A roundabout and ramp work condition visual trigger multi-stage deceleration-turning coordination control method and system

Abstract

The application discloses a kind of visual trigger multi-stage deceleration-turning coordination control method and system of roundabout and ramp working condition, and it is related to vehicle intelligent control technical field.The method identifies road roundabout entrance or ramp feature in real time, when meeting trigger condition, automatically enter multi-stage deceleration process: first, carry out large amplitude emergency deceleration, then carry out smooth deceleration or speed maintenance, finally, restore speed after passing through scene.At the same time of each stage deceleration, according to the curvature of road and target track adjustment steering angle, realize the collaborative control of braking and steering.Meanwhile, system is based on vehicle state and environmental condition dynamic adjustment control parameter, to improve safety and comfort.The application can significantly enhance the stability and active safety performance of vehicle driving under complex road scene.

Description

Technical Field

[0001] This invention relates to the field of vehicle intelligent control technology, specifically to a visually triggered multi-stage deceleration-steering coordinated control method and system for roundabout and ramp conditions. Background Technology

[0002] In existing technologies, vehicle control for complex scenarios such as ramps or roundabouts primarily relies on traditional sensors or pre-defined strategies. For example, some solutions use cameras or lidar to detect ramp angles and sensors to measure the slope to pre-adjust vehicle speed; others employ multi-sensor fusion and fuzzy logic algorithms for environmental perception and path planning in roundabout scenarios. However, these solutions typically focus on environmental recognition and path decision-making, lacking specificity for emergency deceleration and steering control; while coordinated control of steering and braking is more commonly found in collision avoidance or racing technology. Relevant patents, such as the Chinese patent application CN111152776A which discloses a method and system for coordinated steering and braking control of an unmanned Formula One race car, point out that stability issues exist when using steering or braking alone during cornering, emphasizing the coupling characteristics of the steering and braking systems. Wang Qidong et al., in their paper "Research on Emergency Obstacle Avoidance of Intelligent Vehicles Based on Braking-Steering Coordinated Control" published in *Automotive Engineering*, Vol. 41, No. 4, 2019, proposed a braking-steering coordinated control system integrating road and vehicle states for lane change collision avoidance. However, this system focuses on lane change safety, which differs from the phased deceleration requirements in roundabout and ramp scenarios. Currently, there is a lack of a complete solution capable of achieving phased deceleration and coordinated steering control at roundabout entrances or ramp conditions based on visual triggering. Summary of the Invention

[0003] Technical Objective: To address the problems of coarse deceleration control, lack of coordination between longitudinal and lateral control, and insufficient adaptive capability in existing technologies for roundabout and ramp conditions, this invention discloses a visually triggered multi-stage deceleration-steering coordinated control method and system for roundabout and ramp conditions. By visually sensing and triggering condition recognition, a multi-stage target speed sequence is constructed, and the longitudinal deceleration reference value and target steering angle are jointly solved in each stage. With the help of an adaptive parameter adjustment module, the vehicle can pass safely, smoothly, and stably through roundabout and ramp conditions.

[0004] Technical solution: To achieve the above technical objectives, the present invention adopts the following technical solution:

[0005] A visually triggered multi-stage deceleration-steering coordinated control method for roundabout and ramp conditions, specifically including the following steps:

[0006] Step 1: Collect images of the road ahead of the vehicle and vehicle operating status information. The vehicle operating status information includes at least vehicle speed, longitudinal acceleration, lateral acceleration, and yaw rate. Preprocess the road images and extract lane lines, road edges, and traffic sign features.

[0007] Step 2: Calculate the confidence level of the roundabout and the confidence level of the ramp based on the road features, and estimate the road slope angle. When the confidence level of the roundabout reaches the first preset threshold or the absolute value of the road slope angle reaches the second preset threshold, the current working condition is determined as the target working condition and multi-stage deceleration control is triggered.

[0008] Step 3: Based on the target operating condition type, the vehicle's current speed, and the road curvature, construct a target speed sequence consisting of at least three stages and corresponding stage lengths. The target speed sequence includes at least an entry deceleration stage, a speed maintenance or slow deceleration stage within the operating condition, and an exit / exit recovery stage. The target speeds of the entry deceleration stage and the stages within the operating condition satisfy a monotonically non-increasing constraint. The target speed of the exit / exit recovery stage is set to gradually recover to the cruising speed, and the longitudinal acceleration and lateral acceleration do not exceed preset safety limits.

[0009] Step 4: In each control cycle, based on the target speed and road curvature of the current stage, jointly solve the longitudinal deceleration reference value and the target steering angle, apply braking control to the braking actuator and steering control to the steering actuator, so that the vehicle speed gradually decreases according to the target speed sequence during the entry deceleration stage and the working condition stage, and gradually recovers to the cruising speed and turns along the planned path during the exit / exit recovery stage, and exits the multi-stage deceleration control after passing through the roundabout or ramp;

[0010] Step 5: Input the vehicle operating status information and target operating condition characteristics into the adaptive parameter adjustment module, output the gain adjustment coefficients of the longitudinal controller and the lateral controller, and dynamically adjust the proportional, integral and derivative terms of the longitudinal deceleration controller and the steering controller to improve control stability and comfort under different roundabout and slope conditions.

[0011] Preferably, when the target operating condition is a roundabout operating condition, the multi-stage target velocity sequence includes:

[0012] The first stage is the entrance deceleration stage. The target speed is calculated based on the vehicle's initial speed and the radius of the roundabout. It is used to reduce the vehicle speed to a speed that does not exceed the lateral acceleration safety limit before entering the roundabout.

[0013] The second stage is a period of uniform or slow deceleration within the roundabout. The target speed is determined based on the upper limit of the allowable lateral acceleration obtained from the roundabout radius and the road adhesion coefficient.

[0014] In the third stage, the exit recovery phase, the target speed is set to gradually recover to cruising speed based on the length of the straight section at the roundabout exit.

[0015] Preferably, when the target working condition is a ramp working condition, the multi-stage target speed sequence includes:

[0016] The ramp entrance pre-deceleration phase is used to reduce the vehicle speed to a safe ramp passage speed before entering the ramp;

[0017] The speed maintenance phase within the ramp is used to control the longitudinal acceleration of the vehicle within a preset range by combining the road slope angle and vehicle load, thereby achieving an approximately constant speed or slow speed change.

[0018] The ramp exit recovery phase is used to gradually restore the cruising speed after the ramp ends. For uphill conditions, the driving torque is increased to compensate for the gravity component. For downhill conditions, the longitudinal deceleration is controlled to not exceed the preset downhill safe deceleration limit by coordinating engine braking and brake actuators.

[0019] Preferably, step 1 further includes: acquiring wheel speed sensor signals and acceleration and angular velocity signals from the inertial measurement unit, and performing fusion estimation of vehicle speed, yaw rate and sideslip angle through a state estimation model. The state estimation model is an extended Kalman filter model or an unscented Kalman filter model. The fusion estimation results are used for the working condition determination in step 2 and the calculation of longitudinal and lateral control quantities in step 4.

[0020] Preferably, in step 4, the longitudinal deceleration reference value and the target steering angle are calculated by optimizing a cost function, wherein the cost function is:

[0021] ,

[0022] in, For lateral acceleration, To ensure a safe upper limit for lateral acceleration, Let be the rate of change of longitudinal acceleration, and be the first derivative of longitudinal acceleration with respect to time. For the limit of the rate of change of longitudinal acceleration, This represents the change in steering angle between adjacent control cycles. The maximum allowable change in steering angle. The weighting coefficients are used to obtain the longitudinal deceleration reference value and the target steering angle by minimizing the cost function.

[0023] Preferably, the longitudinal deceleration controller employs an incremental control law, and its longitudinal control output increment is:

[0024] ,

[0025] in, Let be the vehicle speed error in the k-th control cycle, and be the difference between the current vehicle speed and the target vehicle speed for the current stage. These are the proportional coefficient, integral coefficient, and derivative coefficient for longitudinal control, respectively.

[0026] The steering controller uses an incremental control law, and its steering control output increment is:

[0027] ,

[0028] in, Let be the steering error in the k-th control cycle, and be the difference between the target steering angle and the actual steering angle. These are the proportional coefficient, integral coefficient, and derivative coefficient for steering control, respectively.

[0029] Preferably, the adaptive parameter adjustment module in step 5 includes a fuzzy rule base and a trainable nonlinear mapping unit. The adaptive parameter adjustment module takes vehicle speed, road slope angle, roundabout radius, lateral acceleration, and longitudinal acceleration change rate as input, and outputs the gain adjustment coefficients of the longitudinal controller and the steering controller. The proportional coefficient, integral coefficient, and derivative coefficient of the longitudinal controller are equal to the product of the reference coefficient and the corresponding adjustment coefficient, respectively. The proportional coefficient, integral coefficient, and derivative coefficient of the steering controller are equal to the product of the reference coefficient and the corresponding adjustment coefficient, respectively, so as to achieve adaptive control for different roundabout and slope conditions.

[0030] A vision-triggered multi-stage deceleration-steering coordination control system for roundabout and ramp conditions, used to implement the vision-triggered multi-stage deceleration-steering coordination control method for roundabout and ramp conditions as described above, comprising:

[0031] The visual perception module is used to acquire images of the road in front of the vehicle and output lane lines, road edges, and traffic sign features.

[0032] The state perception module is used to collect signals from the wheel speed sensor and the inertial measurement unit, and output vehicle speed, longitudinal acceleration, lateral acceleration and yaw rate;

[0033] The working condition recognition module, connected to the visual perception module and the state perception module, is used to determine whether the current working condition is a roundabout or a ramp based on road features and road slope angle, and outputs a multi-stage deceleration trigger signal when the working condition meets the triggering conditions.

[0034] A multi-stage target generation module, connected to the working condition recognition module, is used to generate a target speed sequence consisting of at least three stages and the corresponding stage length based on the working condition type, the vehicle's current speed, and the road curvature when a multi-stage deceleration trigger signal is received.

[0035] The collaborative control module, connected to the multi-stage target generation module and the state perception module, is used to jointly solve the longitudinal deceleration reference value and the target steering angle based on the target speed and road curvature of the current stage in each control cycle, and output control commands to the braking actuator and steering actuator to realize multi-stage deceleration and steering coordinated control.

[0036] The adaptive parameter adjustment module, connected to the state perception module and the working condition recognition module, is used to adjust the gain parameters of the longitudinal controller and the lateral controller in the cooperative control module in real time according to the vehicle's operating status and target working condition characteristics.

[0037] Preferably, the state perception module includes a state estimation unit for fusing signals from wheel speed sensors and inertial measurement units. The state estimation unit estimates vehicle speed, yaw rate, and sideslip angle using an extended Kalman filter model or an unscented Kalman filter model. The cooperative control module constructs a cost function based on the estimation results and determines the longitudinal deceleration reference value and the target steering angle by minimizing the cost function. The cost function is used to constrain the lateral acceleration, longitudinal acceleration rate of change, and steering angle change to not exceed preset safety limits.

[0038] Preferably, the collaborative control module and the adaptive parameter adjustment module are integrated into the vehicle electronic control unit. The vehicle electronic control unit is connected to the vision perception module, the status perception module, the braking actuator, and the steering actuator through the vehicle communication bus. The vehicle electronic control unit realizes the functions of working condition recognition, multi-stage target generation, collaborative control, and adaptive parameter adjustment through software programs.

[0039] Beneficial Effects: The visually triggered multi-stage deceleration-steering coordinated control method and system for roundabout and ramp conditions provided by this invention has the following beneficial effects:

[0040] 1. This invention uses visually recognized roundabout and ramp conditions as triggering conditions, rather than relying solely on single state variables such as vehicle speed or yaw rate; it combines the type of working condition, road geometry, and vehicle status to construct a multi-stage target speed sequence, realizing an overall strategy of rapid deceleration at the entrance, smooth control within the working condition, and speed recovery at the exit, significantly improving throughput efficiency and ride comfort while ensuring safety.

[0041] 2. This invention combines deceleration and steering by using a cost function that includes lateral acceleration and the rate of change of longitudinal acceleration. It simultaneously solves for the longitudinal deceleration reference value and the target steering angle in each control cycle, effectively limiting roll and pitch. This avoids drastic changes in vehicle attitude caused by relying solely on longitudinal deceleration or lateral correction, and improves the overall stability of the vehicle in roundabout and ramp conditions.

[0042] 3. This invention introduces an adaptive parameter adjustment module, which takes features such as vehicle speed, road slope angle, roundabout radius, lateral acceleration and longitudinal acceleration change rate as inputs to dynamically correct the gain of the longitudinal controller and the lateral controller, so that the control parameters can be automatically adjusted for different tire adhesion conditions, different road slopes and different driving needs, thereby improving the robustness and generalization ability of the system under various working conditions. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below.

[0044] Figure 1 This is a block diagram of the system structure of the present invention;

[0045] Figure 2 This is a flowchart of the control method of the present invention;

[0046] Figure 3 This is a schematic diagram of multi-stage target speed and vehicle trajectory under roundabout conditions.

[0047] Figure 4 This is a schematic diagram illustrating the relationship between target speed and time in multiple stages under slope conditions.

[0048] Figure 5 This is a structural diagram of the adaptive parameter adjustment module. Detailed Implementation

[0049] The present invention will now be described more clearly and completely by way of a preferred embodiment in conjunction with the accompanying drawings, but this does not limit the invention to the scope of the described embodiment.

[0050] like Figure 1 As shown in the figure, this embodiment provides a vision-triggered multi-stage deceleration-steering coordination control system for roundabout and ramp conditions, which is installed on a vehicle equipped with an electronically controlled braking actuator and an electronically controlled steering actuator. The system includes: a vision perception module, a state perception module, a working condition recognition module, a multi-stage target generation module, a coordination control module, an adaptive parameter adjustment module, a braking actuator, a steering actuator, and an on-board electronic control unit (ECU).

[0051] The visual perception module includes a forward-facing camera and an image processing unit. The forward-facing camera is installed at the front of the vehicle to capture images of the road ahead. The image processing unit is used to perform algorithms such as denoising, distortion correction, perspective transformation, edge detection, lane line fitting, and traffic sign recognition, and outputs the recognition results of lane line parameters, road edge information, and roundabout-related traffic signs.

[0052] The state perception module includes wheel speed sensors, inertial measurement units, and lateral acceleration sensors, which are used to collect signals such as wheel speed, vehicle longitudinal acceleration, lateral acceleration, and angular velocity. The state perception module is equipped with a state estimation unit, which uses an extended Kalman filter (EKF) model or an unscented Kalman filter (UKF) model to perform fusion estimation of vehicle speed, yaw rate, and sideslip angle.

[0053] The working condition recognition module communicates with the visual perception module and the state perception module to calculate the roundabout confidence and the ramp confidence based on the lane curvature change, traffic sign recognition results and estimated road slope angle, and compares the confidence with preset thresholds. When the roundabout confidence is ≥ the first threshold, a roundabout working condition trigger signal is output, and when the absolute value of the slope angle is ≥ the second threshold, a ramp working condition trigger signal is output.

[0054] The multi-stage target generation module is connected to the working condition recognition module and the state perception module. When it receives a roundabout working condition trigger signal or a ramp working condition trigger signal, it generates a target speed sequence consisting of at least three stages based on the current vehicle speed, road curvature, road slope angle, and vehicle dynamics constraints. and the corresponding stage length The target velocity sequence is then sent to the cooperative control module.

[0055] The collaborative control module communicates with the multi-stage target generation module, state perception module, adaptive parameter adjustment module, braking actuator, and steering actuator. Internally, the collaborative control module includes a longitudinal control submodule and a lateral control submodule.

[0056] The longitudinal control submodule is used to calculate the braking control quantity based on the vehicle speed error and the longitudinal acceleration change rate, and output the braking hydraulic command or the electric braking torque command to the braking actuator.

[0057] The lateral control submodule is used to calculate the target steering angle based on the path curvature and steering angle error, and output steering motor torque command or steering angle command to the steering actuator.

[0058] The collaborative control module has a cost function calculation unit inside, which is used to construct and minimize the cost function that includes the lateral acceleration, the rate of change of longitudinal acceleration and the change of steering angle, so as to obtain the longitudinal deceleration reference value and the target steering angle.

[0059] The adaptive parameter adjustment module is connected to the state perception module and the working condition recognition module. The adaptive parameter adjustment module takes vehicle speed, road slope angle, roundabout radius, lateral acceleration and longitudinal acceleration change rate as inputs, and calculates the gain adjustment coefficients of the longitudinal controller and the lateral controller through fuzzy rules or neural network mapping. It then adjusts the proportional, integral and derivative gains of the longitudinal controller and the lateral controller inside the cooperative control module.

[0060] The vehicle electronic control unit (ECU) can provide computing resources for the above-mentioned functional modules, or the above-mentioned modules can be integrated into the ECU's software.

[0061] See Figure 2 The visually triggered multi-stage deceleration-steering coordinated control method for roundabout and ramp conditions provided in this embodiment includes the following steps S101 to S108:

[0062] S101: Environmental and state perception.

[0063] The visual perception module acquires images of the road ahead at a fixed frame rate, performs denoising and distortion correction on the images, transforms the images into a bird's-eye view through perspective transformation, extracts lane lines and road edges, and identifies traffic signs related to roundabouts. The state perception module acquires wheel speed, longitudinal acceleration, lateral acceleration, and angular velocity signals, and outputs vehicle speed v and yaw rate through the state estimation unit. and sideslip angle .

[0064] S102: Operating condition identification.

[0065] The driving condition recognition module calculates the roundabout confidence level based on lane curvature changes and traffic sign recognition results. The confidence level of a roundabout can be obtained by weighted summation of multiple features, for example:

[0066]

[0067] in, Confidence level for roundabout sign detection Confidence level for matching lane curvature to roundabout geometry. Confidence scores are assigned to road topology. These are the weighting coefficients.

[0068] Meanwhile, the working condition identification module estimates the road slope angle based on the road's longitudinal profile. ,when When the value exceeds the first threshold, it is determined to be a roundabout condition; when When the value exceeds the second threshold, it is determined to be a ramp condition; when both conditions are met at the same time, it can be controlled as a roundabout condition first, or the priority can be set according to application requirements.

[0069] S103: Multi-stage target velocity planning.

[0070] When the situation is determined to be a roundabout, the multi-stage target generation module generates targets based on the current vehicle speed. The estimated radius R of the roundabout and the maximum permissible lateral acceleration. Calculate the target speed limit for passing through the roundabout. :

[0071]

[0072] The deceleration process is divided into three stages:

[0073] Phase 1: Entry deceleration phase, from current vehicle speed Decelerate to ,in Not greater than ;

[0074] Phase Two: Within the roundabout, maintain a constant or gradual deceleration speed, increasing the vehicle speed from... Adjust to , near Or slightly lower than that value;

[0075] Phase Three: Export Recovery Phase, from Gradually return to cruising speed .

[0076] Duration of each stage Determined based on road length and comfort constraints, for example, ensuring that longitudinal acceleration does not exceed a preset limit. At the same time, the rate of change of longitudinal acceleration is limited to no more than .

[0077] like Figure 3 The diagram illustrates the multi-stage target speed and vehicle trajectory under roundabout conditions according to the present invention. The inner circle represents the central island of the roundabout, and the outer circle represents the outer edge of the roundabout's driving lanes. The vehicle's trajectory lies within the lane area between the inner and outer circles. The vehicle enters the roundabout from the entrance direction below the diagram along a straight path. On the straight section before entering the roundabout, the trajectory corresponds to the first stage, the entrance deceleration stage, where the vehicle speed decreases from v0 to v1. The vehicle then travels counter-clockwise around the roundabout lanes, from the south side of the roundabout through the west side to the north side. This trajectory corresponds to the second stage, the inner roundabout stage, where the vehicle speed is adjusted to a safer v2. The counter-clockwise direction is indicated by arrows on the trajectory in the diagram. When the vehicle reaches the exit position on the north side of the roundabout, it travels upwards along a straight path to leave the roundabout, corresponding to the third stage, the exit recovery stage, where the vehicle speed gradually recovers from v2 to v1. cruise .

[0078] When the condition is determined to be a slope condition, the multi-stage target generation module determines the target based on the slope angle. Current vehicle speed Similar to the expected speed range, a target speed sequence is planned for the entry pre-deceleration phase, the speed maintenance phase within the ramp, and the ramp exit recovery phase. For uphill driving, a target speed slightly lower than the cruising speed on flat roads can be maintained within the ramp; for downhill driving, the vehicle speed is limited to a safe downhill speed.

[0079] like Figure 4 The diagram illustrates the multi-stage target speed versus time relationship under slope conditions according to the present invention. The horizontal axis represents time t, and the vertical axis represents vehicle speed v. A segmented curve represents the process of vehicle speed changing with time under slope conditions. The initial stage of the curve corresponds to the pre-deceleration stage at the entrance, where the vehicle speed smoothly decreases from the entrance speed v0 to the pre-deceleration target speed v1. The middle stage corresponds to the speed maintenance stage within the slope, where the vehicle speed is maintained near the target speed v2 within the slope, ensuring longitudinal safety and comfort. The final stage corresponds to the slope exit recovery stage, where the vehicle speed gradually increases from v2 to the recovered cruising speed v. cruise .

[0080] S104: Construct the cost function.

[0081] The collaborative control module determines the vehicle speed (v) and target vehicle speed in each control cycle. Road curvature Construct the cost function using the state estimator:

[0082]

[0083] in This is a cost function used to measure the overall cost of the current control action in terms of lateral safety, longitudinal comfort, and steering comfort. It is lateral acceleration; This is the upper limit for lateral acceleration; This is the rate of change of longitudinal acceleration, i.e., the first derivative of longitudinal acceleration with respect to time; This is the limit value for the rate of change of longitudinal acceleration; This represents the change in steering angle between adjacent control cycles; This is the maximum permissible change in steering angle; These are weighting coefficients used to balance roll, pitch, and ride comfort.

[0084] The cooperative control module uses numerical optimization methods to find the longitudinal deceleration reference value and target steering angle that minimize the cost function J within the feasible longitudinal deceleration and steering angle space.

[0085] S105: Longitudinal control.

[0086] Define the vehicle speed error in the k-th control cycle as: The longitudinal control submodule uses an incremental control law to calculate the braking control output increment:

[0087]

[0088] in This represents the longitudinal control output increment for the k-th control cycle, corresponding to the increase or decrease in braking command (e.g., the increment in braking pressure or braking torque) required in this cycle. k is the sequence number of the discrete control cycle and is a non-negative integer. The actual vehicle speed sampled in the k-th control cycle. The target vehicle speed corresponding to the kth control cycle is generated by the multi-stage target generation module. , This represents the vehicle speed error from the previous cycle and the two cycles prior, used to characterize the trend of error change. The proportional, integral, and derivative reference coefficients for longitudinal control are adjusted by the adaptive parameter adjustment module to obtain an effective gain, which is used to control the braking force or braking torque applied by the braking actuator, so that the vehicle speed gradually follows the target speed sequence.

[0089] S106: Lateral control.

[0090] Define the steering error in the k-th control cycle as: ,in The target steering angle for the k-th control cycle is obtained based on the road curvature and optimization results. This represents the actual steering angle of the vehicle in the k-th control cycle. The lateral control submodule employs an incremental control law:

[0091]

[0092] in This represents the steering control output increment in the k-th control cycle, corresponding to the command increment of the steering actuator (such as a motor). , This represents the steering error of the previous cycle and the two cycles prior. The proportional, integral, and derivative reference coefficients for steering control are adjusted by the adaptive parameter adjustment module to obtain an effective gain, which is used to control the actuation amount of the steering actuator so that the vehicle travels according to the desired curvature.

[0093] S107: Adaptive parameter adjustment.

[0094] like Figure 5 As shown, the adaptive parameter adjustment module uses vehicle speed v and road slope angle... Roundabout radius R, lateral acceleration and longitudinal acceleration rate of change Taking these parameters as input, the longitudinal and lateral gain adjustment coefficients are calculated using fuzzy rules and a nonlinear mapping unit. For example, fuzzy rules might include increasing the steering ratio gain and decreasing the rate of change of the longitudinal braking gain when the vehicle speed is high, the roundabout radius is small, and the lateral acceleration is close to its upper limit. The output of the adaptive parameter adjustment module serves as a multiplicative factor for each gain of the longitudinal and lateral controllers, enabling automatic gain adjustment under different operating conditions.

[0095] In one alternative implementation, the adaptive parameter adjustment module is implemented using a three-layer feedforward neural network. The input layer contains the aforementioned multiple operating condition features, and the output layer consists of the gain adjustment coefficients of the longitudinal and lateral controllers. The network can be trained offline using simulation data and real vehicle test data, and fine-tuned online during operation.

[0096] S108: Exit multi-stage control.

[0097] After the vehicle completes the roundabout or ramp passage, the condition recognition module detects a decrease in roundabout confidence or a return of the slope angle to the normal range, and switches the control mode from multi-stage deceleration-steering coordinated control mode back to conventional cruise control or other driving control modes.

[0098] In a typical roundabout scenario, assuming the roundabout radius is 30m, what is the maximum permissible lateral acceleration? 3m / s 2 The vehicle approaches the roundabout entrance at a speed of 50 km / h, and the condition recognition module determines it to be a roundabout condition based on traffic sign recognition and lane line geometry features.

[0099] Based on the above parameters, the multi-stage target generation module calculates that the safe speed limit for the roundabout is approximately 33 km / h, and divides the deceleration process into three stages: the entry deceleration stage, the roundabout stage, and the exit recovery stage.

[0100] Entry deceleration phase: Reduce vehicle speed from 50km / h to 33km / h;

[0101] Within the roundabout: Maintain a target speed of 30–33 km / h on the roundabout curves;

[0102] Export recovery phase: The speed at the roundabout exit will gradually recover from 33 km / h to 50 km / h.

[0103] The collaborative control module optimizes the system through a cost function, ensuring that the lateral acceleration does not exceed 3 m / s². 2 Under the premise of limiting the rate of change of longitudinal acceleration, the vehicle deceleration and steering processes are made smoother. Measured or simulated results show that, compared to a single deceleration strategy without multi-stage control, the proposed solution reduces the fluctuation of the vehicle's maximum yaw rate and lowers the peak passenger acceleration under the same roundabout conditions, thus improving passenger comfort.

[0104] In a typical ramp scenario, assuming a slope angle of −8° (downhill), a vehicle approaches the ramp entrance at 40 km / h. The condition identification module determines the condition as ramp based on the road longitudinal profile and inertial measurement unit signals.

[0105] The multi-stage target generation module sets the downhill safe speed limit to 35km / h and divides the slope control into an entrance pre-deceleration stage, a speed maintenance stage within the slope, and a slope exit recovery stage. It also sets limits on longitudinal acceleration and longitudinal acceleration change rate to avoid brake fade problems caused by continuous high braking during long downhill sections.

[0106] The collaborative control module coordinates the braking force applied by the engine braking and the braking actuator within the slope to maintain the vehicle speed within the range of 30-35 km / h; the adaptive parameter adjustment module adjusts the longitudinal control gain according to the slope angle and vehicle load status, so as to increase the share of engine braking force under heavy load and reduce the thermal load of the braking system.

[0107] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A roundabout and ramp work condition visual trigger multi-stage deceleration-steering coordination control method, characterized in that, Specifically comprising the following steps: Step 1, collecting the road image in front of the vehicle and the vehicle running state information, the vehicle running state information at least including vehicle speed, longitudinal acceleration, lateral acceleration and yaw rate, pre-processing the road image and extracting lane line, road edge and traffic sign features; Step 2, calculating roundabout confidence and ramp confidence based on road features, and estimating road slope angle, when the roundabout confidence reaches a first preset threshold or the absolute value of the road slope angle reaches a second preset threshold, determining the current working condition as the target working condition and triggering multi-stage deceleration control; Step 3, according to the target working condition type, the current vehicle speed and the road curvature, constructing a target speed sequence composed of at least three stages and the corresponding stage length, the target speed sequence at least including an entrance deceleration stage, a working condition inside speed maintaining or slow deceleration stage and an exit / exit recovery stage; Wherein the target speed of the entrance deceleration stage and the working condition inside stage satisfies the monotone non-increasing constraint, and the target speed of the exit / exit recovery stage is set to gradually recover to the cruising speed, and the longitudinal acceleration and lateral acceleration do not exceed the preset safety limit; Step 4, in each control period, jointly solving the longitudinal deceleration reference value and the target steering angle according to the target speed of the current stage and the road curvature, applying brake control to the brake actuator and applying steering control to the steering actuator, so that the vehicle speed gradually decreases in the entrance deceleration stage and the working condition inside speed maintaining stage according to the target speed sequence, and gradually recovers to the cruising speed in the exit / exit recovery stage and steers along the planned path, and exits the multi-stage deceleration control after passing through the roundabout or the ramp; Step 5, inputting the vehicle running state information and the target working condition features into the adaptive parameter adjustment module, outputting the gain adjustment coefficient of the longitudinal controller and the lateral controller, and dynamically adjusting the proportional term, the integral term and the differential term of the longitudinal deceleration controller and the steering controller, so as to improve the control stability and comfort in different roundabout and ramp working conditions.

2. The visual trigger multi-stage deceleration-steering coordination control method for roundabout and ramp conditions according to claim 1, characterized in that, When the target working condition is a roundabout working condition, the multi-stage target speed sequence includes: The first stage entrance deceleration stage, the target speed of which is calculated according to the initial vehicle speed and the roundabout radius, used to reduce the vehicle speed to a speed not exceeding the lateral acceleration safety limit before entering the roundabout; The second stage, the uniform speed or slow deceleration stage in the roundabout, the target speed of which is determined according to the upper limit of the allowed lateral acceleration obtained from the roundabout radius and the road adhesion coefficient; The third stage, the exit recovery stage, the target speed of which is set to gradually recover to the cruising speed according to the length of the straight line segment at the exit of the roundabout. 3.The visual trigger multi-stage deceleration-steering coordination control method for roundabout and ramp conditions according to claim 1, wherein, When the target working condition is a ramp working condition, the multi-stage target speed sequence includes: The ramp entrance pre-deceleration stage, used to reduce the vehicle speed to the safe passing speed of the ramp before entering the ramp; The ramp inside speed maintaining stage, used to control the vehicle longitudinal acceleration within a preset range in combination with the road slope angle and the vehicle load, to achieve approximately constant speed or slow speed change; The ramp exit recovery phase is used to gradually recover to the cruising speed after the end of the ramp, wherein the gravity component is compensated by increasing the driving torque for uphill working conditions, and the longitudinal deceleration is controlled by coordinating the engine braking and the brake actuator for downhill working conditions, so as to not exceed the preset upper limit of the downhill safety deceleration.

4. The visual trigger multi-stage deceleration-steering coordination control method for roundabout and ramp conditions according to claim 1, characterized in that, Step 1 further comprises: collecting wheel speed sensor signals and acceleration and angular velocity signals of an inertial measurement unit, and performing fused estimation of vehicle speed, yaw rate and side slip angle through a state estimation model, which is an extended Kalman filter model or an unscented Kalman filter model, and the fused estimation result is used for working condition determination in step 2 and longitudinal and lateral control quantity calculation in step 4.

5. The visual trigger multi-stage deceleration-steering coordination control method for roundabout and ramp conditions according to claim 1, characterized in that, The longitudinal deceleration reference value and the target steering angle in step 4 are calculated by optimizing a cost function, and the cost function is ， wherein is the lateral acceleration, is the lateral acceleration safety upper limit, is the longitudinal acceleration rate of change, is the first derivative of the longitudinal acceleration with respect to time, is the longitudinal acceleration rate of change limit, is the adjacent control period steering angle change amount, is the maximum allowable steering angle change amount, is the weight coefficient, the longitudinal deceleration reference value and the target steering angle are obtained by minimizing the cost function.

6. The visual trigger multi-stage deceleration-steering coordination control method for roundabout and ramp conditions of claim 1, wherein, The longitudinal deceleration controller adopts an incremental control law, and the longitudinal control output increment is ， wherein, is the speed error of the kth control cycle, is the difference between the current speed and the target speed of the current stage, are the proportional coefficient, integral coefficient and differential coefficient of the longitudinal control, respectively; The steering controller adopts an incremental control law, and the steering control output increment is ， wherein, is the steering error for the kth control cycle, is the difference between the target steering angle and the actual steering angle, are the proportional coefficient, the integral coefficient and the differential coefficient of the steering control, respectively.

7. The visual trigger multi-stage deceleration-steering coordination control method for roundabout and ramp conditions of claim 1, wherein, The adaptive parameter adjustment module in step 5 includes a fuzzy rule base and a trainable nonlinear mapping unit, and the adaptive parameter adjustment module takes vehicle speed, road slope angle, roundabout radius, lateral acceleration and longitudinal acceleration change rate as inputs, and outputs gain adjustment coefficients of the longitudinal controller and the steering controller, and the proportional coefficient, integral coefficient and differential coefficient of the longitudinal controller are equal to the product of the baseline coefficient and the corresponding adjustment coefficient, and the proportional coefficient, integral coefficient and differential coefficient of the steering controller are equal to the product of the baseline coefficient and the corresponding adjustment coefficient, so as to realize adaptive control for different roundabout and ramp working conditions.

8. A roundabout and ramp work condition visual trigger multi-stage deceleration-steering coordination control system, characterized in that, A visual trigger multi-stage deceleration-steering coordination control method for roundabout and ramp working conditions is implemented, and the method comprises the steps of: A visual perception module is used to collect road images in front of the vehicle and output lane lines, road edges and traffic sign features; A state perception module is used to collect signals of wheel speed sensors and inertial measurement units, and output vehicle speed, longitudinal acceleration, lateral acceleration and yaw rate; A working condition identification module is connected with the visual perception module and the state perception module, and is used to determine whether the current working condition is a roundabout working condition or a ramp working condition based on road features and road slope angles, and output a multi-stage deceleration trigger signal when the working condition meets the trigger condition; A multi-stage target generation module is connected with the working condition identification module, and is used to generate a target speed sequence composed of at least three stages and corresponding stage lengths according to the working condition type, the current vehicle speed and the road curvature when receiving the multi-stage deceleration trigger signal; A cooperative control module is connected with the multi-stage target generation module and the state perception module, and is used to jointly solve a longitudinal deceleration reference value and a target steering angle according to the target speed of the stage to which the current position belongs and the road curvature in each control period, and output control instructions to brake actuators and steering actuators, so as to realize multi-stage deceleration and steering coordination control; An adaptive parameter adjustment module is connected with the state perception module and the working condition identification module, and is used to adjust gain parameters of a longitudinal controller and a lateral controller in the cooperative control module in real time according to vehicle operating states and target working condition features. The ramp exit recovery phase is used to gradually recover to the cruising speed after the end of the ramp, wherein the gravity component is compensated by increasing the driving torque for uphill working conditions, and the longitudinal deceleration is controlled by coordinating the engine braking and the brake actuator for downhill working conditions, so as to not exceed the preset upper limit of the downhill safety deceleration. Step 1 further comprises: collecting wheel speed sensor signals and acceleration and angular velocity signals of an inertial measurement unit, and performing fused estimation of vehicle speed, yaw rate and side slip angle through a state estimation model, which is an extended Kalman filter model or an unscented Kalman filter model, and the fused estimation result is used for working condition determination in step 2 and longitudinal and lateral control quantity calculation in step 4. The longitudinal deceleration reference value and the target steering angle in step 4 are calculated by optimizing a cost function, and the cost function is The longitudinal deceleration controller adopts an incremental control law, and the longitudinal control output increment is The steering controller adopts an incremental control law, and the steering control output increment is The adaptive parameter adjustment module in step 5 includes a fuzzy rule base and a trainable nonlinear mapping unit, and the adaptive parameter adjustment module takes vehicle speed, road slope angle, roundabout radius, lateral acceleration and longitudinal acceleration change rate as inputs, and outputs gain adjustment coefficients of the longitudinal controller and the steering controller, and the proportional coefficient, integral coefficient and differential coefficient of the longitudinal controller are equal to the product of the baseline coefficient and the corresponding adjustment coefficient, and the proportional coefficient, integral coefficient and differential coefficient of the steering controller are equal to the product of the baseline coefficient and the corresponding adjustment coefficient, so as to realize adaptive control for different roundabout and ramp working conditions. A visual trigger multi-stage deceleration-steering coordination control method for roundabout and ramp working conditions is implemented, and the method comprises the steps of: A visual perception module is used to collect road images in front of the vehicle and output lane lines, road edges and traffic sign features; A state perception module is used to collect signals of wheel speed sensors and inertial measurement units, and output vehicle speed, longitudinal acceleration, lateral acceleration and yaw rate; A working condition identification module is connected with the visual perception module and the state perception module, and is used to determine whether the current working condition is a roundabout working condition or a ramp working condition based on road features and road slope angles, and output a multi-stage deceleration trigger signal when the working condition meets the trigger condition; A multi-stage target generation module is connected with the working condition identification module, and is used to generate a target speed sequence composed of at least three stages and corresponding stage lengths according to the working condition type, the current vehicle speed and the road curvature when receiving the multi-stage deceleration trigger signal; A cooperative control module is connected with the multi-stage target generation module and the state perception module, and is used to jointly solve a longitudinal deceleration reference value and a target steering angle according to the target speed of the stage to which the current position belongs and the road curvature in each control period, and output control instructions to brake actuators and steering actuators, so as to realize multi-stage deceleration and steering coordination control; An adaptive parameter adjustment module is connected with the state perception module and the working condition identification module, and is used to adjust gain parameters of a longitudinal controller and a lateral controller in the cooperative control module in real time according to vehicle operating states and target working condition features.

9. The visual trigger multi-stage deceleration-steering coordination control system for roundabout and ramp conditions of claim 8, wherein, The state perception module comprises a state estimation unit for fusing wheel speed sensor and inertial measurement unit signals, which estimates vehicle speed, yaw rate and side slip angle by using an extended Kalman filter model or an unscented Kalman filter model, the cooperative control module constructs a cost function according to the estimation results and determines a longitudinal deceleration reference value and a target steering angle by minimizing the cost function, and the cost function is used to constrain the lateral acceleration, longitudinal acceleration rate of change and steering angle change amount from exceeding preset safety limits.

10. The visual trigger multi-stage deceleration-steering coordination control system for roundabout and ramp conditions of claim 8, wherein, The cooperative control module and the adaptive parameter adjustment module are integrated in a vehicle-mounted electronic control unit, which is connected with the visual perception module, the state perception module, the brake actuator and the steering actuator through a vehicle-mounted communication bus, and the vehicle-mounted electronic control unit realizes the functions of working condition recognition, multi-stage target generation, cooperative control and adaptive parameter adjustment through a software program.

Images