Priori danger degree measurement graph and construction method thereof, driving method and storage medium

By constructing a priori risk degree measurement map, the problem of autonomous driving system judging the degree of road hazard is solved, the detailed evaluation of road safety and bicycle planning control are achieved, and the safety and strategy selection capabilities of autonomous driving system are improved.

CN118991817AActive Publication Date: 2024-11-22GUANGZHOU WERIDE TECH LTD CO
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
CN202410934592.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-12
Publication Date
2024-11-22
Estimated Expiration
2044-07-12

AI Technical Summary

Technical Problem

The autonomous driving system cannot effectively judge the risk level of the road and lacks the judgment ability of human drivers, which affects their safety and policy choices.

Method used

By constructing a priori risk degree measurement map, obtain basic map information and historical driving behavior data of the target area, establish a hazard degree measurement model, nonlinear mapping the risk probability to the hazard degree measurement, generate a priori risk degree measurement map of the target area, use In the bicycle planning control.

Benefits of technology

It has achieved a detailed assessment of the safety level of specific roads and sections, improved the safety and policy selection capabilities of the autonomous driving system, and provided prior hazard warnings and driving style recommendations.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118991817A_ABST
    Figure CN118991817A_ABST
Patent Text Reader

Abstract

The invention discloses a priori danger degree measurement map, a construction method thereof, a driving method and a storage medium. The method comprises the following steps: acquiring basic map information of a target area and storing a first corresponding relation between the basic map information and the basic map information; acquiring historical driving behavior data of the vehicle in the target area and extracting a road safety event from the historical driving behavior data according to a preset rule; obtaining a second corresponding relation between the road safety event and the position information of the target area; establishing a danger degree measurement model according to the second corresponding relation, and obtaining a third corresponding relation between the position information of the target area and danger degree measurement; and summarizing the position information of the target area and the corresponding danger degree measurement according to the third corresponding relationship, and constructing a prior danger degree measurement map of the target area. Therefore, for the automatic driving area, a priori danger degree measurement graph can be obtained, the danger degree can be directly inquired when needed, and the quality of the automatic driving system and the safety degree of the road section can be estimated in a more detailed manner.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention is a patent funded by Shanghai Pujiang Talent Program ("Autonomous Driving Behavior and Trajectory Prediction Algorithm Based on Deep Neural Network" (No. 22PJD087)). The technical content relates to the field of transportation technology, and in particular to a priori danger degree measurement map and its construction method, driving method and storage medium. Background Art

[0002] When human drivers are driving on familiar roads, they usually remember that some special sections require special attention. For example, there are often bicycles turning right at some intersections, and there are few cars on some highways, so they can be more relaxed without worrying about being squeezed in.

[0003] However, for autonomous driving, since there is no human driver, it is impossible to manually judge the danger level of a certain road section. But this degree of danger of a certain road section is precisely an important requirement to ensure the safety of autonomous driving, especially to quantify the risks under various conditions.

[0004] Therefore, how to give autonomous driving the same ability as manual driving to judge the danger level of a road section based on the prior geographical location to help the autonomous driving system choose the appropriate strategy has become an urgent problem to be solved. Summary of the invention

[0005] The main purpose of the present invention is to provide a priori danger degree measurement map and its construction method, driving method and storage medium, which can more finely estimate the quality of the automatic driving system and the safety level of the road section for a specific road.

[0006] To achieve the above object, the present invention provides a method for constructing a priori risk degree measurement map, the method comprising the following steps:

[0007] S1: acquiring basic map information of a target area, and saving a first correspondence between the location information of the target area and the basic map information; wherein the basic map information includes road data;

[0008] S2: Obtain historical driving behavior data of the target vehicle in the target area, and extract road safety events from the historical driving behavior data according to a preset rule; match the road safety event to the corresponding position information according to the position information of the road safety event and the first corresponding relationship, and obtain a second corresponding relationship between the road safety event and the position information of the target area;

[0009] S3: establishing a danger level measurement model according to the second corresponding relationship, obtaining the risk probability of each location information, and nonlinearly mapping it to the danger level measurement, thereby obtaining a third corresponding relationship between the location information of the target area and the danger level measurement;

[0010] S4: Summarizing the location information of the target area and its corresponding danger level measurement according to the third corresponding relationship, and constructing a priori danger level measurement map of the target area.

[0011] In a possible implementation, step S1 includes:

[0012] Obtaining basic map information of the target area;

[0013] Acquire location information of the target area;

[0014] Matching the basic map information of the target area with the location information of the target area, obtaining and saving a first corresponding relationship between the location information of the target area and the basic map information;

[0015] The road data includes at least one of the following: the type of each road, the number of road lanes, the number of intersections, and the curvature of the road;

[0016] and / or the basic map information also includes historical traffic data;

[0017] The historical traffic data includes at least one of the following: historical road traffic volume, average vehicle speed, congestion index, passenger volume, passenger-to-vehicle ratio, truck ratio, and number of traffic accidents.

[0018] In a possible implementation manner, after obtaining the location information of the target area, the method further includes:

[0019] A clustering algorithm is used to perform clustering processing on the location information of the target area to obtain processed location information of the target area.

[0020] In a possible implementation, step S2 includes:

[0021] Acquire historical driving behavior data of the vehicle in the target area;

[0022] The historical driving behavior data includes the vehicle status information at each time point and the vehicle speed and position information around the vehicle; the vehicle status information at each time point includes at least one of the following: position, speed, acceleration and deceleration, and turning angle;

[0023] Extracting road safety events from the historical driving behavior data according to preset rules, wherein the road safety events include self-vehicle safety events and other-vehicle safety events;

[0024] The vehicle safety event includes at least one of the following: sudden braking, sudden steering, honking, flashing high beams, and manual takeover during automatic driving;

[0025] The other vehicle safety event includes at least one of the following: the distance between the vehicle and the other vehicle is less than a preset value, the acceleration or deceleration of the other vehicle is greater than a preset value, dangerous merging and changing lanes, and collision with the other vehicle;

[0026] assigning a significance value to the road safety event;

[0027] According to the basic map information of the road safety event and the first corresponding relationship, the road safety event is allocated to the corresponding position information according to a preset ratio, and according to the weight of the corresponding position information and the importance value of the road safety event, a second corresponding relationship between the road safety event and the position information of the target area is calculated and saved.

[0028] In a possible implementation, step S3 includes:

[0029] Establishing a danger degree measurement model according to the second corresponding relationship;

[0030] Taking the data points of the second corresponding relationship as the pre-fitted distribution, using maximum likelihood estimation to obtain a probability value of each position information;

[0031] According to the probability value of each position information, the risk probability of each position information encountering danger is obtained;

[0032] The risk probability is nonlinearly mapped to the danger level metric to obtain a third corresponding relationship between the location information of the target area and the danger level metric.

[0033] In addition, to achieve the above-mentioned purpose, the present invention also proposes a priori danger level measurement map constructed as described above, wherein the prior danger level measurement map is a summary of the location information of the target area roads and their corresponding danger level measurements, and the prior danger level measurement map includes a third correspondence between the location information of the target area and the danger level measurement.

[0034] In addition, to achieve the above object, the present invention also proposes a driving method using the a priori danger degree measurement map as described above, the method comprising the following steps:

[0035] A1: Obtain a priori danger level measurement map of the target area, and obtain a priori danger level measurement value of the area of ​​interest of the vehicle according to the priori danger level measurement map of the target area;

[0036] A2: Plan and control the vehicle based on the acquired prior danger level measurement value.

[0037] In a possible implementation, step A1 includes the following steps:

[0038] Obtain the priori danger level measurement value of the vehicle's focus area within a preset time in the future, including:

[0039] Acquire first position information of the vehicle; wherein the first position information includes at least one of the following: current coordinates, future expected coordinates;

[0040] Obtaining, according to the priori danger degree measurement map and the first position information of the ego vehicle, information of a region of interest corresponding to the first position information of the ego vehicle;

[0041] The danger level measurement corresponding to the concerned area information is obtained according to the third corresponding relationship, and the danger level measurements corresponding to all the concerned area information are weighted averaged to obtain a priori danger level measurement value of the concerned area of ​​the vehicle within a preset future time.

[0042] In a possible implementation, step A2 includes the following steps:

[0043] Providing a danger warning for a road section whose prior danger degree measurement value is greater than a first preset threshold; and / or,

[0044] Sort by prior danger level metric value and recommend the route with the lowest prior danger level metric value as the driving route; and / or,

[0045] The driving style of the road section whose prior danger degree metric value is greater than the second preset threshold is recommended as conservative; and the driving style of the road section whose prior danger degree metric value is not greater than the second preset threshold is recommended as aggressive.

[0046] In addition, to achieve the above-mentioned objectives, the present invention also proposes a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, it implements the steps of the method for constructing a priori hazard degree measurement map as described above or implements the steps of the driving method using a priori hazard degree measurement map as described above.

[0047] The technical solution of the present invention obtains the basic map information of the target area and saves the first correspondence between the location information of the target area and the basic map information; wherein the basic map information includes road data; obtains the historical driving behavior data of the target vehicle in the target area, and extracts road safety events from the historical driving behavior data according to preset rules; matches the road safety event to the corresponding location information according to the location information of the road safety event and the first correspondence, and obtains the second correspondence between the road safety event and the location information of the target area; establishes a danger degree measurement model according to the second correspondence, obtains the risk probability of each location information, nonlinearly maps it to the danger degree measurement, and obtains the third correspondence between the location information of the target area and the danger degree measurement; summarizes the location information of the target area and its corresponding danger degree measurement according to the third correspondence, and constructs a priori danger degree measurement map of the target area. In this way, the priori danger degree measurement map of the target area can be obtained, and the quality of the automatic driving system and the safety of the road section can be estimated more finely for a specific road.

[0048] The technical solution of the present invention, after constructing the prior danger degree measurement map of the target area, can obtain the prior danger degree measurement map of the target area, and obtain the prior danger degree measurement value of the area of ​​interest of the vehicle according to the prior danger degree measurement map of the target area; and plan and control the vehicle based on the obtained prior danger degree measurement value. In this way, after obtaining the prior danger degree measurement map of the target area, the coordinates of the prior danger degree measurement map can be directly queried to obtain the danger degree when it is needed, so that the quality of the automatic driving system and the safety of the road section can be estimated more finely for a specific road. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 A schematic flow chart of a method for constructing a priori risk degree measurement diagram provided by the present invention.

[0050] Figure 2 A schematic flow chart of step S1 in a method for constructing a priori risk degree measurement diagram provided by the present invention.

[0051] Figure 3 A schematic flow chart of step S2 in a method for constructing a priori risk degree measurement diagram provided by the present invention.

[0052] Figure 4 A schematic flow chart of step S3 in a method for constructing a priori risk degree measurement diagram provided by the present invention.

[0053] Figure 5 A schematic flow chart of a driving method based on a priori danger degree measurement diagram provided by the present invention.

[0054] Figure 6 The figure is a flow chart of step A1 in a driving method based on a priori danger degree measurement map provided by the present invention.

[0055] Figure 7 It is a schematic diagram of the structure of a transportation tool of the hardware operating environment involved in the embodiment of the present invention.

[0056] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0057] In order to make the technical problems, technical solutions and beneficial effects to be solved by the present invention clearer and more understandable, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only for explaining the present invention and are not intended to limit the present invention.

[0058] In the subsequent description, the suffixes such as "module", "component" or "unit" used to represent elements are only used to facilitate the description of the present invention, and have no specific meanings. Therefore, "module", "component" or "unit" can be used in a mixed manner.

[0059] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0060] Please refer to Figure 1 As shown, the present invention provides a method for constructing a priori danger degree measurement diagram, the method comprising:

[0061] Step S1: Acquire basic map information of a target area, and save a first correspondence between the location information of the target area and the basic map information; wherein the basic map information includes road data.

[0062] Step S2: Obtain historical driving behavior data of the target vehicle in the target area, and extract road safety events from the historical driving behavior data according to preset rules; match the road safety events to the corresponding position information according to the position information of the road safety events and the first corresponding relationship, and obtain a second corresponding relationship between the road safety events and the position information of the target area.

[0063] Step S3: Establish a danger level measurement model based on the second corresponding relationship, obtain the risk probability of each location information, and nonlinearly map it to the danger level measurement to obtain a third corresponding relationship between the location information of the target area and the danger level measurement.

[0064] Step S4: Summarize the location information of the target area and its corresponding danger level measurement according to the third corresponding relationship to construct a priori danger level measurement map of the target area.

[0065] In an embodiment of the present invention, basic map information of a target area is obtained, and a first correspondence between the location information of the target area and the basic map information is saved; wherein the basic map information includes road data; historical driving behavior data of the target vehicle in the target area is obtained, and road safety events are extracted from the historical driving behavior data according to a preset rule; the road safety event is matched to the corresponding location information according to the location information of the road safety event and the first correspondence, and a second correspondence between the road safety event and the location information of the target area is obtained; a danger degree measurement model is established according to the second correspondence, and the risk probability of each location information is obtained, and nonlinearly mapped to the danger degree measurement, and a third correspondence between the location information of the target area and the danger degree measurement is obtained; the location information of the target area and its corresponding danger degree measurement are summarized according to the third correspondence, and a priori danger degree measurement map of the target area is constructed. In this way, a priori danger degree measurement map of the target area can be obtained, and the quality of the automatic driving system and the safety of the road section can be estimated more finely for a specific road, and a basis is provided for the subsequent use of the priori danger degree measurement map.

[0066] It should be noted that the definition of the danger level measurement in this implementation is: only the danger level measurement based on geographical location is considered, that is, the real-time interaction of the scene is not considered. In addition, in order to simplify and compress the size of the prior danger level measurement map, this embodiment only draws the area where the autonomous driving can drive: according to different autonomous driving tasks, generally only the hardened road surface existing in the map is included.

[0067] In this embodiment, absolute safety is set to the danger level measurement = 0: that is, there are no other dynamic participants on the road, and there are no static obstacles that may cause collisions. Of course, this situation is obviously impossible in reality, because the road boundary itself is one of the obstacles, and absolute safety 0 exists as a theoretical value. The danger level measurement of absolute danger is set to 1: that is, an accident is bound to occur when driving to a certain point.

[0068] As a specific embodiment, step S1 is for the collection and processing of basic map information, that is, in order to make the calculation of the danger level more accurate, some basic information needs to be collected first.

[0069] Specifically, basic map information of the target area is obtained, wherein the basic map information at least includes road data. Then, for the location information in the target area, the basic map information is matched to form a first correspondence between the location information of the target area and the basic map information.

[0070] That is to say, eventually, step S1 will obtain a first matrix whose length is the number of position information and whose width is the dimension of basic map information.

[0071] In this embodiment, the specific content of step S1 can be further referred to in 2. Figure 2 As shown, the step S1 includes:

[0072] Step 101: Obtain basic map information of the target area.

[0073] In this embodiment, data corresponding to the dimensions required by the danger level measurement theme model are screened from basic map information containing multiple data sources, wherein the basic map information includes road data, and the data source is very rich, based on which the danger level calculation is more accurate.

[0074] During implementation, big data technologies can be used, such as Hadoop clusters to process massive amounts of data and implement data processing processes such as data mining, aggregation, and crawling.

[0075] Road data related to roads is the most important component of basic map information, including but not limited to: the type of each road, the number of road lanes, the number of intersections, road curvature, etc.

[0076] Of course, the basic map information may also include historical traffic data, and big data technology may also be used, such as Hadoop cluster to process massive data to achieve data mining, aggregation, crawling and other data processing processes.

[0077] The historical traffic data includes but is not limited to: historical road traffic volume, average vehicle speed, congestion index, passenger flow, passenger-to-vehicle ratio, truck ratio, number of traffic accidents, etc.

[0078] Step 102: Acquire the location information of the target area.

[0079] In this embodiment, after a target area is defined, each location of interest in the target area is taken as a point of interest, and its point of interest coordinates, i.e., location information, are obtained. It is understandable that the target area may be a city map of a certain area, such as City A. It is understandable that each location of interest may be spaced a certain distance apart, for example, each point may be 10 meters apart.

[0080] Step 103: Match the basic map information of the target area with the location information of the target area to obtain and save a first corresponding relationship between the location information of the target area and the basic map information.

[0081] In this embodiment, after obtaining the basic map information and location information in the target area, for each location of interest, i.e., location information, the basic map information of the target area and the location information of the target area can be matched to obtain and save the first correspondence between the location information of the target area and the basic map information. For example, a vector composed of discrete values ​​(classification: such as narrow roads, side roads, and highways) or continuous values ​​(such as average daily traffic volume) can be given to all basic map information.

[0082] That is to say, finally, based on the steps 101 to 103, a first matrix with a length equal to the number of position information and a width equal to the dimension of basic map information will be obtained.

[0083] Of course, as an optional embodiment, in subsequent use, the location information of the target area obtained in step S1 may be further processed.

[0084] Specifically, a clustering algorithm may be used to perform clustering processing on the location information of the target area to obtain processed location information of the target area.

[0085] For example, a clustering algorithm can be used to cluster the coordinates of each point of interest into M classes, where the size of M can be set according to the amount of data, the verification period can be set to 50-100, and then it can be set to 100-1000.

[0086] Of course, considering the need for continuity, the spatial correlation of coordinates needs to be considered when clustering. In actual use, a complete road, such as from intersection 1 to intersection 2, or the entire intersection 3, is generally required to have uniformly classified coordinates.

[0087] In the early stage of verification, a simple decision tree algorithm can be used for clustering: for example, class B is defined as a highway section with 4 or more lanes and no on / off ramps within 50 meters.

[0088] It should be noted that the location information of the target area before processing can be used later, or the location information of the target area after processing can be used, and this embodiment does not limit this. It can be understood that the difference between the two is only the amount of data calculated. The amount of data used for the location information of the target area before processing is greater than that used for the location information of the target area after processing, and the processing efficiency is lower than that used for the location information of the target area after processing. However, both can achieve the technical effect of the present invention.

[0089] As a specific embodiment, step S2 is for collecting and processing historical driving behavior data.

[0090] Specifically, historical driving behavior data of the target vehicle is obtained, and road safety events are screened out from the historical driving behavior data according to preset rules; then the road safety events are processed to obtain a second corresponding relationship between the road safety events and the location information of the target area.

[0091] That is to say, in the end, step S3 will obtain a second matrix whose length is the number of location information and whose width is the type of road safety incidents; wherein the value of the second matrix is ​​the weighted number of incidents, that is, the number of road safety incidents within a unit space.

[0092] In this embodiment, the specific content of step S2 can be further referred to in 3. Figure 3 As shown, step S2 includes:

[0093] Step 201: Acquire historical driving behavior data of the vehicle in the target area.

[0094] In this embodiment, a sufficient amount of historical driving behavior data needs to be sorted out, wherein the source of these historical driving behavior data can be based on autonomous driving or manual driving, and this embodiment does not impose any limitation on this.

[0095] It can be understood that what is obtained here is the global historical driving behavior data, and the "target vehicle" mentioned here can be any vehicle.

[0096] The historical driving behavior data considered include but are not limited to: the vehicle status information at each time point and the speed and position information of the vehicles around the vehicle; the vehicle status information at each time point includes at least one of the following: position, speed, acceleration and deceleration, and turning angle.

[0097] Step 202: extracting road safety events from the historical driving behavior data according to preset rules, wherein the road safety events include own-vehicle safety events and other-vehicle safety events.

[0098] In this embodiment, events related to road safety can be captured from the historical driving behavior data by setting preset rule conditions, and the road safety events include self-vehicle safety events and other-vehicle safety events.

[0099] The vehicle safety event includes at least one of the following: sudden braking, sudden turning, honking the horn, flashing high beams, and manual takeover during automatic driving.

[0100] For example, typical vehicle safety events include: an emergency brake (e.g., deceleration greater than 1m / s 2 ); a sudden turn (e.g. angular velocity greater than 1 rad / s); a horn sound; a high beam flash; a takeover during the autonomous driving process (i.e. the driver believes that the autonomous driving cannot currently handle the current scenario and there may be safety issues); an important takeover during the autonomous driving process (i.e. after simulation, it is confirmed that safety issues such as collision will occur if the takeover is not carried out).

[0101] The other vehicle safety event includes at least one of the following: the distance between the vehicle and the other vehicle is less than a preset value, the acceleration or deceleration of the other vehicle is greater than a preset value, a dangerous merging lane change, and a collision with the other vehicle.

[0102] For example, other-vehicle-related events include: the distance between the vehicle and other vehicles is too close (e.g., less than 1 meter); the acceleration or deceleration of the other vehicle is too large; a dangerous merging lane change; a collision between other vehicles, etc.

[0103] Step 203: assigning an importance value to the road safety event.

[0104] In this embodiment, each road safety event includes an importance value, and an importance value can be assigned to the road safety event, for example: a deceleration of -8m / s 2 The sudden brake will decelerate by -2m / s 2 The importance of emergency braking is greater.

[0105] Step 204: Allocate the road safety event to the corresponding position information according to a preset ratio based on the basic map information of the road safety event and the first corresponding relationship, and calculate and save the second corresponding relationship between the road safety event and the position information of the target area based on the weight of the corresponding position information and the importance value of the road safety event.

[0106] In this embodiment, each road safety event can be proportionally distributed to the nearest K coordinates. For example, when K=4, the coordinates of a road safety event are within a square surrounded by its four nearest points. Then, according to the distance, part of the weight of the event multiplied by the importance value is added to these four points. The weight distribution according to the distance is based on the principle of proportional distribution. The logic of "the closer the distance, the greater the weight" can be followed. For example, the reciprocal of the distance can be used as the weight.

[0107] That is to say, finally, based on the steps 301 to 304, a second matrix with a length equal to the number of location information and a width equal to the types of road safety events will be obtained; wherein the value of the second matrix is ​​the weighted number of events.

[0108] As a specific embodiment, step S3 is for modeling and calculating the hazard level measurement.

[0109] Specifically, a danger level measurement model may be established based on the second corresponding relationship, and the risk probability of each location information may be calculated, and then nonlinearly mapped to the danger level measurement to obtain a third corresponding relationship between the location information of the target area and the danger level measurement.

[0110] That is to say, eventually, step S3 will obtain a second vector with a length equal to the number of position information; wherein the value of the second vector is a measure of the degree of danger of the position information.

[0111] That is, the second corresponding relationship is (location information, road safety incident), which is converted into the third corresponding relationship (location information, danger level measurement) according to the model.

[0112] In this embodiment, the specific content of step S3 can be further referred to in FIG. Figure 4 As shown, step S4 includes:

[0113] Step 301: Establish a risk level measurement model according to the second corresponding relationship.

[0114] Step 302: taking the data points of the second corresponding relationship as the pre-fitted distribution and using maximum likelihood estimation to obtain a probability value of each position information.

[0115] Step 303: According to the probability value of each location information, the risk probability of each location information encountering danger is obtained.

[0116] Step 304: nonlinearly map the risk probability to a hazard level metric to obtain a third corresponding relationship between the location information of the target area and the hazard level metric.

[0117] In this embodiment, we can model the degree of danger as a Latent Dirichlet Allocation (LDA) topic model. Assume that the probability of a vehicle encountering danger in a scene is composed of multiple possible road safety events X_1, X_2, ..., X_k, and each road safety event is determined by some factors. The occurrence of each factor is related to multiple state information. Of course, in this embodiment, the state information is simply divided into two categories: state information related to spatial coordinates and other irrelevant information. Since this embodiment only discusses spatial information, other state information can be discarded.

[0118] Specifically, the data points of the second correspondence obtained in step S2 can be used as a pre-fitted distribution, and maximum likelihood estimation can be used to obtain the probability value of each position information, and thus the risk probability of encountering danger, which is then nonlinearly mapped to a 0-1 danger level measurement to obtain a third correspondence between the position information of the target area and the danger level measurement.

[0119] That is to say, eventually, based on the steps 301 to 304, a second vector having a length equal to the number of position information will be obtained; wherein the value of the second vector is a measure of the degree of danger of the position information.

[0120] It should be noted that the modeling in step S3 can also directly model the probability into a more general energy model or Gaussian mixture model, and replace its energy / parameters with a deep neural network to improve the accuracy of the model. However, it should be noted that more complex neural networks require more data for training. If the amount of data is not enough, the effect of a simple model may not be achieved.

[0121] As a specific embodiment, the step S4 is to construct a priori risk degree measurement diagram.

[0122] Specifically, after the third corresponding relationship is obtained in step S3, the location information of the target area and its corresponding danger level measurement may be summarized according to the third corresponding relationship, thereby constructing a priori danger level measurement map of the target area.

[0123] It can be understood that the priori danger level measurement map constructed in this embodiment can be an offline calculation process, and it is only necessary to perform online calculation when using the priori danger level measurement map.

[0124] In an embodiment of the present invention, basic map information of a target area is obtained, and a first correspondence between the location information of the target area and the basic map information is saved; wherein the basic map information includes road data; historical driving behavior data of the target vehicle in the target area is obtained, and road safety events are extracted from the historical driving behavior data according to a preset rule; the road safety event is matched to the corresponding location information according to the location information of the road safety event and the first correspondence, and a second correspondence between the road safety event and the location information of the target area is obtained; a danger degree measurement model is established according to the second correspondence, and the risk probability of each location information is obtained, and nonlinearly mapped to the danger degree measurement, and a third correspondence between the location information of the target area and the danger degree measurement is obtained; the location information of the target area and its corresponding danger degree measurement are summarized according to the third correspondence, and a priori danger degree measurement map of the target area is constructed. In this way, a priori danger degree measurement map of the target area can be obtained, and the quality of the automatic driving system and the safety of the road section can be estimated more finely for a specific road, and a basis is provided for the subsequent use of the priori danger degree measurement map.

[0125] The present invention also provides a priori danger level measurement map, which is constructed using the above method. The prior danger level measurement map is a summary of the location information of the target area roads and their corresponding danger level measurements. The prior danger level measurement map includes a third correspondence between the location information of the target area and the danger level measurement.

[0126] It is understandable that after the a priori danger degree measure map is constructed, it can be used.

[0127] Please refer to Figure 5 As shown, the present invention provides a driving method based on a priori danger degree measurement map, the method comprising:

[0128] Step A1: Obtain a priori danger level measurement map of the target area, and obtain a priori danger level measurement value of the area of ​​interest of the vehicle according to the priori danger level measurement map of the target area.

[0129] Step A2: Plan and control the vehicle based on the acquired prior danger level measurement value.

[0130] In an embodiment of the present invention, after obtaining a priori danger level measurement map of the target area, the coordinates of the priori danger level measurement map are directly queried to obtain the danger level when needed, which can more finely estimate the quality of the autonomous driving system and the safety level of the road section for a specific road.

[0131] That is, this embodiment is a method for using a priori danger degree measurement map after drawing it, so that a priori danger warning can be performed based on the danger degree priori danger degree measurement map.

[0132] In this embodiment, the specific content of step A1 can be further referred to in 6. Figure 6 As shown, step S5 includes:

[0133] Step 501: Obtain first position information of the vehicle; wherein the first position information includes at least one of the following: current coordinates, and future expected coordinates.

[0134] Step 502: Obtaining, according to the priori danger degree measurement map and the first position information of the vehicle, area of ​​interest information corresponding to the first position information of the vehicle.

[0135] Step 503: Obtaining the focus area information corresponding to the first position information of the vehicle according to the priori danger degree measurement map and the first position information of the vehicle.

[0136] Specifically, when the prior danger level measurement map of the target area is used, the first position information of the vehicle can be obtained, and the first position information includes at least one of the following: current coordinates, expected future coordinates, that is, the current position or the position of a period of time in the future. Then, according to the prior danger level measurement map and the first position information of the vehicle, the area of ​​interest information corresponding to the first position information of the vehicle is obtained. It should be noted that the acquisition of the area of ​​interest information can be to take a distance less than a certain distance, or to forcibly set the area of ​​a certain road or a certain relationship as the same area, and this embodiment does not limit this. Finally, according to the third corresponding relationship, the danger level measurement corresponding to the area of ​​interest information is obtained, and the danger level measurements corresponding to all the area of ​​interest information are weighted averaged to obtain the prior danger level measurement value of the area of ​​interest of the vehicle within a preset time in the future.

[0137] It is understandable that if it is necessary to calculate the danger level of the scene where the vehicle is at a certain moment, a difference algorithm can be used. For the current coordinates (x, y) of the vehicle, find all points in the prior danger level measurement map whose coordinate distance is less than a certain threshold, such as = 10 meters, and then take the weighted average of the danger level measurements of these points to get the final danger level measurement value of the point.

[0138] Among them, the "weighted average" can take some weighting coefficients as 0, thereby forming any method such as taking the middle value, removing the maximum and minimum to calculate the average, taking the maximum value, etc., which is not limited in this embodiment.

[0139] Of course, it is understandable that if the vehicle itself has a plan or prediction of the future direction of movement, the coordinate points in this plan can be used to make an expected danger level measurement, that is, how high the danger level is in the near future (for example, within 30 seconds). If the future planned trajectory has 100 points, then the sum of the danger level measurements of the 100 points can be used directly.

[0140] In this embodiment, the specific content of step A2 is how to plan and control the vehicle based on the acquired priori danger level measurement value.

[0141] It can be understood that there are many ways to plan and control the vehicle based on the prior danger level measurement map obtained in this embodiment, including issuing danger warnings for road sections with a prior danger level measurement value greater than a first preset threshold; and / or, sorting by prior danger level measurement value, and recommending the route with the lowest prior danger level measurement value as the driving route; and / or, recommending the driving style as conservative for road sections with a prior danger level measurement value greater than a second preset threshold; and recommending the driving style as aggressive for road sections with a prior danger level measurement value not greater than the second preset threshold.

[0142] The following are some examples:

[0143] 1. Reminder to human drivers: For dangerous sections of road, the driver will be prompted during navigation that the current section of road is dangerous, please pay attention.

[0144] 2. For the autonomous driving system, a more aggressive driving style is used on roads with a lower a priori risk, with progress as the priority; on roads with a higher a priori risk, a more conservative driving style is used, with safety as the priority, including active defensive driving and reduced speed.

[0145] 3. Give some proactive signal prompts to the autonomous driving system, such as turning on the high beam and honking the horn appropriately on roads with a high a priori risk level, so as to interact with road participants more explicitly.

[0146] 4. For the transportation department, targeted road clearing or lane improvement will be carried out on sections of road with a priori higher risk.

[0147] 5. For navigation software, if multiple routes have similar travel times, they can be sorted by prior danger level to provide a safer route.

[0148] In the embodiment of the present invention, a priori danger degree measurement map of the target area is obtained, and a priori danger degree measurement value of the area of ​​interest of the vehicle is obtained according to the priori danger degree measurement map of the target area. In this way, after obtaining the priori danger degree measurement map of the target area, the coordinates of the priori danger degree measurement map are directly queried to obtain the danger degree when it is needed. A priori danger warning based on the danger degree priori danger degree measurement map can also be used to estimate the quality of the automatic driving system and the safety of the road section in a more detailed manner for a specific road.

[0149] In addition, in some optional embodiments, the location information, such as coordinates, can be two-dimensional coordinates or three-dimensional coordinates. That is, in order to better distinguish the difference in road danger between peak hours in the morning and evening and off-peak hours, the coordinates can be derived into generalized time-space coordinates, namely (x, y, t), where t can be at intervals of 2 hours, with a total of 12 discrete values ​​throughout the day.

[0150] It is understandable that the processing method of three-dimensional coordinates is basically the same as that of two-dimensional coordinates, and this embodiment will not be described in detail here. However, adding the time dimension will lead to an increase in the number of coordinates, and the collection of historical data needs to be further increased, and the collection amount of each region and time must be basically uniform. Although the amount of historical data has increased, the number of coordinates has not decreased. In order to reduce data requirements, an optional solution is: if the amount of historical data is insufficient, a waypoint can be defined as the center line of the lane of a section of road, with a point every 10 meters, and these waypoints are used as coordinates, and all other coordinates are discarded. This type of method can be used to reduce the number of coordinates. Because: the smaller the number of coordinates, the smaller the amount of history required.

[0151] Reference Figure 7 , Figure 7 A schematic diagram of the structure of a vehicle in the hardware operating environment involved in an embodiment of the present invention.

[0152] like Figure 7 As shown, the vehicle may include: a processor 1001, such as a CPU, a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. Among them, the communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display), an input unit such as a keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as WI-FI, 4G, 5G interface). The memory 1005 may be a high-speed RAM memory, or a stable memory (non-volatile memory), such as a disk memory. The memory 1005 may also be a storage device independent of the aforementioned processor 1001.

[0153] Those skilled in the art will understand that Figure 7 The structure shown in the figure does not constitute a limitation on the vehicle, and may include more or less components than shown in the figure, or combine certain components, or arrange the components differently.

[0154] like Figure 7 As shown, the memory 1005 as a computer storage medium may include an operating system, a network communication module, a user interface module, a driving program based on a priori hazard degree measurement map, and a program for constructing a priori hazard degree measurement map.

[0155] exist Figure 7 In the vehicle shown, the network interface 1004 is mainly used for data communication with an external network; the user interface 1003 is mainly used for receiving user input instructions; the vehicle calls the driving program based on the prior danger degree measurement map stored in the memory 1005 through the processor 1001, and performs the following operations:

[0156] A1: Obtain a priori danger level measurement map of the target area, and obtain a priori danger level measurement value of the area of ​​interest of the vehicle according to the priori danger level measurement map of the target area;

[0157] A2: Plan and control the vehicle based on the acquired prior danger level measurement value.

[0158] In a possible implementation, step A1 includes the following steps:

[0159] Obtain the priori danger level measurement value of the vehicle's focus area within a preset time in the future, including:

[0160] Acquire first position information of the vehicle; wherein the first position information includes at least one of the following: current coordinates, future expected coordinates;

[0161] Obtaining, according to the priori danger degree measurement map and the first position information of the ego vehicle, information of a region of interest corresponding to the first position information of the ego vehicle;

[0162] The danger level measurement corresponding to the concerned area information is obtained according to the third corresponding relationship, and the danger level measurements corresponding to all the concerned area information are weighted averaged to obtain a priori danger level measurement value of the concerned area of ​​the vehicle within a preset future time.

[0163] In a possible implementation, step A2 includes the following steps:

[0164] Providing a danger warning for a road section whose prior danger degree measurement value is greater than a first preset threshold; and / or,

[0165] Sort by prior danger level metric value and recommend the route with the lowest prior danger level metric value as the driving route; and / or,

[0166] The driving style of the road section whose prior danger degree metric value is greater than the second preset threshold is recommended as conservative; and the driving style of the road section whose prior danger degree metric value is not greater than the second preset threshold is recommended as aggressive.

[0167] exist Figure 7 In the vehicle shown, the vehicle also calls the program for constructing a priori danger degree metric map stored in the memory 1005 through the processor 1001, and performs the following operations:

[0168] S1: acquiring basic map information of a target area, and saving a first correspondence between the location information of the target area and the basic map information; wherein the basic map information includes road data;

[0169] S2: Obtain historical driving behavior data of the target vehicle in the target area, and extract road safety events from the historical driving behavior data according to a preset rule; match the road safety event to the corresponding position information according to the position information of the road safety event and the first corresponding relationship, and obtain a second corresponding relationship between the road safety event and the position information of the target area;

[0170] S3: establishing a danger level measurement model according to the second corresponding relationship, obtaining the risk probability of each location information, and nonlinearly mapping it to the danger level measurement, thereby obtaining a third corresponding relationship between the location information of the target area and the danger level measurement;

[0171] S4: Summarizing the location information of the target area and its corresponding danger level measurement according to the third corresponding relationship, and constructing a priori danger level measurement map of the target area.

[0172] In a possible implementation, step S1 includes:

[0173] Obtaining basic map information of the target area;

[0174] Acquire location information of the target area;

[0175] Matching the basic map information of the target area with the location information of the target area, obtaining and saving a first corresponding relationship between the location information of the target area and the basic map information;

[0176] The road data includes at least one of the following: the type of each road, the number of road lanes, the number of intersections, and the curvature of the road;

[0177] and / or the basic map information also includes historical traffic data;

[0178] The historical traffic data includes at least one of the following: historical road traffic volume, average vehicle speed, congestion index, passenger volume, passenger-to-vehicle ratio, truck ratio, and number of traffic accidents.

[0179] In a possible implementation manner, after obtaining the location information of the target area, the method further includes:

[0180] A clustering algorithm is used to perform clustering processing on the location information of the target area to obtain processed location information of the target area.

[0181] In a possible implementation, step S2 includes:

[0182] Acquire historical driving behavior data of the vehicle in the target area;

[0183] The historical driving behavior data includes the vehicle status information at each time point and the speed and position information of the vehicle surrounding the vehicle; the vehicle status information at each time point includes at least one of the following: position, speed, acceleration and deceleration, and turning angle;

[0184] Extracting road safety events from the historical driving behavior data according to preset rules, wherein the road safety events include self-vehicle safety events and other-vehicle safety events;

[0185] The vehicle safety event includes at least one of the following: sudden braking, sudden steering, honking, flashing high beams, and manual takeover during automatic driving;

[0186] The other vehicle safety event includes at least one of the following: the distance between the vehicle and the other vehicle is less than a preset value, the acceleration or deceleration of the other vehicle is greater than a preset value, dangerous merging and changing lanes, and collision with the other vehicle;

[0187] assigning a significance value to the road safety event;

[0188] According to the basic map information of the road safety event and the first corresponding relationship, the road safety event is allocated to the corresponding position information according to a preset ratio, and according to the weight of the corresponding position information and the importance value of the road safety event, a second corresponding relationship between the road safety event and the position information of the target area is calculated and saved.

[0189] In a possible implementation, step S3 includes:

[0190] Establishing a danger degree measurement model according to the second corresponding relationship;

[0191] Taking the data points of the second corresponding relationship as the pre-fitted distribution, using maximum likelihood estimation to obtain a probability value of each position information;

[0192] According to the probability value of each position information, the risk probability of each position information encountering danger is obtained;

[0193] The risk probability is nonlinearly mapped to the danger level metric to obtain a third corresponding relationship between the location information of the target area and the danger level metric.

[0194] In the embodiment of the present invention, a priori danger level measurement map can be obtained for the autonomous driving area, and the danger level can be directly queried when needed, so as to estimate the quality of the autonomous driving system and the safety level of the road section in more detail.

[0195] In addition, an embodiment of the present invention further provides a computer-readable storage medium, on which a driving program based on a priori danger degree measurement map and a program for constructing a priori danger degree measurement map are stored, wherein the driving program based on the priori danger degree measurement map is executed by a processor to implement the following operations:

[0196] A1: Obtain a priori danger level measurement map of the target area, and obtain a priori danger level measurement value of the area of ​​interest of the vehicle according to the priori danger level measurement map of the target area;

[0197] A2: Plan and control the vehicle based on the acquired prior danger level measurement value.

[0198] In a possible implementation, step A1 includes the following steps:

[0199] Obtain the priori danger level measurement value of the vehicle's focus area within a preset time in the future, including:

[0200] Acquire first position information of the vehicle; wherein the first position information includes at least one of the following: current coordinates, future expected coordinates;

[0201] Obtaining, according to the priori danger degree measurement map and the first position information of the ego vehicle, information of a region of interest corresponding to the first position information of the ego vehicle;

[0202] The danger level measurement corresponding to the concerned area information is obtained according to the third corresponding relationship, and the danger level measurements corresponding to all the concerned area information are weighted averaged to obtain a priori danger level measurement value of the concerned area of ​​the vehicle within a preset future time.

[0203] In a possible implementation, step A2 includes the following steps:

[0204] Providing a danger warning for a road section whose prior danger degree measurement value is greater than a first preset threshold; and / or,

[0205] Sort by prior danger level metric value and recommend the route with the lowest prior danger level metric value as the driving route; and / or,

[0206] The driving style of the road section whose prior danger degree metric value is greater than the second preset threshold is recommended as conservative; and the driving style of the road section whose prior danger degree metric value is not greater than the second preset threshold is recommended as aggressive.

[0207] Among them, the program for constructing a priori danger degree measurement diagram implements the following operations when the program is executed by the processor:

[0208] S1: acquiring basic map information of a target area, and saving a first correspondence between the location information of the target area and the basic map information; wherein the basic map information includes road data;

[0209] S2: Obtain historical driving behavior data of the target vehicle in the target area, and extract road safety events from the historical driving behavior data according to a preset rule; match the road safety event to the corresponding position information according to the position information of the road safety event and the first corresponding relationship, and obtain a second corresponding relationship between the road safety event and the position information of the target area;

[0210] S3: establishing a danger level measurement model according to the second corresponding relationship, obtaining the risk probability of each location information, and nonlinearly mapping it to the danger level measurement, thereby obtaining a third corresponding relationship between the location information of the target area and the danger level measurement;

[0211] S4: Summarizing the location information of the target area and its corresponding danger level measurement according to the third corresponding relationship, and constructing a priori danger level measurement map of the target area.

[0212] In a possible implementation, step S1 includes:

[0213] Obtaining basic map information of the target area;

[0214] Acquire location information of the target area;

[0215] Matching the basic map information of the target area with the location information of the target area, obtaining and saving a first corresponding relationship between the location information of the target area and the basic map information;

[0216] The road data includes at least one of the following: the type of each road, the number of road lanes, the number of intersections, and the curvature of the road;

[0217] and / or the basic map information also includes historical traffic data;

[0218] The historical traffic data includes at least one of the following: historical road traffic volume, average vehicle speed, congestion index, passenger volume, passenger-to-vehicle ratio, truck ratio, and number of traffic accidents.

[0219] In a possible implementation manner, after obtaining the location information of the target area, the method further includes:

[0220] A clustering algorithm is used to perform clustering processing on the location information of the target area to obtain processed location information of the target area.

[0221] In a possible implementation, step S2 includes:

[0222] Acquire historical driving behavior data of the vehicle in the target area;

[0223] The historical driving behavior data includes the vehicle status information at each time point and the speed and position information of the vehicle surrounding the vehicle; the vehicle status information at each time point includes at least one of the following: position, speed, acceleration and deceleration, and turning angle;

[0224] Extracting road safety events from the historical driving behavior data according to preset rules, wherein the road safety events include self-vehicle safety events and other-vehicle safety events;

[0225] The vehicle safety event includes at least one of the following: sudden braking, sudden steering, honking, flashing high beams, and manual takeover during automatic driving;

[0226] The other vehicle safety event includes at least one of the following: the distance between the vehicle and the other vehicle is less than a preset value, the acceleration or deceleration of the other vehicle is greater than a preset value, dangerous merging and changing lanes, and collision with the other vehicle;

[0227] assigning a significance value to the road safety event;

[0228] According to the basic map information of the road safety event and the first corresponding relationship, the road safety event is allocated to the corresponding position information according to a preset ratio, and according to the weight of the corresponding position information and the importance value of the road safety event, a second corresponding relationship between the road safety event and the position information of the target area is calculated and saved.

[0229] In a possible implementation, step S3 includes:

[0230] Establishing a danger degree measurement model according to the second corresponding relationship;

[0231] Taking the data points of the second corresponding relationship as the pre-fitted distribution, using maximum likelihood estimation to obtain a probability value of each position information;

[0232] According to the probability value of each position information, the risk probability of each position information encountering danger is obtained;

[0233] The risk probability is nonlinearly mapped to the danger level metric to obtain a third corresponding relationship between the location information of the target area and the danger level metric.

[0234] In the embodiment of the present invention, a priori danger level measurement map can be obtained for the autonomous driving area, and the danger level can be directly queried when needed, so as to estimate the quality of the autonomous driving system and the safety level of the road section in more detail.

[0235] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or system including the element.

[0236] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0237] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes a number of instructions for a terminal device (which can be a mobile phone, computer, server, controller, or network device, etc.) to execute the methods described in each embodiment of the present invention.

[0238] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for constructing a priori danger degree measurement diagram, characterized in that: The method comprises the following steps: S1: acquiring basic map information of a target area, and saving a first correspondence between the location information of the target area and the basic map information; wherein the basic map information includes road data; S2: Obtain historical driving behavior data of the target vehicle in the target area, and extract road safety events from the historical driving behavior data according to a preset rule; match the road safety event to the corresponding position information according to the position information of the road safety event and the first corresponding relationship, and obtain a second corresponding relationship between the road safety event and the position information of the target area; S3: establishing a danger level measurement model according to the second corresponding relationship, obtaining the risk probability of each location information, and nonlinearly mapping it to the danger level measurement, thereby obtaining a third corresponding relationship between the location information of the target area and the danger level measurement; S4: Summarizing the location information of the target area and its corresponding danger level measurement according to the third corresponding relationship, and constructing a priori danger level measurement map of the target area.

2. The method according to claim 1, characterized in that The step S1 comprises: Obtaining basic map information of the target area; Acquiring location information of the target area; Matching the basic map information of the target area with the location information of the target area, obtaining and saving a first corresponding relationship between the location information of the target area and the basic map information; The road data includes at least one of the following: the type of each road, the number of road lanes, the number of intersections, and the curvature of the road; and / or the basic map information also includes historical traffic data; The historical traffic data includes at least one of the following: historical road traffic volume, average vehicle speed, congestion index, passenger volume, passenger-to-vehicle ratio, truck ratio, and number of traffic accidents.

3. The method according to claim 2, characterized in that After obtaining the location information of the target area, the method further includes: A clustering algorithm is used to perform clustering processing on the location information of the target area to obtain processed location information of the target area.

4. The method according to claim 1 or 3, characterized in that: The step S2 comprises: Acquire historical driving behavior data of the vehicle in the target area; The historical driving behavior data includes the vehicle status information at each time point and the speed and position information of the vehicle surrounding the vehicle; the vehicle status information at each time point includes at least one of the following: position, speed, acceleration and deceleration, and turning angle; Extracting road safety events from the historical driving behavior data according to preset rules, wherein the road safety events include self-vehicle safety events and other-vehicle safety events; The vehicle safety event includes at least one of the following: sudden braking, sudden steering, honking, flashing high beams, and manual takeover during automatic driving; The other vehicle safety event includes at least one of the following: the distance between the vehicle and the other vehicle is less than a preset value, the acceleration or deceleration of the other vehicle is greater than a preset value, dangerous merging and changing lanes, and collision with the other vehicle; assigning a significance value to the road safety event; According to the basic map information of the road safety event and the first corresponding relationship, the road safety event is allocated to the corresponding position information according to a preset ratio, and according to the weight of the corresponding position information and the importance value of the road safety event, a second corresponding relationship between the road safety event and the position information of the target area is calculated and saved.

5. The method according to claim 1 or 3, characterized in that: The step S3 comprises: Establishing a danger degree measurement model according to the second corresponding relationship; Taking the data points of the second corresponding relationship as the pre-fitted distribution, using maximum likelihood estimation to obtain a probability value of each position information; According to the probability value of each position information, the risk probability of each position information encountering danger is obtained; The risk probability is nonlinearly mapped to the danger level metric to obtain a third corresponding relationship between the location information of the target area and the danger level metric.

6. A priori risk degree measurement map constructed using the method according to any one of claims 1 to 5, characterized in that: The a priori danger level measurement map is a summary of the location information of the target area roads and the corresponding danger level measurements, and the a priori danger level measurement map includes a third corresponding relationship between the location information of the target area and the danger level measurement.

7. A driving method using the priori danger degree measurement map according to claim 6, characterized in that: The method comprises the following steps: A1: Obtain a priori danger level measurement map of the target area, and obtain a priori danger level measurement value of the area of ​​interest of the vehicle according to the priori danger level measurement map of the target area; A2: Plan and control the vehicle based on the acquired prior danger level measurement value.

8. The method according to claim 7, characterized in that The step A1 comprises the following steps: Obtain the priori danger level measurement value of the vehicle's focus area within a preset time in the future, including: Acquire first position information of the vehicle; wherein the first position information includes at least one of the following: current coordinates, future expected coordinates; Obtaining, according to the priori danger degree measurement map and the first position information of the ego vehicle, information of a region of interest corresponding to the first position information of the ego vehicle; The danger level measurement corresponding to the concerned area information is obtained according to the third corresponding relationship, and the danger level measurements corresponding to all the concerned area information are weighted averaged to obtain a priori danger level measurement value of the concerned area of ​​the vehicle within a preset future time.

9. The method according to claim 7, characterized in that: The step A2 comprises the following steps: Providing a danger warning for a road section whose prior danger degree measurement value is greater than a first preset threshold; and / or, Sort by prior danger level metric value and recommend the route with the lowest prior danger level metric value as the driving route; and / or, The driving style of the road section whose prior danger degree metric value is greater than the second preset threshold is recommended as conservative; and the driving style of the road section whose prior danger degree metric value is not greater than the second preset threshold is recommended as aggressive.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the computer program implements the steps of the method according to any one of claims 7 to 9 or the steps of the method according to any one of claims 1 to 5.

Citation Information

Patent Citations

  • Recommendation speed determination method, device and equipment

    CN116224993A

  • Driving support device

    JP2012008675A

  • Systems and methods for safe route determination

    WO2019232022A1