Intelligent obstacle avoidance method and system for autonomous vehicle

Abstract

An intelligent obstacle avoidance method for an autonomous vehicle, comprising: acquiring driving parameters of an autonomous vehicle (S001); acquiring a minimum obstacle avoidance area of the autonomous vehicle on the basis of the driving parameters of the autonomous vehicle, calculating an obstacle occupancy value of the minimum obstacle avoidance area of the autonomous vehicle at each moment, calculating a directional collision coefficient of the minimum obstacle avoidance area of the autonomous vehicle at each moment, and acquiring an autonomous vehicle occupancy and collision grid map on the basis of the directional collision coefficient of the minimum obstacle avoidance area of the autonomous vehicle (S002); calculating a collision dynamic adjustment coefficient on the basis of the autonomous vehicle occupancy and collision grid map, and calculating an obstacle avoidance search distance of the minimum obstacle avoidance area of the autonomous vehicle (S003); and using the obstacle avoidance search distance of the minimum obstacle avoidance area of the autonomous vehicle to perform obstacle avoidance path dynamic search planning on the autonomous vehicle at each moment (S004), so that the accuracy in the obstacle avoidance process of the autonomous vehicle is ensured. Also disclosed is an intelligent obstacle avoidance system for an autonomous vehicle.

Description

Intelligent obstacle avoidance method and system for unmanned vehicles Technical Field

[0001] The present invention relates to the field of unmanned driving technology, and in particular to an intelligent obstacle avoidance method and system for an unmanned vehicle. Background Art

[0002] Autonomous vehicle technology, based on new energy vehicles, integrates high-precision sensors, transportation, communications, and other diverse fields. With the continued maturity of computer hardware and optimization of artificial intelligence algorithms, autonomous vehicles will be used in a variety of common life scenarios. However, in actual operation, autonomous vehicles face complex road conditions. The random distribution of obstacles can disrupt the vehicle's operation and affect its accuracy in avoiding obstacles.

[0003] In the traditional unmanned vehicle obstacle avoidance path planning process, the grid map cannot accurately represent the distribution of obstacles, which affects the rapid exploration of the random tree algorithm for the obstacle avoidance route search range, resulting in poor accuracy of the unmanned vehicle obstacle avoidance path planning.

[0004] Summary of the Invention

[0005] The present invention provides an intelligent obstacle avoidance method and system for unmanned vehicles to solve the above problems. The technical solutions adopted are as follows:

[0006] In a first aspect, an embodiment of the present invention provides an intelligent obstacle avoidance method for an unmanned vehicle, the method comprising the following steps:

[0007] Obtaining driving parameters of unmanned vehicles;

[0008] Obtaining a minimum obstacle avoidance area of ​​the unmanned vehicle according to the driving parameters of the unmanned vehicle, calculating an obstacle occupancy value of the minimum obstacle avoidance area of ​​the unmanned vehicle at each moment according to the minimum obstacle avoidance area of ​​the unmanned vehicle, calculating a directional collision coefficient of the minimum obstacle avoidance area of ​​the unmanned vehicle at each moment according to the obstacle occupancy value of the minimum obstacle avoidance area of ​​the unmanned vehicle, and obtaining a collision grid map occupied by the unmanned vehicle according to the directional collision coefficient of the minimum obstacle avoidance area of ​​the unmanned vehicle;

[0009] Calculate the collision dynamic adjustment coefficient of the minimum obstacle avoidance area of ​​the unmanned vehicle at each moment based on the collision grid map occupied by the unmanned vehicle, and calculate the obstacle avoidance search distance of the minimum obstacle avoidance area of ​​the unmanned vehicle based on the collision dynamic adjustment coefficient of the minimum obstacle avoidance area of ​​the unmanned vehicle;

[0010] The obstacle avoidance search distance of the minimum obstacle avoidance area of ​​the unmanned vehicle is used to dynamically search and plan the obstacle avoidance path of the unmanned vehicle at each moment.

[0011] Preferably, the unmanned vehicle driving parameters include: the unmanned vehicle driving speed at each moment, the unmanned vehicle driving direction angle, the obstacle distance, the perceptible limit detection distance, and the vehicle body length.

[0012] Preferably, the method for obtaining the minimum obstacle avoidance area of ​​the unmanned vehicle according to the driving parameters of the unmanned vehicle is:

[0013] The result of rounding up the ratio of the unmanned vehicle's perceptible limit detection distance to the unmanned vehicle's body length at each moment is recorded as the first length. In a square area centered on the unmanned vehicle's position at each moment and with the unmanned vehicle's perceptible limit detection distance as its side length, the minimum obstacle avoidance area of ​​the unmanned vehicle is divided with the first length as its side length.

[0014] Preferably, the method for calculating the obstacle occupancy value of the minimum obstacle avoidance area of ​​the unmanned vehicle at each moment according to the minimum obstacle avoidance area of ​​the unmanned vehicle is:

[0015] In the above formula, Oc j Indicates the obstacle occupancy value of the jth minimum obstacle avoidance area, Dt i,j represents the Manhattan distance between the coordinates of the center point of the jth minimum obstacle avoidance area and the position of the unmanned vehicle at the i-th moment, σ i It represents the variance of all distances between different positions within the obstacle avoidance sensing range of the unmanned vehicle at the current time i and the current position of the unmanned vehicle. exp() represents an exponential function with a natural constant as the base. Pd i,j It represents the collision risk index of the jth area in the extreme obstacle avoidance perception range of the unmanned vehicle at the current time i.

[0016] Preferably, the method for obtaining the obstacle occupancy value of the minimum obstacle avoidance area is:

[0017] A preset number of sampling points is set in the minimum obstacle avoidance area of ​​the unmanned vehicle, and the number of sampling points with values ​​of the preset value in the minimum obstacle avoidance area is recorded as the obstacle occupancy value of the minimum obstacle avoidance area.

[0018] Preferably, the method of calculating the collision coefficient of the minimum obstacle avoidance area of ​​the unmanned vehicle at each moment according to the obstacle occupancy value of the minimum obstacle avoidance area of ​​the unmanned vehicle and obtaining the collision grid map occupied by the unmanned vehicle according to the collision coefficient of the minimum obstacle avoidance area of ​​the unmanned vehicle is:

[0019] In the above formula, Norm() represents the normalization function, v i represents the speed of the unmanned vehicle at the current moment i, Dt i,j It represents the Manhattan distance between the coordinates of the center point of the jth minimum obstacle avoidance area and the position of the unmanned vehicle at the i-th moment, n j is the preset experience value, Oc j It represents the obstacle occupancy value of the jth minimum obstacle avoidance area, Oc k It represents the obstacle occupancy value of the kth minimum obstacle avoidance area adjacent to the jth minimum obstacle avoidance area, Min j It represents the minimum obstacle occupancy value in the jth minimum obstacle avoidance area and its adjacent obstacle avoidance areas. It represents the driving direction angle θ of the unmanned vehicle i The directional collision coefficient of the next j-th minimum obstacle avoidance area;

[0020] All minimum obstacle avoidance areas under the driving direction angle of the unmanned vehicle are recorded as the first set. The sum of the collision risk index and the directional collision coefficient of all minimum obstacle avoidance areas in the first set is recorded as the occupied collision value of the corresponding minimum obstacle avoidance area. The collision risk index of all minimum obstacle avoidance areas that do not belong to the first set is recorded as the occupied collision value of the corresponding minimum obstacle avoidance area. The occupied collision values ​​of the minimum obstacle avoidance areas are used to construct the unmanned vehicle occupied collision grid map.

[0021] Preferably, the method for calculating the collision dynamic adjustment coefficient of the minimum obstacle avoidance area of ​​the unmanned vehicle at each moment according to the collision grid map occupied by the unmanned vehicle is:

[0022] In the above formula, Norm() represents the normalization function, Max i,j It represents the maximum occupied collision value in the local collision area centered on the jth minimum obstacle avoidance area in the collision grid map occupied by the unmanned vehicle at time i, Min i,j represents the minimum occupied collision value in the local collision area centered on the jth minimum obstacle avoidance area in the collision grid map occupied by the unmanned vehicle at time i, b represents the total number of different minimum obstacle avoidance areas in the local collision area, Pc i,j It represents the occupied collision value of the jth minimum obstacle avoidance area at the center of the local collision area at time i, Pc i,k It represents the occupied collision value of the kth minimum obstacle avoidance area in the local collision area at time i, csd i,j It represents the collision dynamic adjustment coefficient of the jth minimum obstacle avoidance area at time i, where the local collision area refers to a local area with a preset length centered on each minimum obstacle avoidance area.

[0023] Preferably, the method of calculating the obstacle avoidance search distance of the minimum obstacle avoidance area of ​​the unmanned vehicle according to the collision dynamic adjustment coefficient of the minimum obstacle avoidance area of ​​the unmanned vehicle and using the obstacle avoidance search distance of the minimum obstacle avoidance area of ​​the unmanned vehicle to perform dynamic search planning of the obstacle avoidance path of the unmanned vehicle at each moment is:

[0024] The sum of the first preset constant and the collision dynamic adjustment coefficient of the minimum obstacle avoidance area is recorded as a first sum value, the product of the first sum value and the second preset constant is recorded as a first product, the difference between the first preset constant and the collision dynamic adjustment coefficient of the minimum obstacle avoidance area is recorded as a first difference, and the product of the first difference and the second preset constant is recorded as a second product. When the occupied collision value of the minimum obstacle avoidance area of ​​the unmanned vehicle is greater than or equal to the collision threshold in the unmanned vehicle occupied collision grid map, the obstacle avoidance search distance of the minimum obstacle avoidance area of ​​the unmanned vehicle is set to the first product. When the occupied collision value of the minimum obstacle avoidance area of ​​the unmanned vehicle is less than the collision threshold in the unmanned vehicle occupied collision grid map, the obstacle avoidance search distance of the minimum obstacle avoidance area of ​​the unmanned vehicle is set to the second product.

[0025] The collision grid map occupied by the unmanned vehicle is used as the input of the fast exploration random tree algorithm, and the position of the unmanned vehicle is used as the initial position. A path search is performed in the area with each minimum obstacle avoidance area as the center and the obstacle avoidance search distance as the radius to obtain the obstacle avoidance path curve of the unmanned vehicle at the current time i.

[0026] Preferably, the method for obtaining the collision threshold in the collision grid map occupied by the unmanned vehicle is:

[0027] The collision grid map occupied by the unmanned vehicle is used as the input of the Otsu method to obtain the collision threshold in the collision grid map occupied by the unmanned vehicle.

[0028] In a second aspect, an embodiment of the present invention further provides an intelligent obstacle avoidance system for an unmanned vehicle, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of any one of the above methods when executing the computer program.

[0029] The beneficial effects of the present invention are as follows: the present invention obtains the minimum obstacle avoidance area of ​​the unmanned vehicle through the driving parameters of the unmanned vehicle and calculates the obstacle occupancy value of the minimum obstacle avoidance area of ​​the unmanned vehicle and the collision coefficient of the direction of the minimum obstacle avoidance area of ​​the unmanned vehicle, and at the same time constructs and obtains the unmanned vehicle occupation collision grid map to calculate and characterize the obstacle occupancy in the path of the unmanned vehicle during the driving process, and better represents the obstacle distribution characteristics of the unmanned vehicle during the driving process by analyzing the route obstacle distribution in the driving direction of the unmanned vehicle at different times. Furthermore, the present invention dynamically adjusts the search range of the obstacle dense area and the sparse area in the unmanned vehicle occupation collision grid map, thereby improving the rationality and accuracy of the path search planning during the driving process of the unmanned vehicle. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0031] FIG1 is a schematic flow chart of an intelligent obstacle avoidance method for an unmanned vehicle provided by one embodiment of the present invention;

[0032] FIG2 is a schematic diagram of the obstacle avoidance driving direction of an unmanned vehicle provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0033] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0034] Please refer to FIG1 , which shows a flow chart of an intelligent obstacle avoidance method for an unmanned vehicle provided by one embodiment of the present invention. The method includes the following steps:

[0035] Step S001: Obtain driving parameters of the unmanned vehicle.

[0036] It should be noted that during the driving process of the unmanned vehicle, in order to ensure the safety and reliability of the unmanned vehicle, a variety of sensors will be set up to monitor the operating status of the unmanned vehicle. Therefore, through the sensors, some driving status parameters of the unmanned vehicle can be obtained during the driving process, including the driving speed of the unmanned vehicle at each moment, the driving direction angle of the unmanned vehicle, the obstacle distance, the sensing limit detection distance, and the vehicle body length. The driving speed of the unmanned vehicle at time i is denoted as v i , the driving direction angle of the unmanned vehicle at time i is θ i , assuming that the distances between the unmanned vehicle and n different obstacles at the current moment i are recorded as d from near to far i,1 ,d i,2 ,d i,3 …d i,n , assuming that the perceptible limit detection distance is d max , the length of the unmanned vehicle body is L.

[0037] Step S002: obtaining the minimum obstacle avoidance area of ​​the unmanned vehicle according to the driving parameters of the unmanned vehicle, calculating the obstacle occupancy value of the minimum obstacle avoidance area of ​​the unmanned vehicle at each moment according to the minimum obstacle avoidance area of ​​the unmanned vehicle, calculating the directional collision coefficient of the minimum obstacle avoidance area of ​​the unmanned vehicle at each moment according to the obstacle occupancy value of the minimum obstacle avoidance area of ​​the unmanned vehicle, and obtaining the collision grid map occupied by the unmanned vehicle according to the directional collision coefficient of the minimum obstacle avoidance area of ​​the unmanned vehicle.

[0038] It should be noted that during the driving process of an unmanned vehicle, the unmanned vehicle relies on a high-precision map during the obstacle avoidance process. Therefore, during the safe driving process of the unmanned vehicle, it is first necessary to construct a suitable unmanned vehicle travel and collision grid map based on the driving status of the unmanned vehicle.

[0039] Specifically, assuming that the perceptible limit detection distance in the obstacle detection system of an unmanned vehicle is d max , taking the position of the unmanned vehicle at the current time i as the center, the obstacle avoidance perception range of the unmanned vehicle at the current time i can be obtained, and the square area surrounded by the length L of the unmanned vehicle body is used as the minimum obstacle avoidance area, thereby dividing the obstacle avoidance perception range of the unmanned vehicle into Different minimum obstacle avoidance areas, [*] indicates rounding up to the nearest integer.

[0040] It should be noted that during the driving of an unmanned vehicle, if there is a suspected obstacle within the unmanned vehicle's extreme obstacle avoidance perception range, it is necessary to calculate the possibility of obstacle collision in the unmanned vehicle's minimum obstacle avoidance area for further calculation and analysis.

[0041] In the above formula, Oc j Indicates the obstacle occupancy value of the jth minimum obstacle avoidance area, Dt i,j represents the Manhattan distance between the coordinates of the center point of the jth minimum obstacle avoidance area and the position of the unmanned vehicle at the i-th moment, σ i It represents the variance of all distances between different positions within the obstacle avoidance sensing range of the unmanned vehicle at the current time i and the current position of the unmanned vehicle. exp() represents an exponential function with a natural constant as the base. Pd i,j It represents the collision risk index of the jth area in the extreme obstacle avoidance perception range of the unmanned vehicle at the current time i.

[0042] The collision risk index calculated using the above formula is: within the autonomous vehicle's extreme obstacle avoidance perception range at time i, if an obstacle exists in a small area, the larger the obstacle occupancy value detected by the sensors at the corresponding small area. Furthermore, within the extreme obstacle avoidance perception range centered on the autonomous vehicle at time i, the closer the jth minimum obstacle avoidance area is to the autonomous vehicle's position at time i, the greater the likelihood of a collision, and the calculated collision risk index at the jth area will also be relatively large. Because the obstacle occupancy values ​​detected by the autonomous vehicle's sensors at different times can vary significantly, range normalization is used to normalize the obstacle occupancy values ​​of all autonomous vehicles at different times to the interval [0,1] to avoid interference from computational errors.

[0043] It should be noted that for the jth minimum obstacle avoidance area, N different sampling observation points are set according to Gaussian distribution, and the empirical value of N is 50. The value of each sampling observation point where there is an obstacle is marked as 1, and the value is marked as 0 if there is no obstacle. At the same time, the minimum distance interval between two different sampling observation points is

[0044] It should be noted that, during the driving process of the unmanned vehicle, the driving direction of the unmanned vehicle is different at different times. Therefore, as shown in FIG2 , when the unmanned vehicle is at the driving direction angle θ i In this case, it is necessary to further calculate and analyze the collision risk index at different minimum obstacle avoidance area positions according to the driving direction of the unmanned vehicle to obtain a more accurate collision grid map occupied by the unmanned vehicle.

[0045] Specifically, the driving direction angle of the unmanned vehicle at the time point i is obtained as θ i The minimum obstacle avoidance area at all different positions is recorded as The corresponding directional collision coefficient can be calculated for each different minimum obstacle avoidance area position.

[0046] In the above formula, Norm() represents the normalization function, v i represents the speed of the unmanned vehicle at the current moment i, Dt i,j It represents the Manhattan distance between the coordinates of the center point of the jth minimum obstacle avoidance area and the position of the unmanned vehicle at the i-th moment, n j The empirical value is 4, which means the n nearest neighbor to the jth minimum obstacle avoidance area. j Different minimum obstacle avoidance areas, Oc j It represents the obstacle occupancy value of the jth minimum obstacle avoidance area, Oc k It represents the obstacle occupancy value of the kth minimum obstacle avoidance area adjacent to the jth minimum obstacle avoidance area, Min j It represents the minimum obstacle occupancy value in the jth minimum obstacle avoidance area and its adjacent obstacle avoidance areas. It represents the driving direction angle θ of the unmanned vehicle i The directional collision coefficient of the next j-th minimum obstacle avoidance area.

[0047] The driving direction angle θ of the unmanned vehicle i In this case, if the obstacle occupancy value of the jth minimum obstacle avoidance area is larger, the ratio of the obstacle occupancy value to the minimum obstacle occupancy value is larger. At the same time, if the difference in the obstacle occupancy values ​​calculated between the jth minimum obstacle avoidance area and its adjacent different areas is larger, it means that the corresponding unmanned vehicle driving direction angle θ is larger. i The greater the proportion of obstacles in the next j-th minimum obstacle avoidance area, the higher the possibility of collision. In addition, the faster the current unmanned vehicle is and the closer the distance is, the higher the possibility of collision at the position of the minimum obstacle avoidance area in the corresponding direction. The calculated direction collision coefficient is The larger the value.

[0048] In the above formula, the driving direction angle θ of the unmanned vehicle is i All different minimum obstacle avoidance areas below constitute the set of driving directions of the unmanned vehicle, which is recorded as PC i,j It represents the collision value of the jth minimum obstacle avoidance area occupied by the unmanned vehicle in the collision grid map at the current time i.

[0049] If the jth unmanned driving minimum obstacle avoidance area at current time i is the path passed by the unmanned vehicle under the driving direction angle, then the larger the occupation collision value of the corresponding position in the unmanned vehicle occupation collision grid map at this time, the higher the probability of the unmanned vehicle colliding at the corresponding driving position. Through the above steps, the occupation collision values ​​of different minimum obstacle avoidance areas within the unmanned vehicle's extreme obstacle avoidance perception range at current time i can be calculated, thereby forming the unmanned vehicle occupation collision grid map at current time i.

[0050] Step S003, calculating the collision dynamic adjustment coefficient of the minimum obstacle avoidance area of ​​the unmanned vehicle at each moment according to the collision grid map occupied by the unmanned vehicle, and calculating the obstacle avoidance search distance of the minimum obstacle avoidance area of ​​the unmanned vehicle according to the collision dynamic adjustment coefficient of the minimum obstacle avoidance area of ​​the unmanned vehicle.

[0051] It should be noted that in the collision grid map occupied by the unmanned vehicle, the larger the collision value in different minimum obstacle avoidance areas, the higher the possibility of collision at the corresponding position. Therefore, the obstacle avoidance route of the unmanned vehicle can be dynamically planned according to the collision grid map occupied by the unmanned vehicle at different times.

[0052] In the collision grid map occupied by the unmanned vehicle, the larger the collision value in the corresponding minimum obstacle avoidance area, the higher the possibility of collision at that location. Therefore, in order to determine the threshold of collision conflict during the driving of the unmanned vehicle, the collision grid map occupied by the unmanned vehicle is first used as the input of the Otsu method to obtain the collision threshold in the collision grid map occupied by the unmanned vehicle, which is recorded as T i , the area greater than the collision threshold is considered to have collided, otherwise, the area less than the threshold is considered not to have collided.

[0053] Specifically, in the collision grid map occupied by the unmanned vehicle at time i, a local collision area with a length of 3L can be obtained with the jth minimum obstacle avoidance area as the center. The following calculation can be performed in this local area.

[0054] In the above formula, Norm() represents the normalization function, Max i,j It represents the maximum occupied collision value in the local collision area centered on the jth minimum obstacle avoidance area in the collision grid map occupied by the unmanned vehicle at time i, Min i,j represents the minimum occupied collision value in the local collision area centered on the jth minimum obstacle avoidance area in the collision grid map occupied by the unmanned vehicle at time i, b represents the total number of different minimum obstacle avoidance areas in the local collision area, Pc i,jIt represents the occupied collision value of the jth minimum obstacle avoidance area at the center of the local collision area at time i, Pc i,k It represents the occupied collision value of the kth minimum obstacle avoidance area in the local collision area at time i, csd i,j It represents the collision dynamic adjustment coefficient of the jth minimum obstacle avoidance area at time i.

[0055] The collision dynamic adjustment coefficient can be calculated using the above formula. If the difference between the maximum occupied collision value and the minimum occupied collision value in the local collision area at time i is larger, and the difference between the occupied collision values ​​in different local collision areas and the center position is larger, it means that there are more obstacles in the local collision area centered on the jth minimum obstacle avoidance area, so the collision dynamic adjustment coefficient is also relatively larger, which provides a suitable dynamic search range for the subsequent further search for a suitable unmanned driving obstacle avoidance path.

[0056] In the above formula, L represents the length of the unmanned vehicle, csd i,j It represents the collision dynamic adjustment coefficient of the jth minimum obstacle avoidance area at time i, Pc i,j It represents the occupied collision value of the jth minimum obstacle avoidance area at the center of the local collision area at time i, Pc i,k It represents the occupied collision value of the kth minimum obstacle avoidance area in the local collision area at time i, R i,j It represents the obstacle avoidance search distance of the jth minimum obstacle avoidance area at time i.

[0057] The above formula can be used to calculate the dynamic search range of the jth minimum obstacle avoidance area at time i. If there are many jth minimum obstacle avoidance areas and the collision value occupied in the minimum obstacle avoidance area is greater than the threshold, then in order to obtain the obstacle avoidance path of the unmanned vehicle, the search range needs to be expanded accordingly to improve the rationality of the obstacle avoidance path search selection. Otherwise, the corresponding search range needs to be narrowed to reduce the time complexity of the obstacle avoidance path search.

[0058] Step S004 , using the obstacle avoidance search distance of the minimum obstacle avoidance area of ​​the unmanned vehicle, dynamically search and plan the obstacle avoidance path for the unmanned vehicle at each moment.

[0059] Through the above analysis, the obstacle avoidance search radius of the unmanned vehicle when it travels to different minimum obstacle avoidance areas can be obtained. The collision grid map occupied by the unmanned vehicle at the current time i is used as the input of the rapid exploration random tree algorithm (RTT), and the position of the unmanned vehicle at the time i is used as the initial position. The path search is performed in the area with each minimum obstacle avoidance area as the center and the obstacle avoidance search distance as the radius. The obstacle avoidance path curve of the unmanned vehicle at the current time i can be obtained, and the cubic spline interpolation algorithm is used to smooth the obstacle avoidance path curve of the unmanned vehicle at the current time i to avoid the occurrence of abnormal driving interruptions during the driving process of the unmanned vehicle. The driving state parameters of the unmanned vehicle at different times are used to construct a reasonable obstacle avoidance path curve for the unmanned vehicle to complete the obstacle avoidance process during the driving process of the unmanned vehicle.

[0060] The above-described embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present application, and should all be included in the scope of protection of the present application.

Claims

1. An intelligent obstacle avoidance method for an unmanned vehicle, characterized in that: The method comprises the following steps: Obtain driving parameters of the unmanned vehicle; Obtain the minimum obstacle avoidance area of ​​the unmanned vehicle according to the driving parameters of the unmanned vehicle, calculate the obstacle occupancy value of the minimum obstacle avoidance area of ​​the unmanned vehicle at each moment according to the minimum obstacle avoidance area of ​​the unmanned vehicle, calculate the collision coefficient of the minimum obstacle avoidance area of ​​the unmanned vehicle at each moment according to the obstacle occupancy value of the minimum obstacle avoidance area of ​​the unmanned vehicle, and obtain the collision grid map occupied by the unmanned vehicle according to the collision coefficient of the minimum obstacle avoidance area of ​​the unmanned vehicle; Calculate the collision dynamic adjustment coefficient of the minimum obstacle avoidance area of ​​the unmanned vehicle at each moment according to the collision grid map occupied by the unmanned vehicle, and calculate the obstacle avoidance search distance of the minimum obstacle avoidance area of ​​the unmanned vehicle according to the collision dynamic adjustment coefficient of the minimum obstacle avoidance area of ​​the unmanned vehicle; The obstacle avoidance search distance of the minimum obstacle avoidance area of ​​the unmanned vehicle is used to dynamically search and plan the obstacle avoidance path of the unmanned vehicle at each moment.

2. The intelligent obstacle avoidance method for an unmanned vehicle according to claim 1, characterized in that: The unmanned vehicle driving parameters include: the unmanned vehicle driving speed at each moment, the unmanned vehicle driving direction angle, the obstacle distance, the perceptible limit detection distance, and the vehicle body length.

3. The intelligent obstacle avoidance method for an unmanned vehicle according to claim 2, characterized in that: The method for obtaining the minimum obstacle avoidance area of ​​the unmanned vehicle according to the driving parameters of the unmanned vehicle is: The result of rounding up the ratio of the perceptible limit detection distance of the unmanned vehicle at each moment to the body length of the unmanned vehicle is recorded as the first length. In a square area with the position of the unmanned vehicle at each moment as the center and the perceptible limit detection distance of the unmanned vehicle as the side length, the minimum obstacle avoidance area of ​​the unmanned vehicle is divided with the first length as the side length.

4. The intelligent obstacle avoidance method for an unmanned vehicle according to claim 3, characterized in that: The method for calculating the obstacle occupancy value of the minimum obstacle avoidance area of ​​the unmanned vehicle at each moment according to the minimum obstacle avoidance area of ​​the unmanned vehicle is: In the above formula, Oc j represents the obstacle occupancy value of the jth minimum obstacle avoidance area, Dt i,j represents the Manhattan distance between the coordinates of the center point of the jth minimum obstacle avoidance area and the position of the unmanned vehicle at the i-th moment, σ i It represents the variance of all distances between different positions within the obstacle avoidance perception range of the unmanned vehicle at the current time i and the current position of the unmanned vehicle. exp() represents an exponential function with a natural constant as the base, Pd i,j It represents the collision risk index at the jth area in the extreme obstacle avoidance perception range of the unmanned vehicle at the current time i.

5. The intelligent obstacle avoidance method for an unmanned vehicle according to claim 4, characterized in that: The method for obtaining the obstacle occupancy value of the minimum obstacle avoidance area is as follows: A preset number of sampling points is set in the minimum obstacle avoidance area of ​​the unmanned vehicle, and the number of all sampling points with preset values ​​in the minimum obstacle avoidance area is recorded as the obstacle occupancy value of the minimum obstacle avoidance area.

6. The intelligent obstacle avoidance method for an unmanned vehicle according to claim 4, characterized in that: The method for calculating the collision coefficient of the minimum obstacle avoidance area of ​​the unmanned vehicle at each moment according to the obstacle occupancy value of the minimum obstacle avoidance area of ​​the unmanned vehicle and obtaining the collision grid map occupied by the unmanned vehicle according to the collision coefficient of the minimum obstacle avoidance area of ​​the unmanned vehicle is: In the above formula, Norm() represents the normalization function, v i represents the driving speed of the unmanned vehicle at the current time i, Dt i,j It represents the Manhattan distance between the coordinates of the center point of the jth minimum obstacle avoidance area and the position of the unmanned vehicle at the i-th moment, n j is the preset experience value, Oc j represents the obstacle occupancy value of the jth minimum obstacle avoidance area, Oc k It represents the obstacle occupancy value of the kth minimum obstacle avoidance area adjacent to the jth minimum obstacle avoidance area, Min j It represents the minimum obstacle occupancy value in the jth minimum obstacle avoidance area and its adjacent obstacle avoidance areas. It represents the driving direction angle θ of the unmanned vehicle. i The directional collision coefficient of the next j-th minimum obstacle avoidance area; All minimum obstacle avoidance areas under the driving direction angle of the unmanned vehicle are recorded as the first set, the collision risk index and the direction collision coefficient of all the minimum obstacle avoidance areas in the first set are recorded as the occupied collision value of the corresponding minimum obstacle avoidance area, and the collision risk index of all the minimum obstacle avoidance areas that do not belong to the first set is recorded as the occupied collision value of the corresponding minimum obstacle avoidance area. The occupied collision value of the minimum obstacle avoidance area is used to construct the unmanned vehicle occupied collision grid map.

7. The intelligent obstacle avoidance method for an unmanned vehicle according to claim 6, characterized in that: The method for calculating the collision dynamic adjustment coefficient of the minimum obstacle avoidance area of ​​the unmanned vehicle at each moment according to the collision grid map occupied by the unmanned vehicle is: In the above formula, Norm() represents the normalization function, Max i,j It represents the maximum occupied collision value in the local collision area centered on the jth minimum obstacle avoidance area in the collision grid map occupied by the unmanned vehicle at time i, Min i,j represents the minimum occupied collision value in the local collision area centered on the jth minimum obstacle avoidance area in the collision grid map occupied by the unmanned vehicle at time i, b represents the total number of different minimum obstacle avoidance areas in the local collision area, Pc i,j It represents the occupied collision value of the jth minimum obstacle avoidance area at the center of the local collision area at time i, Pc i,k It represents the occupied collision value of the kth minimum obstacle avoidance area in the local collision area at time i, csd i,j It represents the collision dynamic adjustment coefficient of the jth minimum obstacle avoidance area at time i, wherein the local collision area refers to a local area with a preset length centered on each minimum obstacle avoidance area.

8. The intelligent obstacle avoidance method for an unmanned vehicle according to claim 7, characterized in that: The method of calculating the obstacle avoidance search distance of the minimum obstacle avoidance area of ​​the unmanned vehicle according to the collision dynamic adjustment coefficient of the minimum obstacle avoidance area of ​​the unmanned vehicle and using the obstacle avoidance search distance of the minimum obstacle avoidance area of ​​the unmanned vehicle to perform dynamic search planning of obstacle avoidance paths for the unmanned vehicle at each moment is: The sum of the first preset constant and the collision dynamic adjustment coefficient of the minimum obstacle avoidance area is recorded as a first sum value, the product of the first sum value and the second preset constant is recorded as a first product, the difference between the first preset constant and the collision dynamic adjustment coefficient of the minimum obstacle avoidance area is recorded as a first difference, and the product of the first difference and the second preset constant is recorded as a second product. When the occupied collision value of the minimum obstacle avoidance area of ​​the unmanned vehicle is greater than or equal to the collision threshold in the unmanned vehicle occupied collision grid map, the obstacle avoidance search distance of the minimum obstacle avoidance area of ​​the unmanned vehicle is set to the first product. When the occupied collision value of the minimum obstacle avoidance area of ​​the unmanned vehicle is less than the collision threshold in the unmanned vehicle occupied collision grid map, the obstacle avoidance search distance of the minimum obstacle avoidance area of ​​the unmanned vehicle is set to the second product. The collision grid map occupied by the unmanned vehicle is used as the input of the fast exploration random tree algorithm, and the position of the unmanned vehicle is used as the initial position. A path search is performed in the area with each minimum obstacle avoidance area as the center and the obstacle avoidance search distance as the radius to obtain the obstacle avoidance path curve of the unmanned vehicle at the current time i.

9. The intelligent obstacle avoidance method for an unmanned vehicle according to claim 8, characterized in that: The method for obtaining the collision threshold in the collision grid map occupied by the unmanned vehicle is: The collision grid map occupied by the unmanned vehicle is used as the input of the Otsu method to obtain the collision threshold in the collision grid map occupied by the unmanned vehicle.

10. An intelligent obstacle avoidance system for an unmanned vehicle, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 9 are implemented.

Images