A lane centering driving control method and system

Abstract

The present invention provides a lane centering driving control method and system. By using the perception ability of a camera to establish a lane line geometric model, according to the vehicle wheelbase, vehicle speed, lane line radius, and the lateral distance of the preview point, the steering wheel angle required for turning a corner is obtained, that is, feedforward calculation is performed. Then, according to the lateral deviation and heading deviation between the vehicle and the lane line, the required lateral deviation compensation angle and heading deviation compensation angle are obtained, that is, feedback calculation is performed. Finally, by combining the feedforward and feedback algorithms, the total steering wheel angle is calculated, enabling the vehicle to virtualize the other lane line in the two-in-one or one-and-two scenarios, plan a safe driving trajectory, smoothly control the host vehicle to center and drive into the new lane, seamlessly achieve the connection between lane lines, avoid frequently asking the driver to take over the vehicle and affecting the driving experience; at the same time, avoid the driving risks caused by the vehicle pressing on or straddling the lane line.

Description

Technical Field

[0001] The present invention relates to the field of vehicle electronics technology, and in particular, to a lane centering driving control method and system. Background Art

[0002] Currently, information technologies represented by high-precision sensors, high-computing-power chips, Internet of Things, cloud computing, big data, and artificial intelligence are widely used, which are accelerating the intelligent development of society. The intelligent process in the field of vehicle traffic is also evolving accordingly. Among them, ADAS / AD (Advanced Driver Assistance System / Autonomous Driving) is a key innovation field in the vehicle industry.

[0003] During the evolution of ADAS (Advanced Driver Assistance System) to AD (Autonomous Driving), various sub-functions emerge in an endless stream (from concept proposal to function implementation), such as ACC (Adaptive Cruise Control), LKA (Lane Keeping Assistance), LDW (Lane Departure Warning), TJA / ICA (Traffic Jam Assist / Integrated Cruise Assist), etc. In the traditional LKA function, when passing through a one-to-two or two-to-one road, generally one of the following two strategies is adopted: ① The LKA function directly exits and is taken over by the driver (one-to-two road). ② Virtual lane lines, and continue to drive forward for 1 second to 2 seconds (two-to-one road). For strategy ①, the superiority of assisted driving is not reflected; for strategy ②, it is possible that the function exits during the process of passing through the intersection, or when entering a new lane, there is a situation of crossing the line / riding on the line. Therefore, there is an urgent need for a method to enrich the traditional LKA function and improve the driving experience, so as to reasonably control the vehicle to drive within the lane. Summary of the Invention

[0004] Aiming at the defects in the prior art, the purpose of the present invention is to provide a lane centering driving control method and system. In a two-to-one or one-to-two scenario, based on the lane line geometric model, a side lane line is virtualized, and the feedforward-feedback algorithm is used to control the lateral operation of the vehicle, and smoothly drive into the new lane in the center.

[0005] To achieve the above technical effects, the present invention adopts the following technical solutions:

[0006] According to the first aspect of the present invention, a lane centering driving control method is provided, including the following steps: Step S1. Taking the vehicle centroid as the origin of the coordinate system, the forward direction of the vehicle as the x-axis, and the lateral direction of the vehicle as the y-axis to establish a vehicle coordinate system. The vehicle front camera continuously collects the information of the front lane lines, obtains the lateral offset distance C0, the lane line heading angle C1, and the lane line curvature C2 of each lane line in the vehicle coordinate system, and based on the above parameters, a geometric model of the lane line is established in the vehicle coordinate system: Among them, the variable x represents the distance along the forward direction of the vehicle (unit: meter); y represents the distance in the lateral direction of the vehicle (unit: meter), representing the lateral offset of the lane line relative to the vehicle's centroid. The model is used to describe the shape of the lane line in real time as the vehicle moves forward;

[0007] Step S2. Calculate the change in the heading angle ΔC1 and the change in curvature ΔC2 of the lane lines on both sides of the vehicle in real time. When one side of the lane line disappears and at the same time |ΔC1| > 0.08 or |ΔC2| > 0.01 of the lane line on the other side is detected, it is determined that the vehicle has entered the one-lane-splits-into-two / two-lanes-merge-into-one scenario;

[0008] Step S3. Generate a virtual lane line in the vehicle coordinate system based on C0 of the originally disappeared lane line, C1 and C2 of the lane line on the other side of the vehicle, and the lane line geometric model established in Step S1; Calculate the duration Ti of the virtual lane line according to the real-time vehicle speed V: Ti = 2.5 - V / 120, where Ti ∈ [1.5, 2.5] seconds and V is the real-time vehicle speed in the unit of Km / h; Within the time Ti, if the length L of the newly recognized lane line is ≥ 20 + 40*(2.5 - Ti) meters, it is determined that the newly recognized lane line has an effective length, and Step 4 is executed; otherwise, the driver is prompted to take over the vehicle;

[0009] Step S4. Within the duration Ti of the virtual lane line, execute the feedforward-feedback control algorithm, including: S4.1 Feedforward control: Calculate the front-wheel steering angle δ through the following formula: , where L is the wheelbase of the vehicle; e is the lateral distance of the preview point in the virtual lane line; l d is the straight-line distance of the preview point; At the same time, calculate the actual steering wheel angle : , where r is the steering wheel transmission ratio;

[0010] S4.2 Feedback control: Calculate the lateral deviation compensation angle and the heading deviation compensation angle to eliminate the lateral deviation and the heading deviation of the vehicle:

[0011] ; ;

[0012] Among them, KP1, KI1, KD1 are the coefficients of the lateral deviation PID controller, KP2, KI2, KD2 are the coefficients of the heading deviation PID controller, the variable ε is the real-time detected C0, and the variable β is the real-time detected C1; S4.3 Add the actual steering wheel angle to the lateral deviation compensation angle and the heading deviation compensation angle Weighted fusion to generate the total steering wheel angle , and execute it.

[0013] Optionally, the generation rule of the virtual lane line in step 3 is as follows: when the left lane line disappears, generate the left virtual lane line based on C0 of the original left lane line and C1 and C2 of the right lane line; when the right lane line disappears, generate the right virtual lane line based on C0 of the original right lane line and C1 and C2 of the left lane line.

[0014] Optionally, the straight-line distance l of the preview point in step 4.1 d is calculated by the following formula: ; where k is the gain coefficient; V is the real-time vehicle speed, in km / h; l b is the basic preview distance, which is 10 m.

[0015] Optionally, the lateral distance e of the preview point is calculated by the following method: Let the coordinates of the preview point in the lane line geometric model be (a, b), substitute this coordinate into the virtual lane line geometric model, and get b = C0 + C1 * a + 1 / 2 * C2 * a 2 , where the ordinate |b| of the preview point in the lane line geometric model is the lateral distance e of the preview point. Combining the coordinate value of the preview point with the straight-line distance l d of the geometric relationship formula , calculate the ordinate |b| of the preview point, that is, the lateral distance e of the preview point.

[0016] Optionally, in step 4.3, the total steering wheel angle is calculated according to the following formula: , where A and B are proportionality coefficients, and A + B = 1.

[0017] Optionally, the proportionality coefficient B is positively correlated with the lane line curvature C2.

[0018] According to the second aspect of the present invention, a lane centering driving control system is provided for implementing the above lane centering driving control method, which specifically includes the following components:

[0019] Environmental perception module: integrated with a forward camera, used to collect lane line parameters and vehicle speed information, and transmit the lane line parameters and vehicle speed information to the data processing module;

[0020] Data processing module: equipped with a lane modeling unit, a scene recognition unit and a virtual lane generation unit, used to analyze the data information transmitted by the environmental perception module and construct a road model; where:

[0021] The lane modeling unit is used to construct a geometric model of the lane line according to the lane line parameters provided by the environmental perception module;

[0022] The scene recognition unit is used to identify the changes in the lane lines on both sides of the vehicle to determine whether the vehicle enters the two-in-one / one-in-two scenario;

[0023] The virtual lane generation unit is used to construct virtual lane lines in the vehicle coordinate system according to the lane line parameters of the disappeared lane line and the lane line on the other side in the two lanes on both sides; at the same time, calculate the maintenance time of the virtual lane lines according to the real-time vehicle speed and identify whether there are new lane lines with effective lengths during the duration of the virtual lane lines;

[0024] The control decision-making module: built-in with a feedforward-feedback composite controller and a steering wheel angle conversion algorithm, is used to execute the feedforward-feedback control algorithm during the duration of the virtual lane lines, calculate the actual steering wheel angle, the lateral deviation compensation angle and the heading deviation compensation angle, and at the same time, according to the steering wheel angle conversion algorithm, weight and fuse the actual steering wheel angle, the lateral deviation compensation angle and the heading deviation compensation angle to generate the total steering wheel angle and transmit it to the actuator module for execution;

[0025] The actuator module: includes an electric power steering system EPS, and is responsible for receiving and executing the total steering wheel angle transmitted in the control decision-making module to smoothly drive the vehicle into the new lane.

[0026] Compared with the prior art, the present invention has the following beneficial effects:

[0027] The method and system provided by the present invention, according to the one-in-two and two-in-one lane environments, with the original lane lines and adjacent lane lines as references, virtualize the lane lines on the other side, plan a safe driving trajectory, smoothly control the longitudinal and lateral operations of the vehicle, and seamlessly realize the connection between the lane lines. On the one hand, it can avoid frequently asking the driver to take over the vehicle, affecting the driving experience, and on the other hand, it can also avoid the vehicle from pressing on or straddling the lane lines, resulting in driving risks. Description of the Drawings

[0028] By reading the detailed description of the non-limiting embodiments with reference to the following drawings, other features, purposes and advantages of the present invention will become more obvious:

[0029] Figure 1 It is a schematic diagram of the lane centering driving control method provided in the first embodiment;

[0030] Figure 2 It is a schematic diagram of virtualizing lane lines in the one-in-two scenario in the lane centering driving control method provided in the first embodiment;

[0031] Figure 3 It is a schematic diagram of virtualizing lane lines in the two-in-one scenario in the lane centering driving control method provided in the first embodiment;

[0032] Figure 4Schematic diagram of the lateral trajectory planning of the vehicle based on preview tracking in the first embodiment. Detailed implementation manners

[0033] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are some, but not all, of the embodiments of the present application. Components of the embodiments of the present application generally described and illustrated in the drawings herein may be arranged and designed in a variety of different configurations.

[0034] Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the present application claimed, but is merely representative of selected embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts fall within the scope of protection of the present application.

[0035] It should be noted that like reference numerals and letters denote like items in the following drawings. Therefore, once an item is defined in one drawing, it does not require further definition and explanation in subsequent drawings. In addition, all directional indications (such as up, down, left, right, front, back, bottom, etc.) in the present application are only used to explain the relative positional relationship and movement conditions between components in a specific posture (as shown in the drawings). If the specific posture changes, the directional indications will also change accordingly. Further, the descriptions involving "first", "second", etc. in the application are only for descriptive purposes and should not be construed as indicating or implying their relative importance or implicitly indicating the quantity of the indicated technical features.

[0036] The first embodiment

[0037] As Figure 1 shown, a lane centering driving control method is provided in this embodiment, which is applied to the merging and splitting scenarios of two-in-one / one-into-two lanes, and specifically includes the following steps:

[0038] Step 1. Road environment perception and lane modeling

[0039] Taking the vehicle's center of mass as the origin of the coordinate system, the forward direction of the vehicle as the x-axis, and the lateral direction of the vehicle as the y-axis, a vehicle coordinate system is established. In this coordinate system, the data information of the front lane lines is collected in real time through a forward camera, and the following parameters of each lane line are obtained: a) the lateral offset distance C0 of the lane line in the vehicle coordinate system; b) the angle between the tangent of the lane line and the forward direction (X-axis) of the vehicle in the vehicle coordinate system, that is, the lane line heading angle C1; c) the lane line curvature C2; d) the longitudinal distance DX_n from the proximal end of the lane line to the origin of the vehicle coordinate system; e) the longitudinal distance DX_f from the distal end of the lane line to the origin of the vehicle coordinate system. At most, the information of the following four lane lines is collected simultaneously: the left-left lane line (LL), the left lane line (L), the right lane line (R), and the right-right lane line (RR).

[0040] Based on the above parameters, the following lane line geometric model is established in the vehicle coordinate system: ; where the variable x represents the distance along the forward direction of the vehicle (unit: meter); y represents the distance in the lateral direction of the vehicle (unit: meter), representing the lateral offset of the lane line relative to the vehicle's center of mass. The above model is used to describe the shape of the lane line in real time as the vehicle moves. The origin of the coordinate system is the vehicle's center of mass. When the vehicle moves, the origin of the coordinate system (the vehicle's center of mass) will change continuously. Therefore, each time the camera collects data, the lane line parameters C0, C1, and C2 sensed by the camera each time will change in real time with the movement of the vehicle. For example, when the vehicle is going straight, if the lane line is straight, C1 and C2 may remain relatively stable, while C0 (lateral offset) may change with the position of the vehicle relative to the lane line. If the vehicle is turning, C1 and C2 will also reflect the change in the curvature of the lane line. In addition, if the vehicle deviates from the center of the lane, C0 will increase or decrease, depending on the deviation direction.

[0041] Step 2. Recognition of merge and split scenarios

[0042] As Figure 2 、 Figure 3 shown, in the scenarios of one split into two and two merging into one, one side of the original lane line will disappear, and significant changes will occur in C1 and C2 of one side of the original lane line. When the change amount of the lane line heading angle |ΔC1| or the change amount of the lane line curvature |ΔC2| exceeds a predetermined threshold, it is determined that the vehicle enters the one split into two / two merging into one scenario. Among them, ΔC1 and ΔC2 refer to the real-time change amounts of the lane line heading angle and curvature between two adjacent frames. For example, in the one split into two scenario shown in Figure 2 , the first lane line 1 disappears, and significant changes occur in C1 and C2 of the second lane line 2; in the two merging into one scenario shown in Figure 3 , the fourth lane line 4 disappears, and significant changes occur in C1 and C2 of the sixth lane line 6. The specific steps include:

[0043] During the driving of the vehicle, the change amount ΔC1 of the lane line heading angle and the change amount ΔC2 of the lane line curvature on both sides of the vehicle are calculated in real time. When it is recognized that one side of the lane line disappears and there is a sudden change in the lane line heading angle or lane curvature on the other side (|ΔC1| is greater than 0.08 rad or |ΔC2| is greater than 0.01), it is determined that the vehicle enters a one-to-two / two-into-one scenario.

[0044] Step 3. Generate a virtual lane line

[0045] As Figure 2 、 Figure 3 shown, after it is determined that the vehicle is in a one-to-two or two-into-one scenario, the original lane line on one side disappears, and a virtual lane line (the dashed line part in Figure 2 Figure 3 ) is constructed on the side where the lane line disappears. At the same time, the duration of the virtual lane line is set and the new lane line is recognized. Specifically, it includes:

[0046] 1. According to the above lane line geometric model , a virtual lane line is generated in the vehicle coordinate system. This virtual lane line maintains the lateral reference of the original disappearing lane line, that is, it inherits the original disappearing lane line C0 to ensure lateral continuity; at the same time, the C1 and C2 of the non-disappearing adjacent lane lines are dynamically used as the included angle and curvature of the virtual lane line. For example, in Figure 2 , when the vehicle enters the lane splitting scenario of one-to-two, the first lane line 1 disappears and the second lane line 2 mutates, and a left virtual lane line is generated. The left virtual lane line maintains the C0 of the original left first lane line 1 and uses the C1 and C2 of the second lane line 2. In Figure 3 , when the vehicle enters the two-into-one scenario, the fourth lane line 4 disappears and the sixth lane line 6 mutates, and a right virtual lane line is generated. The right virtual lane line maintains the C0 of the original fourth lane line 4 and uses the C1 and C2 of the sixth lane line 6. The generated virtual lane line satisfies the following geometric model in the vehicle coordinate system: , where the C0 in the geometric model of the virtual lane line uses the C0 value of the original disappearing lane line, and the C1 and C2 dynamically use the C1 and C2 values of the non-disappearing adjacent lane lines.

[0047] 2. Based on the vehicle speed of the host vehicle, the duration of the virtual lane line is set by interpolation so that the camera can recognize the new lane line. In this embodiment, the duration Ti of the virtual lane is determined according to the real-time vehicle speed V, and the duration is calculated according to the following formula: Ti = 2.5 - V / 120. As shown in Table 1, for safety considerations, the duration Ti of the virtual lane line is 1.5 seconds to 2.5 seconds.

[0048] Table 1

[0049] Vehicle speed V (km / h) 0 120 Duration Ti (s) 2.5 1.5

[0050] 3. Identify the effective length of the new lane line:

[0051] During the time of the virtual lane line, a new lane line with an effective length must be recognized ( Figure 2 the third lane line 3 in Figure 3 or the fifth lane line 5 in

[0052] Table 2

[0053] Time (s) 2.5 1.5 Minimum value of the effective length of the new lane (m) 20 60

[0054] In the duration Ti of the virtual lane line, when a new lane line with an effective length is recognized, execute the lateral control decision in step 4 to more smoothly and continuously control the vehicle to enter the new lane. If no new lane line is recognized or the effective length of the new lane line is insufficient during the duration, the lane keeping assist function exits and the driver is prompted to take over.

[0055] Step 4. Execution of the feedforward-feedback composite control algorithm

[0056] Based on the fitted virtual lane line and the real lane, execute the lateral control decision, and the lateral control decision adopts a feedforward + feedback control algorithm to realize the lateral control of the vehicle and keep the vehicle smoothly and safely enter the new lane. Specifically, it includes:

[0057] S4.1 Adopt feedforward control for lateral trajectory planning of the vehicle based on preview tracking as shown in Figure 4 specifically including:

[0058] 1. Set the coordinates of the preview point P on the virtual lane line as P(X r , Y r ), where X r represents the longitudinal distance of the preview point, and Y r represents the lateral distance e of the preview point. The abscissa and ordinate of this preview point satisfy the following lane line geometric equation: Y r =C0 + C1*X r +1 / 2*C2*X r 2 ; The straight-line distance between the preview point and the origin of the vehicle coordinate system is ld , l d is calculated by the following formula: ; where k is the gain coefficient (0.2); V is the real-time vehicle speed (unit: km / h); l b is the basic preview distance (10 m); at the same time, according to Figure 4 the geometric relationship in, it can be known that , combined with the virtual lane line equation Y r = C0 + C1*X r + 1 / 2*C2*X r 2 , the Y r value is calculated, that is, the lateral distance e of the preview point.

[0059] 2. Calculate the front wheel steering angle according to the following formula : ; where L is the wheelbase of the vehicle (unit: m); e is the lateral distance of the preview point; l d is the straight-line distance of the preview point.

[0060] The derivation process of the above formula for calculating the front wheel steering angle is as follows:

[0061] Based on geometric principles, the turning radius can be calculated according to the relationship between the preview point and the vehicle position , and then the front wheel steering angle is obtained through the wheelbase L of the vehicle and the turning radius .

[0062] First, it can be deduced by the sine theorem that , where l d is the straight-line distance of the preview point; is the angle between the preview point and the vehicle's forward direction; is the turning radius.

[0063] Furthermore, it is deduced that: ,

[0064] At the same time, from Figure 4 as shown, it can be obtained that , , L is the wheelbase of the vehicle; e is the lateral distance of the preview point.

[0065] Finally, the formula for calculating the front wheel steering angle is sorted out: .

[0066] 3. Convert the theoretical front wheel steering angle to the actual steering wheel steering angle according to the following formula: , where r is the steering wheel transmission ratio (fixed parameter).

[0067] S4.2 Perform feedback control based on real-time PID correction to eliminate the lateral deviation of the vehicle and the heading deviation , specifically including:

[0068] 1. Real-time monitor the lateral and heading deviation amounts of the current vehicle: Among them, the lateral deviation of the vehicle is C0 detected in real time, and the heading deviation is C1 (the included angle of the lane lines) detected in real time;

[0069] 2. Calculate the lateral deviation compensation angle and the heading deviation compensation angle to eliminate the lateral deviation of the vehicle and the heading deviation :

[0070]

[0071]

[0072] Among them, KP1, KI1, KD1 are the coefficients of the lateral deviation PID controller, and KP2, KI2, KD2 are the coefficients of the heading deviation PID controller. The variable ε is C0 detected in real time, and the variable β is C1 detected in real time.

[0073] S4.3 Weight and fuse the feedforward and feedback quantities to generate the steering wheel angle , and hand it over to the actuator EPS for execution;

[0074] Since the proportions of the feedforward and feedback in the system are different, coefficients (A, B) are required for adjustment. Obtain the total steering wheel angle required by the system according to the following formula :

[0075]

[0076] Among them, A and B are proportionality coefficients, and A + B = 1. The proportionality coefficients are related to the curvature (i.e., C2) - the greater the curvature, the greater B, to eliminate the predictable lateral deviation and heading deviation. Vice versa.

[0077] The Second Embodiment

[0078] This embodiment provides a lane centering driving control system for implementing the lane centering driving control method described in the first embodiment, specifically including the following parts:

[0079] Environmental perception module: Integrates a forward camera for collecting lane line parameters and vehicle speed information; in this module, a two-dimensional coordinate system is established with the vehicle's centroid as the origin, and five parameters of the forward lane line are collected through the forward camera: a) the lateral offset distance C0 of the lane line in the vehicle coordinate system; b) the angle between the tangent of the lane line and the vehicle's forward direction (X-axis) in the vehicle coordinate system, i.e., the lane line heading angle C1; c) the lane line curvature C2; d) the longitudinal distance DX_n from the proximal end of the lane line to the origin of the vehicle coordinate system; e) the longitudinal distance DX_f from the distal end of the lane line to the origin of the vehicle coordinate system, and the above collected parameters are uploaded to the data processing module.

[0080] Data processing module: Equipped with a lane modeling unit, a scene recognition unit, and a virtual lane generation unit, for parsing the data information transmitted by the environmental perception module and constructing a road model; among them:

[0081] The lane modeling unit is used to construct a geometric model of the lane line according to the lane line parameters provided by the environmental perception module: ;

[0082] The scene recognition unit is used to identify changes in the lane lines on both sides of the vehicle to determine whether it enters a two-in-one / one-in-two scenario. The specific recognition content is whether one side of the lane line disappears and the change in the lane line heading angle |ΔC1| or the change in the lane line curvature |ΔC2| on the other side exceeds a preset threshold;

[0083] The virtual lane generation unit is used to construct a virtual lane line in the vehicle coordinate system according to the C0 of the originally disappeared lane line in both lanes and the C1 and C2 of the lane line on the other side, and the virtual lane line satisfies the geometric model of the lane line: ; At the same time, calculate the maintenance time of the virtual lane line according to the real-time vehicle speed and identify whether there is a new lane line with an effective length during the duration of the virtual lane line.

[0084] Control decision-making module: Built-in a feedforward-feedback composite controller and a steering wheel angle conversion algorithm, for executing the feedforward-feedback control algorithm during the duration of the virtual lane line to calculate the actual steering wheel angle and the lateral deviation compensation angle and the heading deviation compensation angle , and at the same time, according to the steering wheel angle conversion algorithm, the actual steering wheel angle and the lateral deviation compensation angle and the heading deviation compensation angle are weighted and fused to generate the total steering wheel angle and it is transmitted to the actuator module for execution.

[0085] Actuator module: Includes an electric power steering system (EPS), responsible for receiving and executing the total steering wheel angle transmitted in the control decision-making module so that the vehicle can smoothly enter the new lane.

[0086] The specific embodiments of the present invention have been described above. Through the above description, relevant staff can make various changes and modifications completely within the scope of not deviating from the technical idea of the present invention.

Claims

1. A lane centering driving control method, characterized in that, It includes the following steps: Step S1. Establish a vehicle coordinate system with the vehicle's center of mass as the origin of the coordinate system, the vehicle's forward direction as the x-axis, and the vehicle's lateral direction as the y-axis. The vehicle's forward-facing camera continuously collects information on the lane lines ahead, obtaining the lateral offset distance C0, the lane line heading angle C1, and the lane line curvature C2 of each lane line in the vehicle coordinate system, and establishing a geometric model of the lane line in the vehicle coordinate system based on the above parameters: , where the variable x represents the distance along the vehicle's forward direction; y represents the distance in the vehicle's lateral direction, representing the lateral offset of the lane line relative to the vehicle's center of mass. The model is used to describe the shape of the lane line in real time as the vehicle moves forward; Step S2. Calculate the change in the heading angle ΔC1 and the change in curvature ΔC2 of the lane lines on both sides of the vehicle in real time. When one lane line disappears and at the same time |ΔC1| > 0.08 or |ΔC2| > 0.01 of the lane line on the other side is detected, it is determined that the vehicle enters the one-lane-to-two-lanes / two-lanes-to-one-lane scenario; Step S3. Generate a virtual lane line in the vehicle coordinate system based on C0 of the original disappearing lane line, C1 and C2 of the lane line on the other side of the vehicle, and the lane line geometric model established in Step S1; calculate the duration Ti of the virtual lane line according to the real-time vehicle speed V: Ti = 2.5 - V / 120, where Ti ∈ [1.5, 2.5] seconds and V is the real-time vehicle speed in the unit of Km / h; within the time of Ti, if the length L of the newly recognized lane line ≥ 20 + 40*(2.5 - Ti) meters, it is determined that the newly recognized lane line has an effective length and Step 4 is executed, otherwise the driver is prompted to take over the vehicle; Step S4. During the virtual lane line duration Ti, execute the feedforward-feedback control algorithm, including: S4.1 Feedforward control: Calculate the front wheel steering angle δ through the following formula: , where L is the wheelbase of the vehicle; e is the lateral distance of the preview point in the virtual lane line; l d is the straight-line distance of the preview point; At the same time, calculate the actual steering wheel angle : , where r is the steering wheel transmission ratio; S4.2 Feedback control: Calculate the lateral deviation compensation rotation angle through the following formula and the heading deviation compensation rotation angle to eliminate the lateral deviation of the vehicle and the heading deviation : ； ； Among them, KP1, KI1, KD1 are the coefficients of the lateral deviation PID controller, KP2, KI2, KD2 are the coefficients of the heading deviation PID controller, the variable ε is the C0 detected in real time, and the variable β is the C1 detected in real time; S4.3 The actual steering wheel angle is weighted and fused with the lateral deviation compensation angle and the heading deviation compensation angle to generate the total steering wheel angle , and execute it.

2. The method according to claim 1, wherein The generation rule of the virtual lane line in Step 3 is: when the left lane line disappears, generate a left-side virtual lane line based on C0 of the original left lane line, C1 and C2 of the right lane line; when the right lane line disappears, generate a right-side virtual lane line based on C0 of the original right lane line, C1 and C2 of the left lane line.

3. The method according to claim 1, wherein The straight-line distance l of the preview point in Step 4.1 d is calculated by the following formula: ; where k is the gain coefficient; V is the real-time vehicle speed in km / h; l b is the basic preview distance, which is 10 m.

4. The method according to claim 3, wherein The lateral distance e of the preview point is calculated by the following method: Assume that the coordinates of the preview point in the lane line geometric model are (a, b). Substitute these coordinates into the virtual lane line geometric model to obtain b = C0 + C1*a + 1 / 2*C2*a 2 , where the ordinate |b| of the preview point in the lane line geometric model is the lateral distance e of the preview point. Combine the coordinate values of the preview point and the geometric relationship formula d of the straight-line distance l to calculate the ordinate |b| of the preview point, that is, the lateral distance e of the preview point.

5. The method according to claim 1, wherein In the step 4.3, the total steering wheel angle is calculated according to the following formula: , where A and B are proportionality coefficients, and A + B = 1.

6. The method according to claim 5, wherein The proportionality coefficient B is positively correlated with the lane line curvature C2.

7. A lane centering driving control system, characterized in that, For implementing the lane centering driving control method described in any one of claims 1-6, it specifically includes the following components: Environmental perception module: integrated with a forward camera, used to collect lane line parameters and vehicle speed information, and transmit the lane line parameters and vehicle speed information to the data processing module; Data processing module: equipped with a lane modeling unit, a scenario recognition unit, and a virtual lane generation unit, used to analyze the data information transmitted by the environmental perception module and construct a road model; where: The lane modeling unit is used to construct a geometric model of the lane line according to the lane line parameters provided by the environmental perception module; The scenario recognition unit is used to identify the changes in the lane lines on both sides of the vehicle to determine whether the vehicle enters the two-lanes-to-one-lane / one-lane-to-two-lanes scenario; The virtual lane generation unit is used to construct a virtual lane line in the vehicle coordinate system according to the lane line parameters of the original disappearing lane line and the lane line on the other side of the vehicle; at the same time, calculate the maintenance time of the virtual lane line according to the real-time vehicle speed and identify whether there is a newly recognized lane line with an effective length within the duration of the virtual lane line; Control decision module: built-in with a feedforward-feedback composite controller and a steering wheel angle conversion algorithm, used to execute the feedforward-feedback control algorithm within the duration of the virtual lane line, calculate the actual steering wheel angle, the lateral deviation compensation angle, and the heading deviation compensation angle, and at the same time, according to the steering wheel angle conversion algorithm, weight and fuse the actual steering wheel angle, the lateral deviation compensation angle, and the heading deviation compensation angle to generate a total steering wheel angle and transmit it to the actuator module for execution; Actuator module: includes an electric power steering system EPS, responsible for receiving and executing the total steering wheel angle transmitted in the control decision module to make the vehicle smoothly drive into the new lane.

Images