Path planning method and system of autonomous vehicle, electronic device and medium

Abstract

The application discloses a path planning method and system of an autonomous vehicle, an electronic device and a medium. The method comprises the following steps: acquiring road surface environment information; perceiving traffic participants on the road surface based on the road surface environment information, and acquiring perception information of each traffic participant; determining the uncertainty of the category perception of the traffic participant based on the category information in the perception information, and acquiring perception uncertainty information for identifying the uncertainty; acquiring the first potential field weight of the traffic participant based on the position information, the category information and the perception uncertainty information; and generating a path planning scheme of the autonomous vehicle by using an autonomous driving system based on the first potential field weight. The method can avoid traffic accidents caused by category perception uncertainty factors, and is beneficial to improving the safety of the autonomous vehicle.

Description

Technical Field

[0001] This application relates to the field of autonomous driving technology, and in particular to a path planning method for autonomous vehicles, a path planning system for autonomous vehicles, electronic devices, and computer-readable storage media. Background Technology

[0002] With the continuous increase in global car ownership, limited road resources have led to frequent traffic congestion and accidents. The development of autonomous vehicle technology offers a feasible solution to this problem. However, in today's complex traffic scenarios, safety is the primary concern for autonomous vehicles, especially intended functional safety, which is also a significant factor limiting their widespread adoption. The limitations of artificial intelligence algorithms, the complexity of dynamic operating environments, and the unintended behaviors of traffic participants will trigger intended functional safety risks in autonomous vehicles.

[0003] Unlike functional safety risks caused by electronic / electrical system failures and information security risks caused by cyberattacks, anticipated functional safety risks refer to risks arising from insufficient functionality, performance limitations, or reasonably foreseeable human misuse of autonomous driving systems. Anticipated functional safety risks comprise two main elements: triggering conditions and insufficient or limited system functionality. Typically, triggering conditions from the operational design domain include extreme weather conditions (such as rain, snow, fog, and strong light) and unexpected road conditions (such as road bumps, road defects, and falling objects from vehicles ahead). Extreme weather conditions can cause perception problems, while unexpected road conditions can affect the decision-making and control functions of autonomous driving, directly leading to anticipated functional safety risks. Insufficient or limited system functionality is mainly due to the limitations of artificial intelligence algorithms, which may lead to missed detections, false detections, and decision-making errors. For example, false detections of traffic participant categories can directly lead to decision-making errors and even safety accidents. Summary of the Invention

[0004] In view of the above-mentioned problems existing in the prior art, this application provides a path planning method for autonomous vehicles, a path planning system for autonomous vehicles, an electronic device, and a computer-readable storage medium, which can avoid traffic accidents caused by category perception uncertainty factors, thereby improving the safety of autonomous vehicles.

[0005] To address the above problems, the technical solution adopted in the embodiments of this application is as follows:

[0006] A path planning method for an autonomous vehicle, comprising:

[0007] Obtain road surface environment information;

[0008] Based on the road surface environment information, traffic participants on the road surface are sensed, and the sensing information of each traffic participant is obtained; wherein, the sensing information includes the location information of the traffic participants and the category information of the traffic participants; the category information includes multiple category identifiers and confidence coefficients corresponding to each category identifier;

[0009] Based on the category information, determine the uncertainty of the category perception of traffic participants, and obtain the perception uncertainty information used to identify the uncertainty;

[0010] Based on the location information, the category information, and the perceived uncertainty information, a first potential field weight of the traffic participant is obtained; wherein, the first potential field weight is used to identify the repulsive field strength of the traffic participant in the artificial potential field algorithm, and at least a portion of the first potential field weight is positively correlated with the perceived uncertainty information.

[0011] Based on the first potential field weights, the autonomous driving system generates a path planning scheme for autonomous vehicles.

[0012] In some embodiments, the step of sensing traffic participants on the road surface based on the road environment information and obtaining sensing information of each traffic participant includes:

[0013] Based on the road environment information, traffic participants on the road surface are perceived, and a perception model with N perception networks is used to output N perception results that correspond one-to-one with the perception networks; wherein, the perception results include the first perception information of one or more traffic participants perceived through the corresponding perception network.

[0014] Cluster analysis is performed on the N perception results to obtain clustering results; wherein, the clustering results include M second perception information corresponding one-to-one with traffic participants, the second perception information being formed by clustering based on one or more first perception information; N and M are both positive integers greater than or equal to 1.

[0015] In some embodiments, determining the uncertainty of the perceived category of traffic participants based on the category information and obtaining perceived uncertainty information for identifying the uncertainty includes:

[0016] The Shannon entropy of the confidence coefficient of each category identifier in the category information is determined, and the sum of the multiple Shannon entropies is obtained as the perceived uncertainty information.

[0017] In some embodiments, determining the uncertainty of the category perception of traffic participants based on the category information and obtaining perception uncertainty information for identifying the uncertainty includes: determining the category perception uncertainty of traffic participants based on the category information through an uncertainty estimation model, and obtaining the perception uncertainty information;

[0018] The uncertainty estimation model is expressed by the following formula:

[0019]

[0020] Where Pen represents perceived uncertainty information; Pc represents the confidence coefficient.

[0021] In some embodiments, obtaining the first potential field weights of traffic participants based on the location information, the category information, and the perceived uncertainty information includes:

[0022] Based on the perceived uncertainty information, traffic participants are divided into a first category of traffic participants, a second category of traffic participants, and a third category of traffic participants; wherein, the perceived uncertainty information of the first category of traffic participants meets a first threshold range, the perceived uncertainty information of the second category of traffic participants meets a second threshold range, and the perceived uncertainty information of the third category of traffic participants meets a third threshold range, wherein the first threshold range is smaller than the second threshold range, and the second threshold range is smaller than the third threshold range.

[0023] Based on the location information and category information of the first type of traffic participants, the first potential field weight of the first type of traffic participants is determined using a preset algorithm;

[0024] Set the first potential field weight of the second type of traffic participants as the target value;

[0025] The first potential field weight of the third type of traffic participant is configured to be greater than the target value and positively correlated with its perceived uncertainty information.

[0026] In some embodiments, the method further includes:

[0027] Based on weather information and road surface environment information, a first constraint condition is determined; wherein, the first constraint condition includes a speed constraint condition and / or a position constraint condition;

[0028] Based on legal and regulatory information, the autonomous vehicle's own information, and the road environment information, a second constraint condition is determined; wherein, the second constraint condition includes a speed constraint condition and / or a position constraint condition.

[0029] Correspondingly, the step of generating a path planning scheme for the autonomous vehicle using the autonomous driving system based on the first potential field weights includes:

[0030] Based on the first potential field weight, the first constraint, and the second constraint, a path planning scheme for autonomous vehicles is generated using the autonomous driving system.

[0031] In some embodiments, determining the first constraint based on weather information and road surface environment information includes:

[0032] Based on the weather information and the road environment information, determine the detection range of the autonomous vehicle's sensors under the current weather conditions;

[0033] Based on the detection distance, a first speed limit for the autonomous vehicle under the current weather conditions is determined.

[0034] In some embodiments, determining the second constraint based on legal and regulatory information, the autonomous vehicle's own information, and the road environment information includes:

[0035] Based on the autonomous vehicle's information and the road environment information, it is determined whether the current driving condition of the autonomous vehicle triggers the supervision triggering conditions of each regulation module; wherein, the regulation module is constructed based on one or more legal provisions in the legal information.

[0036] When the current driving condition of the autonomous vehicle triggers the supervision trigger condition of a regulation module, the second constraint condition is determined by the regulation module based on the autonomous vehicle information and the road environment information.

[0037] In some embodiments, determining whether the current driving condition of the autonomous vehicle triggers the supervision triggering conditions of each regulation module based on the vehicle information and the road environment information includes:

[0038] Based on the road environment information, the current motion state of traffic participants on the road is detected by the autonomous driving system, and the motion state of traffic participants at the next moment is predicted, so as to obtain the current motion information and the motion information at the next moment of each traffic participant.

[0039] Based on traffic sign and marking information, autonomous vehicle information, current motion information of traffic participants, and next-moment motion information, determine whether the current driving status of the autonomous vehicle triggers the supervision trigger conditions of each regulation module.

[0040] A path planning system for an autonomous vehicle includes an autonomous driving system and a safety decision-making system, wherein the safety decision-making system includes a perception self-detection subsystem; wherein:

[0041] The self-detection subsystem is used for:

[0042] Obtain road surface environment information;

[0043] Based on the road surface environment information, traffic participants on the road surface are sensed, and the sensing information of each traffic participant is obtained; wherein, the sensing information includes the location information of the traffic participants and the category information of the traffic participants; the category information includes multiple category identifiers and confidence coefficients corresponding to each category identifier;

[0044] Based on the category information, determine the uncertainty of the category perception of traffic participants, and obtain the perception uncertainty information used to identify the uncertainty;

[0045] Based on the location information, the category information, and the perceived uncertainty information, a first potential field weight of the traffic participant is obtained; wherein, the first potential field weight is used to identify the repulsive field strength of the traffic participant in the artificial potential field algorithm, and at least a portion of the first potential field weight is positively correlated with the perceived uncertainty information.

[0046] The autonomous driving system is used for:

[0047] Based on the first potential field weights, a path planning scheme for autonomous vehicles is generated.

[0048] An electronic device includes at least a memory and a processor, wherein an application is stored on the memory, and the processor implements the method described above when executing the application on the memory.

[0049] A computer-readable storage medium storing computer-executable instructions, wherein executing the computer-executable instructions in the computer-readable storage medium implements the method described above.

[0050] The path planning method for autonomous vehicles in this application embodiment can correct the influence range of the repulsive field of a traffic participant by using a first potential field weight when the perceived category of the traffic participant has high uncertainty. This increases the avoidance distance of the autonomous vehicle from the traffic participant, avoids traffic accidents caused by uncertainty in category perception, and helps improve the safety of autonomous vehicles. Attached Figure Description

[0051] Figure 1 This is a flowchart of the path planning method for autonomous vehicles according to this application;

[0052] Figure 2 This is a flowchart of step S120 in the path planning method for autonomous vehicles of this application;

[0053] Figure 3 This is a scene diagram of step S120 in the path planning method for autonomous vehicles of this application;

[0054] Figure 4 This is a flowchart of step S160 in the path planning method for autonomous vehicles of this application;

[0055] Figure 5 This is a flowchart of step S170 in the path planning method for autonomous vehicles of this application;

[0056] Figure 6 This is a structural block diagram of the path planning system for the autonomous vehicle of this application;

[0057] Figure 7 This is a structural block diagram of the electronic device of this application. Detailed Implementation

[0058] Various embodiments and features of this application are described herein with reference to the accompanying drawings.

[0059] It should be understood that various modifications can be made to the embodiments described herein. Therefore, the above description should not be considered as limiting, but merely as an example of embodiments. Other modifications within the scope and spirit of this application will be apparent to those skilled in the art.

[0060] The accompanying drawings, which are included in and form part of this specification, illustrate embodiments of the present application and, together with the general description of the present application given above and the detailed description of the embodiments given below, serve to explain the principles of the present application.

[0061] These and other features of this application will become apparent from the following description of preferred forms of embodiments given as non-limiting examples, with reference to the accompanying drawings.

[0062] It should also be understood that although this application has been described with reference to some specific examples, those skilled in the art can certainly implement many other equivalent forms of this application.

[0063] The above and other aspects, features and advantages of this application will become more apparent when taken in conjunction with the accompanying drawings and in view of the following detailed description.

[0064] Specific embodiments of this application are described thereafter with reference to the accompanying drawings; however, it should be understood that the claimed embodiments are merely examples of this application, which can be implemented in various ways. Well-known and / or repeated functions and structures are not described in detail to avoid unnecessary or redundant details that could obscure the application. Therefore, the specific structural and functional details claimed herein are not intended to be limiting, but merely serve as the basis and representative basis for the claims to teach those skilled in the art to use this application in a variety of substantially any suitable detailed structures.

[0065] This specification may use the phrases “in one embodiment,” “in another embodiment,” “in yet another embodiment,” or “in other embodiments,” all of which may refer to one or more of the same or different embodiments according to this application.

[0066] This application provides a path planning method for autonomous vehicles, used to plan the driving path of autonomous vehicles and generate path planning schemes. Figure 1 This is a flowchart of the path planning method for an autonomous vehicle according to an embodiment of this application. See also... Figure 1 As shown, the path planning method for autonomous vehicles in this application embodiment may specifically include the following steps.

[0067] S110, obtain road surface environment information.

[0068] Optionally, road environment information can be acquired through onboard sensors or from other electronic devices via network communication. Onboard sensors may include onboard image acquisition devices, LiDAR, millimeter-wave radar, and ultrasonic radar, etc. Based on this, road environment information may include image information acquired by the image acquisition device, point cloud information acquired by LiDAR, detection information from millimeter-wave radar, and detection information from ultrasonic radar, etc.

[0069] Optionally, road environment information may also include weather information, traffic sign and marking information, and real-time road condition information. Weather information can be obtained through meteorological sensors, such as rain sensors, snow depth sensors, and visibility sensors. Weather information can also be obtained from servers or databases that provide weather information services. Traffic sign and marking information can be loaded onto the autonomous vehicle's local storage as part of the map information, and can then be obtained locally from the autonomous vehicle. Traffic sign and marking information and real-time road condition information can also be obtained in real time from servers or databases via network communication. Of course, weather information, traffic sign and marking information, and road condition information may also be excluded from road environment information and treated as independent types of information.

[0070] S120, based on the road surface environment information, sense traffic participants on the road surface and obtain sensing information for each traffic participant. The sensing information includes the location information of the traffic participants and their category information; the category information includes multiple category identifiers and confidence coefficients corresponding to each category identifier.

[0071] Optionally, traffic participants on the road surface can be perceived based on image information acquired by an image acquisition device, point cloud information acquired by lidar, detection information from millimeter-wave radar, and / or detection information from ultrasonic radar to obtain perception information for each traffic participant. Alternatively, traffic participants on the road surface can be perceived using a perception model. That is, road environment information can be used as input information to the perception model, and the perception model can output perception information for each traffic participant. For example, image information can be used as input information to the perception model, and the perception model can output perception information for each traffic participant.

[0072] Optionally, the location information of traffic participants can be two-dimensional or three-dimensional. Taking traffic participants on the road surface perceived based on image information as an example, this location information can be location information in pixel coordinates, image coordinates, camera coordinates, or world coordinates, etc. For example, this location information may include the coordinate information of the smallest bounding rectangle of the traffic participant in pixel coordinates or image coordinates.

[0073] Optionally, traffic participants can include various objects that can appear on the road surface, including but not limited to people, animals, cars, trucks, construction vehicles, bicycles, electric vehicles, guardrails, traffic signs and markings, etc., which will not be listed here. Based on this, category information can include category identifiers to identify the multiple categories to which a traffic participant may belong, and confidence coefficients for each category identifier. Category identifiers can include category names, category numbers, or other information that can uniquely identify the category to which a traffic participant belongs. Confidence coefficients characterize the degree of confidence that a traffic participant belongs to a certain category. This confidence coefficient can include the probability that a traffic participant belongs to a certain category, or it can include other numerical values ​​that can characterize the degree of confidence.

[0074] S130, determine the uncertainty of the category perception of traffic participants based on the category information, and obtain perception uncertainty information used to identify the uncertainty.

[0075] The uncertainty in the perceived category of a traffic participant is the uncertainty in perceiving the category to which a traffic participant belongs. Perceived uncertainty information is used to identify this uncertainty. If it is certain that a traffic participant belongs to a certain category, and the confidence coefficient for that category is significantly higher than the confidence coefficients for other categories, then the perceived category of that traffic participant has low uncertainty, and the value of the perceived uncertainty information is small. If it is uncertain whether a traffic participant belongs to a certain category, and the confidence coefficients corresponding to each category are relatively close, then the perceived category of that traffic participant has high uncertainty, and the value of the perceived uncertainty information is large.

[0076] For example, the category information of a certain traffic participant includes three categories: humans, warning cones, and light poles. The confidence coefficient for humans is 0.5, for warning cones it is 0.45, and for light poles it is 0.05. Because the confidence coefficients for humans and warning cones are very close, it is impossible to accurately determine whether the traffic participant belongs to humans or warning cones. Therefore, in this case, the traffic participant should have a high degree of uncertainty, and the corresponding perceived uncertainty information should have a large value.

[0077] For example, the category information of a traffic participant includes two category labels: cars and trucks. The confidence coefficient for cars is 0.9, and the confidence coefficient for trucks is 0.1. The two confidence coefficients differ significantly, indicating a high degree of certainty that the traffic participant is a car. In this case, the traffic participant should have low uncertainty, and the corresponding perceived uncertainty information should have a small value.

[0078] Optionally, the perceived uncertainty information can be calculated based on the category information using an uncertainty algorithm. That is, the category information can be used as the input information for the uncertainty algorithm, and the uncertainty algorithm can be used to solve for the perceived uncertainty information.

[0079] S140, based on the location information, the category information, and the perceived uncertainty information, obtain the first potential field weight of the traffic participant. The first potential field weight is used to identify the repulsive field strength of the traffic participant in the artificial potential field algorithm, and at least a portion of the first potential field weight is positively correlated with the perceived uncertainty information.

[0080] In artificial potential field algorithms, a repulsive field is used to guide autonomous vehicles to avoid obstacles. The stronger the repulsive field, the larger its influence range, and the greater the avoidance distance for the autonomous vehicle. Conversely, the weaker the repulsive field, the smaller its influence range, and the smaller the avoidance distance for the autonomous vehicle.

[0081] Given the availability of information about perceived uncertainty, the first potential field weights of traffic participants can be calculated jointly based on location information, category information, and perceived uncertainty information. These first potential field weights can be configured to be positively correlated with the influence range of the repulsive field.

[0082] Optionally, the first potential field weights for all traffic participants can be configured to be positively correlated with the perceived uncertainty information. For example, the first potential field weights can be configured to be equal to the sum of the base weights and the additional weights. The base weights of traffic participants can be determined using an artificial potential field algorithm based on location and category information. The additional weights are determined based on the perceived uncertainty information and a scaling factor. The scaling factor is used to adjust the degree of influence of the perceived uncertainty information on the first potential field weights.

[0083] Optionally, the first potential field weights of some traffic participants can be configured to be positively correlated with the perceived uncertainty information. For example, a first threshold range, a second threshold range, and a third threshold range can be pre-configured, wherein the first threshold range is smaller than the second threshold range, and the second threshold range is smaller than the third threshold range. Based on the perceived uncertainty information, traffic participants are divided into three categories: Category 1, Category 2, and Category 3. That is, traffic participants whose perceived uncertainty information meets the first threshold range are identified as Category 1 traffic participants, those whose perceived uncertainty information meets the second threshold range are identified as Category 2 traffic participants, and those whose perceived uncertainty information meets the third threshold range are identified as Category 3 traffic participants. For Category 1 traffic participants, since their category perception is relatively certain, their first potential field weights can be determined based on location information and category information using a pre-set algorithm, such as an artificial potential field algorithm. For Category 2 traffic participants, which have moderate uncertainty, their first potential field weights can be set as target values. This target value can be, for example, the potential field weight of humans, since humans are traffic participants with a high level of protection and have a large influence range. For the third type of traffic participants, the first potential field weight can be configured to be positively correlated with the perceived uncertainty information, based on the target value. For example, the target value can be used as the basic weight, and additional weights can be calculated based on the perceived uncertainty information and the proportional coefficient. Then, the basic weight and the additional weights can be summed to obtain the first potential field weight.

[0084] It should be noted that the above method for calculating the first potential field weight is only an example. In specific implementation, various algorithms can be used to calculate the first potential field weight based on the perceived uncertainty information, as long as the first potential field weight of some traffic participants is positively correlated with the perceived uncertainty information.

[0085] S150, based on the first potential field weights, use the autonomous driving system to generate a path planning scheme for the autonomous vehicle.

[0086] The autonomous driving system is used to plan the driving path of autonomous vehicles based on an artificial potential field algorithm. Upon obtaining a first potential field weight, the autonomous driving system can directly use this first potential field weight as the target potential field weight for the traffic participant, and determine the influence range of the repulsive field of the traffic participant based on this first potential field weight. Alternatively, the autonomous driving system can also calculate a second potential field weight based on location and category information using an artificial potential field algorithm, and use an arbitration algorithm to arbitrate the first and second potential field weights, selecting one of them as the target potential field weight for the traffic participant. Alternatively, the autonomous driving system can also calculate a second potential field weight based on location and category information using an artificial potential field algorithm, and multiply the first and second potential field weights by a scaling factor to obtain the target potential field weight for the traffic participant.

[0087] Given the target potential field weights of traffic participants, the driving path of an autonomous vehicle can be planned based on an artificial potential field algorithm, generating a path planning scheme for the autonomous vehicle. This path planning scheme is used to control the autonomous vehicle to travel along the planned path. Since determining the influence range of the repulsive field using an artificial potential field algorithm, and planning the driving path of an autonomous vehicle using an artificial potential field algorithm, are existing technologies, they will not be elaborated upon here.

[0088] The path planning method for autonomous vehicles in this invention involves: perceiving traffic participants on the road surface based on road environment information, acquiring perception information for each traffic participant; determining the uncertainty of traffic participant category perception based on category information, acquiring perception uncertainty information to identify this uncertainty; acquiring a first potential field weight for the traffic participant based on the location information, the category information, and the perception uncertainty information; the first potential field weight can identify the repulsive field strength in the artificial potential field algorithm and can directly affect the influence range of the repulsive field of the traffic participant; and generating a path planning scheme for the autonomous vehicle using the autonomous driving system based on the first potential field weight. Thus, when the perceived category of a traffic participant has high uncertainty, the influence range of the repulsive field of that traffic participant can be corrected by the first potential field weight, increasing the avoidance distance of the autonomous vehicle for that traffic participant, avoiding traffic accidents caused by category perception uncertainty, and improving the safety of the autonomous vehicle.

[0089] Cooperate Figure 2 As shown, in some embodiments, step S120, which involves sensing traffic participants on the road surface based on the road environment information and obtaining sensing information for each traffic participant, may include the following steps.

[0090] S121, based on the road surface environment information, traffic participants on the road surface are sensed, and a perception model with N perception networks is used to output N perception results corresponding one-to-one with the perception networks. The perception results include first perception information of one or more traffic participants sensed through the corresponding perception network.

[0091] S122, perform cluster analysis on N perception results to obtain cluster results; wherein, the cluster results include M second perception information corresponding one-to-one with traffic participants, the second perception information being formed by clustering based on one or more first perception information; N and M are both positive integers greater than or equal to 1.

[0092] Optional, in combination Figure 3 As shown, N randomly initialized perceptual networks (YOLOv5) can be established using different random number seeds, resulting in a perceptual model with N perceptual networks, which can be represented as E = {O1, O2, O3, ..., O...}. N Each perception network can include three parts: a backbone, a neck, and a detection head. After the perception model is built, it can be trained using a shuffled training dataset. Once validated, the perception model can be used to perceive traffic participants on the road surface. Road environment information can be input into each perception network, and each perception network outputs a perception result. This perception result includes the first perception information of one or more traffic participants. This first perception information may include the location information and category information of the traffic participants. The set of perception results can be represented as P = {P1, P2, P3, ..., P...} N}, where P1, P2, ..., P N Each represents a different perception result.

[0093] Given N sensing results, cluster analysis can be performed on the first sensing information from these N results, grouping the first sensing information of the same traffic participants into one category to form second sensing information. Given M traffic participants, M pieces of second sensing information can be formed. This can be represented by C = {C1, C2, ..., C...} M} represents the clustering results, where C1, C2, ..., C M Each represents a different piece of second-sensory information.

[0094] Optionally, cluster analysis can be performed on the N perception results to obtain the clustering results.

[0095] Step 1: Define the cluster association index Affinity = {IoU & WL}, where IoU represents the positional association (IoU) threshold, and WL represents the class association condition, specifically, the class labels with the highest confidence coefficients being the same. Based on this, the meaning of the cluster association index is that the class labels with the highest confidence coefficients are the same, and the positional association is greater than the IoU threshold. For example, the IoU threshold θ = 0.95.

[0096] Step 2: Receive the input prediction set P = {P1, P2, P3, ..., P...} N}, for each first perceptual information B in the first perceptual result P1. 1，k Establish cluster subset C k .

[0097] Step 3: For the remaining i-th perception result P i Initialize the clustering condition by setting excl_flag = 0 M The meaning of this clustering condition is a perceptual result P. i Multiple first-perception information B i，j They cannot be assigned to the same cluster subset.

[0098] Step 4: For each cluster subset C k Check each perception result P i Each of the first perceived information B i，j Its correlation, if Affinity(B) i，j C k If )≥θ and excl_flag(k)=0, then B i，j Add to cluster subset C k In the middle, the clustering condition is updated to excl_flag(k) = 1.

[0099] Step 5, if the first perceived information B i，j If no existing cluster subset is assigned in step 4, the number of cluster subsets is updated to M = M + 1, and set to B. i，j Establish a new clustering subset C M .

[0100] Step 6: After all the initial perception results have been assigned, output the clustering result C = {C1, C2, ..., C}. M}

[0101] In some embodiments, step S130 involves determining the uncertainty of the category perception of traffic participants based on the category information and obtaining perception uncertainty information for identifying the uncertainty, which may specifically include...

[0102] The Shannon entropy of the confidence coefficient of each category identifier in the category information is determined, and the sum of the multiple Shannon entropies is obtained as the perceived uncertainty information.

[0103] Optionally, an uncertainty estimation model can be constructed, which can be configured to determine the Shannon entropy of the confidence coefficient of each category identifier in the category information, and obtain the sum of multiple Shannon entropies as perceived uncertainty information.

[0104] Specifically, the uncertainty estimation model can be expressed as the following formula.

[0105]

[0106] Where Pen represents perceived uncertainty information; Pc represents the confidence coefficient.

[0107] It should be noted that calculating the uncertainty of traffic participant category perception based on Shannon entropy is only one example. In specific implementation, various uncertainty algorithms can be used to calculate the uncertainty of traffic participant category perception. It should not be understood as being limited to calculating the uncertainty of traffic participant category perception only through Shannon entropy.

[0108] In some embodiments, the method further includes:

[0109] S160, based on weather information and the road surface environment information, determine the first constraint condition; wherein, the first constraint condition includes a speed constraint condition and / or a position constraint condition;

[0110] S170, based on legal and regulatory information, the vehicle information of the autonomous vehicle, and the road environment information, a second constraint condition is determined; wherein, the second constraint condition includes a speed constraint condition and / or a position constraint condition;

[0111] Correspondingly, step S150 involves generating a path planning scheme for the autonomous vehicle using the autonomous driving system based on the first potential field weights, including:

[0112] Based on the first potential field weight, the first constraint, and the second constraint, a path planning scheme for autonomous vehicles is generated using the autonomous driving system.

[0113] In this way, when planning the driving path of autonomous vehicles, the autonomous driving system can not only take into account the safety risks caused by the uncertainty of category perception, but also the expected functional safety risks caused by weather conditions and traffic laws and regulations, which is beneficial to further improving the safety of autonomous vehicles.

[0114] Optionally, in specific implementation, the first constraint condition can be determined based solely on weather information and road environment information, or the second constraint condition can be determined solely on legal and regulatory information, the autonomous vehicle's own information, and the road environment information, or both the first and second constraint conditions can be determined simultaneously.

[0115] In some embodiments, step S160, determining the first constraint condition based on weather information and the road surface environment information, may include the following steps.

[0116] S161, Based on the weather information and the road environment information, determine the detection distance of the autonomous vehicle's sensors under the current weather conditions.

[0117] S162, Based on the detection distance, determine the first speed limit of the autonomous vehicle under the current weather conditions.

[0118] Weather conditions directly impact sensor detection range. In adverse weather, sensor detection range is significantly reduced, and the ability to perceive anticipated functional safety risks is noticeably diminished. For example, rain, snow, or fog affects atmospheric transmissibility, consequently impacting the laser energy received and the detection range of lidar. Setting a primary speed limit based on current weather conditions can significantly reduce anticipated functional safety risks caused by decreased sensor perception capabilities, thereby improving the safety of autonomous vehicles.

[0119] Cooperate Figure 4 As shown, taking lidar as an example, rain and fog models can be pre-built. Current weather conditions can be determined based on road surface and weather information. In rainy weather, the rain model can be selected; in non-rainy weather, the fog model can be selected.

[0120] If a foggy weather model is selected, the detection range S of the lidar can be calculated using the following formula.

[0121]

[0122] Where S represents the detection range of the lidar under the current weather conditions; S standard E represents the standard detection range of a lidar. standard E represents the standard received energy of a lidar system. avg This represents the average received energy of the lidar.

[0123] If a rainy day model is selected, the water film thickness can be calculated based on the rainfall density:

[0124]

[0125] Where d is the water film thickness; L fI is the channel length; I is the rainfall density; S f Indicates the slope of the passage.

[0126] Calculate the road surface adhesion coefficient based on the water film thickness:

[0127] μ rain =0.21e -1.8d +0.4

[0128] Where, μ rain The coefficient of adhesion is denoted by d, and the thickness of the water film is d.

[0129] Subsequently, based on rainfall density and road surface adhesion coefficient, the detection range S of the lidar in rainy weather is output.

[0130] Given the sensor's detection range under current weather conditions, this detection range can be used as a responsibility-sensitive safety distance. Based on this responsibility-sensitive safety distance, the maximum safe speed of the autonomous vehicle can be determined, and this maximum safe speed can be used as the first speed limit for the autonomous vehicle under current weather conditions.

[0131] Optionally, the maximum safe speed of an autonomous vehicle can be calculated using the following function.

[0132] SSD(v lim ) = S

[0133] Where SSD() represents the responsibility-sensitive safety distance function; v lim S represents the maximum safe vehicle speed; S represents the sensor's detection range under current weather conditions.

[0134] It should be noted that although the above example uses lidar as an example to explain in detail the process of calculating the lidar detection range and the first speed limit, in actual implementation, it is not limited to determining the lidar detection range. It is also possible to calculate the detection range of sensors such as image acquisition devices, millimeter-wave radar, and ultrasonic radar, and set the first vehicle speed limit based on the detection range of these sensors.

[0135] In some embodiments, step S170, which determines the second constraint based on legal and regulatory information, the autonomous vehicle's information, and the road environment information, may include the following steps.

[0136] Based on the autonomous vehicle's information and the road environment information, it is determined whether the current driving condition of the autonomous vehicle triggers the supervision trigger conditions of each regulation module; wherein, the regulation module is constructed based on one or more legal provisions in the legal information.

[0137] When the current driving condition of the autonomous vehicle triggers the supervision trigger condition of a regulation module, the second constraint condition is determined by the regulation module based on the autonomous vehicle information and the road environment information.

[0138] Optionally, multiple regulation modules can be pre-constructed based on one or more legal provisions from the legal and regulatory information. The monitoring trigger conditions can be set based on the logical definitions of these one or more laws and regulations. Each regulation module can output a second constraint condition, which is used to constrain the driving behavior of the autonomous vehicle, ensuring that the autonomous vehicle complies with the provisions of these one or more laws and regulations. For example, a lane-changing regulation module can be set up for lane-changing behavior. When the monitoring trigger condition of the lane-changing regulation module is triggered, the module outputs a second constraint condition, ensuring that the autonomous vehicle changes lanes in compliance with laws and regulations. Alternatively, if it is determined that lane-changing by the autonomous vehicle under the current driving conditions may constitute a violation, then lane-changing by the autonomous vehicle can be restricted.

[0139] Optionally, the autonomous vehicle's self-vehicle information may include the autonomous vehicle's location information and driving information. The autonomous vehicle's location information can be represented as (Xe, Ye), where Xe is the autonomous vehicle's coordinate on the X-axis, and Ye is the autonomous vehicle's coordinate on the Y-axis. The X-axis may extend longitudinally along the autonomous vehicle, and the Y-axis may extend laterally along the autonomous vehicle. The autonomous vehicle's driving information may include the autonomous vehicle's speeds Vxe and Vye in the X-axis and Y-axis directions, vehicle length Le, vehicle width We, heading angle HAe, etc. The self-vehicle information may also include the autonomous vehicle's decision-making information, which may include information used to instruct the autonomous vehicle to perform driving operations such as acceleration, lane changing, or turning.

[0140] Optional, in combination Figure 5 As shown, determining whether the current driving condition of the autonomous vehicle triggers the supervision trigger conditions of each regulation module based on the vehicle information and the road environment information may include:

[0141] Based on the road environment information, the current motion state of traffic participants on the road is detected by the autonomous driving system, and the motion state of traffic participants at the next moment is predicted, so as to obtain the current motion information and the motion information at the next moment of each traffic participant.

[0142] Based on traffic sign and marking information, autonomous vehicle information, current motion information of traffic participants, and next-moment motion information, determine whether the current driving status of the autonomous vehicle triggers the supervision trigger conditions of each regulation module.

[0143] Optionally, the autonomous driving system may include a perception and prediction module. This module can detect the current motion state of traffic participants on the road based on road information and predict their motion state at the next moment. For example, images acquired by an image acquisition device can be input into the perception and prediction module, which can then output the current motion information and the motion information at the next moment for the traffic participants.

[0144] The current motion information of a traffic participant may include its current coordinates (Xti and Yti) in the X and Y axes, its velocities (Vxti and Vyti) in the X and Y axes, and its length (Lti) and width (Wti). The motion information for the next moment may include the predicted motion information of the traffic participant at the next moment, such as its coordinates (Xpti and Ypti) in the X and Y axes, and its velocities (Vpxti and Vpyti) in the X and Y axes, etc.

[0145] Optionally, traffic sign and marking information may include various signs and markings located on the road surface, roadside, and above the road. For example, it may include road markings and signs, roadside signs, and signs above the road. Roadside signs may include traffic lights and no-turn signs. Traffic sign and marking information can be pre-loaded into the autonomous vehicle's local storage as part of the map information, or it can be retrieved in real-time from a server or database.

[0146] Having acquired information on traffic signs and markings, the autonomous vehicle's own information, and the current and next-moment motion information of traffic participants, it can be determined whether the current driving condition of the autonomous vehicle triggers the supervision trigger conditions of each regulation module. If one or more regulation modules' supervision trigger conditions are triggered, the corresponding regulation module determines the second constraint condition based on the traffic sign and marking information, the autonomous vehicle's own information, and the current and next-moment motion information of traffic participants.

[0147] Taking the lane-changing regulation module as an example, when it receives decision information for an autonomous vehicle to change lanes to the left, a supervision trigger condition for the lane-changing regulation module is triggered. This trigger condition can be described as follows:

[0148]

[0149] Where Decision_change_leftlane = T represents the decision information for changing lanes to the left, Ego represents the autonomous vehicle, and x tLet d represent the left rear vehicle, s(x) represent the distance of target x along the lane direction from the origin in the coordinate system of the autonomous vehicle, vx(x) be the longitudinal velocity of target x, TTCx be the compliance threshold for the longitudinal collision time between the autonomous vehicle and the left rear vehicle, and d be the distance between the target x and the origin. clmin This is the compliance threshold for the longitudinal distance between an autonomous vehicle and the vehicle to its left rear.

[0150] If the current driving conditions meet the above-mentioned monitoring triggering conditions, the lane-changing regulation module is triggered, and the second constraint condition is output through this module. For example, if the lane-changing regulation module decides to restrict the autonomous vehicle from changing lanes until the distance between it and the vehicle to its left rear is compliant, the second constraint condition can be expressed as:

[0151] The lateral positional constraints for autonomous vehicles are:

[0152] rightline(Ego)+0.5w(Ego)<y(Ego)<leftline(Ego)-0.5w(Ego)

[0153] The lateral position reference for autonomous vehicles is:

[0154] y ref =[rightline(Ego)+leftline(Ego)] / 2

[0155] Where y(x) is the lateral coordinate of target x, w(x) is the lateral width of target x, rightline(x) is the coordinate of the right lane line of the lane where target x is located, and leftline(x) is the coordinate of the left lane line of the lane where target x is located. By setting the regulations module, the compliance of autonomous vehicles can be monitored, which can further improve the safety of autonomous vehicles.

[0156] See Figure 6 As shown, this application embodiment provides a path planning system for an autonomous vehicle, which includes an autonomous driving system 210 and a safety decision system 220, wherein the safety decision system 220 includes a perception self-detection subsystem 221; wherein:

[0157] The self-detection subsystem 221 is used for:

[0158] Obtain road surface environment information;

[0159] Based on the road surface environment information, traffic participants on the road surface are sensed, and the sensing information of each traffic participant is obtained; wherein, the sensing information includes the location information of the traffic participants and the category information of the traffic participants; the category information includes multiple category identifiers and confidence coefficients corresponding to each category identifier;

[0160] Based on the category information, determine the uncertainty of the category perception of traffic participants, and obtain the perception uncertainty information used to identify the uncertainty;

[0161] Based on the location information, the category information, and the perceived uncertainty information, a first potential field weight of the traffic participant is obtained; wherein, the first potential field weight is used to identify the repulsive field strength of the traffic participant in the artificial potential field algorithm, and at least a portion of the first potential field weight is positively correlated with the perceived uncertainty information.

[0162] The autonomous driving system 210 is used for:

[0163] Based on the first potential field weights, a path planning scheme for autonomous vehicles is generated.

[0164] In some embodiments, the self-sensing detection subsystem 221 is specifically used for:

[0165] Based on the road environment information, traffic participants on the road surface are perceived, and a perception model with N perception networks is used to output N perception results that correspond one-to-one with the perception networks; wherein, the perception results include the first perception information of one or more traffic participants perceived through the corresponding perception network.

[0166] Cluster analysis is performed on the N perception results to obtain clustering results; wherein, the clustering results include M second perception information corresponding one-to-one with traffic participants, the second perception information being formed by clustering based on one or more first perception information; N and M are both positive integers greater than or equal to 1.

[0167] In some embodiments, the self-sensing detection subsystem 221 is specifically used for:

[0168] The Shannon entropy of the confidence coefficient of each category identifier in the category information is determined, and the sum of the multiple Shannon entropies is obtained as the perceived uncertainty information.

[0169] In some embodiments, the self-sensing detection subsystem 221 is specifically used for:

[0170] Based on the category information, the perceived uncertainty of traffic participants is determined by an uncertainty estimation model, and the perceived uncertainty information is obtained.

[0171] The uncertainty estimation model is expressed by the following formula:

[0172]

[0173] Where Pen represents perceived uncertainty information; Pc represents the confidence coefficient.

[0174] In some embodiments, the self-sensing detection subsystem 221 is specifically used for:

[0175] Based on the perceived uncertainty information, traffic participants are divided into a first category of traffic participants, a second category of traffic participants, and a third category of traffic participants; wherein, the perceived uncertainty information of the first category of traffic participants meets a first threshold range, the perceived uncertainty information of the second category of traffic participants meets a second threshold range, and the perceived uncertainty information of the third category of traffic participants meets a third threshold range, wherein the first threshold range is smaller than the second threshold range, and the second threshold range is smaller than the third threshold range.

[0176] Based on the location information and category information of the first type of traffic participants, the first potential field weight of the first type of traffic participants is determined using a preset algorithm;

[0177] Set the first potential field weight of the second type of traffic participants as the target value;

[0178] The first potential field weight of the third type of traffic participant is configured to be greater than the target value and positively correlated with its perceived uncertainty information.

[0179] In some embodiments, the safety decision system further includes a weather condition detection subsystem 222 and / or a compliance detection subsystem 223;

[0180] The weather condition detection subsystem 222 is used for:

[0181] Based on weather information and road surface environment information, a first constraint condition is determined; wherein, the first constraint condition includes a speed constraint condition and / or a position constraint condition;

[0182] The compliance testing subsystem 223 is used for:

[0183] Based on legal and regulatory information, the autonomous vehicle's own information, and the road environment information, a second constraint condition is determined; wherein, the second constraint condition includes a speed constraint condition and / or a position constraint condition.

[0184] The autonomous driving system 210 is specifically used for:

[0185] Based on the first potential field weight, the first constraint, and the second constraint, a path planning scheme for autonomous vehicles is generated using the autonomous driving system.

[0186] In some embodiments, the weather condition detection subsystem 222 is specifically used for:

[0187] Based on the weather information and the road environment information, determine the detection range of the autonomous vehicle's sensors under the current weather conditions;

[0188] Based on the detection distance, a first speed limit for the autonomous vehicle under the current weather conditions is determined.

[0189] In some embodiments, the compliance testing subsystem 223 is specifically used for:

[0190] Based on the autonomous vehicle's information and the road environment information, it is determined whether the current driving condition of the autonomous vehicle triggers the supervision triggering conditions of each regulation module; wherein, the regulation module is constructed based on one or more legal provisions in the legal information.

[0191] When the current driving condition of the autonomous vehicle triggers the supervision trigger condition of a regulation module, the second constraint condition is determined by the regulation module based on the autonomous vehicle information and the road environment information.

[0192] In some embodiments, the compliance testing subsystem 223 is specifically used for:

[0193] Based on the road environment information, the current motion state of traffic participants on the road is detected by the autonomous driving system, and the motion state of traffic participants at the next moment is predicted, so as to obtain the current motion information and the motion information at the next moment of each traffic participant.

[0194] Based on traffic sign and marking information, autonomous vehicle information, current motion information of traffic participants, and next-moment motion information, determine whether the current driving status of the autonomous vehicle triggers the supervision trigger conditions of each regulation module.

[0195] See Figure 7 As shown, this application embodiment also provides an electronic device, including at least a memory 301 and a processor 302. The memory 301 stores an application program, and the processor 302 implements the method described in any of the above embodiments when executing the application program on the memory 301.

[0196] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed, implement the method described in any of the above embodiments.

[0197] Those skilled in the art will understand that embodiments of this application can be provided as methods, electronic devices, computer-readable storage media, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware. Furthermore, this application can take the form of a computer program product implemented on one or more computer-readable storage media containing computer-readable program code. When implemented in software, these functions can be stored in a computer-readable medium or transmitted as one or more instructions or code on a computer-readable medium.

[0198] The aforementioned processor can be a general-purpose processor, a digital signal processor, an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The aforementioned PLD can be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL), or any combination thereof. The general-purpose processor can be a microprocessor or any conventional processor, etc.

[0199] The aforementioned memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, like read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0200] The aforementioned readable storage medium may be a magnetic disk, optical disk, DVD, USB, read-only memory (ROM) or random access memory (RAM), etc. This application does not limit the specific form of storage medium.

[0201] The above embodiments are merely exemplary embodiments of this application and are not intended to limit this application. The scope of protection of this application is defined by the claims. Those skilled in the art can make various modifications or equivalent substitutions to this application within its substance and scope of protection, and such modifications or equivalent substitutions should also be considered to fall within the scope of protection of this application.

Claims

1. A path planning method for an autonomous vehicle, characterized in that, include: Obtain road surface environment information; Based on the road surface environment information, traffic participants on the road surface are sensed, and the sensing information of each traffic participant is obtained; wherein, the sensing information includes the location information of the traffic participants and the category information of the traffic participants; the category information includes multiple category identifiers and confidence coefficients corresponding to each category identifier; Based on the category information, determine the uncertainty of the category perception of traffic participants, and obtain the perception uncertainty information used to identify the uncertainty; Based on the location information, the category information, and the perceived uncertainty information, a first potential field weight of the traffic participant is obtained; wherein, the first potential field weight is used to identify the repulsive field strength of the traffic participant in the artificial potential field algorithm, and at least a portion of the first potential field weight is positively correlated with the perceived uncertainty information. Based on the first potential field weights, the autonomous driving system generates a path planning scheme for autonomous vehicles.

2. The method according to claim 1, characterized in that, The process of sensing traffic participants on the road surface based on the road environment information and acquiring sensing information for each traffic participant includes: Based on the road environment information, traffic participants on the road surface are perceived, and a perception model with N perception networks is used to output N perception results that correspond one-to-one with the perception networks; wherein, the perception results include the first perception information of one or more traffic participants perceived through the corresponding perception network. Cluster analysis is performed on the N perception results to obtain clustering results; wherein, the clustering results include M second perception information corresponding one-to-one with traffic participants, the second perception information being formed by clustering based on one or more first perception information; N and M are both positive integers greater than or equal to 1.

3. The method according to claim 1, characterized in that, The step of determining the uncertainty of the perceived category of traffic participants based on the category information and obtaining perceived uncertainty information for identifying this uncertainty includes: The Shannon entropy of the confidence coefficient of each category identifier in the category information is determined, and the sum of multiple Shannon entropies is obtained as the perceived uncertainty information.

4. The method according to claim 1, characterized in that, The step of determining the uncertainty of the category perception of traffic participants based on the category information and obtaining the perceived uncertainty information used to identify the uncertainty includes: determining the category perception uncertainty of traffic participants based on the category information through an uncertainty estimation model, and obtaining the perceived uncertainty information; The uncertainty estimation model is expressed by the following formula: Where Pen represents perceived uncertainty information; Pc represents the confidence coefficient.

5. The method according to claim 1, characterized in that, The step of obtaining the first potential field weights of traffic participants based on the location information, the category information, and the perceived uncertainty information includes: Based on the perceived uncertainty information, traffic participants are divided into a first category of traffic participants, a second category of traffic participants, and a third category of traffic participants; wherein, the perceived uncertainty information of the first category of traffic participants meets a first threshold range, the perceived uncertainty information of the second category of traffic participants meets a second threshold range, and the perceived uncertainty information of the third category of traffic participants meets a third threshold range, wherein the first threshold range is smaller than the second threshold range, and the second threshold range is smaller than the third threshold range. Based on the location information and category information of the first type of traffic participants, the first potential field weight of the first type of traffic participants is determined using a preset algorithm; Set the first potential field weight of the second type of traffic participants as the target value; The first potential field weight of the third type of traffic participant is configured to be greater than the target value and positively correlated with its perceived uncertainty information.

6. The method according to claim 1, characterized in that, The method further includes: Based on weather information and road surface environment information, a first constraint condition is determined; wherein, the first constraint condition includes a speed constraint condition and / or a position constraint condition; Based on legal and regulatory information, the autonomous vehicle's own information, and the road environment information, a second constraint condition is determined; wherein, the second constraint condition includes a speed constraint condition and / or a position constraint condition. Correspondingly, the step of generating a path planning scheme for the autonomous vehicle using the autonomous driving system based on the first potential field weights includes: Based on the first potential field weight, the first constraint, and the second constraint, a path planning scheme for autonomous vehicles is generated using the autonomous driving system.

7. The method according to claim 6, characterized in that, The determination of the first constraint condition based on weather information and road surface environment information includes: Based on the weather information and the road environment information, determine the detection range of the autonomous vehicle's sensors under the current weather conditions; Based on the detection distance, a first speed limit for the autonomous vehicle under the current weather conditions is determined.

8. The method according to claim 6, characterized in that, The second constraint condition, determined based on legal and regulatory information, the autonomous vehicle's information, and the road environment information, includes: Based on the autonomous vehicle's information and the road environment information, it is determined whether the current driving condition of the autonomous vehicle triggers the supervision triggering conditions of each regulation module; wherein, the regulation module is constructed based on one or more legal provisions in the legal information. When the current driving condition of the autonomous vehicle triggers the supervision trigger condition of a regulation module, the second constraint condition is determined by the regulation module based on the autonomous vehicle information and the road environment information.

9. The method according to claim 8, characterized in that, The determination of whether the current driving condition of the autonomous vehicle triggers the supervision trigger conditions of each regulation module based on the vehicle information and the road environment information includes: Based on the road environment information, the current motion state of traffic participants on the road is detected by the autonomous driving system, and the motion state of traffic participants at the next moment is predicted, so as to obtain the current motion information and the motion information at the next moment of each traffic participant. Based on traffic sign and marking information, autonomous vehicle information, current motion information of traffic participants, and next-moment motion information, determine whether the current driving status of the autonomous vehicle triggers the supervision trigger conditions of each regulation module.

10. A path planning system for an autonomous vehicle, characterized in that, It includes an autonomous driving system and a safety decision-making system, wherein the safety decision-making system includes a perception self-detection subsystem; wherein: The self-detection subsystem is used for: Obtain road surface environment information; Based on the road surface environment information, traffic participants on the road surface are sensed, and the sensing information of each traffic participant is obtained; wherein, the sensing information includes the location information of the traffic participants and the category information of the traffic participants; the category information includes multiple category identifiers and confidence coefficients corresponding to each category identifier; Based on the category information, determine the uncertainty of the category perception of traffic participants, and obtain the perception uncertainty information used to identify the uncertainty; Based on the location information, the category information, and the perceived uncertainty information, a first potential field weight of the traffic participant is obtained; wherein, the first potential field weight is used to identify the repulsive field strength of the traffic participant in the artificial potential field algorithm, and at least a portion of the first potential field weight is positively correlated with the perceived uncertainty information. The autonomous driving system is used for: Based on the first potential field weights, a path planning scheme for autonomous vehicles is generated.

11. An electronic device, characterized in that, It includes at least a memory and a processor, wherein an application is stored on the memory, and the processor implements the method as claimed in any one of claims 1 to 9 when executing the application on the memory.

12. A computer-readable storage medium storing computer-executable instructions, characterized in that, The method of any one of claims 1-9 is implemented when the computer-executable instructions in the computer-readable storage medium are executed.

Images