Vehicle path planning and obstacle avoidance method based on CMS assistance

Abstract

The present invention relates to the field of vehicle assisted driving technology, and in particular to a vehicle path planning and obstacle avoidance method based on CMS assistance. The present invention solves the problem of automatic lane change and improves the intelligent driving experience while reducing costs by using the auxiliary path planning function of CMS in the case of low-cost intelligent driving of a front-view integrated device; the auxiliary path planning of the vehicle intelligent driving system more accurately improves the accuracy and safety of the path planning curve; the high-frequency sampling cycle and the identification of side and rear targets are used to further improve vehicle safety, and emergency path planning can assist the intelligent driving system in replanning the route to reduce the risk and severity of collision.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle assisted driving, and in particular to a vehicle path planning and obstacle avoidance method based on CMS assistance. Background Art

[0002] At present, when the assisted driving controller turns on the assisted driving function, if the vehicle starts AEB emergency braking in an emergency situation, or if the vehicle ahead blocks the way, the vehicle will slow down along the original route until the vehicle stops. In the prior art, the electronic rearview mirror (CMS) is rarely used and is generally only used for field of view display. The original intelligent driving solution for the front-view integrated machine is basically a 40ms sampling period. Such an assisted driving system has the following defects:

[0003] 1. The electronic rearview mirror is only used for field of view display and cannot be combined with vehicle control, resulting in low resource utilization and high cost;

[0004] 2. The sampling cycle of the front-view integrated machine is long, and the braking is delayed in emergency situations;

[0005] 3. In the path planning of the forward-looking integrated device, it is difficult to implement lane change path planning, and there are problems such as being unable to make timely avoidance when encountering the risk of collision with targets on the side and rear. Summary of the invention

[0006] The technical problem to be solved by the present invention is: in order to solve the problems existing in the prior art in the above-mentioned background technology, a path planning and obstacle avoidance method based on CMS assistance is provided.

[0007] The technical solution adopted by the present invention to solve the technical problem is: a vehicle path planning and obstacle avoidance method based on CMS assistance, comprising the following steps:

[0008] S1, power-on self-test: When the vehicle is started, the electronic rearview mirror and front-view integrated device perform self-test and complete initialization;

[0009] S2. Auxiliary path planning:

[0010] (a) When the vehicle is driving, the electronic rearview mirror collects road information in real time and displays it to the user synchronously. The driver turns on the lateral and longitudinal joint control function of the front-view integrated device. When the driver turns the steering wheel, the body controller sends the steering information to the front-view integrated device and the electronic rearview mirror;

[0011] (b) After receiving the steering information, the electronic rearview mirror uses the external environment sensor to sample at a sampling period of 60 frames per second to identify the environment and vehicle information of the lane change route, and calculates the collision time based on the road information of the upcoming lane change. TCC; At the same time, the external environment sensor is used to determine whether there is a lane line. If there is a lane line, a lane change path trajectory is planned according to the lane line information, and the planned lane change path trajectory information is sent to the front-view integrated device;

[0012] (c) After the front-view integrated device receives the turning signal sent by the vehicle body controller, if there is a lane line, it fits a planned predicted motion trajectory based on the lane line;

[0013] (d) When the front-view integrated device receives a lane change permission instruction, the lane change starts according to the lane change trajectory of the front-view integrated device. If the road curvature radius is ≥ 1000m, the lane change is performed after optimizing the planning and predicting the motion trajectory;

[0014] S3, obstacle avoidance when changing lanes:

[0015] (i) The electronic rearview mirror and the front-view integrated device collect external presence information and select targets that meet the requirements of collision time, angle range, instantaneous offset, and relative speed based on the external presence information;

[0016] (ii) Calculate the collision time based on the selected targets and use the minimum function to obtain the actual collision time detected by the electronic rearview mirror TTC CMS_act , the actual collision time detected by the front-view camera alone TTC Ssmart_act The actual collision time with the front-view camera TTC act ;

[0017] (iii) Predict the position of the target and calculate the predicted longitudinal distance x_predicted and the predicted lateral distance y_predicted;

[0018] (iv) Determine whether the lateral position of the vehicle satisfies the normal distribution of N~(y_predicted, var), then calculate the target collision probability, and obtain the values ​​of the left and right boundaries. If they reach the set values, it is determined that there is a risk of collision;

[0019] (v) If the rear-view mirror detects a rear-end collision risk, TTC act ≥t1 and TTC Ssmart_act > t1, based on the safety ahead, the front integrated device sends an acceleration request, and at the same time the vehicle moves laterally to a safe distance of L1, flashes the double flashes to remind the following vehicle to brake urgently, reducing the risk of collision;

[0020] Determine whether the electronic rearview mirror is safe to detect the rear. TTC act ≥t1 and TTC Ssmart_actWhen t1 is less than t1, the front-view integrated device sends an emergency braking request and turns on the hazard lights to alert surrounding vehicles. Otherwise, the vehicle drives normally.

[0021] Furthermore, in step S2, the collision time is calculated based on the road information of the upcoming lane change. TCC , the calculation formula is:

[0022] (1),

[0023] In formula (1), v rel is the relative speed between the target vehicle and the ego vehicle in the target lane, a rel is the relative longitudinal acceleration between the target vehicle and the ego vehicle in the target lane, s rel is the relative longitudinal distance between the target vehicle and the ego vehicle in the target lane.

[0024] Furthermore, the calculation formula of the lane change path trajectory planned in step S2 is:

[0025] (2),

[0026] In formula (2), y cms is the predicted road lateral distance of the electronic rearview mirror, x is the predicted longitudinal distance of the electronic rearview mirror, a 0～3 is the coefficient of the forward prediction trajectory equation, a 0 is the lateral deviation of the predicted trajectory, a 1 is the tangent value of the heading angle of the predicted trajectory, a 2 is the curvature of the predicted trajectory / 2, a 3 is the curvature change rate of the predicted trajectory / 6.

[0027] Furthermore, the calculation formula for planning the predicted motion trajectory in step S2 is:

[0028] (3),

[0029] In formula (3), y pre To plan and predict the lateral distance of the motion trajectory, x To plan and predict the longitudinal distance of the motion trajectory, b 0～3 is the coefficient of the planning prediction motion trajectory equation, b 0 is the lateral deviation of the planned predicted motion trajectory, b 1 is the tangent value of the heading angle of the planned predicted motion trajectory, b 2 is the curvature of the planned predicted motion trajectory / 2,b 3 is the curvature change rate of the planned predicted motion trajectory / 6.

[0030] Furthermore, the calculation formula for optimizing the planning and predicting the motion trajectory in step S2 is:

[0031] (4),

[0032] When turning with large curvature, in order to ensure that the vehicle planning path can be controlled stably, the trajectory equation of the path planning is:

[0033] (5),

[0034] In order to cope with the situation of large curvature of the vehicle when turning relative to a straight line, the lateral deviation is compensated, where deta is the compensation factor, specifically:

[0035] When the radius of curvature is between 600 and 1000 m, the value of deta is between 0 and 0.006;

[0036] When the radius of curvature is between 300 and 600 m, the value of deta is between 0.006 and 0.01;

[0037] When the radius of curvature is between 50 and 100 m, the value of deta is between 0.01 and 0.1;

[0038] When the radius of curvature is less than 50m, the path is not specifically planned and the user is reminded not to change lanes automatically.

[0039] Furthermore, the target screening in step S3 needs to meet the following conditions: (1) continuous monitoring of more than n frames, where n is defined by the user; (2) collision time TCC ≥0.1s, (3) within the line of sight of the sensor, (4) the position is within the calibration range, (5) the angle is within the calibration range, (6) the instantaneous offset of the lateral position is within the path of the vehicle, and (7) the relative speed is within the set range.

[0040] Furthermore, the actual collision time detected by the electronic rearview mirror in step S3 is TTC CMS_act , the actual collision time detected by the front-view camera alone TTC Ssmart_act The actual collision time with the front-view camera TTC act The calculation formulas are:

[0041] (6),

[0042] (7),

[0043] (8),

[0044] In formula (6), ttc CMS_lat is the lateral collision time calculated by the electronic rearview mirror, ttc CMS_long is the longitudinal collision time calculated by the electronic rearview mirror; in formula (7), ttc smart_lat The lateral collision time calculated by the front-view camera; ttc smart_long Longitudinal collision time calculated by the forward-looking vehicle.

[0045] Furthermore, the formula for predicting the longitudinal distance x_predicted in step S3 is:

[0046] (9),

[0047] The formula for the predicted longitudinal distance y_predicted is:

[0048] (10),

[0049] In equations (9) and (10), dpsi_ego is the yaw rate of the ego vehicle, x_rel is the longitudinal distance of the target, y_rel is the lateral distance of the target, dt is the period, v_ego is the ego vehicle speed, psi is the target orientation, a_ego is the ego vehicle acceleration, a is the target acceleration, and v is the target speed.

[0050] Furthermore, the calculation formulas for the left and right boundary values ​​in step S3 are:

[0051] left border = -path border – half of the target width,

[0052] right border = path border + half of the target width,

[0053] Among them, path border is a configuration parameter.

[0054] Beneficial effects of the present invention:

[0055] The present invention solves the problem of using the auxiliary path planning function of CMS to solve the problem of automatic lane change and improve the intelligent driving experience while reducing costs in the case of low-cost intelligent driving of the front-view integrated machine; the auxiliary path planning of the vehicle intelligent driving system more accurately improves the accuracy and safety of the path planning curve;

[0056] By utilizing high-frequency sampling cycles and identification of side and rear targets, vehicle safety can be further improved, and emergency path planning can assist the intelligent driving system in replanning the route to reduce collision risk and collision severity. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0058] Figure 1 It is a control block diagram of the CMS-assisted vehicle path planning and obstacle avoidance method of the present invention.

[0059] Figure 2 It is a flow chart of collision risk judgment in the CMS-assisted vehicle path planning and obstacle avoidance method of the present invention.

[0060] Figure 3 It is a schematic diagram of lane changing in a straight road scenario based on the CMS-assisted vehicle path planning and obstacle avoidance method of the present invention.

[0061] Figure 4 It is a schematic diagram of lane changing in a curved road scenario based on the CMS-assisted vehicle path planning and obstacle avoidance method of the present invention. DETAILED DESCRIPTION

[0062] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, which only illustrate the basic structure of the present invention in a schematic manner, and therefore only show the components related to the present invention.

[0063] like Figure 1 As shown, the front-view integrated machine (joint lateral and longitudinal control controller), also called the assisted driving controller, uses the perception mode of a single front-view camera to receive external environmental information, as well as external environmental information preprocessing and lateral and longitudinal planning and control. The electronic rearview mirror controller (CMS) is used for environmental information detection, processing and early warning control functions of the rearview mirror. The left and right screens are used to display external environmental information of the CMS. The left and right environmental sensors are used to replace the original physical rearview mirrors to collect external environmental information. The body controller (BCM) transmits the body status and switch signals to the front-view integrated machine. The large central control screen is used to display relevant information of lateral and longitudinal control and road display information, etc. The vehicle stability system (ESP) controls vehicle acceleration and deceleration and vehicle stability control. The power steering (EPS) is used for steering control of lateral control.

[0064] The external environmental information is collected and transmitted to the CMS through the left and right environmental sensors. The CMS detects and processes the environmental information, and transmits the warning information to the BCM. The BCM transmits the vehicle body status and switch signals to the front view integrated machine, which is processed by the front view integrated machine. The vehicle is controlled to accelerate, decelerate or turn through the ESP and EPS. At the same time, the central control large screen displays the relevant information of the horizontal and vertical control and the road display information, etc.

[0065] The vehicle path planning and obstacle avoidance method based on CMS assistance in this embodiment is specifically implemented as follows:

[0066] First, when the vehicle starts, it performs a power-on self-check. After the self-check is completed, both the CMS and the front view integrated machine have no faults and the initialization is completed. Otherwise, the functions of the CMS and the front view integrated machine cannot be enabled.

[0067] Secondly, the auxiliary path planning function in the turning scenario, taking a right turn as an example: As Figure 3 shown, during the vehicle driving process, the CMS real-time collects the road information and synchronously displays it to the user. The driver enables the horizontal and vertical combined control function of the front view integrated machine. When the driver turns on the left turn, at this time, the body controller BCM will send the turning information to the front view integrated machine and the CMS. After receiving the driver's turning information, the CMS uses the external environmental sensors to identify the environment of the lane-changing route and the vehicle information, and calculates TTC Time-To-Collision (TTC), and the specific mathematical model is as follows:

[0068] (1),

[0069] In formula (1), v rel is the relative vehicle speed between the target vehicle in the target lane and the self-vehicle, a rel is the relative longitudinal acceleration between the target vehicle in the target lane and the self-vehicle, s rel is the relative longitudinal distance between the target vehicle in the target lane and the self-vehicle.

[0070] If the calculated TTC < t1 (t1 is a calibrated value and can be changed according to actual needs), the CMS sends the allowable lane change flag bit flag1 = 0 to the front view integrated machine, and the front view integrated machine does not allow lane change, indicating a lane change risk. Otherwise, the CMS sends the allowable lane change flag bit flag1 = 1 to the front view integrated machine.

[0071] The CMS receives the driver's steering information and uses the external environment sensor to sample at a sampling period of 60 frames per second to determine whether there is a lane line. If there is a lane line, a lane change path trajectory is planned based on the lane line information and the planned lane change path trajectory information is sent to the front-view integrated device. The specific equation is as follows:

[0072] (2),

[0073] In formula (2), y cms is the predicted road lateral distance of the electronic rearview mirror, x is the predicted longitudinal distance of the electronic rearview mirror, a 0～3 is the coefficient of the forward prediction trajectory equation, a 0 is the lateral deviation of the predicted trajectory, a 1 is the tangent value of the heading angle of the predicted trajectory, a 2 is the curvature of the predicted trajectory / 2, a 3 is the curvature change rate of the predicted trajectory / 6.

[0074] After the front-view integrated device receives the turn signal sent by the BCM, if there is a lane line, it will fit a planned predicted motion trajectory based on the lane line. The specific prediction equation is as follows:

[0075] (3),

[0076] In formula (3), y pre To plan and predict the lateral distance of the motion trajectory, x To plan and predict the longitudinal distance of the motion trajectory, b 0～3 is the coefficient of the planning prediction motion trajectory equation, b 0 is the lateral deviation of the planned predicted motion trajectory, b 1 is the tangent value of the heading angle of the planned predicted motion trajectory, b 2 is the curvature of the planned predicted motion trajectory / 2, b 3 is the curvature change rate of the planned predicted motion trajectory / 6.

[0077] When the front-view integrated device receives the lane change permission flag flag1=1, the lane change starts according to the lane change trajectory of the front-view integrated device, and when the road curvature radius reaches more than 1000m, the actual lane change trajectory equation is:

[0078] (4),

[0079] It should be noted here that the car is moving forward, so there is a time difference between the lane change path trajectory planned by the CMS and the predicted motion trajectory planned by the forward-looking integrated device, that is, the lane change path trajectory planned by the CMS is lagging, so the actual lane change trajectory needs to be optimized. Equation (4) is obtained from Equation (2) and Equation (3).

[0080] At the same time, according to its stability, it is ensured that the vehicle can pass stably when turning with large curvature, such as Figure 4 As shown in the figure, when turning with large curvature, in order to ensure that the vehicle planning path can be controlled stably, the trajectory equation of the path planning is:

[0081] (5),

[0082] In order to cope with the situation of large curvature of the vehicle when turning relative to a straight line, the lateral deviation is compensated, where deta is the compensation factor, specifically:

[0083] When the radius of curvature is between 600 and 1000 m, the value of deta is between 0 and 0.006;

[0084] When the radius of curvature is between 300 and 600 m, the value of deta is between 0.006 and 0.01;

[0085] When the radius of curvature is between 50 and 100 m, the value of deta is between 0.01 and 0.1;

[0086] When the radius of curvature is less than 50m, the path is not specifically planned and the user is reminded not to change lanes automatically.

[0087] Then, obstacle avoidance analysis is performed when changing lanes, and target screening should meet the following conditions:

[0088] It must be a mature tracking target (the target is continuously monitored for more than 5 (this value can be defined by yourself) frames);

[0089] TTC It cannot be less than the minimum value, usually 0.1s (calibrable);

[0090] The target location must be within the sensor's FOV (field of view);

[0091] The target cannot be a ground reflection point. This type of target point often has the following characteristics: it initially appears near the center of the target vehicle path and is often located within the range of 25m±2m. This value can be calibrated.

[0092] The target angle should be within ±30°, which is a calibrable value and meets the FOV angle of the CMS;

[0093] The instantaneous offset of the target’s lateral position should be within the ego vehicle’s path;

[0094] The relative speed of the target should be approaching;

[0095] The target should not be one that is laterally far away from the center of the vehicle.

[0096] The CMS and front-view integrated device collect external presence information. If there is a TTC collision risk, it is calculated as follows:

[0097] (6),

[0098] (7),

[0099] (8),

[0100] In formula (6), ttc CMS_lat is the lateral collision time calculated by the electronic rearview mirror, ttc CMS_long is the longitudinal collision time calculated by the electronic rearview mirror; in formula (7), ttc smart_lat The lateral collision time calculated by the front-view camera; ttc smart_long Longitudinal collision time calculated by the forward-looking vehicle.

[0101] Predict the position of the target, calculate the predicted longitudinal distance x_predicted and the predicted lateral distance y_predicted, and the formula for the predicted longitudinal distance x_predicted is:

[0102] (9),

[0103] The formula for the predicted longitudinal distance y_predicted is:

[0104] (10),

[0105] In equations (9) and (10), dpsi_ego is the yaw rate of the ego vehicle, x_rel is the longitudinal distance of the target, y_rel is the lateral distance of the target, dt is the period, v_ego is the ego vehicle speed, psi is the target orientation, a_ego is the ego vehicle acceleration, a is the target acceleration, and v is the target speed.

[0106] Determine whether the lateral position of the vehicle satisfies the normal distribution of N~(y_predicted, var) (N~(Y, Var) is a format of normal distribution, Var is the variance calculated by the filtering algorithm), and then calculate the target collision probability to obtain the values ​​of the left and right boundaries. The calculation formulas for the left and right boundary values ​​are:

[0107] The left border = - path border – half of the target width,

[0108] The right border = path border + half of the target width,

[0109] where path border is a configuration parameter that can be defined by oneself, and a typical value is 1.2 m. When the collision probability is 0.5 (calibratable), it is considered that there is a collision risk.

[0110] As Figure 2 shown, when the CMS detects the TTC, when TTC act < t1, there is a collision risk. At the same time, it is further determined that if it is a rear collision risk detected by the CMS, when TTC act ≥ t1 and TTC Ssmart_act > t1, on the basis of the front being safe, the forward-looking all-in-one machine sends an acceleration request. At the same time, the vehicle moves laterally by a distance of L1 (calibratable according to the actual situation) to the safe side, and at the same time flashes the hazard lights to remind the following vehicle to brake emergently to reduce the collision risk;

[0111] When the forward-looking all-in-one machine in front detects the TTC, when TTC act < t1, there is a collision risk. Then, it is detected whether the rear is safe by the CMS at the same time. When TTC act ≥ t1 and TTC Ssmart_act < t1, then the forward-looking all-in-one machine sends an emergency braking request and turns on the hazard lights to remind the surrounding vehicles; otherwise, the vehicle travels normally.

[0112] Inspired by the ideal embodiments of the present invention described above, through the above description, relevant staff can completely make various changes and modifications without departing from the technical idea of this invention. The technical scope of this invention is not limited to the content in the specification, and its technical scope must be determined according to the scope of the claims.

Claims

1. A CMS-assisted vehicle path planning and obstacle avoidance method, characterized in that: The steps include: S1, power-on self-test: When the vehicle is started, the electronic rearview mirror and front-view integrated device perform self-test and complete initialization; S2. Auxiliary path planning: (a) When the vehicle is driving, the electronic rearview mirror collects road information in real time and displays it to the user synchronously. The driver turns on the lateral and longitudinal joint control function of the front-view integrated device. When the driver turns the steering wheel, the body controller sends the steering information to the front-view integrated device and the electronic rearview mirror; (b) After receiving the steering information, the electronic rearview mirror uses the external environment sensor to sample at a sampling period of 60 frames per second to identify the environment and vehicle information of the lane change route, and calculates the collision time based on the road information of the upcoming lane change. TCC ; At the same time, the external environment sensor is used to determine whether there is a lane line. If there is a lane line, a lane change path trajectory is planned according to the lane line information, and the planned lane change path trajectory information is sent to the front-view integrated device; (c) After the front-view integrated device receives the turning signal sent by the vehicle body controller, if there is a lane line, it fits a planned predicted motion trajectory based on the lane line; (d) When the front-view integrated device receives a lane change permission instruction, the lane change starts according to the lane change trajectory of the front-view integrated device. If the road curvature radius is ≥ 1000m, the lane change is performed after optimizing the planning and predicting the motion trajectory; S3, obstacle avoidance when changing lanes: (i) The electronic rearview mirror and the front-view integrated device collect external presence information and select targets that meet the requirements of collision time, angle range, instantaneous offset, and relative speed based on the external presence information; (ii) Calculate the collision time based on the selected targets and use the minimum function to obtain the actual collision time detected by the electronic rearview mirror TTC CMS_act , the actual collision time detected by the front-view camera alone TTC Ssmart_act The actual collision time with the front-view camera TTC act ; (iii) Predict the position of the target and calculate the predicted longitudinal distance x_predicted and the predicted lateral distance y_predicted; (iv) Determine whether the lateral position of the vehicle satisfies the normal distribution of N~(y_predicted, var), then calculate the target collision probability, and obtain the values ​​of the left and right boundaries. If they reach the set values, it is determined that a risk will occur; (v) If the rear-view mirror detects a rear-end collision risk, TTC act ≥t1 and TTC Ssmart_act > At t1, based on the safety ahead, the front integrated device sends an acceleration request, and at the same time, the vehicle moves laterally to a safe distance of L1, and flashes the double flashes to remind the following vehicle to brake urgently, reducing the risk of collision; Determine whether the rear is safe when the electronic rearview mirror is detected. TTC act ≥t1 and TTC Ssmart_act When the vehicle is less than t1, the front-view integrated device sends an emergency braking request and turns on the double flashes to alert surrounding vehicles. Otherwise, the vehicle drives normally. The actual collision time detected by the electronic rearview mirror in step S3 TTC CMS_act , the actual collision time detected by the front-view camera alone TTC Ssmart_act The actual collision time with the front-view camera TTC act The calculation formulas are: ， ， ， In formula (6), ttc CMS_lat is the lateral collision time calculated by the electronic rearview mirror, ttc CMS_long Longitudinal impact time calculated for electronic rearview mirrors; In formula (7), ttc smart_lat The lateral collision time calculated by the front-view camera; ttc smart_long Longitudinal collision time calculated for the front-view camera 2. The CMS-assisted vehicle path planning and obstacle avoidance method according to claim 1, characterized in that: In step S2, the collision time is calculated based on the road information of the upcoming lane change. TCC , the calculation formula is: ， In formula (1), v rel is the relative speed between the target vehicle and the ego vehicle in the target lane, a rel is the relative longitudinal acceleration between the target vehicle and the ego vehicle in the target lane, s rel is the relative longitudinal distance between the target vehicle and the ego vehicle in the target lane.

3. The CMS-assisted vehicle path planning and obstacle avoidance method according to claim 1, characterized in that: The calculation formula of the lane change path trajectory planned in step S2 is: ， In formula (2), y cms is the predicted road lateral distance of the electronic rearview mirror, x is the predicted longitudinal distance of the electronic rearview mirror, a 0. a 1. a 2 and a 3 are the coefficients of the forward prediction trajectory equation, a 0 is the lateral deviation of the predicted trajectory, a 1 is the tangent value of the heading angle of the predicted trajectory, a 2 is the curvature of the predicted trajectory / 2, a 3 is the curvature change rate of the predicted trajectory / 6.

4. The CMS-assisted vehicle path planning and obstacle avoidance method according to claim 3, characterized in that: The calculation formula for planning and predicting the motion trajectory in step S2 is: ， In formula (3), y pre To plan and predict the lateral distance of the motion trajectory, x To plan and predict the longitudinal distance of the motion trajectory, b 0. b 1. b 2 and b 3 are the coefficients of the planning prediction motion trajectory equation, b 0 is the lateral deviation of the planned predicted motion trajectory, b 1 is the tangent value of the heading angle of the planned predicted motion trajectory, b 2 is the curvature of the planned predicted motion trajectory / 2, b 3 is the curvature change rate of the planned predicted motion trajectory / 6.

5. The CMS-assisted vehicle path planning and obstacle avoidance method according to claim 4, characterized in that: The calculation formula for optimizing the planning and predicting the motion trajectory in step S2 is: ， When turning with large curvature, in order to ensure that the vehicle planning path can be controlled stably, the trajectory equation of the path planning is: ， In order to cope with the situation of large curvature of the vehicle when turning relative to a straight line, the lateral deviation is compensated, where deta is the compensation factor, specifically: When the radius of curvature is between 600 and 1000 m, the value of deta is between 0 and 0.006; When the radius of curvature is between 300 and 600 m, the value of deta is between 0.006 and 0.01; When the radius of curvature is between 50 and 100 m, the value of deta is between 0.01 and 0.1; When the radius of curvature is less than 50m, the path is not specifically planned and the user is reminded not to change lanes automatically.

6. The CMS-assisted vehicle path planning and obstacle avoidance method according to claim 1, characterized in that: The conditions that need to be met for target screening in step S3 are: (1) continuous monitoring of more than n frames, where n is defined by oneself, and (2) collision time TCC ≥0.1s, (3) within the line of sight of the sensor, (4) the position is within the calibration range, (5) the angle is within the calibration range, (6) the instantaneous offset of the lateral position is within the path of the vehicle, and (7) the relative speed is within the set range.

7. The CMS-assisted vehicle path planning and obstacle avoidance method according to claim 1, characterized in that: The formula for the predicted longitudinal distance x_predicted in step S3 is: ， The formula for the predicted longitudinal distance y_predicted is: ， In equations (9) and (10), dpsi_ego is the yaw rate of the ego vehicle, x_rel is the longitudinal distance of the target, y_rel is the lateral distance of the target, dt is the period, v_ego is the ego vehicle speed, psi is the target orientation, a_ego is the ego vehicle acceleration, a is the target acceleration, and v is the target speed.

8. The CMS-assisted vehicle path planning and obstacle avoidance method according to claim 1, characterized in that: The calculation formulas for the left and right boundary values ​​in step S3 are: leftborder=-pathborder – half the width of the target, rightborder = pathborder + half of the target width, Among them, leftborder represents the left boundary value where a collision risk may occur; rightborder represents the right boundary value where a collision risk may occur, and pathborder is a configuration parameter.

Images