driving safety with the robustness of the control logic.

[0046] The above technical solution introduces a confidence-based control allocation strategy to achieve human-machine collaborative driving. This design quantifies the driver's operational intentions into confidence values ​​and divides them into three levels: full driver control, system-assisted control, and full system takeover. A clear and smooth human-machine responsibility division system is established. When the driver's intentions are clear and confident, the system fully respects the driver's leadership. When the driver's intentions are ambiguous or the environment is extremely dangerous, the system decisively takes over to avoid accidents. This effectively solves the long-standing problem in assisted driving systems of when to control the vehicle and when to control it, and significantly improves the system's acceptability and safety.

[0047] In some preferred embodiments, when a driver correction operation is detected during a period when the vehicle is completely under the control of the decision control module, an S-shaped smooth transition curve is used to return control to the driver within a preset time, while controlling the vehicle's acceleration change rate to not exceed a set change rate threshold. Among them, the S-shaped smooth transition curve is a time-domain control quantity change curve generated based on the Sigmoid function, which is used to realize the continuous stepless switching of control authority or control command; the slope of the curve approaches zero in the initial and final stages, and monotonically increases in the middle stage, so that the output acceleration, deceleration or steering torque exhibits a smooth S-shaped change.

[0048] In some preferred embodiments, the handover of control to the driver within a preset time period specifically refers to a gradual control quantity decay process based on an S-shaped smooth transition curve, rather than an instantaneous cutoff of system control commands, i.e., a gradual return. Specifically, when the decision control module determines that it needs to exit the takeover state, the system does not immediately stop outputting braking or acceleration commands. Instead, within a preset time window of 100ms to 150ms, it continuously and steplessly adjusts the acceleration or deceleration output by the system according to a time-domain control quantity change curve generated based on the Sigmoid function: at the initial stage of the handover, the system control quantity maintains its original strength to ensure that the vehicle's dynamics do not change abruptly; subsequently, in the intermediate transition stage, the system control quantity gradually and linearly decays with the monotonically increasing slope of the S-shaped curve, strictly limiting the rate of change of the vehicle's acceleration to 2.5m / s². 3 Within a certain range, until the end of the handover phase, the system control input smoothly returns to zero, and the vehicle state is entirely determined by the driver's input. This gradual return mechanism ensures the continuity of the vehicle's longitudinal force at the moment of control handover, completely eliminating the risk of vehicle jerking, occupant panic, or instability caused by sudden changes in tire adhesion that may result from a step-like switch, thus achieving a seamless handover of human-machine control.

[0049] The aforementioned technical solution further refines the smoothness control method during the handover of control, achieving seamless control switching through an S-shaped smooth transition curve. Recognizing that abrupt control switching can pose safety hazards, a sigmoid function is used to generate the trajectory of control variable changes, ensuring that the rate of acceleration change approaches zero during the start and end phases. This completely eliminates the abruptness of system intervention or withdrawal, strictly limiting the vehicle's acceleration change rate to 2.5 m / s². 3 Within this range, not only is the comfort of the occupants significantly improved, but more importantly, the dynamic stability of the vehicle is ensured at the moment of control transfer, preventing secondary accidents caused by sudden changes in control parameters, and achieving a perfect balance between safety and comfort.

[0050] In some preferred embodiments, to ensure the system's timeliness in emergency situations, the present invention strictly limits the end-to-end latency of the control process. From the environmental perception module acquiring data, the decision control module running algorithms to generate instructions, to the execution module responding with actions, the end-to-end latency of the entire perception, recognition, decision-making, and execution process is controlled within 100ms. The decision control module employs an automotive-grade multi-core processor, performing cyclic calculations at a frequency of 50Hz, ensuring the efficiency of solving complex algorithms in a very short time. This low-latency characteristic enables the system to intervene instantly before the driver perceives the risk (e.g., TTC < 2.5s), providing valuable buffer time for hazard avoidance maneuvers.

[0051] Example 2 An overtaking control method, comprising: Based on the driver's operation data, the driver's operation status is determined. Based on the surrounding vehicle environment information and the driver's operation status, it is determined whether the current situation is a scenario where a vehicle cuts in from the right or a scenario where the vehicle is overtaking, and the determination result is obtained. Execute the corresponding control strategy based on the scenario type in the discrimination result: If the scenario is determined to be a vehicle cutting in from the right and the driver is in an emergency operation state, a monitoring command will be output. If the scenario is determined to be a vehicle cutting in from the right and the driver is in a normal operating state and the predicted collision time is less than a preset threshold, then a braking command is output. If the scenario is determined to be an overtaking scenario and the driver begins to steer before the vehicle has fully overtaken the vehicle being overtaken, an alarm command will be output; if no driver correction operation is detected within a preset time, an acceleration command will be output. Execute corresponding controls based on monitoring commands, braking commands, alarm commands, or acceleration commands.

[0052] Example 3 A computer program product includes a computer program that, when executed by a processor, implements the steps of the method described in Embodiment 2.

[0053] The contents not described in detail in this specification are existing technologies known to those skilled in the art.

Claims

1. An overtaking control system, characterized in that, It includes: The scene discrimination module is used to determine the driver's current operating status based on the driver's operation data. Based on the vehicle environment information around the vehicle and the driver's current operating status, it determines whether the vehicle is currently in a scenario where a vehicle cuts in from the right or in a scenario where the vehicle is overtaking. The decision control module is used to output monitoring instructions when the vehicle is in a scenario where a vehicle cuts in from the right and the driver's current operation state is an emergency operation state. When the vehicle is in a scenario where a vehicle cuts in from the right and the driver is currently operating normally, and the predicted collision time is less than a preset threshold, the braking acceleration required for braking is obtained based on the initial front-to-back distance between the vehicle and the vehicle on the right, the vehicle's speed, the distance between the vehicle and the vehicle on the right, the steering angle of the vehicle on the right, and the speed of the vehicle on the right, and a braking command is output based on the braking acceleration required for braking. When the vehicle is in an overtaking scenario and the driver starts to steer before the vehicle has completely overtaken the overtaken vehicle, an alarm command is output. If no correction operation by the driver is detected within a preset time, the acceleration for the safe overtaking time is output based on the time required for safe overtaking, the initial front-to-back distance between the vehicle and the overtaken vehicle, and the speed of the overtaken vehicle. An acceleration command is then output based on the acceleration for the safe overtaking time.

2. The overtaking control system according to claim 1, characterized in that: When the decision control module determines that a vehicle is cutting in from the right and the driver is in a normal operating state, it calculates the predicted collision time using the following formula. ; ； in, This is the distance between this vehicle and the vehicle on its right. This represents the steering angle of the vehicle on the right. The speed of the vehicle on the right; When the calculated predicted collision time When the braking force is less than the preset threshold, the decision control module outputs a braking command and calculates the braking acceleration required for braking according to the following formula. : ； in, This represents the initial front-to-back distance between this vehicle and the vehicle on its right. This is the speed of the vehicle.

3. The overtaking control system according to claim 1, characterized in that: If no driver correction is detected within the specified time, the acceleration required for a safe overtaking maneuver is output. The time required for a safe overtaking maneuver is first calculated using the following formula. : ； in, For the speed of this vehicle, The distance between the two vehicles. The steering angle of the vehicle being overtaken; Calculate the acceleration required for safe overtaking based on the time required for safe overtaking. : ； in, This represents the initial front-to-back distance between this vehicle and the vehicle being overtaken. The speed of the vehicle being overtaken.

4. The overtaking control system according to claim 1, characterized in that: The specific method for determining whether a vehicle is in a situation where a vehicle is cutting in from the right or overtaking, based on the surrounding vehicle environment information and the driver's current operating status, is as follows: If the vehicle environment information around the vehicle shows that there is a vehicle on the right side of the vehicle that has turned on its turn signal to the left or is actually turning to the left, it is determined to be in a scenario where a vehicle cuts in from the right; if the vehicle environment information around the vehicle shows that the vehicle and the vehicle on the right overlap longitudinally, and the driver's current operation includes increasing the accelerator pedal opening or turning the steering wheel to the left, it is determined to be in a scenario where the vehicle is overtaking.

5. The overtaking control system according to claim 1, characterized in that: The monitoring command is used to increase the monitoring frequency of the left and rear areas defined by the vehicle to a set frequency, and to collect the relative speed and travel distance parameters of target vehicles in the area in real time. When a high-speed approaching vehicle with a relative speed greater than a set speed threshold and a distance less than a set distance threshold is detected in the left or rear area defined by the vehicle, an audible and visual alarm is triggered to prompt the driver to avoid the risk. If no abnormal approaching vehicle is detected in the left or rear area defined by the vehicle, no warning or control intervention is triggered.

6. The overtaking control system according to claim 1, characterized in that: When the decision control module determines that the vehicle is overtaking, if it detects that the driver starts to turn before the vehicle has completely overtaken the overtaken vehicle, it controls the execution module to output an audible and visual alarm command at a set frequency. If no driver correction is detected within the preset time period, an acceleration command will be output. If, during the output of the acceleration command, the increase in accelerator pedal opening is not greater than the set percentage or the steering wheel return angle is not greater than the set angle, it is determined that no driver correction operation has been detected.

7. The overtaking control system according to claim 1, characterized in that: The decision control module is also configured with a confidence-based control allocation strategy: When the confidence level of the driver's operating intention is higher than the first set confidence threshold, the driver has complete control. When the confidence level is between the second set confidence level threshold and the first set confidence level threshold, the vehicle auxiliary control mode is adopted; When the confidence level is lower than the second set confidence threshold and the collision risk level is high, the vehicle is completely taken over by the decision control module.

8. The overtaking control system according to claim 7, characterized in that: When the driver corrects his actions during a period when the vehicle is completely under the control of the decision control module, the control is returned to the driver within a preset time using an S-shaped smooth transition curve, while the vehicle's acceleration change rate is controlled to not exceed the set change rate threshold. Among them, the S-shaped smooth transition curve is a time-domain control quantity change curve generated based on the Sigmoid function. The slope of this curve approaches zero in the initial and final stages and monotonically increases in the middle stage, so that the output acceleration, deceleration or steering torque exhibits a smooth S-shaped change.

9. An overtaking control method, characterized in that, It includes: Based on the driver's operation data, the driver's operation status is determined. Based on the surrounding vehicle environment information and the driver's operation status, it is determined whether the current situation is a scenario where a vehicle cuts in from the right or a scenario where the vehicle is overtaking, and the determination result is obtained. Execute the corresponding control strategy based on the scenario type in the discrimination result: If the scenario is determined to be a vehicle cutting in from the right and the driver is in an emergency operation state, a monitoring command will be output. If the scenario is determined to be a vehicle cutting in from the right and the driver is in a normal operating state and the predicted collision time is less than a preset threshold, then a braking command is output. If the scenario is determined to be an overtaking scenario and the driver begins to steer before the vehicle has fully overtaken the vehicle being overtaken, an alarm command will be output; if no driver correction operation is detected within a preset time, an acceleration command will be output. Execute corresponding controls based on monitoring commands, braking commands, alarm commands, or acceleration commands.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method of claim 9.

Images