KR20230063390A - The safty driving evaluation method for autonomous vehicles supporting mobility for the transportation vulnerable - Google Patents

Abstract

The present invention relates to a safe driving evaluation method for an autonomous vehicle supporting mobility for the transportation vulnerable.
In the present invention, the safe driving evaluation method for autonomous vehicles supporting mobility for the disabled generates safety indicators including risky driving behavior evaluation indicators reflecting road geometry and risk evaluation indicators around autonomous vehicles based on sensor information. In addition, it can notify whether autonomous vehicles are driving dangerously and provide safe driving analysis information.

Description

Safe driving evaluation method for autonomous vehicles supporting mobility for the transportation vulnerable

The present invention relates to an autonomous vehicle system. In particular, it relates to mobility support services for the transportation vulnerable such as the disabled or laborers.

Various researches and developments are being conducted for autonomous vehicles. Technologies related to self-driving vehicles are divided into 6 levels of technology development, from level 0 to level 5, depending on the degree of self-driving. Level 4 and Level 5 are classified as Highly Autonomous Vehicles (HAVs) that can operate without a driver intervening in driving. Prior Patent Document 1 discloses a control method for an autonomous vehicle. When an autonomous vehicle operates solely dependent on information received from an external server, it cannot actively cope with various unexpected situations that may occur in the driving environment. Prior Patent Document 1 proposes a method that can be dealt with. Whether the self-driving vehicle actively responds to situations or the server controls the motion control of the self-driving vehicle, technologies related to autonomous driving are mostly focused on research on passenger vehicles and vans. However, public buses can also operate autonomously, and in this case, safety and reliability inevitably become more important. Nevertheless, according to the prior art, there is no optimized solution related to such self-driving public buses. Thus, the present inventors have completed the inventions of Prior Patent Document 2 and Prior Patent Document 3.

However, even in the field of autonomous driving, mobility support for the transportation vulnerable will not be the same as for ordinary people. The inventors of the present invention completed the present invention after long research on safe mobility support for the transportation vulnerable, which can comprehensively reflect the driving situation of autonomous vehicles while considering the structure of the road.

Prior Patent Document 1: Republic of Korea Patent Publication No. 10-2019-0078105

Prior Patent Document 2: Korean Patent Publication No. 10-2141988

Prior Patent Document 3: Korean Patent Publication No. 10-2204199

The present invention provides a method for effectively monitoring and evaluating the safety of an autonomous vehicle supporting mobility for the disabled. Since the transportation vulnerable has less responsiveness to unexpected situations, autonomous vehicles transporting the transportation vulnerable require a more sophisticated solution that goes beyond the usual risk/safety assessment. The present invention proposes such a solution unknown to the world.

Meanwhile, other unspecified objects of the present invention will be additionally considered within the scope that can be easily inferred from the following detailed description and effects thereof.

In order to achieve the above object, the safety driving evaluation method of an autonomous vehicle supporting mobility for the disabled of the present invention has a service server:

(a) While communicating with N (N is an integer of 1 or more) autonomous vehicles through a wireless network, sensor information of autonomous vehicles is collected in real time, sensor-based risk assessment data is calculated, and sensor-based evaluation indicators are created;

(b) For M (M is an integer greater than 1) cells defined as independent zones on the driving road and assigned IDs, risky driving behaviors based on the road geometry are classified in advance into a plurality of grades, and stored in the system of the service server. In a registered situation, when an autonomous vehicle enters a specific cell, the risky driving behavior is calculated according to the risky driving behavior grade based on the road geometry of the cell to generate an evaluation index for the risky driving behavior;

(c) transmitting the sensor-based evaluation index and the risky driving behavior evaluation index to any one or more of an autonomous vehicle, a device of a system monitoring the autonomous vehicle, and a mobile device collecting information about the autonomous vehicle. It is characterized by including.

In the safe driving evaluation method of an autonomous vehicle supporting mobility for the disabled according to a preferred embodiment of the present invention, the risky driving behavior is any one or more of a rapid acceleration type, a rapid deceleration type, a sharp turn type, and a rapid lane change type. It is good to correspond to

The present invention can further improve driving safety of autonomous vehicles. In particular, it can contribute to the technology development of popularized self-driving vehicles by promoting the use and convenience of self-driving vehicles that transport the transportation vulnerable.

On the other hand, even if the effects are not explicitly mentioned here, it is added that the effects described in the following specification expected by the technical features of the present invention and their provisional effects are treated as described in the specification of the present invention.

1 schematically shows an example of a system configuration according to a preferred embodiment of the present invention.
2 shows the configuration of the service server 100 according to a preferred embodiment of the present invention in more detail.
3 schematically shows a system logic configuration according to a preferred embodiment of the present invention.
4 schematically illustrates a process of generating risky driving behavior evaluation indexes based on a road geometry according to a preferred embodiment of the present invention.
5 is a diagram conceptually illustrating the determination of risky driving behavior by road geometry through the relationship between cells and risk groups.
※ It is revealed that the accompanying drawings are illustrated as references for understanding the technical idea of the present invention, and thereby the scope of the present invention is not limited.

Hereinafter, with reference to the drawings, look at the configuration of the present invention guided by various embodiments of the present invention and the effects resulting from the configuration. In the description of the present invention, if it is determined that a related known function may unnecessarily obscure the subject matter of the present invention as an obvious matter to those skilled in the art, the detailed description thereof will be omitted.

1 schematically shows an example of a system configuration according to a preferred embodiment of the present invention.

The service server 100 is the subject of the method of evaluating the driving safety while monitoring the driving of the autonomous vehicle and generating and notifying the evaluation result value. The service server 100 includes one or more server devices and is composed of a system including one or more hardware and software devices. A detailed configuration thereof will be described later.

In the network system of the present invention, N autonomous vehicles 10 (N is an integer of 1 or more) communicate with the service server 100. Preferably, the self-driving vehicle 10 is an electric vehicle. The self-driving vehicle 10 is provided with a wireless communication modem capable of exchanging data with the service server 100 through a wireless communication network, a control device capable of controlling vehicle operation, and a display device. In the best mode of the present invention, the self-driving vehicle 10 performs mobility support for the disabled. The transportation vulnerable refers to passengers who are physically unable to respond promptly in an emergency or dangerous situation. For example, it may include the disabled, the elderly, children, pregnant women, and the like. When these transportation vulnerable people get on the autonomous vehicle 10, the danger to them will be much greater than that of ordinary people. The present invention configures a system to effectively prevent and control it. However, only in the case of this best mode, the technical scope described in the claims of the present invention is not limited and interpreted.

The service server 100 receives sensor information from the autonomous vehicle 10 while communicating with the autonomous vehicle 10 . This sensor information is information that can be obtained by Vision, Lidar, Radar, and GPS sensors, and specifically includes object recognition/determination, vehicle location, vehicle speed and direction angle, and longitudinal and lateral acceleration information.

In addition, the safety indicator generator 110 of the service server 100 may calculate a time required for a collision with a surrounding object according to the current speed standard of the self-driving vehicle 10 . In addition, through data processing described below, autonomous vehicle safety evaluation may be performed and safety indicators may be generated.

The service server 100 of the present invention builds one or more databases 180 and 190. The database 180 refines operation and sensor information of the autonomous vehicle 10 and manages an integrated database. The database 190 may store data related to driving roads on which autonomous vehicles drive. In addition, it is added that a database can be added and operated as needed.

The safety index generated by the safety indicator generator 110 is a safety evaluation index for each autonomous vehicle 10 in operation, and is notified to the management server 200 that manages the autonomous vehicle.

The management server 200 that manages the self-driving vehicle is composed of one or more hardware/software devices, and is delivered to pre-registered sites such as a dashboard, a manager operating terminal, a web homepage, and a web mobile.

2 shows the configuration of the service server 100 according to a preferred embodiment of the present invention in more detail.

The safety indicator generator 110 includes a road geometry-based dangerous driving behavior evaluation module 111, a sensor information-based vehicle safety evaluation evaluation module 112, an autonomous vehicle safety index generation module 115, and an information provision management module 116 , a warning notification management module 117, a statistical analysis management module 118, and a process management module 119.

The road geometry-based dangerous driving behavior evaluation module 111 is preferably composed of modules that execute a 4-step process for road geometry-based evaluation. There are modules constituting road geometry cells, a module that classifies dangerous driving behavior risk groups into 5 stages, a module that determines risky driving behavior occurrence cells due to road geometry, and a module that calculates risky driving behavior based on road geometry. , which will be described in detail again in this specification.

As described above, the sensor information-based vehicle safety evaluation evaluation module 112 analyzes the object recognition/determination received from the sensor information of the self-driving vehicle supporting mobility for the disabled, the location/speed/directional angle of the vehicle, and the longitudinal and lateral acceleration information of the vehicle. Therefore, the safety evaluation of the autonomous vehicle is performed by calculating the time required for collision with surrounding objects according to the current speed of the autonomous vehicle.

The self-driving vehicle safety index generation module 115 generates a safety index of the autonomous vehicle by using the road geometry-based dangerous driving behavior evaluation index and the sensor information-based vehicle safety evaluation index.

The information provision management module 116 provides statistical analysis of the safety level of the self-driving vehicle supporting mobility for the disabled, including road geometry-based risky driving behavior evaluation index and sensor information-based safety evaluation index, to the dashboard, autonomous vehicle display, and web mobile means. can be provided through

The warning notification management module 117 notifies the self-driving vehicle of safety evaluation information including the above-described risk driving behavior evaluation information based on the road geometry and safety risk evaluation around the self-driving vehicle.

The statistical analysis management module 118 performs statistical analysis on the safety of the autonomous vehicle by utilizing road geometry-based risky driving behavior evaluation information and sensor information-based safety evaluation information.

The process management module 119 manages the state of a process constituting a safety evaluation system based on autonomous vehicle risk driving behavior evaluation and sensor information.

The data storage 180 stores and manages structured, unstructured, and big data collected from autonomous vehicles using a big data system.

Distributed messaging system 120 facilitates real-time processing of large-volume data for evaluation of risky driving behavior of autonomous vehicles and safety evaluation of autonomous vehicles, and applies a large-capacity distributed messaging system to increase performance through this. .

The V2X server 130 collects operation and sensor information of the autonomous vehicle in real time through the LTE communication method, and notifies the autonomous vehicle of risky driving behavior evaluation indicators and safety indicators reflecting the road geometry of the current autonomous vehicle. . Preferably, SAE J2735, which is a standard protocol for autonomous vehicle and V2X communication, can be applied. In addition, data received from the autonomous vehicle is decoded and transmitted to the distributed messaging system 120 . And it evaluates risky driving behavior and delivers safety indicators to autonomous vehicles.

3 schematically illustrates a system logic configuration according to a preferred embodiment of the present invention.

The system of the present invention is composed of four layers. Layer 1 is related to data collection, layer 2 is related to data management, layer 3 is related to the creation of safety indicators, and layer 4 is related to information provision.

Layer 1 receives data from the vision, lidar, radar, and GPS sensors of the autonomous vehicle, and collects data such as object recognition/judgment, vehicle location, vehicle speed, vehicle direction angle, and longitudinal and lateral acceleration information.

Layer 2 is a data management layer that refines collected data to refine autonomous vehicle operation and sensor information and manages an integrated database.

Layer 3 creates risky driving behavior evaluation indicators that reflect the road geometry and safety indicators that generate risk indicators around autonomous vehicles based on sensor information.

In layer 4, the safety indicators generated in layer 3 are provided to the information providing medium. Such an information providing medium includes a situation board, a display of an autonomous vehicle, a web mobile, and the like.

As reviewed in FIGS. 2 and 3 , the service server 100 of the present invention provides safety indicators generated by the self-driving vehicle 10 and/or the management server 200 . But these are two safety indicators. First, it is a risk driving behavior evaluation index based on the road geometry, and second, it is safe driving evaluation information related to the surroundings of the vehicle based on sensor information collected in real time from the autonomous vehicle. Let's look at the former first.

4 schematically shows a process of generating risky driving behavior evaluation indexes based on road geometry according to a preferred embodiment of the present invention.

First, a road geometric structure cell is configured (S10). These cells, which are defined as zones independent of each other in the driving road map, are composed of M (M is an integer greater than 1) grid. In addition, each cell is registered and managed by giving an ID. The cell used for calculating the speeding index of the eTAS system can be used, and the service server can configure a grid cell separately.

Next, risk groups for risky driving behavior are classified in order to identify road geometrical structural characteristics of the cell (S20). To this end, we first analyzed the driver's risky safety behavior, not the self-driving vehicle. The number of risky driving behaviors for each trip was calculated using 8,660 bus trip data registered in Seoul as the target data for analysis. After calculating the average and standard deviation of the number of risky driving behaviors in all trips and identifying the distribution of risky driving behaviors for each trip through a standard normal distribution table, risk groups for risky driving behaviors were classified into a total of 5 stages using these data. . It is shown in Table 1 below.

[Table 1]

The average number of risky driving behaviors was about 8 cases. As shown in Table 1 above, the composition of risk group 1 to risk group 5 includes risk group 1, which corresponds to the safety level where risky driving behaviors rarely occur, risk group 2-3, which occurs relatively averagely, and risky driving behaviors The most frequently occurring risk grade was set at 4 to 5 levels.

Next, a cell risk area due to the road geometry is determined (S30).

Dangerous driving behavior is considered to be due to the driver's individual driving pattern. However, it may appear to be due to the road geometry. This can be distinguished by comparing the risky driving behavior occurrence statistics between the road geometry cell and the driver. If there is a lot of risky driving behavior in a specific area, it may be because there are many dangerous drivers in that area, but it may also be caused by the geometric structure of the road. The risky driving behavior occurrence determination based on the road geometry can be derived by comparing the risky driving behavior occurrence statistics for each cell and the risky driving behavior occurrence statistics for each individual driver.

More specifically, the step S30 is performed by calculating the number of risky driving behaviors for each cell and risk group. For each cell, the ratio of the risk group to the number of risky driving behaviors is calculated to calculate the risk level by the risk group for each cell. As for the risk level, it can be determined that the risk level is high if risky driving behaviors occur in all risk groups in the corresponding cell, and the risk level can be determined as low if risky driving behaviors occur only in some risk groups. See Table 2.

[Table 2]

Table 2 shows the result of determining the number of dangerous driving behaviors and the degree of risk according to the risk group by selecting two cells each in which a sudden course change/sudden overtaking type and a sudden left/right turn/quick U-turn type appear frequently. It can be seen that the risky driving behaviors generated in the four cells of Table 2 are generally distributed in all stages of the risk group.

As shown in FIG. 5, there is a possibility of risky driving behavior due to the road geometry within the cell, and cells in which risky driving behavior occurs in most of the risk group are investigated.

If it is judged to be a dangerous area by the road geometry of the cell and is judged to be a dangerous area only in the high-risk group, the cell is determined to be affected by drivers' risky driving behavior. However, if risky driving behavior occurs from most risk groups, it is judged to be influenced by the road geometry. The hatched portion in the drawing corresponds to this.

Table 3 shows cell configuration information in which risky driving behaviors frequently occur due to road geometry. Risk type 1 is an area where sudden left and right turns/refueling turns occur, and type 2 has a lot of sudden course changes/sudden overtaking.

[Table 3]

Next, risky driving behavior based on the road geometry is evaluated (S40).

Data were collected for a week as bus trip data registered with the Seoul area operation code via the 4 cells introduced in Table 3. Then, the data before and after applying the road geometry were compared and analyzed. The results are shown in Table 4.

[Table 4]

Cell 1 and Cell 2 are connected road areas, and even if the vehicle is operated at a normal speed, when the current calculation standard is applied, it is judged to be an area where sudden left and right turns and refueling turns may occur due to road geometrical structural factors. Dangerous driving behavior is calculated by applying the risky driving behavior calculation standard of this cell to the turning angle range of sudden left and right turns as +- 60° → +- 80°. Then, as shown in Table 4 above, as a result, it can be confirmed that the number of sudden left and right turns and refueling turns is reduced.

Cells 3 and 4 are intersection areas, and the driver must change the direction angle of the vehicle when passing these intersections to maintain the lane while driving, so it is judged to be an area where sudden change of route and sudden overtaking may occur due to road geometrical factors. do. Dangerous driving behaviors are calculated by applying the risky driving behavior calculation criteria of this cell to the range of turning angles for sudden course changes and sudden overtaking as +- 8° → +- 12°. As a result, it can be confirmed that the number of sudden course changes and sudden overtakings is reduced.

That is, it can be confirmed that the occurrence frequency of sudden course change, sudden overtaking, sudden right turn, sudden left turn, and refueling turn among the risky driving behavior types is reduced after applying the risk driving behavior calculation criteria reflecting the road geometry.

The types of risky driving behavior can be defined as shown in Table 5 below.

[Table 5]

In the present invention, the road geometry is reflected in the basic criteria for determining risky driving behavior in Table 5. The speed and heading angle thresholds of predefined grid cells are applied differently according to the geometric structure of the cell.

The risk driving behavior evaluation index calculation formula is as follows.

DRI _k : k is an evaluation index for risky driving behavior of autonomous vehicles, Pa is the number of occurrences of rapid acceleration, Pd is the number of occurrences of sudden deceleration, Pr is the number of occurrences of sudden turn, and Pc is the number of occurrences of rapid lane change. α, β, γ, and δ mean weights, and their values can be exemplified in Table 6 below.

[Table 6]

The closer the DRI _k value is to 100, the safer the k self-driving vehicle can be. On the other hand, as k is closer to 0, the self-driving vehicle can be judged to be dangerous.

We have previously described that the service server 100 of the present invention generates two types of safety indicators. The risky driving behavior evaluation index based on the road geometry was reviewed above. Now, let's take a look at how to evaluate safe driving related to the vehicle's surroundings based on sensor information collected in real time from the autonomous vehicle. This is an example of a process for longitudinal and lateral safety of an autonomous vehicle.

As described above, the service server collects data such as object recognition information, GPS location, speed, direction angle, vertical and horizontal acceleration from the autonomous vehicle (S100). The collected data is refined (S110). In the data refinement process, prior to the Vision/Lidar/Radar/GPS data analysis of autonomous vehicles supporting mobility for the mobility handicapped, mistakes in the investigation or input process are found and corrected. For example, processes such as minimizing loss and distortion of information caused by sensor malfunctions and limitations of information effective range through discovery of inconsistent data, correction of data inconsistency, and correction of missing values, removal of input errors, and smoothing of data noise are performed. Data noise smoothing refers to a process of smoothing out undesirable irregular fluctuations due to noise during data extraction by attenuating or removing such fluctuations or discontinuities.

Next, a data set to be analyzed is constructed (S120). This is real-time data accumulated from autonomous vehicle sensors, and can be composed of object recognition data collected from Vision/Lidar/Radar, GPS location/velocity/directional angle, and longitudinal and lateral acceleration data. In order to construct a sensor information-based safety evaluation analysis data set, it is desirable to observe the change in performance by time period (second unit) and compose continuous data that can quantify the performance.

Then, longitudinal and lateral acceleration safety evaluation is performed (S130). In order to evaluate the risk of longitudinal and lateral acceleration of autonomous vehicles supporting mobility for the transportation vulnerable, safety evaluation can be performed based on the evaluation criteria specified in the ISO standard. The evaluation content, evaluation method, collected data, and evaluation basis are shown in Table 7 below.

[Table 7]

Finally, collision safety with surrounding objects is evaluated (S140). The service server may calculate the expected time to collision (TTC) by analyzing the relative speed and the distance between vehicles while the autonomous vehicle is operating. The time to collision (TTC) is a value obtained by dividing the distance (m) from the vehicle behind the vehicle by the speed (m/s) of the autonomous vehicle. It is possible to calculate the speed of an autonomous vehicle for collision avoidance, and if the TTC is less than 2 seconds, it can be judged as a collision risk. In addition, the longitudinal separation distance from the vehicle behind the vehicle can be determined as a safety distance of at least 4m.

In this way, through sensor information collected in real time from the self-driving vehicle, it is possible to generate a safety index by comparing the speed of the currently operating self-driving vehicle with the collision avoidance speed to evaluate the degree of collision safety.

Safety indicators can be expressed in various ways. In some embodiments, an indicator indicating 'safe' and an indicator indicating 'dangerous' will be included. In particular, the service server performs a warning procedure when the latter index is met (S150).

The safety index evaluation value that the service server collects, analyzes, and warns of data may be transmitted to one or more of the self-driving vehicle, the device of the self-driving vehicle monitoring system, and the mobile device collecting information about the self-driving vehicle. there is.

For reference, the safety driving evaluation method of an autonomous vehicle supporting mobility for the disabled according to an embodiment of the present invention can be implemented in the form of program commands that can be executed through various computer means and recorded on a computer readable medium. . The computer readable medium may include program instructions, data files, data structures, etc. alone or in combination. Program instructions recorded on the medium may be specially designed and configured for the present invention, or may be known and usable to those skilled in computer software.

Examples of computer readable media include magnetic media such as hard disks, floppy disks and magnetic tapes, optical recording media such as CD-ROMs and DVDs, magneto-optical media such as floptical disks, and ROMs, RAMs, A hardware device specially configured to store and execute program instructions, such as flash memory, may be included. Examples of program instructions include high-level language codes that can be executed by a computer using an interpreter, as well as machine code such as those produced by a compiler. The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the present invention, and vice versa.

The protection scope of the present invention is not limited to the description and expression of the embodiments explicitly described above. In addition, it is added once again that the scope of protection of the present invention cannot be limited due to obvious changes or substitutions in the technical field to which the present invention belongs.

Claims (2)

The service server (a) collects sensor information of autonomous vehicles in real time while communicating with N autonomous vehicles (where N is an integer greater than 1) through a wireless network, calculates sensor-based risk assessment data, and generates sensor-based evaluation indicators. do,
(b) For M (M is an integer greater than 1) cells defined as independent zones on the driving road and assigned IDs, risky driving behaviors based on the road geometry are classified in advance into a plurality of grades, and stored in the system of the service server. In a registered situation, when an autonomous vehicle enters a specific cell, the risky driving behavior is calculated according to the risky driving behavior grade based on the road geometry of the cell to generate an evaluation index for the risky driving behavior;
(c) transmitting the sensor-based evaluation index and the risky driving behavior evaluation index to any one or more of an autonomous vehicle, a device of a system monitoring the autonomous vehicle, and a mobile device collecting information about the autonomous vehicle. Characterized in that it comprises, a safe driving evaluation method of an autonomous vehicle supporting mobility for the disabled.
According to claim 1,
The risky driving behavior corresponds to any one or more of a rapid acceleration type, a rapid deceleration type, a sharp turn type, and a rapid lane change type, a safe driving evaluation method for an autonomous vehicle supporting mobility for the disabled.

Images