阅读说明：本技术 一种基于直线圆弧路径的全自动泊车控制方法及控制器 (Full-automatic parking control method and controller based on straight-line arc path ) 是由 程凯 王俊毅 刘威 丁会利 郑迎 赵丁莲 于 2021-09-01 设计创作，主要内容包括：本发明提供了一种基于直线圆弧路径的全自动泊车控制方法,包括：根据泊车路径,确定下一段泊车路径的期望方向盘角度和提前转向距离；在每个周期,基于位置偏差和航向偏差,对当前段的泊车路径的期望方向盘角度进行修正,得到当前周期的目标方向盘角度；直到车辆在当前段的泊车路径的剩余距离小于提前转向距离,则将下一段的泊车路径更新作为当前段的泊车路径；根据当前周期的实际方向盘角度和目标方向盘角度来得到当前周期的当前目标角度并将其输出。本发明还提供了相应的控制器。本发明通过先提前转向然后对路径进行自修正,避免了转向不足导致无法修正方向的问题,从而提高全自动泊车的转向精度。(The invention provides a full-automatic parking control method based on a straight line circular arc path, which comprises the following steps: determining an expected steering wheel angle and an early steering distance of the next parking path according to the parking path; in each period, correcting the expected steering wheel angle of the parking path of the current segment based on the position deviation and the course deviation to obtain the target steering wheel angle of the current period; updating the next section of parking path as the current section of parking path until the remaining distance of the vehicle in the current section of parking path is less than the steering distance in advance; and obtaining and outputting the current target angle of the current period according to the actual steering wheel angle and the target steering wheel angle of the current period. The invention also provides a corresponding controller. According to the invention, the steering is performed in advance and then the path is self-corrected, so that the problem that the direction cannot be corrected due to insufficient steering is avoided, and the steering precision of full-automatic parking is improved.)

1. A full-automatic parking control method based on a straight circular arc path is characterized by comprising the following steps:

step S1: determining an expected steering wheel angle and an early steering distance of the next parking path according to the segmented parking path in the planning path information;

step S2: in each period, based on the position deviation and the course deviation of the path tracking of the vehicle in the current period, correcting the expected steering wheel angle of the parking path in the current segment to obtain the target steering wheel angle sigma of the current period_t(ii) a Until the vehicle is parked in the current segmentIf the remaining distance of the path is less than the early steering distance, updating the parking path of the next segment as the parking path of the current segment, and returning to the step S1;

step S3: according to the actual steering wheel angle of the current cycle and the target steering wheel angle σ of the current cycle in the step S2_tTo obtain the current target angle delta of the current period_outAnd the current target angle delta of the current period is used_outAnd (6) outputting.

2. The method for controlling full-automatic parking according to claim 1, wherein the step S1 includes:

step S11: obtaining Ackermann rotation angles corresponding to different steering wheel angles in advance according to a vehicle test and a calculation formula of the Ackermann rotation angles;

step S12: when the next parking path is a straight line, the expected steering wheel angle of the next parking path is 0 degree; when the next parking path is an arc, obtaining an ackermann corner according to the radius of the arc of the parking path and a calculation formula of the ackermann corner, and obtaining a steering wheel angle corresponding to the ackermann corner as an expected steering wheel angle of the next parking path according to the calibration result of the step S11;

step S13: and determining the advanced steering distance s of the next parking path according to the expected steering wheel angle of the next parking path, the current vehicle speed, the delay response time of the EPS controller and the corner response time of the EPS controller.

3. The method for controlling full-automatic parking according to claim 2, wherein the step S11 includes:

step S111: on open flat ground, a driver fixes the steering wheel angle alpha, the vehicle runs for one circle at idle speed, the radius of the running track of the vehicle rotating for one circle is measured by a tape measure to be used as the radius R of the circular arc of the parking path, and the ackermann corner corresponding to the steering wheel angle alpha is calculated by the calculation formula of the ackermann corner

Step S112: continuously changing the steering wheel angle alpha and repeating the step S111 until all the steering wheel angles alpha in the rotatable range of the steering wheel are tested, and completing calibration at the moment;

the calculation formula of the Ackerman angle is as follows:

wherein R is the radius of the circular arc of the parking path, L is the vehicle wheel base,is Ackerman corner.

4. The method for controlling full-automatic parking according to claim 2, wherein in step S13, when the desired steering wheel angle of the next parking route is the maximum steering wheel angle, the method further depends on the arc length threshold l_shTo assist in determining the advance steering distance s of the next parking path; arc length threshold l_shThe test method comprises the following steps: fixing a steering wheel on a left dead or right dead position, gradually increasing the vehicle speed from 0 to the maximum parking speed, after the vehicle runs for a section of arc length, enabling the turning radius of the track to be larger than the turning radius corresponding to the turning angle of the steering wheel, and measuring the arc length distance from a starting point to a real vehicle track point of which the turning radius begins to be larger;

the advance steering distance s of the next parking path is as follows:

s＝v*(|σ_e-σ_c|/ω)+v*t_sr+C*K*(l-l_sh)，

where v is the current vehicle speed, t_srIs the hysteresis response time of the EPS controller; sigma_eA desired steering wheel angle for the next parking path segment; sigma_cThe actual steering wheel angle of the current period, and omega is the steering wheel angular velocity;c is a selection parameter, when the expected steering wheel angle of the next parking path is the maximum steering wheel angle and the arc length thereof exceeds an arc length threshold value l_shIf so, setting C as 1, otherwise, setting C as 0; l is the arc length of the next parking path when the expected steering wheel angle is the maximum steering wheel angle; l_shIs an arc length threshold; k is a calibration parameter.

5. The method for controlling full-automatic parking according to claim 2, wherein in step S2, the correcting the desired steering wheel angle of the parking path at the current stage includes:

step S21: in each period, judging whether the position deviation and the course deviation tracked by the current path of the vehicle are both smaller than a control error threshold value;

step S22: determining and outputting the target steering wheel angle of the current period according to the judgment result of the step S21;

in step S22, if at least one of the position deviation and the heading deviation of the current path tracking of the vehicle is greater than the control error threshold, the target steering wheel angle in the current cycle is:

σ＝σ_e+K_p*e_pos+K_h*e_head，

where σ is the target steering wheel angle σ for the current cycle_tAn intermediate variable of (d); sigma_tTarget steering wheel angle, σ, for the current cycle_eDesired steering wheel angle of parking path for current segment, e_posAs the position deviation of the vehicle in the current cycle, e_headIs the course deviation of the vehicle in the current period, K_pIs a coefficient of positional deviation, K_hIs the heading deviation term coefficient, σ_maxIs the maximum steering wheel angle;

and when the position deviation and the course deviation of the vehicle in the path tracking of the current period are both smaller than the set threshold value, directly outputting the expected steering wheel angle of the parking path of the current segment or the target steering wheel angle of the previous period as the target steering wheel angle of the current period.

6. The method for controlling full automatic parking according to claim 1, wherein in step S3, the current target angle δ of the current cycle is_outComprises the following steps:

in the formula: delta_outIs the current target angle of the current period; delta_cIs the actual steering wheel angle for the current cycle; delta_tIs the target steering wheel angle for the current cycle; and delta is a control step size.

7. The full-automatic parking control method based on the straight circular arc path according to claim 1, further comprising step S1': determining a longitudinal distance deviation s from a current vehicle pose and a parking path in the planned path information_eAnd according to the longitudinal distance deviation s_eOutputting the target vehicle speed of the current period;

the step S1' includes:

step S11': in each period, determining the advance braking distance s according to the delayed response time of the vehicle chassis control system and the current starting braking speed_f；

Step S12': calculating and judging the longitudinal distance deviation s in each period_eDistance s from advance brake_fSize; when the longitudinal distance deviation s_eNot more than advance braking distance s_fThen, the target vehicle speed of the current period is recorded as 0, and the target vehicle speed and the target control deceleration a of the current period are output_d(ii) a Otherwise, the longitudinal distance deviation s_e>Advance braking distance s_fDetermining and outputting the target vehicle speed v of the current period according to the maximum vehicle speed and the minimum vehicle speed for parking_t。

8. The method for controlling full-automatic parking according to claim 7, wherein the advanced braking distance s is_JComprises the following steps:

in the formula, s_fTo advance the braking distance, t_rIs a delayed response time of the vehicle chassis control system; v. of_dThe current starting braking speed; a is_dControlling deceleration for a target;

when the longitudinal distance deviation s_e>Advance braking distance s_fTarget vehicle speed v of the current cycle_tComprises the following steps:

in the formula: v. of_tA target vehicle speed; a is_pCalibrating the acceleration; s_eIs the longitudinal distance deviation; v. of_maxFor maximum speed of parking, v_minA minimum vehicle speed for parking; v. of_oIs the target vehicle speed v_tThe intermediate variable of (1).

9. A full-automatic parking controller based on a straight-line circular arc path is characterized by comprising a target corner control module and an EPS corner control module connected with the target corner control module;

the target steering angle control module is configured to perform the steps of:

step S1: determining an expected steering wheel angle and an early steering distance of the next parking path according to the parking path in the planned path information;

step S2: in each period, correcting the expected steering wheel angle of the parking path in the current period based on the position deviation and the course deviation of the path tracking of the vehicle in the current period to obtain the target steering wheel angle in the current period; updating the parking path of the next segment as the parking path of the current segment until the remaining distance of the parking path of the vehicle in the current segment is less than the steering advance distance, and returning to the step S1;

the EPS corner control module 2 is set to be according to the actual steering wheel angle of the current period and the target steering wheel angle sigma of the current period_tTo obtain the current target angle delta_outAnd the current target angle delta of the current period is used_outAnd (6) outputting.

10. The fully-automatic parking controller based on the straight-line circular arc path as claimed in claim 9, further comprising a speed planning module configured to determine a longitudinal distance deviation s according to the current vehicle pose and the parking path in the planned path information_eAnd according to the longitudinal distance deviation s_eAnd outputting the target vehicle speed of the current period.

Technical Field

The invention belongs to the field of automobile auxiliary driving, and particularly relates to a full-automatic parking control method and a controller.

Background

The full-automatic parking system can effectively avoid safety accidents in the process of manual parking and is also a key technology for automatically driving the last kilometer in the future. The full-automatic parking system mainly comprises parking space scanning, parking path planning and vehicle control. The purpose of vehicle control is to enable the vehicle to smoothly travel along a parking path and eventually stop at a specified target point. The embedded controller is a part of a full-automatic parking system, the calculation force of the used embedded controller is limited, and at present, a path formed by a straight line and an arc is often generated only through a parking path planning module to be used as a parking path, and a vehicle moves along the path.

However, in the current fully automatic parking system, the commonly used vehicle control methods all need smooth paths, and the sudden change of curvature in the parking path formed by straight lines and circular arcs generated by the current parking path planning module can cause uneven steering and even divergence.

Disclosure of Invention

The invention aims to provide a full-automatic parking control method and a controller based on a straight-line circular arc path so as to improve the steering precision of full-automatic parking.

In order to achieve the above object, the present invention provides a full-automatic parking control method based on a straight circular arc path, comprising:

s1: determining an expected steering wheel angle and an early steering distance of the next parking path according to the segmented parking path in the planning path information;

s2: in each period, based on the position deviation and the course deviation of the path tracking of the vehicle in the current period, correcting the expected steering wheel angle of the parking path in the current segment to obtain the target steering wheel angle sigma of the current period_t(ii) a Updating the parking path of the next segment as the parking path of the current segment until the remaining distance of the parking path of the vehicle in the current segment is less than the early steering distance, and returning to the step S1;

s3: according to the actual steering wheel angle of the current cycle and the target steering wheel angle σ of the current cycle in the step S2_tTo obtain the current target angle delta of the current period_outAnd the current target angle delta of the current period is used_outAnd (6) outputting.

The step S1 includes:

s11: obtaining Ackermann rotation angles corresponding to different steering wheel angles in advance according to a vehicle test and a calculation formula of the Ackermann rotation angles;

s12: when the next parking path is a straight line, the expected steering wheel angle of the next parking path is 0 degree; when the next parking path is an arc, obtaining an ackermann corner according to the radius of the arc of the parking path and a calculation formula of the ackermann corner, and obtaining a steering wheel angle corresponding to the ackermann corner as an expected steering wheel angle of the next parking path according to the calibration result of the step S11;

s13: and determining the advanced steering distance s of the next parking path according to the expected steering wheel angle of the next parking path, the current vehicle speed, the delay response time of the EPS controller and the corner response time of the EPS controller.

The step S11 includes:

s111: on open flat ground, a driver fixes the steering wheel angle alpha, the vehicle runs for one circle at idle speed, the radius of the running track of the vehicle rotating for one circle is measured by a tape measure to be used as the radius R of the circular arc of the parking path, and the ackermann corner corresponding to the steering wheel angle alpha is calculated by the calculation formula of the ackermann corner

S112: continuously changing the steering wheel angle alpha and repeating the step S111 until all the steering wheel angles alpha in the rotatable range of the steering wheel are tested, and completing calibration at the moment;

the calculation formula of the Ackerman angle is as follows:

wherein R is the radius of the circular arc of the parking path, L is the vehicle wheel base,is Ackerman corner.

In step S13, when the desired steering wheel angle of the next parking path is the maximum steering wheel angle, the arc length threshold l is further determined_shTo assist in determining the advance steering distance s of the next parking path; arc length threshold l_shThe test method comprises the following steps: fixing a steering wheel on a left dead or right dead position, gradually increasing the vehicle speed from 0 to the maximum parking speed, after the vehicle runs for a section of arc length, enabling the turning radius of the track to be larger than the turning radius corresponding to the turning angle of the steering wheel, and measuring the arc length distance from a starting point to a real vehicle track point of which the turning radius begins to be larger;

the advance steering distance s of the next parking path is as follows:

s＝v*(|σ_e-σ_c|/ω)+v*t_sr+C*K*(l-l_sh)，

where v is the current vehicle speed, t_srIs the hysteresis response time of the EPS controller; sigma_eA desired steering wheel angle for the next parking path segment; sigma_cThe actual steering wheel angle of the current period, and omega is the steering wheel angular velocity; c is a selection parameter, when the expected steering wheel angle of the next parking path is the maximum steering wheel angle and the arc length thereof exceeds an arc length threshold value l_shIf so, setting C as 1, otherwise, setting C as 0; l is the desired steering wheel angle is the maximum steering wheel angleThe arc length of the next section of parking path in the hour; l_shIs an arc length threshold; k is a calibration parameter.

In step S2, the correcting the desired steering wheel angle of the parking path of the current segment specifically includes:

s21: in each period, judging whether the position deviation and the course deviation tracked by the current path of the vehicle are both smaller than a control error threshold value;

s22: determining and outputting the target steering wheel angle of the current period according to the judgment result of the step S21;

in step S22, if at least one of the position deviation and the heading deviation of the current path tracking of the vehicle is greater than the control error threshold, the target steering wheel angle in the current cycle is:

σ＝σ_e+K_p*e_pos+K_h*e_head，

where σ is the target steering wheel angle σ for the current cycle_tAn intermediate variable of (d); sigma_tTarget steering wheel angle, σ, for the current cycle_eDesired steering wheel angle of parking path for current segment, e_posAs the position deviation of the vehicle in the current cycle, e_headIs the course deviation of the vehicle in the current period, K_pIs a coefficient of positional deviation, K_hIs the heading deviation term coefficient, σ_maxIs the maximum steering wheel angle;

and when the position deviation and the course deviation of the vehicle in the path tracking of the current period are both smaller than the set threshold value, directly outputting the expected steering wheel angle of the parking path of the current segment or the target steering wheel angle of the previous period as the target steering wheel angle of the current period.

In the step S3, the current target angle δ of the current cycle_outComprises the following steps:

in the formula: delta_outIs the current target angle of the current period; delta_cIs the actual steering wheel angle for the current cycle; delta_tIs the target steering wheel angle for the current cycle; and delta is a control step size.

The full-automatic parking control method based on the straight circular arc path further includes step S1': determining a longitudinal distance deviation s from a current vehicle pose and a parking path in the planned path information_eAnd according to the longitudinal distance deviation s_eOutputting the target vehicle speed of the current period;

step S1' includes:

s11': in each period, determining the advance braking distance s according to the delayed response time of the vehicle chassis control system and the current starting braking speed_f；

S12': calculating and judging the longitudinal distance deviation s in each period_eDistance s from advance brake_fSize; when the longitudinal distance deviation s_eNot more than advance braking distance s_fThen, the target vehicle speed of the current period is recorded as 0, and the target vehicle speed and the target control deceleration a of the current period are output_d(ii) a Otherwise, the longitudinal distance deviation s_e>Advance braking distance s_fDetermining and outputting the target vehicle speed v of the current period according to the maximum vehicle speed and the minimum vehicle speed for parking_t。

The advanced braking distance s_fComprises the following steps:

in the formula, s_fTo advance the braking distance, t_rIs a delayed response time of the vehicle chassis control system; v. of_dThe current starting braking speed; a is_dControlling deceleration for a target;

when the longitudinal distance deviation s_e>Advance braking distance s_fTarget vehicle speed v of the current cycle_tComprises the following steps:

in the formula: v. of_tA target vehicle speed; a is_pCalibrating the acceleration; s_eIs the longitudinal distance deviation; v. of_maxFor maximum speed of parking, v_minA minimum vehicle speed for parking; v. of_oIs the target vehicle speed v_tAn intermediate variable of (d);

on the other hand, the invention provides a full-automatic parking controller based on a linear arc path, which comprises a target corner control module and an EPS corner control module connected with the target corner control module; the target steering angle control module is configured to perform the steps of: s1: determining an expected steering wheel angle and an early steering distance of the next parking path according to the parking path in the planned path information; s2: in each period, correcting the expected steering wheel angle of the parking path in the current period based on the position deviation and the course deviation of the path tracking of the vehicle in the current period to obtain the target steering wheel angle in the current period; updating the parking path of the next segment as the parking path of the current segment until the remaining distance of the parking path of the vehicle in the current segment is less than the steering advance distance, and returning to the step S1; the EPS corner control module 2 is set to be according to the actual steering wheel angle of the current period and the target steering wheel angle sigma of the current period_tTo obtain the current target angle delta_outAnd the current target angle delta of the current period is used_outAnd (6) outputting.

The full-automatic parking controller based on the straight-line circular arc path further comprises a speed planning module, wherein the speed planning module is set to determine a longitudinal distance deviation s according to the current vehicle pose and the parking path in the planning path information_eAnd according to the longitudinal distance deviation s_eAnd outputting the target vehicle speed of the current period.

The full-automatic parking control method based on the straight line and arc path is based on the characteristics of full-automatic parking and low-speed driving and the characteristic that the straight line and the arc form the parking path, and the problem that the direction cannot be corrected due to insufficient steering is avoided by firstly steering in advance and then self-correcting the path, so that the steering precision of full-automatic parking is improved. In addition, according to the characteristic that the speed of the parking path at the gear shifting point and the end point is necessarily 0, the target speed is calculated through the reverse estimation of the length of each section of track, and therefore the vehicle is controlled to run stably.

Drawings

Fig. 1 is a flowchart illustrating the operation of steering control of a steering wheel in a full-automatic parking control method based on a straight circular arc path according to the present invention.

Fig. 2 is a work flow diagram of speed planning of the full-automatic parking control method based on a straight circular arc path.

FIG. 3 is a schematic diagram of an error model of a parking process.

Fig. 4 is a general block diagram of the full-automatic parking controller based on a straight circular path according to the present invention.

Detailed Description

As shown in fig. 1 and 2, the full-automatic parking control method based on the straight circular arc path of the present invention includes the following steps:

step S1: as shown in fig. 1, according to the segmented parking path in the planned path information, determining the expected steering wheel angle and the ahead steering distance of the next parking path;

the whole parking path is segmented, and under the condition of the same gear, the track with the same radius is a segment. In the initialized state (i.e., the vehicle does not start traveling), the next parking path is the first parking path. It should be noted that the expected steering wheel angle and the early steering distance of the next parking route do not need to be repeatedly calculated, and only the expected steering wheel angle and the early steering distance of the next parking route need to be calculated when the vehicle just reaches the start point of the route.

Therefore, the full-automatic parking control method can advance the expected steering wheel angle corresponding to the distance from the steering wheel to the next parking path, and avoids errors caused by insufficient steering.

The step S1 specifically includes:

step S11: obtaining Ackermann rotation angles corresponding to different steering wheel angles in advance according to a vehicle test and a calculation formula of the Ackermann rotation angles so as to realize vehicle calibration;

the step S11 specifically includes:

step S111: on open flat ground, the driver fixes the steering wheel angle alpha, lets the vehicle run for one turn at idle speed, measures the radius of the running track of the vehicle for one turn by using a tape as the radius R of the circular arc of the parking path, and calculates the Ackerman corner corresponding to the steering wheel angle alpha by using the calculation formula of the Ackerman corner (namely formula (1))

The calculation formula of ackermann angle is as follows:

wherein R is the radius of an arc of a parking path, L is the vehicle wheel base,is Ackerman corner. Wherein, R is obtained according to the planning path information; the vehicle wheel base L is measured on the vehicle and is directly predefined in the algorithm.

Step S112: the steering wheel angle α is continuously changed and step S111 is repeated until all the steering wheel angles α within the rotatable range of the steering wheel are tested, at which point the calibration is completed. Thus, a series of steering wheel angles alpha and Ackerman angles can be obtainedThe corresponding relationship of (1).

All steering wheel angles alpha in the steering wheel rotating range comprise a plurality of steering wheel angles alpha which are equally spaced between a target steering wheel angle corresponding to a left dead-beat steering wheel and a target steering wheel angle corresponding to a right dead-beat steering wheel. In the present embodiment, the interval between two adjacent steering wheel angles α is 30 °, that is, the steering wheel angle α is changed every 30 °.

Step S12: when the next parking path is a straight line, the expected steering wheel angle of the next parking path is 0 degree; when the next parking path is an arc, the ackermann angle is obtained according to the radius of the arc of the parking path and the calculation formula of the ackermann angle, and the steering wheel angle corresponding to the ackermann angle is obtained as the expected steering wheel angle of the next parking path according to the calibration result of the step S11.

Step S13: and determining the advanced steering distance s of the next parking path according to the expected steering wheel angle of the next parking path, the current vehicle speed, the delay response time of the EPS controller and the corner response time of the EPS controller.

In addition, since the vehicle turns at the maximum steering wheel angle when a certain parking path in the planned path information is an arc corresponding to the maximum steering wheel angle (i.e., an arc of minimum radius), the vehicle travels over a certain arc length threshold l due to the understeer characteristic of the vehicle_shThen, the tracking error (including the position deviation and the heading deviation of the path tracking) exceeds the allowable range, and the error cannot be reduced by self-correction because the steering wheel is already at the limit. Therefore, the invention adds the arc length threshold l into the calculation formula of the advance steering distance_shThe extra distance calculation term of (2) is used to properly increase the advance steering distance to offset the error due to understeer, within the control error range.

Therefore, in step S13, when the desired steering wheel angle of the next parking path is the maximum steering wheel angle, the arc length threshold l is also used_shTo assist in determining the advance steering distance s of the next parking path; wherein, the arc length threshold value l_shThe value of (A) is obtained by real vehicle test, and the test method comprises the following steps: the steering wheel is fixed on the left dead or right dead position, the vehicle speed is gradually increased from 0 to the maximum parking speed, and the vehicle movesAfter the vehicle runs for a section of arc length, the turning radius of the track is larger than the turning radius corresponding to the steering wheel corner, and the arc length distance from the starting point to the actual track point with the turning radius starting to be larger is measured.

Therefore, the calculation formula of the advance turning distance s of the next parking path is as follows:

s＝v*(|σ_e-σ_c|/ω)+v*t_sr+C*K*(l-l_sh) (2)

where v is the current vehicle speed, t_srThe delay response time of the EPS controller is obtained by EPS real vehicle test; sigma_eThe expected steering wheel angle of the next parking path is obtained according to the planned path information; sigma_cThe actual steering wheel angle in the current period is represented as omega, the steering wheel angular velocity is represented as the maximum steering wheel angular velocity, and the maximum steering wheel angular velocity can be obtained by acquiring a signal of the steering wheel angle through an actual vehicle and differentiating the signal; c is a selection parameter, when the expected steering wheel angle of the next parking path is the maximum steering wheel angle and the arc length thereof exceeds an arc length threshold value l_shIf so, setting C as 1, otherwise, setting C as 0; l is the arc length of the next parking path when the expected steering wheel angle is the maximum steering wheel angle; l_shIs an arc length threshold; k is a calibration parameter. The calibration parameters are obtained by continuously adjusting in the real vehicle test process.

Step S2: in each fixed period, based on the position deviation and the course deviation of the path tracking of the vehicle in the current period, correcting the expected steering wheel angle of the parking path in the current segment to obtain the target steering wheel angle sigma of the current period_t(ii) a Updating the parking path of the next segment as the parking path of the current segment until the remaining distance of the parking path of the vehicle in the current segment is less than the early steering distance, and returning to the step S1;

after the central point of the rear axle of the vehicle runs through the path connection point, a corresponding control strategy is needed to ensure that the vehicle runs in the track error range, so the step S2 of the full-automatic parking control method based on the straight-line circular-arc path of the invention corrects the position deviation and the course deviation of the path tracking of the vehicle in the current period around the expected steering wheel angle.

It should be noted that the whole full-automatic parking control method based on the straight circular arc path of the present invention is operated periodically, that is, a result containing a plurality of parameters is output at intervals (i.e., a period), rather than only one parameter being output periodically. In the present embodiment, the period is empirically determined in practice, and a value in the range of 10 to 50ms is suitable.

An error model of the parking process is shown in fig. 3.

In fig. 3, G is a rear axle center point of the vehicle, AB is a planned parking path including a straight line and an arc, O is a point closest to G in the parking path, i.e., a control target point on the parking path, and a position deviation defining path tracking is a distance between the rear axle center point G of the vehicle and the control target point O, θ and θ_dRespectively representing the included angles between the tangent lines of the longitudinal axis of the vehicle and the parking path at the control target point O and the X axis, and defining the course deviation of path tracking as theta-theta_dAssuming that the vehicle is traveling backward, the length from the control target point O to the end point a of the path (i.e., the shift point or the end point) is defined as a longitudinal distance deviation s_e(ii) a The path end point A is the end point of a certain parking path; the longitudinal distance deviation is used for speed planning and calculating the target vehicle speed.

In step S2, the correcting the desired steering wheel angle of the parking path of the current segment specifically includes:

step S21: in each period, judging whether the position deviation and the course deviation tracked by the current path of the vehicle are both smaller than a control error threshold value; the control error threshold is set manually.

Step S22: determining and outputting the target steering wheel angle of the current period according to the judgment result of the step S21;

in step S22, if at least one of the position deviation and the heading deviation of the current path tracking of the vehicle is greater than the control error threshold, the target steering wheel angle in the current cycle is:

σ＝σ_e+K_p*e_pos+K_h*e_head (3)

where σ is the target steering wheel angle σ for the current cycle_tThe function of equation (4) is clipping; sigma_tTarget steering wheel angle, σ, for the current cycle_eDesired steering wheel angle of parking path for current segment, e_posFor the deviation of the position of the vehicle in the current cycle for controlling the vehicle to travel close to the reference path, e_headThe course deviation of the vehicle in the current period is used for controlling the course angle of the vehicle to be consistent with the course angle of the reference path control point, and meanwhile, the condition that the system causes the vehicle to run in a snake shape due to overshoot is avoided, K_pIs a coefficient of positional deviation, K_hIs the heading deviation term coefficient, σ_maxThe maximum steering wheel angle.

In step S22, when both the position deviation and the heading deviation of the vehicle in the path tracking in the current cycle are smaller than the set threshold, control is not performed (i.e., the steering wheel angle does not need to be corrected again), and the desired steering wheel angle of the parking path in the current segment or the target steering wheel angle in the previous cycle is directly output as the target steering wheel angle in the current cycle. When the target steering wheel angle of the previous period exists in the parking path of the current segment, outputting the target steering wheel angle of the previous period; otherwise, the desired steering wheel angle for the parking path of the current segment is output. Therefore, the stability of the steering wheel control is ensured.

The above-mentioned steps S1, S2 are both performed by the target steering angle control module 100, so that the target steering wheel angle of the current cycle is output by the target steering angle control module 100, and thus the target steering wheel angle is output by the target steering angle control module 100 every cycle.

Step S3: according to the actual steering wheel angle of the current cycle and the target steering wheel angle σ of the current cycle in the step S2_tTo obtain the current target angle delta of the current period_outAnd comparing the current period with the current periodFront target angle delta_outAnd (6) outputting. Wherein, the current target angle delta of the current period is determined_outTo an EPS controller (electric power steering system controller) to control the vehicle at the current target angle delta_outThe vehicle is steered to travel.

Wherein, the actual steering wheel angle of the current period is provided by a rotation angle sensor of the vehicle. Therefore, the rotating speed of the steering wheel can be controlled within a certain steering step range, and the failure of the EPS due to the fact that the difference between the current angle and the target angle is too large is prevented. In the present embodiment, the step S3 is executed by an EPS corner control module 2.

Current target angle delta of current cycle_outComprises the following steps:

in the formula: delta_outThe parameter is the current target angle of the current period and is the output result of the whole full-automatic parking control method based on the straight-line circular arc path; delta_cActual steering wheel angle, δ, for the current cycle_tIs the target steering wheel angle for the current cycle; and delta is a control step size.

Wherein the control step size Δ needs to be as large as possible while the specification parameters of the EPS currently used by the vehicle must be satisfied. Because the larger the control step size delta is, the more real-time performance of control can be ensured, but the larger the control step size delta is, the more easy the control step size delta is to cause downtime of the EPS module, a balance needs to be taken between the control step size delta and the control step size delta, and under the condition that the EPS module is guaranteed not to be downtime, the larger the value is, the better the value is, and the value of the control step size delta can be determined according to real vehicle tests in advance.

The test method for the control step Δ is as follows: the target steering wheel angle is periodically sent to the EPS module through a CANoe (controller area network) communication tool, the sending period is the same as the value of the period of the full-automatic parking control method based on the straight-line circular arc path, the target steering wheel angle sent each time is increased by a control step delta on the basis of the last sending, the control step delta is gradually increased from 10 degrees until the EPS module crashes, and the control step delta before the crashes is the required control step delta.

It should be noted that the influence of the control step Δ on the advance steering distance s is already taken into account by the steering wheel angular velocity ω, so the calculation of the advance steering distance s does not separately add a correction term for the control step Δ.

The above-mentioned step S3 is executed by the EPS corner control module 2, so that the current target angle δ of the current cycle is output by the EPS corner control module 2_out。

Thus, the steering wheel angle control design is realized collectively by step S1, step S2, and step S3.

In addition, the full-automatic parking control method based on the straight circular arc path of the present invention may further include step S1': as shown in FIG. 2, a longitudinal distance deviation s is determined from the current vehicle pose and parking path in the planned path information_eAnd according to the longitudinal distance deviation s_eAnd outputting the target vehicle speed of the current period. Thus, a speed planning design is realized.

The step S1' is performed by the speed planning module 3, and is performed simultaneously with the above steps S1-S3. The step S1' specifically includes:

step S11': in each period, determining the advance braking distance s according to the delayed response time of the vehicle chassis control system and the current starting braking speed_f；

According to the characteristic that the speed of a parking path at a gear shifting point (namely a position of switching a forward gear into a reverse gear or a position of switching the reverse gear into the forward gear) and an end point is 0, the parking running is regarded as a process of uniform deceleration, and the target speed is calculated by backward deducing each section of track length. To ensure a more accurate stop at the target point, the braking must be initiated a distance ahead, taking into account the delayed response time of the vehicle chassis control system.

Thus, the braking distance s is advanced_fComprises the following steps:

in the formula, s_fTo advance the braking distance, t_rIs the delayed response time of the vehicle chassis control system, which is obtained by vehicle testing; v. of_dFor the current start of braking speed, the current start of braking speed v_dThe current vehicle speed is transmitted to the speed planning module 3 by the vehicle through the CAN in each period and is obtained as the current starting braking vehicle speed, and the current vehicle speed is one of the input parameters of the controller; a is_dThe deceleration is controlled for the target. Target control deceleration a_dThe maximum braking deceleration used for sending to the chassis is a calibrated value under the condition of ensuring the smoothness of the brake, and is obtained by testing a driver on the vehicle.

Step S12': calculating and judging the longitudinal distance deviation s in each period_eDistance s from advance brake_fSize; when the longitudinal distance deviation s_eNot more than advance braking distance s_fThen, the target vehicle speed of the current period is recorded as 0, and the target vehicle speed and the target control deceleration a of the current period are output_dControlling vehicle braking; otherwise, the longitudinal distance deviation s_e>Advance braking distance s_fDetermining and outputting the target vehicle speed v of the current period according to the maximum vehicle speed and the minimum vehicle speed for parking_tTo control the vehicle to run.

Wherein when the longitudinal distance deviation s_e>Advance braking distance s_fTarget vehicle speed v of the current cycle_tComprises the following steps:

in the formula: v. of_tIs a target vehicle speed, a_pIn order to calibrate the acceleration, the acceleration sensor is,calibrating acceleration a_pWith the target control deceleration a as described above_dAre the same value of s_eIs a longitudinal distance deviation, i.e., the distance from the control target point O on the parking path to the path end point a (i.e., shift point or end point); v. of_maxFor maximum speed of parking, v_minThe parking minimum vehicle speed, the parking maximum vehicle speed and the parking minimum vehicle speed are artificially given; v. of_oIs the target vehicle speed v_tEquation (8) is for clipping.

During parking, the target vehicle speed and the target control deceleration a of the current period_dAnd the output is output to an ESC controller (body electronic stability controller) of the automobile chassis through a CAN bus so as to control the running of the automobile. The actual acceleration and the target acceleration have a certain access, but the access is not greatly influenced in the acceleration process, and the deceleration process is ensured by an ESC module of the vehicle.

Fig. 4 shows a fully-automatic parking controller 100 based on a straight-line arc path according to the present invention, which is implemented based on the above-mentioned fully-automatic parking control method based on a straight-line arc path.

The overall block diagram of the full-automatic parking controller 100 is shown in fig. 1, and the input parameters of the full-automatic parking controller 100 are the current vehicle pose, the current vehicle speed, the actual steering wheel angle of the current period, and the planned path information. Wherein the planned path information includes a parking path and a desired steering wheel angle corresponding to each parking path.

The full-automatic parking controller 100 comprises a target rotation angle control module 1, an EPS rotation angle control module 2 connected with the target rotation angle control module 1, and a speed planning module 3.

The input parameters of the target turning angle control module 1 include the current vehicle pose, the current vehicle speed, the actual steering wheel angle of the current period, and the planned path information.

The target steering angle control module 1 is configured to determine a target steering wheel angle σ for a current cycle based on a current vehicle pose, a current vehicle speed, an actual steering wheel angle for the current cycle, and planned path information_t. In particular toIn other words, the target steering angle control module 1 is configured to perform the following steps: step S1: determining an expected steering wheel angle and an early steering distance of the next parking path according to the parking path in the planned path information; step S2: in each period, correcting the expected steering wheel angle of the parking path in the current period based on the position deviation and the course deviation of the path tracking of the vehicle in the current period to obtain the target steering wheel angle in the current period; and updating the parking path of the next segment as the parking path of the current segment until the remaining distance of the parking path of the vehicle in the current segment is less than the advanced steering distance, and returning to the step S1. Wherein the target steering wheel angle σ_tThe specific calculation method of (3) is as described above, and the calculation formulas are formula (3) and formula (4) above.

The input parameters of the EPS corner control module 2 are the actual steering wheel angle of the current period and the target steering wheel angle of the current period, and the output parameters are the current target angle.

The EPS corner control module 2 is set to be according to the actual steering wheel angle of the current period and the target steering wheel angle sigma of the current period_tTo obtain the current target angle delta_outAnd the current target angle delta of the current period is used_outAnd (6) outputting. Current target angle delta_outThe specific calculation method of (3) has been described in detail above, and the calculation formula is as shown in formula (5) above.

Therefore, the EPS controller is controlled to steer within a certain steering step length (namely control step length delta) by adding the EPS corner control module 2, so that the steering wheel rotating speed is controlled, and the failure of the EPS controller due to the fact that the difference value between the actual steering wheel angle of the current period and the target steering wheel angle provided by the target corner control module 1 is too large is prevented.

The speed planning module 3 is configured to output the target speed of the current period according to the current vehicle pose and the parking path in the planned path information, thereby performing speed planning design.

Thus, the fully automatic parking controller 100 of the present invention can adjust the current target angle δ_outAnd the current target speed is sent to the automobile chassis through a CAN bus to controlAnd (5) controlling the vehicle to run. Specifically, the fully-automatic parking controller 100 sets the current target angle δ_outThe EPS controller is sent to an EPS controller of a vehicle chassis, is a controller of the vehicle and is responsible for steering a steering wheel to a current target angle; the full-automatic parking controller 100 sends the current target vehicle speed to an ESC (electronic stability controller) on the vehicle chassis, where the ESC is a module of the vehicle and is responsible for driving the vehicle according to the current target vehicle speed.

In summary, the full-automatic parking control method based on the linear arc path provides a steering control method for steering in advance and then self-correcting aiming at the characteristic that the curvature of the linear arc path planned by the full-automatic parking is discontinuous, and simultaneously provides a speed planning method for reversely deducing the target speed through the length of each section of path according to the characteristic that the speed at a gear shifting point and an end point is necessarily zero.

The above embodiments are merely preferred embodiments of the present invention, which are not intended to limit the scope of the present invention, and various changes may be made in the above embodiments of the present invention. All simple and equivalent changes and modifications made according to the claims and the content of the specification of the present application fall within the scope of the claims of the present patent application. The invention has not been described in detail in order to practice the invention.