KR101889607B1 - A Method and Apparatus for Decision and Control for Robust Platooning - Google Patents

Abstract

The present invention is based on a scenario in which a stable clustering operation that overcomes the limitations of platoon technology, which is a representative clustering technique, is present, and that there is a limitation of autonomous driving and convoy technology, is overcome, The present invention relates to a judgment and control method and a control and apparatus for driving a stable running on a scenario, including a driving algorithm for switching between different driving modes; The running algorithm includes a vehicle synchronization step S11 for synchronizing the clouded database with a storage device of a vehicle through wired / wireless communication, a driving technology activation step S12 for activating a driving technology suitable for a driving vehicle and a road condition, A real time vehicle control step S13 for controlling the movement of the vehicle in real time on the basis of a real time driving and a real time judgment and a control algorithm by recognizing the state around the road to the destination, (S15), if the vehicle is not maintained, determining a traveling mode to be switched on the basis of a determination algorithm mounted on the vehicle. If the vehicle is not maintained, the traveling mode is maintained (S16) To the cloud server through communication (S17).

Description

TECHNICAL FIELD [0001] The present invention relates to a judgment / control method for robust community driving,

More particularly, the present invention relates to a method and apparatus for judging and controlling a driving of a cluster, and more particularly, it relates to a method of controlling a driving force of a platoon, which is robust and stable, Control method and a determination / control device for a cluster running to overcome the limitations of the Convoy technology and the platooning technology to perform a stable running.

In recent years, research and development to provide intelligent vehicle services including autonomous driving have been carried out. However, until now, standardization and product development centering on general vehicles have been progressing, and public transportation (buses, trains, airplanes, Are expected to be expanded and considered.

The specification and related technology related to intelligent vehicle service is called "IAVE" under the name of ITS-G5, mainly in Europe, and the basic technology is "Vehicle to everything (V2X) communication .

In addition, a standard technology for providing enhanced V2X service is being developed centering on 3GPP, and consideration is also being given to linking with existing ITS-G5 and WAVE. Proximity service (ProSe) .

Autonomous driving technology is one of the representative intelligent vehicle service technologies, focusing on near-field communication using laser and radar, and it is expected to be complementary by utilizing infrastructure-based and bi-directional communication.

According to the publication No. 10-2015-0055052 (a method, a measuring apparatus and a control unit for adjusting the vehicle-end control), at least the first driver 210A and the second driver 210B, (400) for adjusting a control algorithm including at least one driver-dependent parameter as a control algorithm for controlling a control of a vehicle end (200) including a second vehicle (220B) And a measurement unit 230. The method of the present invention includes the steps of measuring 401 at least one body property of the first driver 210A based on measurements 401 performed at step 401, (Step 402) of determining a driver's stress level, and adjusting (step 403) the control algorithm to match the stress level of the first driver determined in step 402. "

Among the intelligent vehicle service technologies, there are three types of technology, depending on the level of cooperation among the vehicles. One is automobile-centered self-propelled vehicle, the other is convoy technology, and the other is , and platooning (platooning).

Convoy technology is a technology that keeps a common movement pattern and formation, so that each vehicle can control itself without a vehicle acting as a leader. Platoon technology, like convoy technology, maintains a common movement pattern and formation But there is a vehicle that acts as a leader, and the rest of the vehicle is basically controlled to follow the leader vehicle without intervention of the driver.

Among platooning technologies, platooning is basically an efficient way to reduce the distance between vehicles, which can reduce mass transportation and fuel consumption and reduce the number of drivers required. However, the existing platoon technology has a limitation in its application Many complementary technologies are required.

In the present invention, it is an object of the present invention to provide a control method and apparatus for stable running of a cluster which can overcome limitations of autonomous driving, convoy technology and platoon technology, which are representative driving techniques.

Yet another object of the present invention is to propose a method and apparatus for data conversion on a scenario basis, and propose a method and apparatus for achieving a robust and stable running for a scenario in which data is created.

Vehicle driving technology is largely divided into cognition, judgment, and control technology.

Specifically, the recognition technology basically refers to a technique of recognizing the situation around the vehicle and the unexpected situation based on the sensor technology.

Decision technology refers to technology that specifically determines how to deal with situations learned through cognitive technology. It is designed in the form of an algorithm.

Finally, control technology means controlling the machinery involved in the movement of the vehicle in accordance with the judgments. In the present vehicle, the steering wheel, the accelerator pedal, and the brake control of the vehicle correspond to this. In particular, when machine learning techniques are applied, judgment and control techniques may be applied in an integrated fashion.

If the steering wheel, accelerator pedal, brake, etc. of the vehicle disappear from the vehicle in the future, this will be the mechanism related to controlling the movement of the vehicle.

Based on scenarios based on data on the limitations of vehicle driving technologies such as autonomous driving, Convoy, and Platoon, a decision algorithm that determines the transition from on-going driving technology to other driving technology and 'control' 'Technology to overcome limitations by operating with other vehicle driving technology.

Here, the different vehicle driving techniques to be switched may specifically denote an intra-driving mode change in the current mode of operation of the vehicle, and may be an inter- driving mode change.

An example of the former is that a cluster vehicle operating as a platoon in one group can be operated in detail mode in a platoon technology operating in two separate groups as a platoon, It may be possible to switch to self-driving in which the vehicle groups travel individually.

The control method for the cluster running according to the present invention, taking the above circumstances into account, includes a running algorithm for switching between different running modes; The running algorithm includes a vehicle synchronization step S11 for synchronizing the clouded database with a storage device of a vehicle through wired / wireless communication, a driving technology activation step S12 for activating a driving technology suitable for a driving vehicle and a road condition, A real time vehicle control step S13 for controlling the movement of the vehicle in real time on the basis of a real time traveling and a real time judgment algorithm by recognizing the state around the road to a destination, a step S14 for judging whether or not the current traveling mode is maintained, A step S15 of determining a driving mode to be switched based on a judgment algorithm mounted on the vehicle, a step S16 of maintaining the current driving mode if the driving mode is maintained, To the cloud server (S17).

In addition, the control device for driving the crowd of the present invention can optimize the decision and control algorithm for switching from the autonomous driving, convoy and platoon techniques, And a machine learning algorithm for storing the machine learning algorithm.

The control device is characterized in that it is mounted on a stopping point, a line peripheral device (RSU) and a signal lamp so as to be able to indicate the driving technology of the vehicle through a base station or terminal type equipment mounted on a stopping point, a line peripheral device (RSU) .

There is an algorithm and related parameters that are provided in real time for the vehicles to run, and optimization is performed at the same time, and the central cloud server logically has two or more storage devices, A parameter obtained through optimization including data and learning, and the like, and the other storage device is a storage device in which parameters to be transmitted to each vehicle are stored.

And a mechanical device and a storage device corresponding to a virtual vehicle replacing each vehicle.

In the Convoy community driving control, the group type constituting the inter-vehicle cluster is controlled in the form of a large type, a wide type, or a combination of a large type and a wide type.

In the case of the Platoon communities, the group type constituting the community between the vehicles is controlled in the form of a large form, a wide form, or a combination of a form and a form.

Convoy When traveling in a cluster, a cluster of one group is divided into two or more convoy groups and controlled to travel in a cluster.

When Platoon communities are driven, the grouping of one group is controlled so as to run in two or more platoon groups.

Further comprising a central cloud server for performing algorithm optimization, and when the algorithm is machine learning, algorithm learning is performed in a central cloud server.

Scenarios 1 through 8 are included in the algorithm optimization stored in the cloud server.

When the algorithm is optimized for the cloud server, data to be considered is used as input data in the form of scenario data or data related to the mechanical movement of the vehicle.

When the algorithm is optimized for the cloud server, the data to be considered is used as output data in the form of scenario data or data related to the mechanical movement of the vehicle.

According to the preferred effect of the present invention, there is an advantage of realizing a stable community driving that can overcome limitations of autonomous driving, convoy technology and platoon technology which are typical driving technology, and scenarios in which there is a limitation of autonomous driving, convoy technology and platoon technology And it has an advantage of improving robustness and stability by supplementing it, and has an advantage of stabilizing running for a scenario in which data is created for each scenario and data is generated.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a flowchart showing a control method for a cluster running according to the present invention; FIG.
Fig. 2 is an example of autonomous running. Fig.
Figure 3 is an example of a convoy.
Figure 4 is another example of platooning.
Figure 5 is a first example of a platooning pattern;
Figure 6 is a second example of a platooning pattern.
Figure 7 is a third example of a platooning pattern.
Figure 8 is a fourth example of a platooning pattern.
Figure 9 is a fifth example of a platooning pattern.
10 is a block diagram showing a configuration of a device for communal running according to the present invention;
11 is a block diagram showing a configuration of a device for communal running according to the present invention.
12 is a block diagram showing a configuration of a central cloud apparatus according to the present invention;

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be described in detail with reference to the accompanying drawings.

First of all, there should be stored the data for each of the limiting scenarios for autonomous driving, convoy, and platoon corresponding to the vehicle driving technology.

Here, the data includes data obtained by a camera sensor that recognizes the surroundings of the vehicle and whether an obstacle is present in a person or a broken vehicle, and data obtained by a laser / radar sensor recognizing distance / speed / acceleration.

In other words, it contains basically three-dimensional spatial information related data related to the environment around the running vehicle.

Such data is shared not only by the data acquired by the individual vehicle but also by the wired / wireless communication to the server that has been clouded, so that the vehicles are stored in the form of integrated data, and each vehicle has a separate storage unit 110n (120n) Data can be transmitted and received by synchronizing the server 100 periodically or eventually. Basically, data corresponding to the perception of the surroundings of the vehicle is transmitted from the storage units 110n and 120n of the individual vehicle to the cloud server 100, and an algorithm corresponding to determination and control is transmitted from the cloud server 100 Lt; RTI ID = 0.0 > a < / RTI >

In addition, in recognition of the surroundings of the vehicle, a method of recognizing the surroundings of the vehicle more stably by utilizing direct communication between the vehicles and communication between the vehicle and the vehicle, in addition to recognizing based on the sensor devices on which the individual vehicle is mounted .

In a decision-making and control-algorithm algorithm that determines the vehicle driving technology to be switched with perceived data and activates the mechanical movement of the vehicle, , And the running of each vehicle is operated according to the following method.

Each vehicle is equipped with a platform and engine that can drive software corresponding to the determination and " control " algorithms. In addition, parameters required for operating the algorithm are downloaded through the cloud server 100. In this case, when the algorithm is a machine learning algorithm, the main parameters including the number of nodes, the number of layers, and the weights existing in each layer may correspond to each other.

Data acquired for each scenario by each vehicle is stored in the cloud server 100 via wired / wireless communication. Here, the data includes the identification ID, the position, the speed, and the acceleration of the subject vehicle and the surrounding vehicle with data related to the circumstance including the subject vehicle. It may also include parameters related to the handle, accelerator pedal, and vehicle (detail) mode associated with the mechanical movement of the vehicle.

Algorithm optimization can be performed in real time in the central cloud server 100. In the case of using a machine learning algorithm, learning is performed in the central cloud server 100, and more optimized parameters are transmitted to each vehicle through cloud synchronization. At this time, the cloud server 100 must have a judgment and control algorithm parameter storage unit 203 for storing the parameters updated each time, and the judgment and control algorithm parameter storage unit 203 ) ≪ / RTI > In addition, it is required to have a vehicle and a surrounding situation data storage unit 201 for storing scenario related data transmitted from each vehicle periodically or eventually.

The central cloud server 100 operates in two major forms. Specifically, the algorithm optimization unit 202 is divided into an operation for updating and synchronizing the determination and 'control' algorithm through the algorithm optimization unit 202, and an operation for activating an algorithm for applying the algorithm to an actual vehicle. In the latter case, it mainly involves transmitting the optimized parameters stored in the judgment and control algorithm parameter storage 203 of the cloud server 100 to each vehicle via the individual vehicle transfer wired and wireless communication unit 204 .

Each vehicle decides whether to switch from the current driving skill to another driving skill through a judgment and 'control' algorithm. In this case, the term " other driving technology " may mean an intra-driving mode change in the current driving technology, and an inter-driving mode change other than the current driving technology It may mean.

As we have seen above, the scenarios facing each vehicle vary, but for the sake of clarity, specific scenarios are exemplified.

Scenario 1

As shown in Fig. 2, the vehicle A, the vehicle B, and the vehicle C on the highway are autonomously driven individually.

In such a situation, it is necessary in terms of energy efficiency to switch the autonomous driving mode to the cluster driving mode.

Scenario 2

As shown in Fig. 3, there are vehicles A to E on a general road, and the vehicles A to C run in a convoy manner, and the vehicles D and E travel in a manual operation mode .

In such a situation, since the vehicles A to C running in the communities have the obstacle 30 in the front and a plurality of vehicles in the other lane, it is necessary from the standpoint of stability to switch to the autonomous traveling mode individually.

Scenario 3

As shown in FIG. 4, the vehicles A to C are running in a platooning manner on a general road, and the vehicle D is running in a manual operation mode in the other lane.

It is necessary to switch the traveling mode of the vehicle A to the vehicle C in a convoy or an autonomous mode in a situation where the vehicle D is in a position close to the vehicle C with the obstacle in front of the vehicle A.

Scenario 4

As shown in FIG. 5, the scenarios in which the vehicles A to C on the highway are running in a general platooning type cluster.

Scenario 5

As shown in FIG. 6, unlike the conventional platooning in which a group is formed in a large scale, a group is formed by a horizontal line, and a platooning cluster is driven. In such a scenario, You can keep it.

Scenario 6

As shown in Fig. 7, when the vehicle serving as a leader in scenario 4 is vehicle A, while the vehicle acting as a leader in scenario 6 is vehicle C, Soon, platooning is a scenario in which a vehicle that plays a leadership role in the course of a cluster run is switched.

Scenario 7

As shown in FIG. 8, a case in which the vehicles A to C proceeding by forming a group of platooning communities traveling in a large scale are divided into two independent groups and travel in a community.

Scenario 8

As shown in FIG. 9, a case in which the vehicles A to C are divided into two independent groups and travels in a cluster is formed by forming a group of platooning communities by a horizontal line.

On the other hand, in the cloud synchronization, wired / wireless communication is utilized between the cloud server and the vehicles. In particular, when utilizing wireless communication, the following methods can be utilized.

In other words, in transmitting the parameters for operating the algorithm from the cloud server to the various groups of vehicles that make up the cluster driving, the application of the efficient communication method can be considered through the individual vehicle transfer wired / wireless communication unit 204.

For example, the star vehicle transfer wired / wireless communication unit 204 transmits the parameters for the algorithm operation to the individual vehicles belonging to the various groups including the autonomous running including the autonomous running in the unicast, broadcast or multicast manner. In this case, when at least one vehicle of each group receives the group driving when platooning is carried out, the vehicle that transmits the parameters obtained from the cloud server 100 through direct communication among the vehicles in the group, All acquire the corresponding parameters. At this time, direct communication includes unicast, broadcast, or multicast.

In the present invention, a device for switching between different traveling modes is proposed, and a traveling algorithm as shown in FIG. 1 is performed based on the device.

As shown in Fig. 1, the driving algorithm can be divided into seven steps.

That is, a vehicle synchronization step S11 for synchronizing a clouded database with a storage device of a vehicle through wired / wireless communication, a driving technology activation step S12 for activating a driving technology suitable for the vehicle and road conditions, A real time vehicle control step S13 for controlling the movement of the vehicle in real time on the basis of a real time driving and a real time judgment algorithm, a step S14 for judging whether or not the current driving mode is maintained, If it is determined that the current driving mode is not to be maintained, the driving mode to be switched is determined based on the determination algorithm mounted on the vehicle (S15). If it is determined in step S14 that the current driving mode is to be maintained, (S16), the data stored in the storage device of the vehicle is transmitted to the cloud server through wired / wireless communication It includes a step (S17) for sending. At this time, S17 can be sent immediately after acquiring data, and can be transmitted periodically or eventually to the cloud server.

The components for carrying out the traveling algorithm as described above will be described as follows.

As shown in FIGS. 10 and 11, storage units 110a and 110b in which scenarios having respective limitations for the autonomous driving, convoy, and platoon corresponding to the vehicle driving technology are stored as data in the respective vehicles. (110n) are mounted.

It is also possible to use the vehicle driving technology even through the base station or terminal type equipment mounted on a stopping point, a road side unit (RSU), a signal lamp, etc. instead of the storage units 110a, 110b, In this case, the present invention is also applicable to a storage device in which the above data is stored in a stopping point, a line peripheral device (RSU) and a traffic light.

12, the central cloud server 100 is provided with a vehicle and surrounding state data storage unit 201, an algorithm optimization unit 202, a determination and control algorithm parameter storage unit 203, And a communication unit 204 may be provided.

The vehicle and surrounding situation data storage unit 201 stores information such as a vehicle ID, a location, a speed, an acceleration, and the like. It is not necessary that data acquired through a physical vehicle necessarily be provided, It is possible.

The algorithm optimizer 202 learns based on scenario-related data in the case of machine learning, optimizes the algorithm through real-time updating, and thus more optimized parameters can be delivered to each vehicle. The determination and control algorithm parameter storage unit 203 includes the number of nodes, the number of layers, and the weight per layer in the case of machine learning. It may also include parameters related to the handle, accelerator pedal, and vehicle (detail) mode associated with the mechanical movement of the vehicle.

The individual vehicle transfer wired / wireless communication unit 204 may transmit the optimized algorithm to the individual vehicle via wired / wireless communication by unicasting, broadcasting, multicasting, or a mixture of these methods.

At this time, since there is an algorithm and related parameters to be provided in real time for driving the vehicles, and optimization can be performed at the same time, the central cloud server 100 may logically have two or more storage means. Here, one storage means is storage means for storing scenario related data and parameters obtained through optimization including optimization during the optimization, and the other storage means is a storage means for storing parameters to be transmitted to the respective vehicles It is means.

100; A cloud server 110; platoon group 1
110a; A storage unit 110b of the vehicle 1; The storage unit of the vehicle 2
110n; A storage unit 120 of the vehicle n; platoon group 2
120a; A storage unit 120b of the vehicle 1; The storage unit of the vehicle 2
120n; The storage unit of the vehicle n
201; A vehicle and surrounding situation data storage unit 202; Algorithm Optimization
203; The judgment and control algorithm parameter storage unit
204; Individual vehicle delivery Wired /

Claims (13)

delete CLAIMS 1. A control device for a cluster running,
The control unit includes a storage and retrieval unit for storing a machine learning algorithm for optimizing the determination and control algorithm by learning, an autonomous drive, convoy and platoon technology, ; And a mechanical device and a storage device corresponding to a virtual vehicle replacing each vehicle are mounted in the vehicle.
The method of claim 2,
The control device is characterized in that it is mounted on a stopping point, a line peripheral device (RSU), and a signal lamp so that the vehicle traveling technology can be instructed through a base station or terminal type equipment mounted on a stopping point, a line peripheral device (RSU) And a control device for driving the cluster.
The method of claim 2,
Wherein said optimization means comprises logically two or more storage means for optimizing algorithms and associated parameters provided in real time for running said vehicles, wherein one storage means comprises means for optimizing, A storage means in which a variable is stored, and another storage means includes a central cloud server having storage means in which parameters to be transmitted to respective vehicles are stored.
delete The method of claim 2,
Wherein the control unit controls the group form of the inter-vehicle communities in the form of a large form, a long form, or a combination of a long form and a long form in the control of the Convoy group.
The method of claim 2,
Wherein the control unit controls the group form of the inter-vehicle clusters in the form of a descending, transverse, or combination of descending and descending.
The method of claim 2,
Wherein the control unit controls the plurality of convoy groups to be divided into two or more convoy groups when the vehicle is traveling in the Convoy group.
The method of claim 2,
Wherein the plurality of platoon groups are grouped into two or more platoon groups when the platoon is driven.
The method of claim 2,
Further comprising a central cloud server for performing algorithm optimization, and in the case where the algorithm is machine learning, algorithm learning is performed in a central cloud server.
The method of claim 10,
And the scenarios 1 to 8 for algorithm optimization stored in the cloud server.
The method of claim 10,
Wherein the data to be considered is used as input data in the scenario type data or the data related to the mechanical movement of the vehicle when the algorithm is optimized in the cloud server.
The method of claim 10,
Wherein the data to be considered is used as output data when the algorithm is optimized for the cloud server, the data being related to the data of the scenario type or the mechanical movement of the vehicle.

Images