Path planning method and device and unmanned equipment

Abstract

The invention provides a path planning method and device and unmanned equipment, and relates to the technical field of automatic driving. The path planning method comprises the following steps: acquiring state information of unmanned equipment at a starting moment and a destination moment; according to the state information of the unmanned equipment at the starting moment and the destination moment, generating a reference path of the unmanned equipment by adopting a curve fitting mode; determining a passable area according to the distance between the point on the reference path and the obstacle; in the passable area, solving a path planning model constructed by using an artificial potential field mode to obtain a solution of a driving path of the unmanned equipment, the path planning model taking a minimum sum of a gravitation index item and a repulsive force index item corresponding to points on the driving path as a target function, the gravitation index item being related to path smoothness, and the repulsive force index item being related to the path smoothness; the repulsive force index item is related to driving safety. Through the method, the real-time performance of path planning can be improved, and the stability of the generated path and the driving safety can be improved at the same time.

Description

Technical Field

[0001] The present disclosure relates to the field of autonomous driving technology, and in particular to a path planning method, device and unmanned driving equipment. Background Art

[0002] Unmanned driving equipment is used to automatically transport people or objects from one location to another. Unmanned driving equipment collects environmental information through sensors on the equipment and completes automatic transportation. Unmanned delivery vehicles controlled by unmanned driving technology for logistics transportation have greatly improved the convenience of production and life and saved labor costs.

[0003] In the field of autonomous driving, commonly used path planning algorithms include sampling, numerical optimization, search, curve fitting, etc. Among them, sampling, numerical, and search methods have high computational complexity and take a long time to plan paths, and are not suitable for driving scenarios with high real-time requirements. Curve fitting methods, such as the method of using polynomial functions to generate paths, are easier to express and process on computers, so this method has lower computational complexity and implementation cost than other complex algorithms, and is more suitable for driving scenarios with high real-time requirements.

[0004] In the related art, there are the following problems when generating paths based on curve fitting method: First,

[0005] Paths generated using high-order polynomials may introduce unnecessary oscillations, especially when the curvature of the path changes greatly or there are sharp turns. The oscillations are more obvious, which may affect the stability of the unmanned driving equipment and the comfort of the passengers. Secondly, when generating paths, driving safety and planning real-time are also issues that need to be considered. Summary of the invention

[0006] The present invention provides a path planning method, device and unmanned driving equipment.

[0007] According to a first aspect of the present disclosure, a path planning method is provided, comprising: obtaining status information of an unmanned driving device at a starting time and a destination time; generating a reference path of the unmanned driving device by using a curve fitting method according to the status information of the unmanned driving device at the starting time and the destination time; determining a traversable area according to the distance between a point on the reference path and an obstacle; within the traversable area, solving a path planning model constructed by using an artificial potential field method to obtain a solution for the driving path of the unmanned driving device, wherein the path planning model takes minimizing the sum of a gravitational index item and a repulsive index item corresponding to a point on the driving path as an objective function, the gravitational index item is related to path smoothness, and the repulsive index item is related to driving safety.

[0008] In some embodiments, the gravity index item is determined according to the following method: determining a first section vector between the point on the driving path and an adjacent path point located before the point; determining a second section vector between the point on the driving path and an adjacent path point located after the point; and determining the gravity index item corresponding to the point on the driving path based on the difference vector between the first section vector and the second section vector.

[0009] In some embodiments, determining the gravity index item corresponding to the point on the driving path based on the difference vector between the first road section vector and the second road section vector includes: using the modulus of the difference vector as the gravity corresponding to the point on the driving path; and determining the gravity index item based on the gravity corresponding to the point on the driving path.

[0010] In some embodiments, the path planning model also includes: a constraint condition that the gravity corresponding to the point on the driving path is less than or equal to a set function item, wherein the set function item is positively correlated with the average road section length corresponding to the reference path, and the set function item is negatively correlated with the minimum turning radius of the unmanned driving device.

[0011] In some embodiments, the repulsion index item is determined according to the following method: determining the distance between the point on the driving path and the nearest obstacle; when the distance is less than a distance threshold, determining the repulsion index item corresponding to the point on the driving path according to the ratio of the distance to the distance threshold, wherein the repulsion index item is negatively correlated with the ratio; when the distance is greater than or equal to the distance threshold, setting the repulsion index item to 0.

[0012] In some embodiments, determining the repulsion index item corresponding to the point on the driving path based on the ratio of the distance to the distance threshold includes: calculating the product of the maximum repulsion and the ratio; taking the difference between the maximum repulsion and the product as the repulsion corresponding to the point on the driving path; and determining the repulsion index item corresponding to the point on the driving path based on the repulsion corresponding to the point on the driving path.

[0013] In some embodiments, determining the traversable area based on the distance between the point on the reference path and the obstacle includes: selecting a plurality of reference points on the reference path;

[0014] A circle with the reference point as the center and the distance between the reference point and the nearest obstacle as the radius is used as the passable sub-area corresponding to the reference point; and the passable area is constructed according to the passable sub-areas corresponding to the multiple reference points.

[0015] In some embodiments, the curve fitting method is a polynomial fitting method, and generating a reference path for the unmanned driving device includes: generating multiple candidate paths using polynomial fitting based on state information of the unmanned driving device at a starting time and a destination time; determining the path cost of each of the multiple candidate paths based on a path cost evaluation function, the path cost evaluation function including at least one of a first cost item positively correlated with the degree of acceleration change corresponding to the path, a second cost item positively correlated with the acceleration corresponding to the path, a third cost item positively correlated with the curvature corresponding to the path, and a fourth cost item positively correlated with the speed corresponding to the path; selecting a candidate path with the smallest path cost from the multiple candidate paths as the reference path for the unmanned driving device.

[0016] In some embodiments, the generating multiple candidate paths by polynomial fitting according to the state information of the unmanned driving device at the starting time and the destination time includes: fitting the lateral trajectory of the unmanned driving device by a first polynomial function according to the state information of the unmanned driving device at the starting time and the destination time to obtain multiple lateral candidate trajectories; fitting the longitudinal trajectory of the unmanned driving device by a second polynomial function according to the state information of the unmanned driving device at the starting time and the destination time to obtain multiple longitudinal candidate trajectories; constructing the multiple candidate paths according to the multiple lateral candidate trajectories and the multiple longitudinal candidate trajectories.

[0017] In some embodiments, constructing the multiple candidate paths according to the multiple lateral candidate trajectories and the multiple longitudinal candidate trajectories includes: combining the multiple lateral candidate trajectories and the multiple longitudinal candidate trajectories to obtain a candidate path set; selecting the multiple candidate paths from the candidate path set according to a validity judgment condition, and the validity judgment condition includes at least one of the following: a maximum acceleration corresponding to the candidate path is lower than an acceleration threshold, a maximum speed corresponding to the candidate path is lower than a speed threshold, a maximum curvature corresponding to the candidate path is lower than a curvature threshold, and there is no interference between the candidate path and an obstacle.

[0018] According to a second aspect of the present disclosure, a path planning device is provided, comprising: a module for executing the path planning method as described above.

[0019] According to a third aspect of the present disclosure, a path planning device is provided, comprising: a memory; and a processor coupled to the memory, wherein the processor is configured to execute the path planning method as described above based on instructions stored in the memory.

[0020] According to a fourth aspect of the present disclosure, an unmanned driving device is provided, comprising: a path planning device as described above.

[0021] According to a fifth aspect of the present disclosure, a computer-readable storage medium is provided, on which computer instructions are stored, and when the instructions are executed by a processor, the path planning method as described above is implemented.

[0022] According to a sixth aspect of the present disclosure, there is provided a computer program product having computer instructions stored thereon, which implement the path planning method as described above when the instructions are executed by a processor.

[0023] Other features and advantages of the present disclosure will become apparent from the following detailed description of exemplary embodiments of the present disclosure with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The accompanying drawings, which constitute a part of the specification, illustrate embodiments of the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0025] Figure 1 is a flowchart of a path planning method according to some embodiments of the present disclosure;

[0026] Figure 2 is a schematic diagram of a process of generating a reference path based on a curve fitting method according to some embodiments of the present disclosure;

[0027] Figure 3 is a schematic diagram of a process of performing path optimization based on an artificial potential field method according to some embodiments of the present disclosure;

[0028] Figure 4 is a schematic diagram of the structure of a path planning device according to some embodiments of the present disclosure;

[0029] Figure 5 is a schematic structural diagram of a path planning device according to other embodiments of the present disclosure;

[0030] Figure 6 is a schematic diagram of the structure of a computer system according to some embodiments of the present disclosure;

[0031] Figure 7 is a schematic diagram of the structure of an unmanned driving device according to other embodiments of the present disclosure;

[0032] Figure 8 It is a schematic diagram of the three-dimensional structure of an unmanned driving device according to some embodiments of the present disclosure.

[0033] The present disclosure can be more clearly understood from the following detailed description with reference to the accompanying drawings. DETAILED DESCRIPTION

[0034] Various exemplary embodiments of the present disclosure will now be described in detail with reference to the accompanying drawings. It should be noted that the relative arrangement of components and steps, numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present disclosure unless otherwise specifically stated.

[0035] At the same time, it should be understood that for the convenience of description, the sizes of the various parts shown in the drawings are not drawn according to the actual proportional relationship.

[0036] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the present disclosure, its application, or uses.

[0037] Technologies, methods, and apparatus known to ordinary technicians in the relevant field may not be discussed in detail, but where appropriate, such technologies, methods, and apparatus should be considered part of the authorization specification.

[0038] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limiting. Therefore, other examples of the exemplary embodiments may have different values.

[0039] It should be noted that like reference numerals and letters refer to similar items in the following figures, and therefore, once an item is defined in one figure, it need not be further discussed in subsequent figures.

[0040] In order to make the objectives, technical solutions and advantages of the present disclosure more clearly understood, the present disclosure is further described in detail below in combination with specific embodiments and with reference to the accompanying drawings.

[0041] In order to solve the problems existing in the path planning algorithm in the related art, the present disclosure proposes a path planning method, device and unmanned driving equipment, which can improve the real-time performance of path planning while improving the smoothness and driving safety of the generated path.

[0042] Figure 1 FIG. 1 is a flow chart of a path planning method according to some embodiments of the present disclosure. Figure 1 As shown, the path planning method includes steps S11 to S14.

[0043] Step S11: Obtaining the status information of the unmanned driving device at the starting time and the destination time.

[0044] The path planning method can be executed by the unmanned driving device, or by other devices that communicate with the unmanned driving device (such as a vehicle management server). The unmanned driving device refers to a device with an unmanned driving function, such as an unmanned vehicle, a robot, etc.

[0045] In step S11, there are multiple ways to obtain the state information. In some examples, the state information is obtained by receiving the state information of the unmanned driving device at the start time and the destination time from other devices or other modules.

[0046] For example, the state information of the unmanned driving device at the starting time can be received from the positioning sensor on the unmanned driving device, and the state information of the unmanned driving device at the destination time can be obtained from the user input instruction.

[0047] For another example, the status information of the unmanned driving equipment at the starting time and the destination time can be received from the upper module of the path planning device.

[0048] Among them, the status information of the unmanned driving device at the starting time may include the position, speed and acceleration of the unmanned driving device at the starting time; the status information of the unmanned driving device at the destination time may include the position, speed and acceleration of the unmanned driving device at the destination time.

[0049] Step S12: Generate a reference path of the unmanned driving device using a curve fitting method according to the state information of the unmanned driving device at the starting time and the destination time.

[0050] The curve fitting method may be a polynomial fitting method or a fitting method based on other forms of curve functions.

[0051] In some examples, the curve fitting method is a polynomial fitting method. Figure 2 The process shown generates a reference path for the unmanned vehicle.

[0052] In other examples, the curve fitting method is a multi-order Bezier curve fitting method. In these examples, multiple candidate paths can be generated based on the multi-order Bezier curve fitting method, and then the path cost of each candidate path is evaluated. Next, a reference path of the unmanned driving device is determined from the multiple candidate paths according to the path cost.

[0053] Step S13: Determine the traversable area according to the distance between the point on the reference path and the obstacle.

[0054] In some examples, step S13 includes: selecting multiple reference points from the reference path; using a circle with the reference point as the center and the distance between the reference point and the nearest obstacle as the radius as the passable sub-area corresponding to the reference point; and constructing a passable area based on the passable sub-areas corresponding to the multiple reference points.

[0055] For example, n-1 points are set at equal intervals between the starting point and the end point of the reference path, and these n-1 points are used as reference points. For each reference point, a circle is drawn with the distance between the reference point and the nearest obstacle as the radius, and the circle is used as the passable sub-area corresponding to the reference point. Then, the entire passable sub-area corresponding to each reference point is used as the passable area of ​​the unmanned driving device. Where n is an integer greater than 2.

[0056] For example, n-1 points are set between the starting point and the end point of the reference path, and these n-1 points are used as reference points. For each reference point, a circle is drawn with the distance between the reference point and the nearest obstacle as the radius, and the circle is used as the passable sub-area corresponding to the reference point. Then, the minimum circumscribed rectangle of the entire passable sub-area corresponding to each reference point is used as the passable area of ​​the unmanned driving device. Where n is an integer greater than 2.

[0057] In the disclosed embodiment, a reference path is first generated based on a curve fitting method, and a traversable area for the unmanned driving device is generated based on the reference path, so that subsequent path planning based on APF is facilitated within a more accurate and safer narrow range, thereby helping to improve planning real-time performance and driving safety.

[0058] Step S14: In the traversable area, the path planning model constructed by using the artificial potential field method is solved to obtain a solution for the driving path of the unmanned driving device.

[0059] The basic idea of ​​artificial potential field (APF) is to regard the movement of unmanned driving equipment in the planning space as the movement of particles in a virtual force field. Obstacles or threat areas produce repulsive forces on unmanned driving equipment, and target positions produce attractive forces on unmanned driving equipment. Different from the traditional APF-based path planning algorithm, the embodiment of the present disclosure constructs a traversable area based on the existing reference path, and performs APF path planning in the traversable area. The traversable area serves as a candidate solution space for points on the driving path.

[0060] Among them, the path planning model constructed by the artificial potential field method takes the minimum sum of the gravitational index item and the repulsive index item corresponding to the points on the driving path as the objective function. Among them, the gravitational index item is related to the path smoothness, and the repulsive index item is related to driving safety. In addition, in the specific implementation, the path planning model can further include the following constraints: constraints on gravity.

[0061] For example, the constraint condition for gravity is: the gravity corresponding to the point on the driving path is less than or equal to the set function term. The set function term is positively correlated with the average section length corresponding to the driving path, and the set function term is negatively correlated with the minimum turning radius of the unmanned driving device. In addition, the path planning model can further include constraints on the position and orientation of the starting point and end point of the driving path.

[0062] In some examples, the gravity index item can be determined in the following manner: the gravity index item corresponding to the point on the driving path is determined according to the modulus of the difference vector between the first segment vector and the second segment vector, wherein the first segment vector refers to a vector formed by the segment between the point on the driving path and the adjacent path point before the point, and the second segment vector refers to a vector formed by the segment between the point on the driving path and the adjacent path point after the point.

[0063] For example, the modulus of the difference vector between the first road section vector and the second road section vector is used as the gravity corresponding to the point on the driving path; and the gravity index item is determined according to the gravity corresponding to the point on the driving path.

[0064] Further, in this example, determining the gravity index item according to the gravity corresponding to the point on the driving path may include: using the square of the gravity as the gravity index item. Alternatively, gravity may be directly used as the gravity index item.

[0065] In the disclosed embodiment, the point on the driving path is subjected to the gravitational force of two adjacent points to make the path straight. When the point on the path moves in the direction of gravity, the gravitational force decreases. When the gravitational force is equal to zero, the point on the driving path and the two adjacent points are on the same straight line, which is the "ideal smoothness". It can be seen that by calculating and characterizing the gravitational index item in the above manner, the path smoothness can be further improved.

[0066] In some examples, the repulsion index item can be determined in the following manner: determine the distance between a point on the driving path and the nearest obstacle; when the distance between the point on the driving path and the nearest obstacle is less than a distance threshold, determine the repulsion index item corresponding to the point on the driving path according to the ratio of the distance to the distance threshold, and the repulsion index item is negatively correlated with the ratio; when the distance between the point on the driving path and the nearest obstacle is greater than or equal to the distance threshold, set the repulsion index item to 0. The obstacle information can be obtained by reading map information.

[0067] For example, after determining that the distance between a point on the driving path and the nearest obstacle is less than a distance threshold, the repulsion index item is determined in the following manner: calculate the product of the ratio between the distance and the distance threshold and the maximum repulsion; use the difference between the maximum repulsion and the product as the repulsion corresponding to the point on the driving path; and determine the repulsion index item corresponding to the point on the driving path based on the repulsion corresponding to the point on the driving path.

[0068] Further, in this example, determining the repulsion index item according to the repulsion may include: taking the square of the repulsion as the repulsion index item. Alternatively, the repulsion may be directly taken as the repulsion index item.

[0069] In the disclosed embodiment, the point on the driving path is subjected to the repulsive force of the nearest obstacle to move the unmanned driving device away from the nearest obstacle. It can be seen that by calculating and characterizing the repulsive force index item in the above manner, driving safety can be further improved.

[0070] After constructing the APF-based path planning model, the model is solved in the traversable area to obtain the solution of the driving path.

[0071] In the disclosed embodiment, the reference path of the unmanned driving device is first planned based on the curve fitting method, and then the passable area is constructed based on the reference path, and the path planning model constructed based on the APF is solved in the passable area, thereby realizing the combination of the curve fitting method and the artificial potential field method. The above combination method is used to obtain the solution of the driving path, which can improve the stability and driving safety of the planned path while ensuring a certain degree of planning real-time performance.

[0072] Figure 2 FIG. 1 is a flow chart of generating a reference path based on a curve fitting method according to some embodiments of the present disclosure. Figure 2 As shown, generating a reference path based on a curve fitting method includes steps S121 to S123.

[0073] Step S121: Generate multiple candidate paths through polynomial fitting based on the state information of the unmanned driving device at the starting time and the destination time.

[0074] In some examples, step S121 includes: step a1, fitting the lateral trajectory of the unmanned driving device using a first polynomial function according to the state information of the unmanned driving device at the starting time and the destination time to obtain multiple lateral candidate trajectories; step a2, fitting the longitudinal trajectory of the unmanned driving device using a second polynomial function according to the state information of the unmanned driving device at the starting time and the destination time to obtain multiple longitudinal candidate trajectories; step a3, constructing multiple candidate paths based on the multiple lateral candidate trajectories and the multiple longitudinal candidate trajectories.

[0075] The lateral direction refers to the direction perpendicular to the vehicle's travel direction, and the longitudinal direction refers to the direction parallel to the vehicle's travel direction. The lateral candidate trajectory and the longitudinal candidate trajectory of the vehicle are planned separately.

[0076] In step a1, the first polynomial function may be a quintic polynomial. For example, the quintic polynomial may be specifically expressed as follows.

[0077]

[0078] Where a0, a1, a2, a3, a4, and a5 represent the parameters of the first polynomial function. The lateral velocity term can be obtained by first-order and second-order differentiation of the polynomial. and the lateral acceleration term . Then, the following formula group for lateral displacement term, lateral velocity term and lateral acceleration term can be obtained.

[0079]

[0080] Next, the above formula can be solved according to the lateral state information at the starting time and the destination time to obtain multiple lateral candidate trajectories of the unmanned driving device.

[0081] For example, the current lateral displacement, lateral velocity and lateral acceleration of the vehicle are used as the state information at the starting time, and then, according to the lateral displacement d(0), lateral velocity d(1) and lateral acceleration d(2) of the unmanned driving device at the starting time (t=0), the and lateral acceleration , determine the parameters a0, a1 and a2. These three parameters are shown in the following formula.

[0082]

[0083] Furthermore, the lateral speed of the unmanned driving device at the destination time (t=T) can be and lateral acceleration is zero, and the lateral displacement at the destination moment d target Sampling is performed within a set range of values ​​(for example, a range of values ​​determined by the road width). Parameters a3, a4, and a5 can be determined based on the state information of the unmanned driving device at the destination time. These three parameters are shown in the following formula.

[0084]

[0085] Since the lateral displacement at the destination time can take multiple possible values ​​and the destination time T can take multiple possible values, multiple lateral candidate trajectories can be determined.

[0086] In step a2, the second polynomial function may be a quartic polynomial. For example, the quartic polynomial may be specifically expressed as follows.

[0087]

[0088] Where b0, b1, b2, b3, b4 represent the parameters of the second polynomial function. The longitudinal velocity term and the longitudinal acceleration term can be obtained by first-order and second-order differentiation of the polynomial. Furthermore, the following can be obtained for the longitudinal displacement term s(t) and the longitudinal velocity term and the longitudinal acceleration term Formula group.

[0089]

[0090] Next, the above formula can be solved according to the longitudinal state information at the starting time and the destination time to obtain multiple longitudinal candidate trajectories of the unmanned driving device.

[0091] For example, the current longitudinal displacement, longitudinal velocity and longitudinal acceleration of the vehicle are used as the state information at the starting time, and then, according to the longitudinal displacement s(0), longitudinal velocity s(1) and longitudinal acceleration s(2) of the unmanned driving device at the starting time (t=0), the and longitudinal acceleration , determine the parameters b0, b1 and b2. These three parameters are shown in the following formula.

[0092]

[0093] Furthermore, the longitudinal acceleration of the unmanned driving device at the destination time (t=T) can be is zero, and the longitudinal velocity at the destination Sampling is performed within the set value range. According to the state information of the unmanned driving device at the destination time, parameters b3 and b4 can be determined. These two parameters are shown in the following formula.

[0094]

[0095] Since the longitudinal speed at the destination time can take multiple possible values ​​and the destination time T can take multiple possible values, multiple longitudinal candidate trajectories can be determined.

[0096] In step a3, the horizontal candidate trajectory and the vertical candidate trajectory are combined to obtain multiple candidate trajectories.

[0097] In some embodiments, after step S121 and before step S122, the following step is further included: screening the multiple candidate paths obtained in step S121 to select a candidate path that meets the validity determination condition. The validity determination condition includes at least one of the following: the maximum acceleration corresponding to the candidate path is lower than the acceleration threshold, the maximum speed corresponding to the candidate path is lower than the speed threshold, the maximum curvature corresponding to the candidate path is lower than the curvature threshold, and there is no interference between the candidate path and the obstacle.

[0098] For example, the validity judgment conditions include the following four conditions: the maximum acceleration of the candidate path is lower than the maximum acceleration that the vehicle can withstand; the maximum speed of the candidate path is lower than the maximum speed limit; the maximum curvature of the candidate path is lower than the maximum curvature; and the candidate path must not collide with other vehicles.

[0099] In the disclosed embodiment, by screening the multiple candidate paths obtained in step S121 according to the validity judgment condition, the number of candidate paths for subsequent path cost evaluation can be reduced, and the real-time performance of path planning can be improved. Moreover, by screening the candidate paths according to the validity judgment condition, it is helpful to select a more stable and safe candidate path.

[0100] Step S122: determining the path cost of each of the plurality of candidate paths according to the path cost evaluation function.

[0101] Among them, the path cost evaluation function includes at least one of a first cost item positively correlated with the degree of acceleration change corresponding to the path, a second cost item positively correlated with the acceleration corresponding to the path, a third cost item positively correlated with the curvature corresponding to the path, and a fourth cost item positively correlated with the speed corresponding to the path.

[0102] For example, the path cost evaluation function can be expressed as follows:

[0103]

[0104] in, J i represents the path cost evaluation function, Indicates i The first cost term of the candidate paths, Indicates i The second cost term of the candidate paths is Indicates i The third cost term of the candidate paths, Indicates i The fourth cost term of the candidate path is w jerk , w a , wk , w v represents the weight coefficient, s i (t) Indicates i The longitudinal displacement of the candidate paths, d i (t) Indicates i The lateral displacement of the candidate paths, Indicates i The acceleration of the candidate paths, Indicates i The curvature of the candidate paths, Indicates i The speed of the candidate paths, V limit is the speed threshold, and T represents the time.

[0105] In the above path cost evaluation function, the first cost item reflects the degree of change of trajectory acceleration. The smaller the item is, the higher the driving comfort is. The second cost item represents the acceleration cost of the trajectory, and the third cost item represents the curvature cost of the trajectory. The second and third cost items are evaluations of the dynamics of the unmanned driving equipment and are related to the continuity of the path. The fourth cost item represents the speed cost of the trajectory, which reflects the speed target of the unmanned driving equipment.

[0106] In the embodiment of the present disclosure, when evaluating the cost of the candidate path, the smoothness and continuity of the path are introduced as evaluation criteria. This improvement makes the generated path more reasonable and efficient, and overcomes the defects of insufficient smoothness and continuity of the path generated based on the polynomial programming method in the related art.

[0107] Step S123: Select a candidate path with the smallest path cost from multiple candidate paths as a reference path for the unmanned driving device.

[0108] In the disclosed embodiment, through the above process, a more stable and safe initial path (i.e., reference path) can be planned. This facilitates the subsequent more accurate determination of the traversable area, thereby not only reducing the amount of calculation for subsequent APF path planning and improving the real-time performance of path planning, but also helping to improve the stability and safety of the subsequent path obtained based on APF planning.

[0109] Figure 3 FIG. 1 is a flow chart of performing path optimization based on an artificial potential field method according to some embodiments of the present disclosure. Figure 3 As shown, the path optimization based on the artificial potential field method includes steps S141 to S142.

[0110] Step S141: constructing a path planning model based on APF.

[0111] In some examples, step S141 includes: step b1, defining the gravitational force exerted on points on the driving path based on APF; step b2, defining the repulsive force exerted on points on the driving path based on APF; and step b3, constructing a path planning model based on gravitational force and repulsive force.

[0112] In step b1, gravity can be defined according to the following formula:

[0113]

[0114] in, F att,k represents the gravitational force on the kth point on the driving path; F k-1 Indicates that the kth point on the driving path is affected by the k-1 The gravitational force of a point, F k-1 It can be represented by the first segment vector between the kth point and the k-1th point; F k Indicates that the kth point on the driving path is affected by the k+1 The gravitational force of a point, F k It can be represented by the second segment vector between the k+1th point and the kth point; P k Represents the coordinates of the kth point on the driving path; P k-1 Represents the coordinates of the k-1th point on the driving path; P k+1 Represents the coordinates of the k+1th point on the driving path.

[0115] A point on the driving path is affected by the gravitational force generated by the two adjacent points. The gravitational force acts to straighten the driving path, thereby making the path smooth. Although the driving path is constrained within the traversable area, the gravitational force may cause the path to approach obstacles. Therefore, it is necessary to use repulsion to keep the path away from obstacles.

[0116] In step b2, the repulsive force can be defined according to the following formula:

[0117]

[0118] in, F rep,k represents the repulsive force on the kth point on the driving path, K rep represents the maximum repulsive force, d(P k ,Ok ) represents the distance between the kth point on the driving path and the nearest obstacle. d th Indicates the distance threshold.

[0119] Although gravity makes the path smooth, some situations (such as U-turns) can cause the path to be pulled off the tip. Therefore, the optimization problem is subject to the turning radius constraint. Assuming that the points on the path are uniformly distributed and "sufficiently" dense on the path, there are some effective approximations: (1) ; (2) The steering angle of the unmanned driving device at the kth point on the driving path Smaller; (3) When the above assumptions are met, the relationship between gravity and curvature radius can be obtained as follows:

[0120]

[0121] in, F att,k represents the gravitational force on the kth point on the driving path, R k Indicates the turning radius corresponding to the kth point on the driving path.

[0122] Furthermore, let It is approximately equal to the average length d of the section between two adjacent points in the reference path (i.e. the average section length), then we can get:

[0123]

[0124]

[0125] Among them, Q k+1 represents the coordinates of the k+1th point on the reference path, Q k represents the coordinates of the kth point on the reference path, and n represents the number of points on the reference path.

[0126] Therefore, the constraints that the gravitational force on the points on the driving path needs to satisfy are:

[0127]

[0128] in, R min Indicates the minimum turning radius of the unmanned equipment.

[0129] In addition, in order to constrain the position and direction of the starting point and the end point of the driving path, constraints on the starting point and the end point can be added to the APF-based path planning model. Furthermore, the path planning model constructed in step b3 can be expressed as:

[0130]

[0131] In the above formula, is the objective function in the path planning model based on APF, which means that the optimization goal is to minimize the sum of the gravitational index item and the repulsive index item corresponding to the points on the driving path; the other contents in the above formula represent the constraints in the path planning model based on APF; P1 represents the first point on the driving path, Q1 represents the first point on the reference path, P2 represents the second point on the driving path, Q2 represents the second point on the reference path, P n-1 represents the n-1th point on the driving path, Q n-1 represents the n-1th point on the reference path, P n represents the nth point on the driving path, Q n represents the nth point on the reference path; P k is the variable to be solved, specifically representing the kth point on the driving path; C k Indicates P k The value range of is determined by the traversable area.

[0132] In addition, in some other examples, the objective function in the path planning model based on APF can also be expressed as: .

[0133] Step S142: solving the path planning model.

[0134] In some examples, the path planning model is solved based on the Sequential Least Squares Programming (SLSQP) method to obtain the optimal solution for the driving path. The SLSQP algorithm is a sequential quadratic programming algorithm based on nonlinear constrained gradient optimization. It gradually approaches the optimal solution by converting the original problem into a series of quadratic programming sub-problems and iteratively solving these sub-problems. By using the above algorithm to solve the path planning model, the solution speed can be improved, thereby improving the real-time performance of path planning.

[0135] In the disclosed embodiment, by obtaining a reference path based on a curve fitting method and combining it with an APF-based path planning algorithm, it is possible to improve the real-time performance of path planning while improving the smoothness and driving safety of the generated path.

[0136] Figure 4 is a schematic diagram of the structure of a path planning device according to some embodiments of the present disclosure. Figure 4As shown, the path planning device 40 is used to execute the path planning method as described above, including an acquisition module 410, a first planning module 420, a determination module 430, and a second planning module 440.

[0137] The acquisition module 410 is configured to acquire the status information of the unmanned driving device at the starting time and the destination time.

[0138] The first planning module 420 is configured to generate a reference path for the unmanned driving device by using a curve fitting method according to the state information of the unmanned driving device at the starting time and the destination time.

[0139] Regarding how to generate a reference path for an unmanned driving device, please refer to the relevant description of the aforementioned embodiment.

[0140] The determination module 430 is configured to determine the traversable area according to the distance between the point on the reference path and the obstacle.

[0141] Regarding how to determine the passable area, please refer to the relevant description of the above embodiment.

[0142] The second planning module 440 is configured to solve the path planning model constructed by using the artificial potential field method in the traversable area to obtain a solution for the driving path of the unmanned driving device.

[0143] Among them, the path planning model takes the minimum sum of the gravitational index item and the repulsive index item corresponding to the points on the driving path as the objective function. The gravitational index item is related to the path smoothness, and the repulsive index item is related to driving safety.

[0144] In the disclosed embodiment, the above device can improve the real-time performance of path planning while improving the smoothness and driving safety of the generated path.

[0145] Figure 5 Schematic diagram of the structure of the path planning device according to other embodiments of the present disclosure. Figure 5 As shown, the path planning device 50 includes a memory 510 and a processor 520 coupled to the memory 510. The memory 510 is used to store instructions for executing the path planning method corresponding to the embodiment. The processor 520 is configured to execute the path planning method in any of the embodiments of the present disclosure based on the instructions stored in the memory 510.

[0146] Figure 6 FIG. 1 is a schematic diagram of the structure of a computer system according to some embodiments of the present disclosure. Figure 6 As shown, the path planning device can be embodied in the form of a general-purpose computing device. The computer system 60 includes a memory 610, a processor 620, and a bus 630 that connects different system components.

[0147] The memory 610 may include, for example, a system memory, a non-volatile storage medium, etc. The system memory may store, for example, an operating system, an application program, a boot loader, and other programs. The system memory may include a volatile storage medium, such as a random access memory (RAM) and / or a cache memory. The non-volatile storage medium may store, for example, instructions for executing at least one of the corresponding embodiments of the path planning method. The non-volatile storage medium may include, but is not limited to, a disk memory, an optical memory, a flash memory, etc.

[0148] The processor 620 may be implemented by a general purpose processor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistors, and other discrete hardware components. Accordingly, each module such as the acquisition module, the first planning module, etc. may be implemented by a central processing unit (CPU) running instructions in a memory for executing corresponding steps, or may be implemented by a dedicated circuit for executing corresponding steps.

[0149] The bus 630 may use any of a variety of bus architectures, including, but not limited to, an Industry Standard Architecture (ISA) bus, a Micro Channel Architecture (MCA) bus, and a Peripheral Component Interconnect (PCI) bus.

[0150] The interfaces 640, 650, 660 in the computer system 60 and the memory 610 and the processor 620 can be connected via a bus 630. The input / output interface 640 can provide a connection interface for input / output devices such as a display, a mouse, and a keyboard. The network interface 650 provides a connection interface for various networked devices. The storage interface 660 provides a connection interface for external storage devices such as a floppy disk, a USB flash drive, and an SD card.

[0151] Figure 7 is a schematic diagram of the structure of an unmanned driving device according to some embodiments of the present disclosure; Figure 8 Schematic diagram of the three-dimensional structure of unmanned driving equipment according to some embodiments of the present disclosure. Figure 7 and Figure 8 The unmanned driving device provided in the embodiment of the present disclosure is described.

[0152] As attached Figure 7 and 8 As shown, the unmanned driving device can be an unmanned vehicle, including a chassis module 710, an automatic driving module 720, a cargo box module 730, and a remote monitoring streaming module 740. In addition, in a specific implementation, the unmanned driving device can also be a robot or other equipment.

[0153] In some embodiments, the chassis module 710 mainly includes a battery, a power management device, a chassis controller, a motor driver, and a power motor. The battery provides power for the entire unmanned vehicle system, and the power management device converts the battery output into different voltage levels that can be used by each functional module and controls power on and off. The chassis controller receives motion instructions issued by the automatic driving module and controls the steering, forward, backward, braking, etc. of the unmanned vehicle.

[0154] In some embodiments, the autonomous driving module 720 includes a core processing unit (Orin or Xavier module), a traffic light recognition camera, front, rear, left, and right surround cameras, a multi-line laser radar, a positioning module (such as Beidou, GPS, etc.), and an inertial navigation unit. The camera and the autonomous driving module can communicate with each other, and GMSL link communication can be used to increase the transmission speed and reduce the wiring harness.

[0155] In some embodiments, the autonomous driving module 720 includes the path planning device in the above embodiments.

[0156] In some embodiments, the remote monitoring streaming module 730 is composed of a front monitoring camera, a rear monitoring camera, a left monitoring camera, a right monitoring camera and a streaming module, which transmits the video data collected by the monitoring camera to the backend server for viewing by the backend operator. The wireless communication module communicates with the backend server through an antenna, and the backend operator can remotely control the unmanned vehicle.

[0157] The cargo box module 740 is a cargo carrying device for the unmanned vehicle. In some embodiments, a display interaction module is also provided on the cargo box module 740, and the display interaction module is used for the interaction between the unmanned vehicle and the user. The user can perform operations such as picking up, storing, and purchasing goods through the display interaction module. The type of cargo box can be changed according to actual needs. For example, in a logistics scenario, the cargo box can include multiple sub-boxes of different sizes, and the sub-boxes can be used to load goods for delivery. In a retail scenario, the cargo box can be set as a transparent box so that users can intuitively see the products for sale.

[0158] The unmanned driving equipment of the disclosed embodiment can improve the real-time performance of path planning while improving the smoothness and driving safety of the generated path.

[0159] Here, various aspects of the present disclosure are described with reference to flowcharts and / or block diagrams of methods, devices, and computer program products according to embodiments of the present disclosure. It should be understood that each frame of the flowchart and / or block diagram and the combination of frames can be implemented by computer-readable program instructions.

[0160] These computer-readable program instructions may be provided to a processor of a general-purpose computer, a special-purpose computer or other programmable device to produce a machine, so that the processor executes the instructions to produce means for implementing the functions specified in one or more blocks in the flowchart and / or block diagram.

[0161] These computer-readable program instructions may also be stored in a computer-readable memory, which cause the computer to work in a specific manner to produce an article of manufacture, including instructions for implementing the functions specified in one or more blocks in the flowchart and / or block diagram.

[0162] The present disclosure can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects.

[0163] So far, the path planning method, device and unmanned driving equipment according to the present disclosure have been described in detail. In order to avoid obscuring the concept of the present disclosure, some details known in the art are not described. Based on the above description, those skilled in the art can fully understand how to implement the technical solution disclosed here.

Claims

1. A path planning method, comprising: Obtaining status information of the unmanned driving device at the starting time and the destination time; Generate a reference path of the unmanned driving device by using a curve fitting method according to the state information of the unmanned driving device at the starting time and the destination time; Determining a traversable area based on the distance between the point on the reference path and the obstacle; In the passable area, a path planning model constructed using an artificial potential field method is solved to obtain a solution to the driving path of the unmanned driving device, wherein the path planning model takes the minimization of the sum of the gravitational index item and the repulsive index item corresponding to the points on the driving path as an objective function, the gravitational index item is related to the path smoothness, and the repulsive index item is related to driving safety.

2. The path planning method according to claim 1, wherein: The gravity index item is determined according to the following method: Determine a first route segment vector between a point on the driving path and an adjacent route point preceding the point; Determine a second route segment vector between a point on the driving path and an adjacent route point located after the point; According to the difference vector between the first section vector and the second section vector, a gravity index item corresponding to the point on the driving path is determined.

3. The path planning method according to claim 2, wherein: The determining, according to the difference vector between the first road section vector and the second road section vector, of the gravity index item corresponding to the point on the driving path comprises: Using the modulus of the difference vector as the gravitational force corresponding to the point on the driving path; The gravity index item is determined according to the gravity corresponding to the point on the driving path.

4. The path planning method according to claim 3, wherein: The path planning model also includes: The constraint condition is that the gravity corresponding to the point on the driving path is less than or equal to a set function term, wherein the set function term is positively correlated with the average section length corresponding to the reference path, and the set function term is negatively correlated with the minimum turning radius of the unmanned driving device.

5. The path planning method according to any one of claims 1 to 4, wherein: The repulsion index item is determined according to the following method: Determine the distance between a point on the driving path and the nearest obstacle; When the distance is less than a distance threshold, determining a repulsion index item corresponding to the point on the driving path according to a ratio of the distance to the distance threshold, wherein the repulsion index item is negatively correlated with the ratio; When the distance is greater than or equal to the distance threshold, the repulsion index item is set to 0.

6. The path planning method according to claim 5, wherein: Determining the repulsion index item corresponding to the point on the driving path according to the ratio of the distance to the distance threshold includes: calculating the product of the maximum repulsive force and the ratio; taking the difference between the maximum repulsive force and the product as the repulsive force corresponding to the point on the driving path; According to the repulsive force corresponding to the point on the driving path, a repulsive force index item corresponding to the point on the driving path is determined.

7. The path planning method according to claim 1, wherein: Determining the traversable area according to the distance between the point on the reference path and the obstacle includes: Selecting a plurality of reference points from the reference path; A circle with the reference point as the center and the distance between the reference point and the nearest obstacle as the radius is used as the passable sub-area corresponding to the reference point; A traversable area is constructed according to the traversable sub-areas corresponding to the multiple reference points.

8. The path planning method according to claim 1, wherein: The curve fitting method is a polynomial fitting method, and generating a reference path for the unmanned driving device includes: Generate multiple candidate paths using polynomial fitting according to the state information of the unmanned driving device at the starting time and the destination time; Determining a path cost of each of the plurality of candidate paths according to a path cost evaluation function, the path cost evaluation function comprising at least one of a first cost item positively correlated with a degree of acceleration change corresponding to the path, a second cost item positively correlated with an acceleration corresponding to the path, a third cost item positively correlated with a curvature corresponding to the path, and a fourth cost item positively correlated with a speed corresponding to the path; A candidate path with the smallest path cost is selected from the multiple candidate paths as a reference path for the unmanned driving device.

9. The path planning method according to claim 8, wherein: The generating of multiple candidate paths by polynomial fitting according to the state information of the unmanned driving device at the starting time and the destination time includes: Fitting the lateral trajectory of the unmanned driving device using a first polynomial function according to state information of the unmanned driving device at a starting time and a destination time to obtain a plurality of candidate lateral trajectories; According to the state information of the unmanned driving device at the starting time and the destination time, a second polynomial function is used to fit the longitudinal trajectory of the unmanned driving device to obtain a plurality of longitudinal candidate trajectories; The plurality of candidate paths are constructed according to the plurality of transverse candidate trajectories and the plurality of longitudinal candidate trajectories.

10. The path planning method according to claim 9, wherein: The constructing the plurality of candidate paths according to the plurality of horizontal candidate trajectories and the plurality of longitudinal candidate trajectories comprises: Combining the plurality of lateral candidate trajectories and the plurality of longitudinal candidate trajectories to obtain a candidate path set; The multiple candidate paths are selected from the candidate path set according to validity judgment conditions, and the validity judgment conditions include at least one of the following: the maximum acceleration corresponding to the candidate path is lower than the acceleration threshold, the maximum speed corresponding to the candidate path is lower than the speed threshold, the maximum curvature corresponding to the candidate path is lower than the curvature threshold, and there is no interference between the candidate path and the obstacle.

11. A path planning device, comprising: Used to execute the path planning method described in any one of claims 1 to 10.

12. A path planning device, comprising: Memory; as well as A processor coupled to the memory, wherein the processor is configured to execute the path planning method according to any one of claims 1 to 10 based on instructions stored in the memory.

13. An unmanned driving device, comprising: A path planning device as claimed in claim 11 or 12.

14. A computer-readable storage medium having computer instructions stored thereon, which, when executed by a processor, implement the path planning method as described in any one of claims 1 to 10.

15. A computer program product having computer program instructions stored thereon, wherein the instructions are executed by a processor to implement the path planning method according to any one of claims 1 to 10.

Images