Path generation method and apparatus, robot, and storage medium

Abstract

The application belongs to the field of robots and relates to a path generation method, which comprises the following steps: acquiring environment information, constructing a scene map, extracting road features according to the scene map, determining a driving road and a road type corresponding to the driving road based on the road features, and generating a target path in the driving road according to the road type and a path generation rule corresponding to the road type. In the process of robot field survey, the method realizes automatic path generation, improves work efficiency and the accuracy of path generation.

Description

[0001] The present application relates to the technical field of robots, and in particular to a path generation method and device, a robot, and a storage medium.

[0002] In an outdoor environment, automatic surveying of road information by a self-moving robot has become an important means of modern engineering and traffic planning. One of the core functions of an automatic surveying robot is path generation.

[0003] Currently, an outdoor surveying process generally includes the steps of image acquisition, boundary setting, path point setting, vehicle path drawing, vehicle driving direction and speed attribute setting, and initialization point and target point addition, and the process is deployed on a robot for debugging. When drawing a path, path points need to be manually added, a path is then added between two adjacent points, and finally a corresponding route is selected to add different speed attributes. Since a large number of path points and attributes need to be manually set, this process usually requires a large amount of time and is prone to errors. In addition, in the process of manually adding attributes, there may be cases of missing or incorrectly setting attributes, which requires a large amount of time to query the problem, resulting in low work efficiency.

[0004] Embodiments of the present application provide a path generation method, device, robot, and storage medium, which can overcome the deficiencies of the prior art, automatically generate a path, and improve work efficiency and path generation accuracy.

[0005] To achieve the above object, in a first aspect, embodiments of the present application provide a path generation method, comprising: acquiring environment information and constructing a scene map; extracting road features according to the scene map, determining a driving road and a road type corresponding to the driving road based on the road features; and generating a target path in the driving road according to the road type and a path generation rule corresponding to the road type.

[0006] In some embodiments, the road features include one or more of a boundary line, a lane line, a pedestrian crossing line, and a speed reduction zone of the road;

[0007] In some embodiments, the step of extracting road features according to the scene map and determining a driving road and a road type corresponding to the driving road based on the road features comprises:

[0008] extracting one or more of a boundary line, a lane line, a pedestrian crossing line, and a speed reduction zone of the road according to the scene map to determine the range of the driving road;

[0009] determining the road type according to the range of the driving road;

[0010] ​​​The driving road is divided into one or more road segments, and a road type corresponding to each road segment is determined according to the one or more road segments, wherein the road type includes a straight road, a curved road, or an intersection.

[0011] In some embodiments, the step of generating the target path in the driving road according to the road type and the path generation rule corresponding to the road type includes determining the road type and the path generation plan according to the shape, length, and width of the driving road, and generating the target path in the driving road based on the path generation rule.

[0012] In some embodiments, the step of determining the road type and the path generation plan according to the shape, length, and width of the driving road, and generating the target path in the driving road based on the path generation rule includes:

[0013] If the shape of the driving road is linear, the driving road is determined to be a straight road;

[0014] If the length of the straight road is greater than a first length threshold, and the width of the straight road is greater than a first width threshold, a first path generation rule is determined, and the target path is generated in the driving road based on the first path generation rule;

[0015] If the length of the straight road is greater than the first length threshold, and the width of the straight road is less than or equal to the first width threshold, a second path generation rule is determined, and the target path is generated in the driving road based on the second path generation rule;

[0016] If the length of the straight road is less than or equal to the first length threshold, and the width of the straight road is greater than the first width threshold, a third path generation rule is determined, and the target path is generated in the driving road based on the third path generation rule;

[0017] If the length of the straight road is less than or equal to the first length threshold, and the width of the straight road is less than or equal to the first width threshold, a fourth path generation rule is determined, and the target path is generated in the driving road based on the fourth path generation rule.

[0018] In some embodiments, the step of determining the road type and the path generation plan according to the shape, length, and width of the driving road, and generating the target path in the driving road based on the path generation rule includes:

[0019] If the shape of the driving road is arc-shaped, the driving road is determined to be a curved road;

[0020] According to the bending degree of the curved road and the width of the curved road, a path generation plan is determined, and the target path is generated in the driving road based on the path generation rule.

[0021] In some embodiments, the step of generating the target path in the travel road comprises setting a travel direction and a travel speed for the target path when the target path is generated in the travel road.

[0022] In some embodiments, after the step of generating the target path in the travel road according to the road type and the path generation rule corresponding to the road type, the method further comprises:

[0023] The target path is tested to optimize the position, travel direction and travel speed of the target path.

[0024] To achieve the above object, in a second aspect, the embodiments of the present application further provide a path generation device, comprising:

[0025] The construction module is configured to acquire environment information and construct a scene map.

[0026] The determination module is configured to extract road features according to the scene map, determine a travel road and a road type corresponding to the travel road based on the road features.

[0027] The generation module is configured to generate a target path in the travel road according to the road type and a path generation rule corresponding to the road type.

[0028] To achieve the above object, in a third aspect, the embodiments of the present application further provide a robot, comprising a processor and a memory connected with the processor in communication; the memory stores computer program instructions, and the computer program instructions, when invoked by the processor, cause the processor to execute the path generation method according to any one of the first aspect.

[0029] To achieve the above object, in a fourth aspect, the embodiments of the present application further provide a storage medium, which stores computer executable instructions, and the computer executable instructions, when executed by at least one processor, cause the at least one processor to execute the path generation method according to any one of the first aspect.

[0030] The path generation method provided by the embodiments of the present application can acquire environment information, construct a scene map, extract road features according to the scene map, determine a travel road and a corresponding road type, generate a target path in the travel road according to the road type and a path generation rule corresponding to the road type, and thus realize automatic generation of the target path. Compared with the prior art, the automatic generation of the target path can more accurately make decisions and plan the path, reduce the time and effort consumed by manual addition of path points and path attributes, and improve work efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0031] One or more embodiments are illustrated by way of example with reference numerals in the accompanying drawings. These illustrations do not constitute a limitation on the embodiments. Elements with the same reference numerals in the drawings are denoted as similar elements. Unless otherwise stated, the figures in the drawings are not to be limited by scale.

[0032] Figure 1 A schematic diagram of the structure of a robot provided in an embodiment of this application;

[0033] Figure 2 This is a schematic diagram of the structure of a path generation device provided in an embodiment of this application;

[0034] Figure 3 A flowchart illustrating the path generation method provided in this application embodiment;

[0035] Figure 4 for Figure 3 A schematic diagram of a sub-process of step S102 in the method;

[0036] Figure 5 for Figure 3 A schematic diagram of a sub-process of step S103 in the method.

Detailed Implementation Methods

[0037] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and thoroughly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0038] Furthermore, the technical features involved in the various embodiments of this application described below do not conflict with each other and can be combined with each other.

[0039] Furthermore, although functional modules are divided in the device diagram and the logical order is shown in the flowchart, in some cases, the module division may differ from that in the device, or the execution order in the flowchart may differ from the steps shown or described. In addition, the terms "first," "second," and "third" used in this document do not limit the data or execution order; they are merely used to distinguish identical or similar items with essentially the same function and purpose.

[0040] Unless otherwise defined, technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in the description herein is for describing particular embodiments only and is not intended to be limiting of the application. It is to be understood that, within the scope of the application, "at least one", "one or more", "multiple", "two or more" refer to one or more than one. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0041] Before the detailed description of the present application, the terms and phrases involved in the embodiments of the present application are explained, and the terms and phrases involved in the embodiments of the present application are applicable to the following explanations:

[0042] (1) Machine learning, machine learning is a branch of artificial intelligence, which realizes the automation of tasks by letting computer systems automatically learn and improve from data. The goal of machine learning is to enable computer systems to automatically improve performance based on historical data or experience without explicit programming. In machine learning, the system learns features from a large amount of data and predicts future results through these features. This learning process is usually achieved by using statistical and optimization algorithms. Machine learning is widely used in various fields, including natural language processing, computer vision, data mining, automation control, biomedicine and finance. Machine learning can be applied to the path generation method proposed in the present application to automatically learn road features and path generation rules. Specifically, a supervised learning method can be used to learn from training data with known road information, and automatically build a road feature extraction model and a path generation model. The road feature extraction model can be used to extract feature information such as road boundaries, ranges and driving directions from environmental information; and the path generation model can automatically calculate the decision target path based on road features and path generation rules. In practical applications, machine learning can help operators more accurately identify road features and generate target paths, thereby improving work efficiency.

[0043] (2) The robot artificial intelligence online process usually includes the following steps: image acquisition, boundary setting, path point setting, vehicle path drawing, vehicle driving direction and speed attribute setting, and initialization point and target point adding, etc. And deploy it on the robot for debugging, finally, the robot provides various robot services in the map.

[0044] The path generation method of the embodiments of the present application can be applied to a kind of Figure 1The robot 10 shown includes at least one processor 11 and at least one memory 12, wherein the processor 11 is in communication connection with the memory 12 through a bus, the processor 11 can include a central processing unit (CPU), a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) device, etc. The processor 11 can also be implemented as a combination of computing devices, for example, a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors in combination with a DSP core, or any other such configuration; the memory 12 as a kind of non-volatile computer readable storage medium, which can be used to store non-volatile software programs, non-volatile computer executable program instructions, the memory 12 can include a high-speed random access memory, and can also include a non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, or other non-volatile solid-state storage device.

[0045] As can be understood by those skilled in the art, the above is only an example of the hardware structure of the robot 10, and in actual application, more components can also be provided for the robot 10 according to actual functional needs, of course, one or more components can also be omitted according to functional needs. For example, the robot can also include sensor, camera, radar and other hardware devices for sensing the environment and obtaining related data.

[0046] Figure 2 An example structure of the path generation device 20 is shown, which includes a construction module 21, a determination module 22 and a generation module 23; wherein:

[0047] The construction module 21 is configured to obtain environment information and construct a scene map;

[0048] The determination module 22 is configured to extract road features from the scene map, determine a driving road and a road type corresponding to the driving road based on the road features;

[0049] The generation module 23 is configured to generate a target path in the driving road according to the road type and a path generation rule corresponding to the road type.

[0050] The road features include one or more of the boundary line, lane line, pedestrian crossing line and deceleration strip of the road.

[0051] The determination module 22 is further configured to extract one or more of the boundary line, lane line, pedestrian crossing line and deceleration strip of the road from the scene map to determine the range of the driving road; determine the road type according to the range of the driving road; divide the driving road into one or more road segments, and determine the road type corresponding to each road segment according to the one or more road segments, wherein the road type includes a straight road, a curved road or an intersection.

[0052] The generation module 23 is further configured to determine a road type and a path generation rule according to the shape, length and width of the driving road, and generate the target path in the driving road based on the path generation rule.

[0053] The generation module 23 is further configured to determine that the driving road is a straight road if the shape of the driving road is linear, determine a first path generation rule if the length of the straight road is greater than a first length threshold and the width of the straight road is greater than a first width threshold, generate the target path in the driving road based on the first path generation rule, determine a second path generation rule if the length of the straight road is greater than the first length threshold and the width of the straight road is less than or equal to the first width threshold, generate the target path in the driving road based on the second path generation rule, determine a third path generation rule if the length of the straight road is less than or equal to the first length threshold and the width of the straight road is greater than the first width threshold, generate the target path in the driving road based on the third path generation rule, determine a fourth path generation rule if the length of the straight road is less than or equal to the first length threshold and the width of the straight road is less than or equal to the first width threshold, and generate the target path in the driving road based on the fourth path generation rule.

[0054] The generation module 23 is further configured to determine that the driving road is a curved road if the shape of the driving road is curved, determine a path generation rule according to the bending degree of the curved road and the width of the curved road, and generate the target path in the driving road based on the path generation rule.

[0055] The generation module 23 is further configured to set a driving direction and a driving speed for the target path when generating the target path in the driving road.

[0056] The generation module 23 is further configured to test the target path to optimize the position, driving direction and driving speed of the target path.

[0057] In some embodiments, the robot 10 provided in the present application can include the path generation device 20 provided in the present application, and the processor 11 and the memory 12 of the robot can provide computing power support and data storage for the execution of the path generation device 20. Specifically, the processor 11 is configured to control the automatic driving of the robot and collect environmental information, execute a machine learning algorithm based on a scene map and road feature extraction, and determine a driving road and a corresponding road type according to the extracted features. Then, the processor generates a target path in the driving road according to a path generation rule corresponding to the road type. The memory 12 is configured to store data such as environmental information, a scene map, road features, road types, path generation rules and a target path.

[0058] The path generation method provided in the embodiments of the present application can be executed by the robot 10 described above, Figure 3As shown, the path generation method S100 includes the following steps S101-S103:

[0059] S101, acquire environment information and construct a scene map.

[0060] In a specific application environment, the robot 10 needs to manually demarcate the travel area of the robot 10 by the operator during the survey, and the robot 10 automatically travels within the limited area or is controlled by the operator to travel the robot, and acquires the environment information around the robot 10 through the information acquisition device and constructs a scene map.

[0061] In some embodiments, the robot 10 can use various sensors and technologies to collect the environment information around the robot, for example, the robot 10 is configured with sensors such as laser radar, depth camera, camera, GPS, inertial measurement unit, etc. Based on the SLAM (Simultaneous Localization and Mapping) algorithm, the robot can collect information such as objects, obstacles, terrain, roads, etc. around the robot through the sensors, and then integrate these information to construct a scene map around the robot. This process can use various algorithms such as scan matching, filter, etc. to improve the accuracy and robustness of the map. The constructed scene map can provide basic data and information for the robot to learn and extract road features.

[0062] S102, extract road features according to the scene map, and determine the travel road and the road type corresponding to the travel road based on the road features.

[0063] In some embodiments, the robot 10 can use machine learning algorithms to process the scene map obtained in step S101 to identify and extract the road in the scene. The machine learning model can use deep learning, convolutional neural network, etc. Through repeated training on the scene data obtained in step S101, the machine learning model can accurately identify the road features of the road.

[0064] Among them, the road features can include one or more of the boundary line of the road, the lane line, the pedestrian crossing line, and the deceleration zone. When the robot 10 automatically travels within the specified operating range for survey, the road features are continuously extracted, which can be used to establish a road model through machine learning algorithms. The model can determine the current road being traveled according to the position, direction and surrounding environment of the robot 10, etc. In addition, the robot 10 acquires traffic signs and other information on the road being traveled, and combines the above road features to automatically determine the type of the current travel road through machine learning algorithms according to the road features. The road type includes straight road, curved road or intersection.

[0065] In some embodiments, the step S102 of "extracting road features from the scene map, determining the driving road and the road type corresponding to the driving road based on the road features" specifically includes the following steps S1021-S1023:

[0066] S1021, extracting one or more of the boundary line of the road, the lane line, the pedestrian crossing line, and the deceleration strip from the scene map to determine the range of the driving road.

[0067] The range of the driving road can be the shape, length, and width of the road, etc.

[0068] In order to enable the robot 10 to accurately determine the road type, it is necessary to first extract road features from the scene map, including road boundary lines, lane lines, pedestrian crossing lines, deceleration strips, etc., and then determine the range of the driving road based on these features. Specifically, these information can be collected by hardware devices such as sensors, and processed and extracted by image processing algorithms. On this basis, the machine learning algorithm is used to classify and identify one or more of the road boundary line, lane line, pedestrian crossing line, and deceleration strip, and the range of the driving road can be determined.

[0069] S1022, determining the road type according to the range of the driving road.

[0070] S1023, dividing the driving road into one or more road segments, and determining the road type corresponding to each road segment according to the one or more road segments, wherein the road type includes a straight road, a curved road, or an intersection.

[0071] In some embodiments, the robot 10 needs to divide the driving road into one or more road segments in order to analyze and plan each road segment. For each road segment, the robot 10 needs to classify it according to the road type determination method described in the above step S1022 to determine its corresponding road type. After dividing the road segments and determining the road types, the robot can calculate the target path according to the type of each road segment and the corresponding path generation rule.

[0072] S103, generating a target path in the driving road according to the road type and the path generation rule corresponding to the road type.

[0073] The method of step S103 includes determining the road type and the path generation rule according to the shape, length, and width of the driving road, and generating the target path in the driving road based on the path generation rule.

[0074] Specifically, the robot determines a road type according to a shape, a length and a width of a driving road, generates a target path in the driving road based on the road type and a preset path generation rule. The path generation rule is determined according to the road type, for example, a straight road, a curved road or an intersection, different rules need to be used to generate the path, and the path generation rule includes a path point interval, a distance from a right boundary of the road, a driving speed and a driving direction. After determining the path generation rule, the position of the path point is calculated according to the path point interval of the path generation rule, so as to generate the path point in the driving road, and the adjacent two path points are connected to form a path, and the distance from the right boundary of the road, the driving speed and the driving direction are set for the path, and finally the target path is generated.

[0075] A path generation rule is exemplarily shown in Table 1 as follows.

[0076]

[0077]

[0078] Table 1

[0079] In some embodiments, the path generation rule shown in Table 1 can be used, and it can be understood that Table 1 is only an exemplary path generation rule, and in specific implementation, the type or value in Table 1 can be adjusted as needed according to the environmental situation. The above steps of determining the road type and the path generation rule according to the shape, the length and the width of the driving road, and generating the target path in the driving road based on the path generation rule include the following steps S1031-S1033:

[0080] S1031, if the shape of the driving road is linear, it is determined that the driving road is a straight road.

[0081] If the shape of the driving road is linear, it is determined that the driving road is a straight road, the path generation rule is determined according to the length and the width of the straight road, and the target path is generated in the driving road based on the path generation rule. Specifically,

[0082] If the length of the straight road is greater than a first length threshold value, and the width of the straight road is greater than a first width threshold value, a first path generation rule is determined, path points are generated in the driving road at a first interval based on a path point interval of the first path generation rule, adjacent two path points are connected to generate the target path;

[0083] If the length of the straight road is greater than a first length threshold value, and the width of the straight road is less than or equal to a first width threshold value, a second path generation rule is determined, path points are generated in the driving road at a first interval based on a path point interval of the second path generation rule, adjacent two path points are connected to generate the target path;

[0084] If the length of the straight path is less than or equal to the first length threshold value and the width of the straight path is greater than the first width threshold value, a third path generation rule is determined, and path points are generated in the travel road at a second interval based on a path point interval of the third path generation rule, and the target path is generated by connecting adjacent two path points.

[0085] If the length of the straight path is less than or equal to the first length threshold value and the width of the straight path is less than or equal to the first width threshold value, a fourth path generation rule is determined, and path points are generated in the travel road at a second interval based on a path point interval of the fourth path generation rule, and the target path is generated by connecting adjacent two path points.

[0086] S1032, if the shape of the travel road is arc-shaped, the travel road is determined to be a curve.

[0087] If the shape of the travel road is arc-shaped, the travel road is determined to be a curve, and a path generation rule is determined according to the curvature degree of the curve and the width of the curve, and the target path is generated in the travel road based on the path generation rule. The curvature degree of the curve can be represented by the maximum curvature of the road segment.

[0088] Specifically,

[0089] If the curvature degree of the curve is greater than the first curvature degree and the width of the curve is greater than or equal to the first width threshold value, a fifth path generation rule is determined, and path points are generated in the travel road at a third interval based on a path point interval of the fifth path generation rule, and the target path is generated by connecting adjacent two path points.

[0090] If the curvature degree of the curve is greater than the first curvature degree and the width of the curve is less than the first width threshold value, a sixth path generation rule is determined, and path points are generated in the travel road at a third interval based on a path point interval of the sixth path generation rule, and the target path is generated by connecting adjacent two path points.

[0091] If the curvature degree of the curve is less than the second curvature degree and the width of the curve is greater than or equal to the first width threshold value, a seventh path generation rule is determined, and path points are generated in the travel road at a fourth interval based on a path point interval of the seventh path generation rule, and the target path is generated by connecting adjacent two path points.

[0092] If the curvature degree of the curve is less than the second curvature degree and the width of the curve is less than the first width threshold value, an eighth path generation rule is determined, and path points are generated in the travel road at a fourth interval based on a path point interval of the eighth path generation rule, and the target path is generated by connecting adjacent two path points.

[0093] The second curvature degree is greater than the first curvature degree, the first interval is greater than the second interval, the second interval is greater than the third interval, and the third interval is greater than the fourth interval.

[0094] S1033, when generating the target path in the travel road, setting a travel direction and a travel speed for the target path.

[0095] When generating the target path in the travel road, setting a travel direction, a travel speed and a distance from the right boundary of the road for the target path according to the path generation rule in Table 1, wherein the travel direction refers to the direction of the robot when traveling along the travel road, which is consistent with the direction indicated by the ground indication sign. The ground indication sign is identified based on the information obtained by extracting the road features previously, such as the arrow sign set at the position of the lane line or traffic sign. Meanwhile, according to the road type and the road features extracted previously, the travel speed and the distance from the right boundary of the road on the target path can also be set, specifically,

[0096] If it is determined to be the first path generation rule, the travel direction is consistent with the direction indicated by the ground indication sign, the travel speed is the first speed, and the distance from the right boundary of the road is the first distance;

[0097] If it is determined to be the second path generation rule, the travel direction is consistent with the direction indicated by the ground indication sign, the travel speed is the second speed, and the distance from the right boundary of the road is the second distance;

[0098] If it is determined to be the third path generation rule, the travel direction is consistent with the direction indicated by the ground indication sign, the travel speed is the first speed, and the distance from the right boundary of the road is the second distance;

[0099] If it is determined to be the fourth path generation rule, the travel direction is consistent with the direction indicated by the ground indication sign, the travel speed is the second speed, and the distance from the right boundary of the road is the second distance;

[0100] If it is determined to be the fifth path generation rule, the travel direction is consistent with the direction indicated by the ground indication sign, the travel speed is the first speed, and the distance from the right boundary of the road is the third distance;

[0101] If it is determined to be the sixth path generation rule, the travel direction is consistent with the direction indicated by the ground indication sign, the travel speed is the second speed, and the distance from the right boundary of the road is the third distance;

[0102] If it is determined to be the seventh path generation rule, the travel direction is consistent with the direction indicated by the ground indication sign, the travel speed is the first speed, and the distance from the right boundary of the road is the fourth distance;

[0103] If it is determined to be the eighth path generation rule, the travel direction is consistent with the direction indicated by the ground indication sign, the travel speed is the first speed, and the distance from the right boundary of the road is the fourth distance.

[0104] In some embodiments, the road type can also include intersections, speed bumps and crosswalks, which can be identified by the aforementioned machine learning for road feature determination. When the target path of the straight road or the curved road includes intersections, speed bumps and crosswalks, the driving direction and the distance from the right boundary of the road are still set according to the path generation rules of the straight road or the curved road, and when the driving speed of the target path is set, special settings need to be made according to these special road conditions to ensure the safety of the robot when passing through these areas. For example, at the intersection, the robot needs to appropriately reduce the vehicle speed according to the indication of the traffic light or the traffic sign. At the speed bump and the crosswalk, the robot also needs to appropriately reduce the speed to ensure the smoothness and safety when passing through these areas. These specially set driving speeds can be defined in the path generation rules, as shown in Table 1,

[0105] If it is determined as the ninth path generation rule, the driving speed is the fourth speed;

[0106] If it is determined as the tenth path generation rule, the driving speed is the sixth speed;

[0107] If it is determined as the eleventh path generation rule, the driving speed is the fifth speed.

[0108] The first boundary distance is greater than the second boundary distance. The first speed is greater than the second speed, the second speed is greater than the third speed, the third speed is greater than the fourth speed, the fourth speed is greater than the fifth speed, and the fifth speed is greater than the sixth speed.

[0109] In some embodiments, after the step of generating the target path in the driving road according to the road type and the path generation rule corresponding to the road type in step S103, the method can further include:

[0110] S104, testing the target path to optimize the position, driving direction and driving speed of the target path.

[0111] Testing and optimizing the generated target path can help determine whether the robot driving on the generated target path meets the actual situation and safety requirements. During the test, an initial point and a target point can be selected from the path points, and the robot automatically generates an optimal path from the initial point to the target point. The optimal path is a target path set with the shortest path distance from the initial point to the target point, which can be simulated using some simulators or tested in a real environment. During the test, it can be observed whether the robot driving path and driving strategy are consistent with the expectations, and the generated target path is modified and optimized according to the test results to achieve better driving effect.

[0112] In the optimization process, machine learning and other technologies can be used to train the model to enable the robot to more intelligently plan the path and control the driving. For example, reinforcement learning and other technologies can be used to enable the robot to continuously learn and optimize the driving path and driving strategy in real-time environment. The path point position, driving direction and driving speed and other parameters can also be artificially modified in simulation to better adapt to various complex road conditions and environments.

[0113] By implementing the above embodiments, the driving road and the corresponding road type are determined by acquiring environmental information, constructing a scene map, extracting road features from the scene map, generating a target path in the driving road according to the road type and the path generation rule corresponding to the road type. Compared with the prior art, the method uses scene map and machine learning based technology to automatically generate the target path, reduces human intervention and operation, and improves the efficiency of path planning.

[0114] The embodiments of the present application also provide a computer readable storage medium, which stores computer executable instructions. The computer executable instructions are executed by one or more processors to enable the one or more processors to perform the path generation method in any of the method embodiments described above, for example, to perform the method steps S101 to S103 described above. Figure 1 The one processor 11 in the computer system 100 can enable the one or more processors to perform the path generation method in any of the method embodiments described above, for example, to perform the method steps S101 to S103 described above. Figure 3 The one processor 11 in the computer system 100 can enable the one or more processors to perform the path generation method in any of the method embodiments described above, for example, to perform the method steps S101 to S103 described above.

[0115] The above products have the corresponding beneficial effects of the methods described in the embodiments, and the technical details and effects not described in detail in the product embodiments can be referred to the methods provided in the embodiments of the present application.

[0116] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; under the idea of the present application, the technical features in the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other changes of different aspects of the present application as described above. In order to be brief, they are not provided in detail; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement to part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A path generation method characterized by, The method comprises the following steps: acquiring environmental information and constructing a scene map; extracting road features from the scene map, determining a driving road and a road type corresponding to the driving road based on the road features; generating a target path in the driving road according to the road type and a path generation rule corresponding to the road type; the step of generating the target path in the driving road according to the road type and the path generation rule corresponding to the road type comprises: determining the road type and the path generation rule according to the shape, length and width of the driving road, and generating the target path in the driving road based on the path generation rule; the step of determining the road type and the path generation rule according to the shape, length and width of the driving road, and generating the target path in the driving road based on the path generation rule comprises: if the shape of the driving road is linear, determining that the driving road is a straight road; if the length of the straight road is greater than a first length threshold and the width of the straight road is greater than a first width threshold, determining a first path generation rule, and generating the target path in the driving road based on the first path generation rule; if the length of the straight road is greater than the first length threshold and the width of the straight road is less than or equal to the first width threshold, determining a second path generation rule, and generating the target path in the driving road based on the second path generation rule; if the length of the straight road is less than or equal to the first length threshold and the width of the straight road is greater than the first width threshold, determining a third path generation rule, and generating the target path in the driving road based on the third path generation rule; if the length of the straight road is less than or equal to the first length threshold and the width of the straight road is less than or equal to the first width threshold, determining a fourth path generation rule, and generating the target path in the driving road based on the fourth path generation rule.

2. The method of claim 1, wherein, the road features comprise one or more of the boundary line, the lane line, the pedestrian crossing line and the speed bump of the road; the step of extracting the road features from the scene map and determining the driving road and the road type corresponding to the driving road based on the road features comprises: extracting one or more of the boundary line, the lane line, the pedestrian crossing line and the speed bump of the road from the scene map to determine the range of the driving road; determining the road type according to the range of the driving road; dividing the driving road into one or more road segments, and determining the road type corresponding to each road segment according to the one or more road segments, wherein the road type comprises a straight road, a curved road or an intersection.

3. The method of claim 1, wherein, the step of determining the road type and the path generation rule according to the shape, length and width of the driving road, and generating the target path in the driving road based on the path generation rule further comprises: if the shape of the driving road is arc-shaped, determining that the driving road is a curved road; determining the path generation rule according to the bending degree of the curved road and the width of the curved road, and generating the target path in the driving road based on the path generation rule.

4. The method according to any one of claims 1 to 3, characterized in that, the step of generating the target path in the driving road comprises: setting the driving direction and the driving speed of the target path when generating the target path in the driving road.

5. The method of claim 1, wherein, The method further comprises, after the step of generating the target path in the driving road according to the road type and the path generation rule corresponding to the road type: testing the target path to optimize the position, driving direction and driving speed of the target path.

6. A route generating apparatus characterized by comprising: The method comprises: a construction module configured to acquire environment information and construct a scene map; a determination module configured to extract road features from the scene map and determine a driving road and a road type corresponding to the driving road based on the road features; a generation module configured to generate a target path in the driving road according to the road type and a path generation rule corresponding to the road type; The step of generating the target path in the driving road according to the road type and the path generation rule corresponding to the road type comprises: determining the road type and the path generation rule according to the shape, length and width of the driving road, and generating the target path in the driving road based on the path generation rule; The step of determining the road type and the path generation rule according to the shape, length and width of the driving road, and generating the target path in the driving road based on the path generation rule comprises: if the shape of the driving road is linear, determining that the driving road is a straight road; if the length of the straight road is greater than a first length threshold and the width of the straight road is greater than a first width threshold, determining a first path generation rule and generating the target path in the driving road based on the first path generation rule; if the length of the straight road is greater than the first length threshold and the width of the straight road is less than or equal to the first width threshold, determining a second path generation rule and generating the target path in the driving road based on the second path generation rule; if the length of the straight road is less than or equal to the first length threshold and the width of the straight road is greater than the first width threshold, determining a third path generation rule and generating the target path in the driving road based on the third path generation rule; if the length of the straight road is less than or equal to the first length threshold and the width of the straight road is less than or equal to the first width threshold, determining a fourth path generation rule and generating the target path in the driving road based on the fourth path generation rule.

7. A robot, characterized in that The method comprises: a processor and a memory connected to the processor in communication; The memory stores computer program instructions, which, when invoked by the processor, cause the processor to execute the path generation method of any one of claims 1-5.

8. A storage medium, characterized by The storage medium stores computer executable instructions, which, when executed by at least one processor, cause the at least one processor to execute the method of any one of claims 1-5.

Images